University of Southern California Dissertations and Theses

The role of local actors in international development: how incentive types affect foreign aid effectiveness

(USC Thesis Other)

The role of local actors in international development: how incentive types affect foreign aid effectiveness

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE ROLE OF LOCAL ACTORS IN INTERNATIONAL DEVELOPMENT:
HOW INCENTIVE TYPES AFFECT FOREIGN AID EFFECTIVENESS
by
Jennifer Rogl a
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulllment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
POLITICAL SCIENCE AND INTERNATIONAL RELATIONS
December 2020
Copyright 2020 Jennifer Rogl a
Acknowledgements
This research was made possible through the generous support of the USC Provost's Fel-
lowship, the U.S. Department of Education's Fulbright-Hays Doctoral Dissertation Research
Abroad (DDRA) Fellowship, the USC Research Enhancement Fellowship, USC Center for
International Studies grants, and summer funding from the USC School of International
Relations and the Sherwin Family Advanced Graduate Student Research Support Fund.
I would like to express my deepest gratitude to so many people, far beyond what I can
t here, but I will try. To my chair Carol Wise, whose red pen has opened up so many new
opportunities. Thank you for adopting me, ghting for me, and always having my back.
To my committee member Gerry Munck, whose advice and enthusiasm for my project kept
me motivated, and my committee member Je Nugent for sharing his vast experience in
development economics and love of Costa Rica with me.
To my Fulbright-Hayes host institution advisor Eli ecer Vargas at CATIE in Turrialba,
Costa Rica, for welcoming me with open arms and giving me so many opportunities to inter-
act with students here. To Alejandro Imbach, for our hours of conversations, your books, your
candid thoughts, your advice on building my measurement tools, and sharing your wealth
of expertise with me. To Isabel Guti errez Montes, who introduced me to new development
theory and provided me a home in Posgrado at CATIE. To my research assistants, Riley
Mallon, Jes ua Reyna M endez, Claudia Rojas Rios, Andr es Madrid, Natalia Ruiz Guevara,
Hannah Rae Warren, Juan Mart n Guti errez Trejos - estar e perdida sin ustedes, tal como
ayudantes y como amig@s. To my fellow doctoradas, Laura Escarraga and Mar a Lourdes
Maldonada, for our mutual support with translations, research methods, spaces for writing,
and reminders to take breaks. To the individuals who helped me track down contacts and
gave advice for visiting eld sites even when it did not relate to their projects: Miguel Ci-
fuentes Jara, Lilibeth Matute Ir as, Adri an Vargas Cambronero. To the wonderful people
at the US Embassy in Costa Rica who helped facilitate my Fulbright-Hays stay: Carolina,
Shirley, Merlyn. To every single one of my interviewees, I am greatly indebted.
To Veri Chavarin, Graduate Advisor and Assistant Director of the POIR Program, whose
vast support allowed me to achieve so much more than just a Ph.D. during my years at USC.
As many POIR students would say, I would have been lost without her. Additionally, I would
like to deeply thank previous POIR Faculty Director Saori Katada; Linda Cole in the School
of International Relations; and Muhammed Ibrahim, Director General of CATIE. To Alison
Dundes Renteln, who helped get me to a doctoral program in the rst place. To Ange-Marie
Hancock Alfaro, to bring me on as an RA and for your thoughtful advice.
To my POIR cohort: I am beyond lucky to have gone through this with you. To Stephanie
Kang and Suzie Mulesky, our semesterly all-day feedback sessions and chats made my work
innitely better. To Victoria Chonn Ching, my go-to expert on economics questions and the
caretaker of my life while out in the eld. To Whitney Hua and Nola Haynes, my co-ghters
for rights and representation in everything we do. To Therese Anders, Mao Suzuki, and Gabi
Cheung, for their thoughtful support of my work and for some amazing travel buddies. To
my POIR companions overall, I chose this program because of the support and friendship I
saw between students, and I am so glad to say it was genuine in the end.
ii
To my CATIE familia: Las experiencias que tuve con ustedes durante dos a~ nos han sido
algunas de las m as poderosas de mi vida. No me gustar a vivir en la burbuja del CATIE
con nadie m as; compartir los altibajos de la vida, intercambiar conocimientos y costumbres
culturales, aprender lecciones, viajar, recuperarse, compartir comidas, bailar en el ranchito,
y todo lo dem as con ustedes, ha sido uno de los grandes honores de mi vida. Me fui no solo
con amistades, sino con herman@s, muchos de l@s cuales s e que volver e a ver.
To my friends and family. I do not know where I would be without you. To my Mom,
for not letting me give up my Ph.D. dream. To my best friend Hannah, sharing life with you
all over again during this chapter has been one of the most special parts of this experience.
To Devin, who provided an escape on and o campus. To my Helenes - after all these years,
our friendships have only grown stronger, especially to Tasha and Kristine for giving me a
base for eldwork in the Philippines, and to Erin, Megan, Kristine and their families for
always housing me in D.C. To Brittany for stepping in to help me balance my life when I
most needed it. And to the dozens and dozens of amazing people who supported me during
a very dicult time as I tried to balance many intense events in 2020 as I nished my Ph.D.,
I do not think I could ever thank you profusely enough.
To Eli an. You were with me in my eldwork, you presented with me at conferences, you
defended my dissertation with me, and you'll be with me on my graduation day. I promise
to honor my and everyone's love for you by always working towards a better world for you.
iii
Table of Contents
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1 The Elusiveness of Development and the Role of Foreign Aid . . . . . . . . . . . . 1
1.1 Research Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Justication and Implications . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Organization of the Present Study . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 The Changing Theories and Practices of Development . . . . . . . . . . . . . 5
1.4.1 The Evolution of Development Theory . . . . . . . . . . . . . . . . . 6
1.4.2 The Evolution of Development Practice . . . . . . . . . . . . . . . . . 14
2 The Billion Dollar Question: How do we Measure Aid Eectiveness? . . . . . . . . 30
2.1 Micro-Level Measures: Aid at the Project Level . . . . . . . . . . . . . . . . 32
2.1.1 What is a Project? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.1.2 How are Projects Designed? . . . . . . . . . . . . . . . . . . . . . . . 35
2.1.3 How is Project Evalution Dened? . . . . . . . . . . . . . . . . . . . . 42
2.1.4 How are Projects Evaluated? . . . . . . . . . . . . . . . . . . . . . . 46
2.1.5 What do we Measure in Project Evaluation? . . . . . . . . . . . . . . 51
2.1.6 Who Evaluates Success and Failure? . . . . . . . . . . . . . . . . . . 53
2.1.7 What do Project Evaluations Look Like? . . . . . . . . . . . . . . . . 58
2.1.8 What Are Project Evaluations Used For? . . . . . . . . . . . . . . . . 59
2.1.9 How Common Are Project Evaluations? . . . . . . . . . . . . . . . . 63
2.2 Macro-Level Measures: Aid at a National Level . . . . . . . . . . . . . . . . 65
2.3 The Relationship Between Development Theory and Measurement . . . . . . 68
2.4 The Micro-Macro Paradox and Interpreting Results . . . . . . . . . . . . . . 70
3 A Theory of Incentive Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1 Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.1.1 Local Actors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.1.2 Buy-In . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2 Theories of Institutions and Incentives . . . . . . . . . . . . . . . . . . . . . 76
3.2.1 Principal-Agent Conceptions of Aid . . . . . . . . . . . . . . . . . . . 76
iv
3.2.2 The Octangle and Aid Delivery . . . . . . . . . . . . . . . . . . . . . 80
3.2.3 Team Production as a Clarifying Lens . . . . . . . . . . . . . . . . . 85
3.3 A Theory of Incentive Types . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.3.1 Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.0.1 Macro Context: Choosing Country Cases . . . . . . . . . . . . . . . . 92
4.0.2 Micro Context: Choosing Development Aid Projects Within Country
Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
4.1 Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4.1.1 Dependent Variable: Project Outcomes . . . . . . . . . . . . . . . . . 99
4.1.2 Independent Variable: Incentive Types for Local Actors . . . . . . . . 110
4.2 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
4.2.1 Project Identication . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.2.2 Archival Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.2.3 Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.2.4 Site Visit and Interviews . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.2.5 Compiling and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 118
4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5 The Case of Costa Rica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
5.1 Case Selection: Costa Rica . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
5.1.1 Macro Context: Costa Rica as a Case of Development `Success' . . . 121
5.1.2 Micro Context: Project Selection . . . . . . . . . . . . . . . . . . . . 133
5.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
5.2.1 Proyecto Manglares . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.2.2 Proyecto Carbono . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.2.3 AC3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6 The Case of Mauritius . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
6.1 Case Selection: Mauritius . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
6.1.1 Macro Context: The \Mauritian Miracle" of Development . . . . . . 160
6.1.2 Comparing Costa Rica and Mauritius . . . . . . . . . . . . . . . . . . 170
6.1.3 Micro Context: Project Selection . . . . . . . . . . . . . . . . . . . . 173
6.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
6.2.1 SWMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
7 Conclusion: The Power of Nonmaterial Incentives in Aid Eectiveness . . . . . . . 185
7.1 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
7.2 Further Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
7.2.1 Development Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
7.2.2 Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
7.2.3 Theories of Institutions and Incentives . . . . . . . . . . . . . . . . . 188
7.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
7.3.1 Country Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
v
7.3.2 Sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
7.3.3 Time and Resource Constraints . . . . . . . . . . . . . . . . . . . . . 190
7.3.4 Implementers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
7.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
A Field Interview Responses on Aid Evaluation . . . . . . . . . . . . . . . . . . . . 193
A.1 Local Actors: Recipient Government and Implementers . . . . . . . . . . . . 193
A.2 Donors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
B List of Project Documents Collected . . . . . . . . . . . . . . . . . . . . . . . . 197
B.1 Costa Rica: Proyecto Manglares . . . . . . . . . . . . . . . . . . . . . . . . . 197
B.1.1 Documents and Academic Articles . . . . . . . . . . . . . . . . . . . 197
B.1.2 Interviews and Meetings . . . . . . . . . . . . . . . . . . . . . . . . . 202
B.2 Costa Rica: Proyecto Carbono . . . . . . . . . . . . . . . . . . . . . . . . . . 202
B.2.1 Documents and Academic Articles . . . . . . . . . . . . . . . . . . . 202
B.2.2 Interviews and Meetings . . . . . . . . . . . . . . . . . . . . . . . . . 211
B.3 Costa Rica: AC3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
B.3.1 Documents and Academic Articles . . . . . . . . . . . . . . . . . . . 212
B.3.2 Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
B.3.3 Interviews and Meetings . . . . . . . . . . . . . . . . . . . . . . . . . 217
B.4 Mauritius: SWMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
B.4.1 Documents and Academic Articles . . . . . . . . . . . . . . . . . . . 218
B.4.2 Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
B.4.3 Interviews and Meetings . . . . . . . . . . . . . . . . . . . . . . . . . 219
C Tool to Measure Institutionalization of Project Outcomes . . . . . . . . . . . . . 221
D Tool to Measure Incentive Types . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
vi
List of Tables
1.1 The Evolution of Development Theories and their Impacts on Aid Approaches 13
2.1 Aid Project Terminology in LogFrame and RBM Approaches . . . . . . . . . 36
2.2 Examples of Aid Project Evaluation Terminology Used by Donors . . . . . . 52
2.3 Selected Responses Concerning Local Actors Evaluation of Projects . . . . . 56
4.1 Dierences in Traditional vs. Process-Based Development Project Measure-
ments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.1 Summary of Aid Projects Chosen . . . . . . . . . . . . . . . . . . . . . . . . 134
5.2 Summary of Results: Project Impacts and Incentives . . . . . . . . . . . . . 145
5.2 Summary of Results: Project Impacts and Incentives - Continued . . . . . . 146
6.1 Examples of Development Bodies in the Mauritian Government . . . . . . . 168
6.2 Summary of Selected Development Indicators for Costa Rica and Mauritius . 174
vii
List of Figures
2.1 The Basic LogFrame Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.2 Table of Contents Examples from UK DFID Project Evaluations . . . . . . . 60
2.3 Aggregated Results Example from World Bank Project Evaluations . . . . . 61
2.4 Questions from the USAID Decision Tree for Selecting the Evaluation Design 64
3.1 Typical Actors in the Foreign Aid Implementation Chain . . . . . . . . . . . 74
3.2 Principal-Agent Relationships in the Aid Implementation Chain . . . . . . . 78
3.3 Actors in the Development Aid Implementation Octangle . . . . . . . . . . . 82
4.1 Main Characteristics of Foreign Aid . . . . . . . . . . . . . . . . . . . . . . . 97
4.2 Basics of an Institutionalization Measure to Assess Project Outcomes . . . . 106
5.1 Network Map of Actors Involved in OLAFO . . . . . . . . . . . . . . . . . . 137
5.2 Network Map of Actors Involved in Proyecto Carbono . . . . . . . . . . . . . 140
5.3 Network Map of Actors Involved in AC3 . . . . . . . . . . . . . . . . . . . . 144
5.4 Photos from Proyecto Manglares Field Site Observation . . . . . . . . . . . . 150
5.5 Photos from Proyecto Carbono Field Site Observation . . . . . . . . . . . . . 153
5.6 Photos from AC3 Field Site Observation . . . . . . . . . . . . . . . . . . . . 155
6.1 Worldwide Governance Indicator Scores for Costa Rica and Mauritius (2018) 173
6.2 Network Map of Actors Involved in SWMP . . . . . . . . . . . . . . . . . . . 179
6.3 Photos from SWMP Field Site Observation . . . . . . . . . . . . . . . . . . . 184
viii
Abstract
What role do local project actors play in development aid outcomes? Theories of aid ef-
fectiveness still center around macro-level variables and country-level outcomes, and suer
from profound measurement problems, producing many contradictory results. In contrast,
the goal of my dissertation is to analyze the micro-level variables that shape aid project
outcomes in a targeted country. What is the dynamic between relevant institutions, the
organizations that enforce them, and local project actors? I posit that the type of incentives
used by institutions to induce local actor `buy-in' into foreign-funded development projects
relate directly to the institutionalization of outcomes, namely that nonmaterial incentives
will lead to longer-lasting project outcomes. To study this phenomenon, I conducted archival
research, eld observations and interviews on the implementation and outcomes of four for-
eign aid projects executed between 1992-2016 in Costa Rica and Mauritius and developed a
new measurement tool for project outcomes and incentive types. Analysis of the data shows
support for the theory that nonmaterial incentives lead to longer-lasting project impacts,
though the eect is mediated by which local actor was targeted. This study allows us to bet-
ter understand the conditions under which aid is eective in the long-term, to conceptualize
and operationalize aid eectiveness beyond the macro-level, and oers clear policy impacts
and innovative scientic evaluation tools in the arena of international aid and development.
ix
Chapter 1
The Elusiveness of Development and the Role of
Foreign Aid
The purpose of this dissertation is to understand how analyzing development aid at the
project-level - and particularly the role that local actors play in this type of international co-
operation - can help us better answer the question posed in much of the aid and development
literature: under what conditions is aid eective? The reasons for development success or
failure are undertheorized at the micro-level, and one major gap in the literature is the role
that local aid beneciaries play in development outcomes. The aid eectiveness literature
has yet to fully consider the importance of individual inputs by each actor or the incentives
that drive their performance. The literature has focused instead on country-level outcomes
and tends to assume that a donor's input is the key variable for the success of an aid project.
It is time to bring in critical local actors - dened here as any actor who is from or whose
interests lie in the recipient country - and the essential role they play in this process.
My focus here will be on aid for development. Ocial development assistance (ODA)
is government aid meant to fund economic development and welfare projects in the form
of grants, concessional loans with a grant portion of at least 25%, or technical assistance
(OECD, 2016).
1
My approach is two-fold: rst, I make the case for how the eld of de-
velopment has changed in the past few decades and why research and measuring results
at the project-level can help us resolve con
icting information about the current state of
development. Second, I propose we analyze aid projects by building a new theory that
1
It is important to emphasize that the ocial denition of development aid does include certain types of loans,
not only grants.
1
adapts previous frameworks to make an argument linking local actor incentive types to the
duration of project outcomes, which I empirically analyze drawing on cases in Costa Rica
and Mauritius. Through this research endeavor, I contribute to the literature by helping us
move beyond macro-level studies of aid eectiveness to the micro-level, and generate theory
concerning the circumstances under which aid project impacts will last. Relatedly, I generate
propositions about the causal mechanisms that are beneath the level of the state, and look
beyond the economic realm for the politics informing these interactions and ultimately aid
eectiveness.
I conducted eld research on the implementation and long-term outcomes of three foreign
aid projects executed between 1992-2016 in Costa Rica and Mauritius. Representing several
sectors, donors, and initiation approaches, I analyze the relationship between the incentives
used to obtain the buy-in of local actors and the outcomes of these aid projects. Results
show support for my theory that the presence of nonmaterial
2
incentives lead to longer-lasting
project impacts, in contrast to the conventional wisdom that economic incentives alone are
necessary and sucient to achieve desired project results. The strength of the cooperative
interactions between on-the-ground implementing organizations and the social incentives
such relationships create has fostered an environment for longer-lasting project outcomes,
which may also provide a basis for eventually understanding the link between micro-level
projects and macro-level outcomes on the aggregate national level. This endeavor allows us
to better understand the conditions under which aid is eective in the long-term, and gives
us a way to conceptualize and operationalize aid eectiveness beyond macro-measures.
1.1 Research Question
The relationship between incentives and the duration of outcomes is clear in seminal work
by Gibson et al. (2005): \The sustainability of aid depends on how incentives structure
2
I use the terms nonmaterial and nonmonetary interchangeably in this dissertation up through the section on
operationalization of my variables, where I only use nonmaterial after.
2
interactions between or among the key corporate and individual actors" [63]. Yet, what
types of local actor incentives lead to longer lasting development project outcomes? The
conventional wisdom derived from past and present mainstream development theories such
as modernization and neoliberalism is that monetary incentives are necessary and sucient
motivators for local actors to buy into development aid projects initiated by foreign donors.
These theories put an emphasis on economic growth as a driving factor of development at
the national level, thus, lack of material capital is construed as a country's limiting factor
to development. This has in turn in
uenced project design to include monetary bonuses
or other material goods like land, seeds, equipment, etc., as lack of capital is also assumed
to be the reason local actors cannot achieve their development goals at the project level.
More recent and less mainstream theories have put emphasis on the importance of other
nonmonetary aspects of development, though none have fully accounted yet for the recent
practitioner emphasis on incentivizing local buy-in through participatory (i.e. nonmonetary)
means over the past two decades.
Are the more mainstream development theories right - are economic incentives the most
powerful motivators for local actors in development cooperation? Or are the newer, less
mainstream theories and practitioners right - do other sorts of nonmonetary incentives pro-
vide more powerful motivators and/or long-lasting results? This study attempts to reconcile
these con
icting messages by bringing in literature on incentives and institutions to answer
this important development question, ultimately hypothesizing that nonmaterial incentives
will lead to longer-lasting outcomes over other incentive types.
1.2 Justication and Implications
This is a timely project. At least 10 articles have been published on questions of foreign
aid in the past two years in the top IR journal alone, International Organization, yet many
of these authors still express frustration with poor measures and oer innovative ways to
crack the code on aid eectiveness. Most authors cannot entirely escape some of their own
3
criticisms.
3
Finding better ways to accurately measure aid eectiveness has become increas-
ingly important, as donor funding grows more scarce and stakeholder concerns intensify.
The micro-level allows us to evaluate independent actions by locals that may reveal fur-
ther information about project results, potentially giving us the ability to resolve con
icting
results.
A concern with the `local' turn in development practice has also entered into topical
studies on development across disciplines. Practices have quickly shifted attention towards
local actors { much faster than the generation of theories to explain this shift. Are local
actors decisive in project outcomes, and if so, why? What are their roles within projects,
and how are they organized? Does this focus on local actors result in more eective aid?
We have little information that allows for rigorous study across cases at the project-level. In
particular, we know little about local actors or project beneciaries, despite the clear power
they hold within projects. However, political science and IR provide unique advantages for
studying these actors, as an understanding of power and its dierent facets can provide unique
insights across cultural and geographical contexts. This may also allow us to eventually scale
up and provide linkages to the macro level, as will be discussed in Section 4.1.1.
The results of my study have important implications for development research. Com-
paring project experiences in macro-environments can: 1) highlight the value of looking at
aid projects outside of the typical principal-agent framework; 2) illuminate the power of
aid recipients in shaping state-level outcomes; and, 3) inform policymakers on key aspects
of project design and thus promote the more eective use of funds. Based on my initial
research, I am convinced that aid recipients wield considerable power during aid project
implementation. However, we need to know considerably more about this major variable
when making macro-level inferences about the conditions under which aid is eective.
My research joins a growing body of IR scholarship that recognizes the independent causal
power of local actors like aid recipients, even in interactions with huge power asymmetries.
3
See: Bermeo (2016), Dietrich (2016).
4
Rosenau (1990), for example, stimulated an ongoing argument that the role of individual
actors will increasingly matter within the world politics arena. This notion informs my own
thinking, which is that an analysis of the impact of everyday people on development is now
key for understanding macro-level outcomes. My innovative exploration of project politics
and incentives at the micro-level combines approaches within the elds of economics, IR,
and development studies, rendering this a vital interdisciplinary approach to the study of
global problems.
1.3 Organization of the Present Study
This study is organized into seven chapters. Chapter 1 introduces the research project and
the changes in development studies and practice over the past few decades that have moti-
vated this research. Chapter 2 reviews the literature on aid eectiveness, namely how it has
been measured, along with a discussion on the politics and controversies surrounding mea-
surement practices and academic research on the topic. Chapter 3 introduces a theoretical
approach to study aid eectiveness at the micro level across cases that involves local actor
perceptions and experiences. Chapter 4 presents my methodological approach, including two
innovative measurements I created for the purposes of this study: an incentive classication
system and a project outcome measurement tool with its base in evaluating the institution-
alization of change. Chapters 5 and 6 present my two country case studies, Costa Rica and
Mauritius, which are widely considered to be development successes since the 1980s and thus
provide a stable macro context to facilitate the study of project-level variables. Chapter 7
concludes the discussion, and considers the implications of my results.
1.4 The Changing Theories and Practices of Development
It is important to position this research within the current state of development practice and
studies. Lessons learned by donors and recipient countries have led to many changes in the
5
practice of development, and theories have continuously evolved to explain unexpected devel-
opment results. There has been a normatively driven move by donors away from giving lump
sums of aid to governments, termed budgetary support, and away from donor-identied and
donor-driven projects. Instead, aid has taken a more `local' turn. Participatory approaches
have become more common, where beneciaries identify their biggest challenges and aid is
awarded to recipient-led projects through a request for proposal (RFP) application process.
These new approaches have preceded changes in broad academic theories, which tend to be
more responsive to than predictive of development in the real world. While not every devel-
opment organization evolves at the same rate, mixed aid eectiveness results and the entry
of social entrepreneurship, microlending, and increasingly powerful NGOs such as the Gates
Foundation, Hewlett Foundation, and Vulcan into the aid arena has forced many donors
to rethink their practices long before researchers build new theories. I will rst discuss the
evolution of development theory, and then the changes in development practice over the
past two decades. I will nish this section by situating this study in the current scheme of
development debates through a political lens.
1.4.1 The Evolution of Development Theory
Since its beginnings, development theory has primarily focused on the macro level: theories
that attempt to explain development patterns at the national or international level. These
theories also drive development strategies, policy, practice, and projects at the local level. In
disbursing development aid, donors are typically assumed to be the only actors with power.
In international relations (IR), we think about power as the ability to get another actor
to intentionally act in a way you want them to, even if not in their interest.
4
This does
not, however, mean that development theories - including those outside of the mainstream
discipline - have no stance on whether local actors have power in development interactions,
even if not explicitly stated. I will brie
y summarize the main development theories and
4
This denition is based o of seminal work by Dahl (1957) and Keohane and Nye (1973).
6
comment on how power in general is conceived of in these paradigms, providing the basis
for my theoretical discussion in Chapter 3 on why principal-agent conceptions of aid imple-
mentation are problematic when evaluating aid eectiveness (see Table 1.1 for a summary
of these theories, their relationship with real world events, and some key authors).
Modernization Theory
We have seen innovations in the who, what, when, where and why of development theory
almost every decade, starting in the 1960s when formal aid agencies were created in the post-
World War II rebuilding period. This era gave rise to modernization theory, which attempted
to explain how `traditional' societies became `modern.' One of the most in
uential authors
emerging in this period was Rostow (1959), whose proposed stages of growth for a developing
country that were based on the experience of wealthy countries. Countries began in Stage 1
as `traditional' societies concentrated on subsistence farming and bartering for trade. They
reached Stage 2 with the introduction of specialization and the reinvestment of surpluses,
leading to industrialization in Stage 3. Economic diversication and reduced imports were
hallmarks of Stage 4, and Stage 5 was the pinnacle of modernity, marked by rising income,
mass consumption, and a service economy.
Modernization theory in
uenced donor decisions about the kinds of technology and other
industrial innovations recipient countries needed in order to reach subsequent stages of de-
velopment. Thus, donors were always considered to be the most powerful actors in foreign
aid interactions, as they were considered development experts by virtue of having already
traversed the modernization trajectory. However, critics pointed out that modernization as-
sumed such a path was the only one toward development, and that it was a desirable path for
all. This theory proved to be far removed from the experiences of still developing countries
at the time, especially those that had been subjugated to colonialism, leaving space for a
new theory that better considered these power dynamics.
7
Dependency Theory
In the 1970s, dependency theory established that there was a group of core countries, and
a group of peripheral countries, where raw materials from the periphery
ow to the core to
be processed. The core manufactured goods back to the periphery, thus making the core
rich at the expense of the periphery (e.g. Smith 1981). This approach better accounted
for the extractive economic relationships established between core and periphery countries
under colonial rule, yet it presented periphery states as permanent and passive victims of
such relationships. This overlooked the autonomy and power that developing countries could
assert over their own trajectory. Donors, typically located in the core countries at this time,
were still the implied initiators of aid programs as they remained dominant over recipient
countries - typically found in the periphery. Critics additionally assert this paradigms over-
values the role of government intervention in the market, and encourages donor-initiated
aid supporting infant industries in developing nations meanwhile basic needs of the local
population remain unmet.
Sustainable Development
Within both the modernization and dependency paradigms, the denition of development
had been largely focused on economic growth. Yet growing awareness of the environmental
costs of development, raised by authors such as Hardin (1968) as the tragedy of the commons,
exposed the weaknesses of this denition. The concept of sustainable development arose as
a response in the 1970s and 1980s, and promoted the balancing of economic, socio-cultural,
and environmental interests in development initiatives. This new paradigm was eventually
consolidated in the 1987 Bruntland Report released by the United Nations World Commis-
sion on Environment and Development (1987). The sustainable development approach was
also the rst mainstream theory that implied a role for local actors. This is clear from a
former World Commission on Environment and Development (WCED) public hearing that
generated the now-famous Bruntland Report:
8
\I think this Commission should give attention on how to look into the question of
more participation for those people who are the object of development. Their basic
needs include the right to preserve their cultural identity, and their right not to be
alienated from their own society, and their own community. So the point I want to
make is that we cannot discuss environment or development without discussing political
development. And you cannot eradicate poverty, at least not only by redistributing
wealth or income, but there must be more redistribution of power" (29).
Of special note here is the recognition that power is not well-distributed in the world of
development nancing, and that local actors have been overlooked in previous approaches.
The report later makes mention of the need for local communities to participate in decision-
making processes. As such, the sustainable development approach marks the rst time that
local actors are credited with the right to exercise power in the context of international
development interactions. In
uential ethnographic work by political scientist Scott (1985)
at this time supported this spotlight beneciary power; locals do not only assert power
through overt rebellions but through everyday resistance acts such as refusing to cooperate
or refusing to change cultural practices that still have important relational impacts.
Neoliberalism
Also in the late 1980s, the development world saw a competing framework gain rapid popu-
larity known as neoliberalism, where privatization, deregulation, trade and capital account
liberalization was promoted over government intervention in the market. Neoliberalism is
rooted in classical free market economic theories and embedded in the bylaws of the Bret-
ton Woods institutions: the World Bank, the International Monetary Fund (IMF), and the
treaty that provided the foundation for the eventual World Trade Organizations (WTO).
During Latin America's debt-burdened `lost decade' of the 1980s, the IMF prevailed in
prescribing market reforms. Neoliberalism triggered a turn away from import substitution
industrialization (ISI) and the shift to an export-led development model.
As with modernization and dependency theory, foreign aid was presented as donor-
initiated in that donors decided which projects or programs were needed in recipient coun-
9
tries, while recipient countries accepted those projects. One of the most well-known ne-
oliberal events of this period was the Washington Consensus in 1989. Published by John
Williamson (see Williamson 2008), recipient countries were expected to undergo a list of ten
market reforms in order to be eligible for nancial assistance, known as structural adjustment
programs. Thus, donors became the clear bearers of power under the neoliberal framework.
However, the Washington Consensus had disastrous results for many countries, sending
many further into debt and giving rise to a critical re-evaluation of development approaches.
It did not account for non-economic aspects of development, which proved overly simplistic.
Prescribed austerity measures sti
ed economies in crisis, and structural adjustment programs
proved to function dierently across dierent cultural contexts.
Human Development
In the 1990s, economist Amartya Sen's ideas about re-orienting the goal of development
away from economic growth and towards human well-being gained popularity, even winning
him a Nobel Prize in Economic Sciences by 1998 (e.g. Sen 1999). Commonly referred
to as the human development paradigm, these ideas were adopted by the United Nations,
and precipitated a normative shift in the objectives of development, as seen in the later-
developed Millennium Development Goals (MDGs). These considered social objectives and
a Human Development Index (HDI) was designed as an alternative to previous GDP-based
measurements of development success.
Like the sustainable development approach, local actors are presented as worthy of hold-
ing power in the development space. For example, Sen (1999, xiii) states that individuals
should be \...seen as active agents of change, rather than as passive recipients of dispensed
benets." Nussbaum (2000, 32) echoes this point through a case analysis of India: \If we
consider each person as worthy of regard, as an end and not just a means, we cannot in any
simple way praise Gujarat's rapid economic growth, which has left many powerless people
behind and caused many self-employed women to lose their livelihoods." Eventually the two
10
authors joined forces to create an academic organization called the Human Development
& Capabilities Association, which promotes participatory research methods and innovative
measurements of development. However, few countries or aid agencies actually measure
changes in human well-being despite pervasive human development rhetoric in the eld
(Fukuda-Parr, 2012). Even the MDG upgrade, the UN's Sustainable Development Goals
(SDGs) agreed upon in 2015, which are arguably a fusion of the sustainable and human
development approaches, still have measurements that are a poor t for assessing the attain-
ment of these goals or are missing measurements all together.
5
In fact, these UN goals have
been criticized for disempowering individuals by leaving out a goal to specically improve
subpar governance (Clemens 2012; Abtew 2017).
Alternative Development
While human development thinking has been integrated into units within international orga-
nizations like the UN and the World Bank, more profound critiques of development method
within such entities have arisen in the 21st century. Since the 2000s, critical analyses of de-
velopment have been gaining popularity, promoting a turn away from aid and towards other
types of investment. Outspoken critics such as Easterly (2006) and Moyo (2010) lambaste
current models of foreign aid as not only inappropriate for achieving global development
goals, but for making matters worse in some recipient countries. These authors argue that
aid has created perverse incentives for those involved in the development industry, and that
it creates dependency in recipient states that limits the possibilities for national autonomous
growth. Foreign direct investment (FDI) is instead suggested as a way to ensure that recip-
ients have incentives to realize a return on investment for the donor. Many have noted the
dierences between traditional Western aid and Chinese aid, which acts more as FDI than
ocial development assistance (ODA) in practice, yet we are still awaiting the results of this
5
See information on SDG measurements at: https://sustainabledevelopment.un.org/hlpf. For example, as of
February 11, 2019, the most recent SDG progress report on their website listed is from 2017, and stated [2]: \Some
targets are not re
ected at this stage, owing to the lack of data, or because they are measured by indicators that are
still undergoing methodological development."
11
alternative development approach.
Other critics assert that true development is generated from within a target popula-
tion, rather than through outside aid eorts. Cross-disciplinary and even cross-development
paradigms suggest that more must be done to empower local communities to help themselves
and insist that participatory planning processes must take place to ensure sustainability (e.g.
Frediani 2006, Escobar 2012, Imbach 2016). Alternative forms of - and approaches to - aid
and development place increasing decision power in the hands of locals and these have gained
popularity in recent mid-level development theory. Gutierrez-Montes et al. (2009) argue for
the importance of system-level approaches to better understand the eects of locally-driven
development eorts. For example, the sustainable livelihood approach (SLA)
6
proposes that
not only must basic needs be met, but that recipients should be empowered to survive eco-
nomic shocks (Chambers and Conway, 1992). Thus, development should focus on creating
this security mechanism. The community capitals framework proposes seven dierent types
of capital available to communities and analyzes how to draw on their capital strengths
when investing in sustainable change, rather than focus on building up what the community
is lacking (Flora et al., 2004). Gutierrez-Montes et al. (2009) additionally posit that during
the project planning process these approaches must also be used with participatory meth-
ods, also referred to as participatory action research (PAR). For example, SLA and CCF
approaches might help you understand the range of potential project goals on the table,
while participatory methods will allow you to choose the goals that best help communities
improve conditions for sustainable change. Participatory methods have also been promoted
in other approaches such as the community-based approach (CBA) and community-based
participatory research (CBPR) (Wallerstein and Duran, 2003).
6
Bold type indicates dierent mid-level theories under the Alternative Development paradigm.
7
A \!" denotes \leads to."
8
The Heavily Indebted Poor Countries Initiative attempted to address the continuing debt burden for the poorest
countries, as use of practical approaches based on previous development theories had in part exacerbated this burden.
See: http://www.worldbank.org/en/topic/debt/brief/hipc
12
Table 1.1: The Evolution of Development Theories and their Impacts on Aid Approaches
Name Years Theory
7 Associated
Authors
Dening Real World
Events
Project
Initiators
Modernization 1960s Economic growth
! democracy !
development
Solow (1956);
Rostow (1959);
Lipset (1959);
Moore (1966)
Formal creation of foreign aid
programs after Marshall Plan;
aid given to Cold War allies;
popularization of ISI economic
policies resulting in aid for in-
fant industries; aid given for
economic growth as thought to
lead to democracy
Donors
Dependency 1970s Economy indepen-
dent of the colo-
nizer ! develop-
ment
Gunder Frank
(1967); Evans
(1979); Cardoso
and Faletto
(1979); Smith
(1981)
Increased government inter-
vention in market; aid for ISI
and infant industries; grow-
ing developing economies but
short-sighted country borrow-
ing; World Bank turn towards
poverty reduction over recon-
struction
Donors
Sustainable
Development
1980s Balance between
economic growth
and environmen-
tal impact !
development
Hardin (1968);
IUCN (1980);
Ostrom (1990);
Tryzna (1995);
Comeliau
(2000)
Tragedy of the commons
popularized; Bruntland Re-
port; questioning of `limitless'
growth assumptions; increas-
ing environmental regulations
at expense of economic growth
Donors
+Locals
Neoliberalism 1980s Trade liberalization
! economic growth
! development
Hayek (1960);
Friedman
(1962); Soto
(2000);
Williamson
(2008)
Response to `lost decade' of
development; Washington
Consensus/aid conditionality;
trade over aid; Structural
Adjustment Programs (SAPs)
Donors
Human
Development
1990s Increase choices
people have !
wellbeing improves
! development
Sen (1999); ul
Haq (1999);
Nussbaum
(2000); Alkire
(2005)
HIPC Initiative;
8
growing in-
equality; creation of Mil-
lennium Development Goals
(MDGs) for developing coun-
tries; creation of HDI and
other multidimensional mea-
sures in response to criticism
of GDP focus
Locals
Alternative
Development
2000s
and
on
Improving the
livelihoods and
assets that people
already have !
development
Escobar (2012);
Chambers and
Conway (1992);
Flora et al.
(2004); Easterly
(2006); Moyo
(2010)
Popularization of micro-
nance; creation of Sustainable
Development Goals (SDGs) for
all countries; new frameworks
e.g. sustainable livelihood ap-
proach (SLA) and community
capitals framework (CCF)
leading funding focus towards
community-initiated projects
and investment over aid; move
towards multidisciplinary
training for development
practitioners
Locals
13
||
Development theories have evolved over time in terms of opening up space for local ac-
tors to assert power - or at least claim some agency in interactions with powerful donors.
Nevertheless, theories that ascribe power to local actors still tend to remain outside of the
mainstream, such as the many mid-level theories that make up the Alternative Develop-
ment approach. Academic papers using sustainable, human, and alternative development
approaches rarely make it into the top international relations (IR) journals. For example,
ten articles alone have been published on questions of foreign aid in the past two years in
the top IR journal, International Organization. None have explicitly mentioned these ap-
proaches, and economic growth prevails as the main denition of development used by these
authors. This emphasis on economic growth has led to project design that privilege eco-
nomic incentives over all other incentives types. Yet the shifts we have seen in development
practice towards acknowledging the power of local actors and their nonmonetary motivators
are prevalent and easily identiable. I will now take an in-depth look at these changes in de-
velopment practice we have seen over the past two decades to highlight the current emphasis
by donors on obtaining local buy-in through nonmonetary means.
1.4.2 The Evolution of Development Practice
The missions and operations of a number of aid agencies have been heavily in
uenced by
development theories and academic research. Take the United States for instance, the largest
bilateral donor in the world. The Millennium Challenge Corporation (MCC), a prominent
branch of U.S. bilateral foreign assistance begun in 2004 by President G.W. Bush, was
created with the neoliberal Washington Consensus as its main reference point, even though
development critiques and alternative theories had already begun to take hold at this time.
For example, the MCC awards aid based on donor-determined eligibility rules, similar to the
aid conditionality policies enacted via the Washington Consensus.
More specically, MCC awards competitive grants based on a recipient country's score
14
on 20 policy indicators, and its approach is based on evidence from an in
uential and now
widely controversial macro-level academic study by Burnside and Dollar (2000) that linked
increased aid eectiveness as measured by GDP growth in countries with good governance
policies, that will be discussed further in Chapter 2. Attached to a September 2014 letter
9
from the MCC is a working paper by two internal researchers that states: \In an in
uential
paper, Burnside and Dollar (2000) used historical data on aid, growth, and various measures
of governance to argue that aid, when disbursed to countries that are governed eectively,
causes more growth...While this conclusion has since been questioned by other research,
such as Easterly, Levine and Roodman (2004),
10
it was a powerful statement in support
of the MCC's eligibility rules" (Johnson et al., 2014, 10). A second paper, a U.S. Agency
for International Development (USAID) Economic Brief
11
written by Crosswell (2015), also
praises Burnside and Dollar (2000) while including this footnote: \For excellent reviews of
the issues surrounding the Burnside/Dollar study and other eorts to gauge aid eectiveness,
see Roodman (2007b) and Arndt, Channing, Jones and Tarp (2010)" (Crosswell, 2015, 2).
Although both the MCC and USAID acknowledge that the original Burnside & Dollar study
may have been discredited, it continues to in
uence the MCC's approach to determining
aid eligibility. However, the MCC does strongly encourage a participatory approach once
aid is awarded, and identies actual projects and uses of aid in consultation with the recip-
ient country government. They are adamant that local ownership is key to the success of
development aid, apparently in
uenced by the more recent turn to alternative development
theories and approaches.
Donors across the board have similarly made a local turn in their development aid prac-
tices. This is driven by the belief that this will increase aid eectiveness, and enable donors
to satisfy increasing transparency demands in the global community with fewer negative
9
The full letter can be accessed at: https://oig.usaid.gov/sites/default/les/audit-reports/m-000-14-005-s.pdf.
Based on a working paper rst by (Johnson and Zajonc, 2006).
10
See: Easterly et al. (2004).
11
The full brief can be accessed at: https://www.usaid.gov/sites/default/les/documents/1870/Econ%20Briefs%
20-%20Crosswell Selectivity 12 7 15 socedits%20nal%20nal soc%20v%20nal%20nal%20nal.pdf
15
repercussions from any negative project evaluations. Why? Gibson et al. (2005) provide
a succinct history of this belief. In the 1990s, practitioners and critics of aid came to the
realization that the heavy involvement of donors in every aspect of development projects
meant that there were few incentives for recipients to become stakeholders and make the
commitments necessary for projects to bear positive long-term results. This learning pro-
cess resulted in an emphasis on ownership, where donor agencies claimed that if recipients
`own' their development assistance, project results would be more sustainable because the
recipient would be more invested in their outcomes. Institutions began to be redesigned to
provide incentives to this end. Terms like partnerships and responsibility gained popularity
with donor agencies. Unfortunately, in practice, there have been problems with this line of
thinking: the terms remained ill-dened and unclear for local aid project actors. Moreover,
this rhetoric tends to ignore power structure realities intrinsic to development projects, as
well as who holds accountability for project results. In short, donors now believe that giving
some power to local actors is good for the project and for development, which should be
contemplated by modern IR researchers of development, as this study will do. I will turn
now to illustrate how donors have changed their practices in order to cede power to local
actors since the early 2000s.
Changing Transparency Norms
The emphasis on ownership paralleled another important movement at the time in the
international aid community, one with the goal of increased transparency of aid's results.
Although the eectiveness of development aid has been a concern since formal aid programs
began post-World War II, aid and specically investments in rigorous project evaluation had
not been a priority for most donor organizations until about 2000 due to resource constraints
and a focus on policy compliance (Hirschman 2015; Carlsson et al. 1994; Savedo et al. 2006;
OECD 2013; Independent Evaluation Group 2016). Calls by international organizations for
better evaluation peaked with a World Bank report, Assessing Aid: What Works, What
16
Doesn't, and Why (World Bank, 1998). The report was a call to action for both academic
researchers and practitioners, encouraging both to more thoroughly explore the conditions
for aid eectiveness and build future agendas based on previous results: \Agencies should
be evaluated on the extent to which they have allocated their resources, both of nance and
knowledge, to stimulate the policy reforms and institutional changes that improve people's
lives" (World Bank, 1998, x).
The World Bank's call also coincided with a turn in the social sciences toward the adop-
tion of large-N, quantitative research methods. Many of the seminal aid studies on eec-
tiveness came about in the late 1990s and early 2000s in top economic journals and took a
large-N, quantitative approach to determine aid's impact on economic growth. Such stud-
ies of aid eectiveness were not only commonplace in academic journals at the time, but
the topic of aid additionally had broad public appeal. The studies became well-known in
academia, practice, and in popular media, and their dismal results - typically showing that
aid was by and large ineective - began to attract much attention from aid stakeholders.
12
While many aid agencies were built under distinct development paradigms (see Table
1.1 for a summary of these theories and key researchers), the need to increase eectiveness
led many agencies to turn to new theoretical models. This triggered a wave of alternative
development theories that sharply criticized earlier paradigms and placed greater emphasis
on community involvement. Changes were implemented across the board in development
practice, including experimentation with new agencies, training, modes of project initiation,
and delivery methods over the last two decades, all of which focus on more holistic environ-
mental considerations and adaptability to context (Honig and Gulrajani, 2018). Evaluation
and transparency have skyrocketed in importance, and there is a growing trend towards
evidence-based policymaking in countries around the world. Policymakers are increasingly
scrutinizing foreign aid by demanding evaluation.
13
Almost every major multilateral, state
12
See these studies as analyzed in Roodman (2007a).
13
For example, H.R. 3766, the Foreign Aid Transparency and Accountability Act, became U.S. federal law in July
2016, requiring close monitoring and evaluation of foreign aid projects, a focus on outcomes, and increased project
transparency by publishing information on government websites. For the full text, see: https://www.congress.gov/
17
and nonstate donor agency - and even some traditional recipient countries - have a webpage
related to evaluation and/or aid transparency, including the World Bank, the European
Commission, the Governments of Canada, the Philippines, and the United Kingdom, US-
AID, the Gates Foundation, and the Gavi Vaccine Alliance.
14
It is not just talk. Rather,
the number of evaluations performed is climbing rapidly, as is concern for their distribution
and quality (Cameron et al. 2016; Sabet and Brown 2018; Rogl a 2020).
Changing Professional Training
After World War II, development professionals were generally trained as economists (Kothari,
2016). Around the 1960s-70s, the eld of development studies was created, which promoted a
more interdisciplinary approach. However, by the 1980s, events like the Latin American debt
crisis and advocacy of strong neoliberal economic policies in response once again brought
economists to the forefront of development practice. The 1990s saw an explicit feedback ef-
fect, as economists like Sen (1999) called for a more human-centered development focus. Sen
and others insisted that the Washington Consensus policy package appeared to exacerbate
problems, as did new reports about the ineectiveness of aid in the 1990s. Thus, institu-
tions returned to the idea of interdisciplinary training for professionals in the development
studies eld. Governments oered competitive funding opportunities to development studies
programs in an eort to increase the quality and quantity of trained professionals (Kothari,
2016).
Yet ineective aid results persisted. The lack of sustainable and inclusive development
was highlighted by the many ineective results appearing in scientic research. In 2007,
bill/114th-congress/house-bill/3766
14
See: World Bank at http://www.worldbank.org/en/research/dime, European Commission at
http://ec.europa.eu/europeaid/policies/eu-approach-aid-eectiveness/aid-transparency en, Government of
Canada at https://www.international.gc.ca/world-monde/issues development-enjeux developpement/priorities-
priorites/aid transparency-transparence aide.aspx?lang=eng, the Government of the Philippines at http:
//www.gov.ph/faith/, the Government of the U.K. at https://www.gov.uk/government/publications/2010-to-
2015-government-policy-overseas-aid-transparency/2010-to-2015-government-policy-overseas-aid-transparency,
USAID at https://www.usaid.gov/results-and-data/progress-data/transparency, the Gates Foundation at
https://docs.gatesfoundation.org/documents/guide-to-actionable-measurement.pdf, Gavi Vaccine Alliance at
http://www.gavi.org/about/ghd/iati/
18
the John D. and Catherine T. MacArthur Foundation made a multi-million dollar push to
create a network of Master's in Development Practice (MDP) programs which presently
includes 32 higher education institutions.
15
The discipline of development practice is dis-
tinguished by its focus on participatory approaches to development. This means that the
beneciaries are incorporated into the project design process by identifying needs that lie
at the intersection of natural resources, health, and social sciences, and management prac-
tices. The International Commission on Education for Sustainable Development Practice
was convened, releasing a nal report in October 2008 that criticized current approaches
to development that focused on creating discipline-specialized professionals who are unable
to implement multi-disciplinary solutions on the ground, and fail to foster participatory
planning processes (International Commission on Education for Sustainable Development
Practice, 2008). Hence, they recommended the creation of interdisciplinary graduate pro-
grams to train development professionals in all aspects of development project planning.
This includes the ability to take a local approach and work using participatory methods.
This report not only led to the aforementioned 32 master's programs at higher education
institutions found in 19 countries on six continents, but also partnerships with numerous
public and private collaborating organizations. There are also spin-o degree and contin-
uing education programs at even more institutions, all focused on providing training on
participatory methods, and empowering local actors in the development project process.
Changing Rhetoric
This push to better train development professionals was re
ective of a broader change in
the rhetoric of many prominent development organizations since the 1990s, a discourse that
privileged local knowledge and solutions.
16
New terminology introduced during this time
15
See: http://mdpglobal.org/
16
See references to participatory approaches in documents from: the Austrian Development Agency in
2008 (https://www.oecd.org/development/evaluation/dcdndep/47069197.pdf), JICA in 2001 (https://www.jica.go.
jp/jica-ri/IFIC and JBICI-Studies/english/publications/reports/study/topical/participatory/pdf/par.pdf), and US-
AID in 1996 (https://pdf.usaid.gov/pdf docs/PNABS539.pdf)
19
included phrases such as participatory approaches, community-based approaches (CBA), and
community-driven development (CDD), which have since entered into mainstream devel-
opment vocabulary. For example, a Google Scholar search for the terms development +
participatory approach returns about 18,000 results from their very rst database records
up until 2007 (when the MacArthur Foundation sponsored a push to change the academic
training of development practitioners, see Section 1.4.2 for a description), yet even more
results, 19,400, appear in just nine years from 2008 to October 2017.
17
Major development organizations have not only adopted new terminology but attempted
to dene and standardize it. In 2003, the United Nations Food and Agriculture Organization
commissioned guidelines on how to incorporate participatory approaches into its projects.
18
The World Bank increased its spending on CDD projects from about USD 325 million in 1996,
to USD 2 billion in 2003 - and these are conservative estimates (Mansuri and Rao, 2003).
Groupe URD, an NGO dedicated to evaluating practices and policy in the humanitarian
sector, produced a handbook in 2004 for humanitarian workers that pushed participatory
approaches during times of crises.
19
In 2004, the Asian Development Bank (ADB) released a report evaluating the new partic-
ipatory approaches it had been implementing since 1996. The ADB, a multilateral regional
organization, has been responding to criticisms of previous projects, including the imple-
mentation of those that were irrelevant to their local context and project results that were
unsustainable in the long-term. The ADB refers to these two problems as \conventional" in
its report, reinforcing these common criticisms of foreign aid projects across all donors (Asian
Development Bank, 2004, iii). While the ADB found that participatory approaches on their
own were not necessarily more eective in the short-term, they did nd that an extensive
17
This dierence could be due to fewer digital listings before 2007. Nevertheless, the dierence is still stark given
that the Internet has been commonly used since 2000 and many journals have digitized their publications several
decades back meaning results from earlier years were included in the search. Search performed in October 2017.
18
See Participatory Development: Guidelines on Beneciary Participation in Agricultural and Rural Development
available at: http://www.fao.org/3/a-ad817e.pdf
19
See Participation Handbook for Humanitarian Workers: Involving Crisis-Aected People in a Humanitarian
Response available at: http://www.urd.org/Participation-Handbook
20
understanding of local contexts during the design phase did help improve the relevance and
sustainability of project outcomes. Participatory approaches remain favored among interna-
tional organizations, for example, the World Health Organization issued guidelines in 2015
promoting such an approach.
20
Changing Approaches to Project Initiation
This turn to local actors in ODA projects has led to a change in donor approaches to project
initiation in the past two decades. I have identied three main types of approaches in terms
of who initiates a project: donor-initiated, participatory, and community-initiated. Donor-
initiated refers to projects that have been identied as priorities by donors. The donor
typically contracts an implementing organization to carry out the project, which can be a
group within the donor itself, a donor country NGO, the recipient government, a recipient
country NGO, or a multilateral agency. Approval from the recipient government is typically
sought for these projects, although donors may try to circumvent the local government,
for example, if corruption is suspected or if the donor seeks to serve its own con
icting
strategic interests through aid-giving. Local beneciaries are rarely consulted about the
necessity or relevance of these projects, though they are the targets of project activities and
their participation is necessary to achieve a desired project impact. Development theories
up through the 1980s typically implied that the donor knew best about what the recipient
country needed, thus the approach of many aid agencies was to initiate projects they felt
were in recipients' best interest, informed by development theories and also the previous
experiences of donors. Strategic aid also falls in this category, which is aid that is used as a
political tool by donors - for example, as a trade for votes at the UN General Assembly or
to security-based allies.
In contrast, projects that take a participatory approach involve a consultation on project
20
See Taking a Participatory Approach to Development and Better Health: Examples from the Regions for Health
Network available at: http://www.euro.who.int/ data/assets/pdf le/0007/294064/Taking-participatory-approach-
development-health-malmo-skane.pdf
21
priorities between the donor and local actors. These projects are informed by the approaches
elaborated in Section 1.4.1, such as the sustainable livelihoods approach (SLA), community
capitals framework (CCF), community-based approach (CBA), and so on. The main idea is
that the donor and local actors - usually project beneciaries - are partners in development.
Donors hereby hope to create local ownership and subsequently favorable long-term results.
Local actors at the country-level may also request projects in consultation with their coun-
terparts in the donor country. However, Jerve (2002) argues that there is a trade-o involved
with local ownership: if a recipient country or beneciary community has strong ownership
over a project, then the `equal' partnership with the donor is diminished, as one side has
more power. Partnerships are tricky, since the incentives of each side are often unaligned.
Thus, this approach to development comes with inherent tensions between donor and recip-
ient goals (Gibson et al., 2005). Nevertheless, it remains a popular way for donors to work
in their target communities or sectors while increasing the chances that a given project's
changes will last due to its relevance to beneciaries.
Community-initiated projects are popular in development assistance, where donors create
small usually competitive funds or grants for communities. NGOs usually respond to these
calls, which operate similarly to any grant request for proposal (RFP) found in the academic
world. This approach has long been used by private donors, but bilateral donors in particular
have been testing out this approach over the past two decades. For example, the U.S.
government established the Development Grants Program in late 2007, which sought to
provide small grants of no more than USD 2 million to both U.S. and local NGOs that served
the purposes of the Foreign Assistance Act of 1961, which rst created the development
giant USAID.
21
Japan's aid program has actually followed such a philosophy, known as its
\request-based principle," since their aid program began as form of reparation after World
21
See Section 674 of the Consolidated Appropriations Act, 2008 at: https://www.gpo.gov/fdsys/pkg/PLAW-
110publ161/pdf/PLAW-110publ161.pdf. This was later modied to the Small Grants Program with slightly dif-
ferent requirements, see Section 7080 of the Consolidated and Further Continuing Appropriations Act, 2015 at:
https://www.congress.gov/113/plaws/publ235/PLAW-113publ235.pdf. USAID currently calls this program local-
works although they are only targeting three to ve USAID missions as of October 2017, see: https://www.usaid.
gov/partnership-opportunities/ngo/localworks.
22
War II (Shimomura et al., 2016, 7). This means they wait until requests are made, based on
their faith that partner countries know what is good for them. Yet, even this practice has
seen some change since 2003: Japan is making a more intensive eort to actively participate
in dialogues with partner countries to steer holistic development, although they maintain
they do not want to get involved in the domestic politics of recipient countries.
We also see community-initiated approaches amongst private sector donors. For example,
micronance, pioneered back in the 1970s in Bangladesh by Nobel Prize-winning economist
Muhammad Yunus via the Grameen Bank, gives small loans to applicants who do not qualify
for regular bank loans. The individual receiving the loan uses the funds as they wish and is
expected to pay the loan back quickly. Although many criticisms have since surfaced with
regards to micronance, this approach has rapidly expanded with the turn to alternative
development theories that focus on community-driven projects. Global NGOs are even
springing up to crowdsource micronance loans, including Kiva (founded in 2005), Zopa
(founded in 2005), MicroPlace (founded in 2006), MYC4 (founded in 2006), Bankers Without
Borders (an arm of the Grameen Bank founded in 2008), Vittana (founded in 2008), and
Zaf en (founded in 2010). This trend has spilled over to the multilateral realm to such
entities as the International Finance Corporation (IFC), which is the branch of the World
Bank Group that focuses on the private sector in developing countries and is one of the
largest funders of micronance, committing USD 519 million in 2014.
22
There was ample evidence of current emphases on participatory and community-initiated
approaches in my interviews with aid practitioners in the Philippines, Washington, D.C.,
and Costa Rica. Some of the most illuminating passages include:
\...they bring them together in dening the needs of the community. And it's very
much { it's supposed to be community based, how they call it, community driven project
identication, and it's very interesting as a model to look at, because they really { that's
where I think you see the local cultural norms in
uencing decisions being made. Yet
when like you know, what they come up with, for example we're asking, `Okay, how did
you do your needs assessment?' And they realize they can do prioritization of the needs
22
See their stance and statistics on micronance at: http://www.ifc.org/wps/wcm/connect/Industry EXT
Content/IFC External Corporate Site/Industries/Financial+Markets/MSME+Finance/Micronance/
23
and they say, `You know, water and sanitation are the number one priority because we
don't have access to safe water, we have to walk ve kilometers to get access to safe
water and then when it's the dry season, and the water is not good anymore, so it cost
a lot of money.' So at the end of the day...what they realized through that process was
social cohesion was enhanced between the various sectors and ethnic groups as a result
of that working together." - Bilateral Practitioner
\Well especially with [agency], we very much { we focus on participation, you know,
so right from the very beginning when we enter a community, we pretty much { well
we do our initial survey and discussion for example with a partner agency. Most of the
time we are in countries upon invitation of a ministry or a department and so we kind
of discuss with them. That is apart from looking at the global indicators, you know,
Human Development Index for example. So any type of work that we do, it starts with
whether the community owns the problems and therefore are they participating in the
determination of solutions or interventions that they think would help them resolve the
issues that they are in. And generally where we see a lot of participation and ownership
of the communities, then there is success and...we feel that we are successful." - NGO
Practitioner
\...in any project you have to be considering what their people's perspectives are, what
their practices are, what their perspectives are, what their political views are, I mean,
just you know, the whole range... if you don't then why would they { [it would make]
any number of problems, I mean, one { you want people you know in a project, to kind
of get them to do something that they're currently not doing, or to change the way
they're doing something. And if they don't see the value of that or if you don't show
the respect to try to adjust whatever it is that you're doing, then you have to gure out
what their values and preferences are, then why should they do it. I suppose if they
didn't really want to do it, but also if it doesn't make sense to them...it might even be
in their interest, but then they see these people come from outside, they try to force
this down our throats, and we're just going to resist it because that's where people are.
And similarly the government, I mean, it's very dicult to do programming where the
government is really strongly opposed to what we're doing." - Bilateral Aid Practitioner
This not to say that donors have ceded all their power in development practice. When
donors promote participatory and community-initiated approaches, they are still operat-
ing within a paradigm whereby they have chosen to structure these interactions. That is,
community members must still assimilate to donors' interests. It is actually in the donor's
interest for the community to initiate and take local ownership of projects. On the one
hand, not all locals are interested in any development projects. For example, a member of
the tourism board from the Mexican state of Quintana Roo shared in a 2017 conference panel
that although they have created some protected land areas for the sake of contributing to
24
sustainable development through conservation and targeted tourism to these areas, the areas
are not actually being protected because no one is interested in taking on the managerial
role for these areas.
23
On the other hand, the idea of participatory may not be familiar or
relevant to some communities. In some places, social hierarchies are determined before birth,
and there is little room for change. The idea that communities must innovate and be en-
trepreneurial in order to uphold donor standards of development may be completely outside
the interests of a given community. Thus, donors do still have power in these interactions,
but are also willing to allow for greater exercise of recipient power.
Changing Measurements
While project models have changed considerably, evaluation measurement and indicators
have been much slower to change. In their 2010 book, Mis-Measuring our Lives: Why GDP
Doesn't Add Up, famed economists Joseph Stiglitz, Amartya Sen, and Jean-Paul Fitoussi
argue that there are too many problems associated with measuring a country's development
via its GDP. In other words, alternative proposals are needed. As members of a broader
special committee convened by former French President Nicolas Sarkozy, their goal was to
illuminate on how we empirically do and normatively should measure the concept of progress.
The committee focused on the issues surrounding GDP as a measurement tool for human
development and proposed the broader concept of quality of life, including measurement
of environmental indicators and sustainable development. Stiglitz et al. (2010) suggested
how to expand metrics to include the voices and experiences of those who are aected by
development initiatives. Again, creating space to incorporate the power of local actors was
brought to the fore. This book originated through the practice side of development rather
than the academic side; the French government funded the committee of high level researchers
and publication of results.
23
Presentation at the 6th People, Planet, Peace Conference in San Jos e, Costa Rica, October 2017.
25
Stiglitz et al. (2010) were by no means the rst to voice concern over the use of GDP
as a main measure of development. The creator of the GDP indicator - Simon Kuznets,
an economist and Nobel Prize winner like Sen and Stiglitz - had warned from its very
inception in the 1930s that GDP growth should not be equated with a nation's improvement
in well-being or welfare.
24
The 1996 United Nations Human Development Report points
to ve dierent scenarios where economic growth is actually harmful to development such
as growth without job creation, or growth at the expense of the poor or cultural minority
groups.
25
Sen (1970) also famously demonstrated that states cannot pursue even minimal
levels of freedom and economic eciency at the same time, meaning that GDP growth
cannot be used as the measurement of development in self-proclaimed liberal societies.
26
Additional researchers have warned against the common perception that economic growth
is limitless, and the linked assumption that development is only occurring if indicators show
economic growth (Comeliau 2000; Stiglitz 2007). Previous work by Easterly et al. (2004)
and Roodman (2007a) elaborate on the methodological dangers of using macro-measures
such as GDP when studying development eectiveness. More recently, political economist
Lorenzo Fioramonti has argued against the concept in his last two major books (Fioramonti,
2013; Fioramonti, 2017). New measures have indeed appeared on the scene such as the HDI,
though work is ongoing for alternative concrete measurement tools.
24
For more on the history of GDP see Nature: http://www.nature.com/news/development-time-to-leave-gdp-
behind-1.14499; The New York Times: http://www.nytimes.com/2010/05/16/magazine/16GDP-t.html? r=0; For-
eign Policy: http://foreignpolicy.com/2011/01/03/gdp-a-brief-history/; Harvard Kennedy School Review: http:
//harvardkennedyschoolreview.com/counting-what-counts-gdp-redened/
25
See pp. 2-4 for all ve scenarios at: http://hdr.undp.org/sites/default/les/reports/257/hdr 1996 en complete
nostats.pdf
26
This is Sen's famous work on the Pareto principal and liberalism. He states: \Nevertheless it turns out that
a principle re
ecting liberal values even in a very mild form cannot possibly be combined with the weak Pareto
principle, given an unrestricted domain. If we do believe in these other conditions, then the society cannot permit
even minimal liberalism. Society cannot then let more than one individual be free to read what they like, sleep the
way they prefer, dress as they care to, etc., irrespective of the preferences of others in the community. What is the
moral? It is that in a very basic sense liberal values con
ict with the Pareto principle" (Sen, 1970, 157).
26
Changing Delivery Methods
Budgetary Support Up until the past two decades, an important and highly controversial
method of aid distribution has been budgetary support. This entails funds given directly to
recipient governments to implement as seen t, though general parameters are still given
by the donor such as sector or project type. This aid is almost exclusively given by other
governments or multilateral agencies (as opposed to private donors) and considered fungible,
as the funds can be used for recipient government priorities instead of specic projects. This
type of distribution method tends to be overestimated in the academic literature as a common
form of delivery, as many macro-level measures assume most aid is fungible (Bermeo, 2016).
A 2006 OECD report estimated that only ve percent of all ODA in the world is delivered
via budgetary support, and an updated 2014 report lowered the estimate to just three to
four percent (OECD 2006; Ministry of Foreign Aairs of Denmark 2014). Its controversy
stems from the common perception that it is the distribution method most vulnerable to
corruption, as there is no clear accounting or transparency about how the money is used.
This method was meant to empower recipient country governments to use the money as
they best saw t - as long as it contributed to the overall donor goals. But there have been
many studies which argue that it only empowers corrupt government leaders, who leverage
the fungibility of this aid to funnel the funds into pet projects or directly into their own
pockets (e.g. Moyo 2010). Donors have thus changed their delivery preferences over time
from budget support to project aid, which gives more in
uence to project beneciaries.
Project A much more common mode of aid distribution is the aid project. A project is
a synonym for a development intervention. USAID, the largest aid agency of the largest
aid donor, the United States,
27
oers the best denition: a project is \a set of executed
interventions, over an established timeline and budget, intended to achieve a discrete devel-
27
By amount, see: http://www.oecd.org/dac/stats/aid-at-a-glance.htm
27
opment result by resolving an associated problem."
28
A donor chooses a problem - typically
in consultation with the recipient country - and develops a detailed plan to address it within
a dened time frame and budget. Projects enumerate activities with specic and measurable
goals that are carried out in the recipient country. Although far less ecient (at least in
theory) than budgetary support, multilateral and bilateral project aid is far more common
(Cordella and Dell'Ariccia, 2007). Jain (2007) reasons this is to avoid corruption, ensure
donor country contractors receive implementation contracts, and restrict aid use in the case
of divergent donor and recipient interests. However, Jain also points out that the nonfun-
gibility assumption of project aid is questionable, as it may just replace previously allotted
government spending in a particular sector, freeing up domestic government funds to be used
elsewhere. Nevertheless, the traditional perception in practice and academic studies is that
donors have the most control over this type of aid.
Program or Sector Aid Another tool is an aid program, dened by the OECD as \a
time bound intervention involving multiple activities that may cut across sectors, themes
and/or geographic areas" (OECD, 2004, 26). Program approaches are called dierent names
by dierent donors, for example, a sector approach
29
and sector wide approaches (SWAps)
are two of the most common synonyms.
30
While project aid tends to focus on a particular
activity, is considered non-fungible, and typically oers a donor more control over how funds
are spent, program aid tends to focus on an overall sector or goal that involves interventions
in various sub-sectors and can be fungible or non-fungible. This is because programs usually
combine both project aid and budgetary support to accomplish their goals in a particular
area. The idea is that coordination across various types of interventions will be more eective
than discrete projects that are done in isolation, but only if local capacity is strong enough
28
See Slide 2 in: http://usaidprojectstarter.org/sites/default/les/resources/pdfs/the-logical-framework-an-
introductory-slideshow 0.pptx
29
See terminology used by the European Commission on International Cooperation and Development at https:
//ec.europa.eu/europeaid/aid-delivery-methods-dierent-approaches-aid en
30
See report prepared for the UN Population Fund (UNFPA) at https://www.unfpa.org/sites/default/les/pub-
pdf/swap-unfpa2005eng.pdf and by the Overseas Development Institute (ODI) at https://www.odi.org/sites/odi.org.
uk/les/odi-assets/publications-opinion-les/2071.pdf
28
to properly implement them (Brown et al., 2001). However, there is no clear evidence that
program approaches are actually eective.
31
||
Development practice encompasses a complex web of factors that are challenging to study.
It is not surprising that development theory has lagged behind changes to development
practices over the past few decades, especially given that many academic theories tend to
rely on linear theoretical models that can oversimplify the way development works in the
real world. Understanding this gap between research and practice forms the inspiration
for this dissertation, specically the important role practitioners clearly believe local actors
play in development projects. This chapter highlights the importance of understanding local
actors and their motivations beyond assumptions that economic incentives reign supreme
at all levels in development, setting the basis for my independent variable. Meanwhile, the
next chapter goes in-depth into how we measure - or could measure - the outcomes of these
development project practices to establish the foundation for my dependent variable.
31
See links in previous footnotes to the UNFPA and OFI reports on SWAps. More recent evaluations of program
aid are scarce.
29
Chapter 2
The Billion Dollar Question: How do we Measure Aid
Eectiveness?
As elaborated in Chapter 1, donors over the past two decades are focusing more on the
needs, decisions, and power of local actors, often putting aid requests and implementation
management directly in their hands. But despite this transformation, the way we study and
measure the eectiveness of development aid seems to have changed very little over this same
period. We make a variety of assumptions in our measurements, for example that most aid
is given as general budgetary support to governments to do what they will with it. In fact,
only three to four percent of all ocial development aid since about 2002 is disbursed in
this form (Ministry of Foreign Aairs of Denmark, 2014). We often assume that each actor
in the aid implementation chain shares the same interests - and that these coincide with
those of the donors. Yet anyone who has worked on the ground on a development project
will tell you this is rarely the case. Local actors, again, those whose base or home is in the
recipient country, have their own varied interests that do not necessarily align with those of
the donor. If local interests are not met, then recipients have little incentive to facilitate the
implementation of a project, lessening the chances of aid eectiveness.
The political use of indicators is a critical but often overlooked consideration in devel-
opment theory. Measurement is certainly a critical aspect of development, but how do we
measure the right things? For example, in a 2017 interview I conducted, a multilateral o-
cial in charge of compiling and analyzing cross-country data described cases where countries
chose to exclusively focus on improving their score on a standardized indicator like GDP
because of the positive signal it sends to the international community. It is no secret that
30
GDP growth can mask a low score on citizens' wellbeing, which could actually be worsening.
Such political incentives in part explain the slow rate of replacing GDP as a development
indicator as discussed in Section 1.4.2, illustrating a classic example of Goodhart's Law:
when an indicator becomes the goal, it is no longer an appropriate indicator (Hoskin, 1996).
Since GDP has so often been used to measure economic wellbeing, the development goal of
many countries became GDP growth and it has been dicult to nd an agreed-upon replace-
ment (Fioramonti, 2013). It was no secret that a number of countries were unhappy when
the UN unveiled one such new indicator in 1990 { the Human Development Index (HDI) {
which combined measures of education, health, and per capita income.
1
The new Human
Development Reports clearly highlighted discrepancies between GDP gains and individual
wellbeing levels.
My research is motivated by this puzzle: how do we determine whether aid has been ef-
fective? After more than 50 years of aid eectiveness research, how do we measure \success"?
The existing literature demonstrates little agreement on how aid has aected recipients. A
recent paper by Rocha de Siqueira (2016) argues that \good enough" numbers - those that
\nd `enough' correlation" in studies - are now the norm in development policy, and most
international organizations nd such numbers to be an acceptable step forward, even if they
may be severely
awed (1). Rocha de Siqueria suggests that which errors are accepted,
by whom, and when, are political choices, meaning much of what we would call objective
development data instead re
ects power dynamics in the international system rather than
the current state of international development. Given the sheer number of indicators to
choose from, development data are created, made available, and analyzed as a result of po-
litical choices made by actors and organizations both in practice and in academia. Here I
will review such previous measurement attempts at the micro-level and then macro-levels.
In Chapter 4, I propose a new process-based measurement that can help us move towards
better measurement and avoid the trap of \good enough" numbers.
1
Per discussion with a multilateral specialist in February 2017.
31
2.1 Micro-Level Measures: Aid at the Project Level
2.1.1 What is a Project?
As explained earlier in Section 1.4.2, essentially a project is when a donor chooses a problem -
typically in consultation with the recipient country - and develops a detailed plan to address
that issue within a dened time frame and budget.
2
Project aid is an intervention that tends
to focus on a particular activity and is non-fungible, whereas budgetary support is fungible
and generalized, and program aid tends to focus on an overall sector or goal that involves
interventions in various sectors and can be fungible or non-fungible. The focus here will
be on aid projects for two reasons: they can be discretely measured, and they constitute
a majority of bilateral and private donor aid, as well as some multilateral aid. It is worth
understanding a few other project-related terms that often come up:
Sector
Sector classications dier across donors, and not all aid sectors are represented equally
across donors. Depending upon a donor's goals for their aid spending, certain donors may
privilege one category over another. For example, non-democratic countries typically do not
give money for democratization projects, and NGOs devoted to environmental concerns may
not work on health projects.
USAID breaks down its aid programs into nine sectors: Agriculture, Commodity Assis-
tance (food aid), Economic Growth, Education, Governance, Health and Population, Hu-
manitarian, Infrastructure, and Other. The Other category includes interventions related to
the environment and climate change, multisector projects, administrative concerns of donors,
debt-related actions, and general budgetary support (fungible aid given directly to recipients
2
Again, drawing on the denition from USAID: a project is \a set of executed interventions, over an established
timeline and budget, intended to achieve a discrete development result by resolving an associated problem." See
Slide 2 in: http://usaidprojectstarter.org/sites/default/les/resources/pdfs/the-logical-framework-an-introductory-
slideshow 0.pptx
32
to spend). The OECD
3
separates ODA into: Social Infrastructure, Economic Infrastructure,
Multi-Sector, Humanitarian Aid, Production, Administrative Costs of Donors, Refugees in
Donor Countries, Debt Relief, Programme Assistance, and Unspecied. In their terms, Pro-
gramme Assistance is another term for Budget Support, and Production refers to a country's
economic activities.
4
The OECD's Social Infrastructure category combines several distinct
USAID categories, such as Education, Health and Population, and Governance. Meanwhile
its Economic Infrastructure category includes transportation, storage, communications, and
energy generation and supply. For the purposes of this research, I will work with OECD
categories, given the OECD's status as a leader and repository of global information on
foreign aid and specically ODA activities.
Technical Assistance
Technical assistance is also a commonly used method of aid delivery, where trained experts in
the area of focus are sent to assist with initiatives in a recipient country. They may provide
consultation, training, or perform individual activities, but their work is meant to impart
some knowledge to locals. The most eective technical assistance nds ways to transfer
knowledge to those in the recipient country. I do not identify technical assistance as its own
delivery method because these experts are usually sent as part of an aid project or program,
but rather classify it as a type of funding (the other two types are grants, and concessional
loans with at least a 25% grant portion). This type of aid is of course nonfungible, since the
assistance is sent in the form of the expertise and education of an actual technician.
Emergency Aid
What I term here as emergency aid is an umbrella term for funds that are given during
times of crisis, and not necessarily part of a donor's planned portfolio of ocial development
3
For a regularly updated list of OECD Sector Classication Codes, see: http://www.oecd.org/dac/stats/
purposecodessectorclassication.htm
4
For more information on the Production sector as dened by the OECD, see: https://stats.oecd.org/mei/default.
asp?lang=e&subject=1
33
assistance. The types of aid given during emergencies fall under this category, which is
outside the focus of this study. I include food aid here, also called commodity assistance,
which gives actual food to recipient country citizens during natural disasters or in times
of con
ict. I also include aid given in a crisis under the broader category of humanitarian
assistance, which, according to the UN,
5
encompass situations \due to natural and man-
made disasters in areas beyond the relief capacity of national authorities alone." Not all
humanitarian assistance is given during emergencies, thus planned humanitarian aid is also
considered its own sector to which both project and program aid may be given.
Research and Capacity-Building Grants
A number of aid organizations oer grants and fellowship to promote research and capacity-
building in locations around the world. This is a form of aid, though not easily categorizable,
as oftentimes the funds will go directly to the researcher to support a stay abroad or to a
local organization to perform research or otherwise implement as seen t to support stability
and growth. This type of aid is also typically outside the scope of this dissertation, but
bears mentioning. This is because the relationships developed through such aid may come
into play in the future, as these researchers are often tapped to perform project evaluations
or NGOs called on to implement donor projects in that location.
Who are Typical Project Actors?
There are three main types of donors: bilateral, multilateral, and private. Bilateral donors
are state agencies - any branch of a country's government involved in economic or even
military assistance. Multilateral donors refer to IGOs like the World Bank, United Nations,
Organization of American States, or any other organization that involves multiple member
states. Money is given to the IGO and the IGO decides collectively how to distribute that
pot of funds, so that the funds no longer become any one state's money. Private aid refers to
5
See the rst paragraph of the UN webpage on delivering humanitarian aid for this denition: http://www.un.
org/en/sections/what-we-do/deliver-humanitarian-aid/
34
aid given by any non-governmental organization (NGO) or enterprise such as as the Gates
Foundation.
Projects have many stakeholders beyond just the donor. As stated earlier, there are
typically four main actor categories involved in individual projects (see Figure 3.1): donors,
implementers, recipients, and beneciaries. Although not included in the diagram, academics
and/or evaluators can be a fth player in projects as designated researchers, but they are
omitted from Figure 3.1 given that they participate in a small number of actual projects
relative to all projects implemented around the world.
The global aid regime as a whole has many more constituents who are interested in aid
outcomes, even if not direct stakeholders: the tax-paying public for bilateral donors, IGOs
who are indirectly interested in aid such as the WTO, transnational NGOs who are indirectly
interested in aid transparency like human rights giant Amnesty International, and any other
actor who is directly aected by aid funds or who simply cares about the global aid regime
or international development.
2.1.2 How are Projects Designed?
There have been two traditional schools of aid project planning and evaluation: the log-
ical framework analysis (LogFrame or LFA), and results-based management (RBM). The
archetype for project-level evaluation across approaches is the LogFrame. Conceived of in
the late 1960s by USAID, traditionally one of the most powerful aid agencies in the world,
the LogFrame was later adopted by many other international agencies (Lawson, 2013). It
classies changes occurring on the ground from start to nish by dividing a project into
four stages: activities, outputs, purpose, and goals. Other similar tools have also been cre-
ated, such as the Kellogg Foundation's Logic Model which is divided into ve stages, and
Bennett's hierarchy which has seven stages (Davies, 2004). The idea behind these tools is
that a project is broken down into a chain of causality where project inputs facilitate activi-
ties/tasks, which in turn produce outputs. Together these foster local eects among project
35
beneciaries, which then contribute to regional eects (Crawford, 2004, 178).
The LogFrame approach has been the most commonly used by far, given it has been
adapted by many agencies globally since its creation in the 1960s through USAID. However,
RBM was popularized in the 1980s and 90s through Canadian aid practices. The conceptual
categories each school employs are essentially the same but use dierent terms, as compared
in Table 2.1 and elaborated further in Section 2.1.5. Both can be used for the variety of aid
delivery methods available, including projects. While there are other evaluation schools that
have been employed in various types of development projects, such as adaptive management,
prototyping, and project-cycle management, these two schools have gained the most traction
across sectors and agencies (Stem et al., 2005).
Although the two approaches are similar, and sometimes are used in tandem by agencies
(Stem et al., 2005), there are some important dierences between them (Binnendijk, 2000).
The LogFrame focuses on one activity and assumes change is linear and logical, as its very
name implies. Yet the project structure alone does not necessarily ensure project results,
since change is often nonlinear and it is dicult to adapt this approach to local contexts.
RBM approaches instead put the emphasis on achieving results. But a focus on project
results may cloud our understanding of the process by which those results were achieved,
not to mention the multiple steps necessary to create those results. This is the main weakness
of the RBM approach.
Table 2.1: Aid Project Terminology in LogFrame and RBM Approaches
Stages listed in chronological order per Binnendijk (2000), CIDA (2000)
LogFrame Stages RBM Stages
Inputs -
Activities Activities
Outputs Outputs (Short-Term Results)
Purpose Outcomes (Medium-Term Results)
Goal Impact
There is also a body of literature that discusses the relationship between the idea of a
theory of change and the LogFrame and RBM approaches. A theory of change or ToC is
36
commonly used to describe a case-by-case theory of how to move from a set of identied
activities to the change desired. Because the LogFrame approach assumes linearity between
project activities and outcomes, it often doesn't consider the other environmental conditions
necessary to produce a desired change, or the cases where change is only possible when a
complex set of activities take place. Conversely, a ToC generally looks holistically at the
project and its environment. Yet scholars have argued that the LogFrame is simply one
type of ToC (e.g. Davies 2004). Presumably every project plan is created with some theory
of change in mind, because it maps out activities meant to reach its ultimate goal, even if
the chain of causality is not as clear or complete as it could be. Similar criticisms could be
leveled at RBM-based projects, as these pay even less attention to the chain of causality,
as argued above. There are abundant critiques of current micro-level measures employed
by LogFrame and RBM approaches that are equally serious, as elaborated later in Section
2.1.2.
LogFrame Approach
Nevertheless, it bears working through the LogFrame approach in more depth, as it has been
the main approach used globally for decades to guide aid project design. As mentioned, the
LogFrame was created in 1969 by USAID as its rst project evaluation tool, and many
multilateral, bilateral and NGOs later adopted the framework as their measurement tool of
choice (USAID, 2012).
6
The LogFrame attempts to show the dierence between activity
outcomes and the overall goal or impact of a project (see Figure 2.1 for an illustration).
The idea is that we work backwards from our goal to design a project, all the way down to
specic activities/inputs we will perform on the ground:
The Goal
7
is the overall impact of the project or the desired result that the project
6
Given a donor agency in the U.S. originally created this tool, I rely heavily on U.S.-based explanations of how
the LogFrame works in this analysis. It is nevertheless important to note that many other donors have slightly
adjusted this framework based on their own needs and experiences, such as the U.K.'s Department for International
Development, the Inter-American Development Bank, and so on.
7
For the purposes of this analysis, I will use upper-case letters to denote project design and evaluation elements,
to distinguish from more general uses of these terms.
37
Figure 2.1: The Basic LogFrame Matrix
Source: (USAID, 2012, 3)
will promote;
The Purpose is the key result the project aims to produce in service of its Goal (complex
projects have Sub-Purposes);
Outputs are the deliverables we measure that contribute to the project's Purpose; and
Inputs are the activities performed and resources needed to achieve our desired deliv-
erables/Outputs.
Each of these stages are operationalized during the project design phase (USAID, 2012).
Below, I will turn to the issue of gender equality to better exemplify how agencies work
backwards to design aid projects:
38
Goal Achieve gender equality in the United States.
Purpose Address the wage gap so that equally qualied men and women make the same
amount in performing the same job.
Outputs
1. Equalize annual wages for men and women in similar jobs within an
industry, including consideration of other demographic characteristics such as
age, ethnicity, and education.
2. Equalize the percentage of men and women in similar jobs within an
industry, especially in industries that are male-dominated.
3. Improve job satisfaction levels especially for women in male-dominated
industries.
4. Improve family leave policies so all employees have
exibility to care for
family members.
Inputs
1. (a) Perform audits on randomly-selected employers and ne those who
cannot justify disparate wages between males and females in similar jobs to
an independent commission.
(b) Give tax breaks to employers that implement wage transparency, so
employees can monitor their own pay equality within an organization.
2. Provide women scholarships to study male-dominated subjects in college.
3. Oer company training on gender discrimination in the workplace, and
create anonymous channels to report discriminatory incidents.
4. Give tax breaks to companies that oer family leave policies, in line with
recommendations developed by a commission of experts on this subject.
Regardless of whether the specic activities mentioned above are feasible or useful, the
purpose of this exercise is to give a better idea of how the LogFrame works. Note that the
project Goal is very broad. The Purpose begins to narrow down the Goal into a specic
objective that can be assessed. In this example, the Goal is to achieve gender quality in
the U.S. and the Purpose is to eliminate the gendered wage gap. Yet we could have chosen
a project Purpose of reducing violence against women, or improving the representation of
women among the country's political leaders; both Purposes would still work towards the
project Goal of gender equality. Likewise, in order to achieve our Purpose of wage parity,
we might have targeted other Outputs such as providing pay for full-time family caregiving,
or lowering the barriers to job market re-entry after full-time family caregiving, etc. The
important point here is that stakeholders involved in each project must make design decisions
that come with trade-os at every level.
Some of these trade-os may not be apparent to stakeholders during the project design
39
phase. One well-documented example involves the anti-malaria mosquito net.
8
Nets keep
out a number of mosquito-borne diseases, but in water-side areas deeply aected by hunger,
their more appealing use is a substitute for expensive shing nets. Yet unlike shing nets,
mosquito nets are treated with insecticide and their holes are much smaller, meaning more
underwater life is caught and exposed to deadly chemicals. This is leading to overshing and
disrupted water ecosystems in these areas with no change in malaria rates. The livelihood
of those who sh for a living is also in danger. The project Goal was to reduce malaria
rates, and one of the Outputs was to distribute free insecticide-treated nets. It was hard
to imagine beforehand the nets would be used for shing instead, or the unintended and
negative consequences this shing would incur. In many ways, every aid project is like an
experiment; no matter how much data we have, we will never be able to predict exactly
how a project will translate from one country, culture, climate, etc., to another. What does
matter, however, is that the project evaluation process catches these problems, corrects for
them, and informs future eorts to make aid more eective.
Other Approaches
Additional design evaluation methods usually respond to deciencies in LogFrame-type
(LFA) approaches. For example, Bell and Aggleton (2012) call for better use of anthro-
pological methods such as ethnography, in response to evaluation concerns that focus on
the donor's initial desired outcomes. Judging a project's eectiveness by whether or not it
achieved the initial goals of the donor oftentimes ignores the realities on the ground that
manifest during implementation. Thus, a project that was at rst evaluated as having a pos-
itive impact might in the end be found ineective because its goals were poorly conceived,
too ambitious, or unrealistic in the rst place.
LFA approaches have been criticized for their linear focus, depicting a one-way path
from inputs to results. The world does not always work in a linear nor predictable way, and
8
A 2015 New York Times article provided a succinct exposition of the problem: http://www.nytimes.com/2015/
01/25/world/africa/mosquito-nets-for-malaria-spawn-new-epidemic-overshing.html
40
changes that occur usually result from complex and non-linear occurrences (Ramalingam,
2013). While Davies (2004) argues that there are some environments where a linear view
of change can be accommodated, he acknowledges that the story LFA approaches tell from
an input to a goal is imprecise, and these approaches do not include feedback loops between
donors/implementers and beneciaries.
LFA approaches also cannot deal well with complex changes that occur in the beneciary
communities before and after the project, a concern that also plagues macro-level measures
which oversimplify complex changes in order to compare across locations. Crawford (2004)
recommends constructing synthetic indices to better measure a project's eectiveness and
more succinctly communicate those changes to stakeholders. While such indices have their
drawbacks, given that these composite measures are harder to interpret and can be costly to
construct, they do allow creation of a measure that looks at clear progress on a single scale
that can combine multiple project eects. For example, if a donor identies three anticipated
outcomes in order to combat pollution, a synthetic index can more concisely show overall
movement towards a goal even if great progress is only being made on two of the three
outcomes, or moderate progress on all three outcomes. If the point is to reduce pollution in
the beneciary community, not all three outcomes may be necessary, as long as some overall
threshold of progress identied by the index is reached.
Finally, LFA approaches tell us little about how politics and practical realities function
within the actual implementation of a project. By removing the implementers from evalu-
ation, we lose valuable information about how politics and personal relationships might be
aecting desired outcomes. Heeks and Stanforth (2014) call for the use of an Actor Theory
Network so that researchers can identify key actors and interactions can be better analyzed.
The actors' ability to create a project support network from both the inside and outside can
be an important variable in project outcomes. These authors also make a point critical to the
present study. In their words, \. . . any ongoing development project can never completely
succeed or fail because its designs and its networks are continuously being re-formed" (Heeks
41
and Stanforth, 2014, 27). This means that aid eectiveness is in fact a moving target. Suc-
cess is not necessarily dened in absolute terms, but rather as progress towards goals that
are constantly changing based on who is involved.
Randomized control trials (RCTs) have received much notoriety in the study of develop-
ment projects via an economics lens, especially with the awarding of the 2019 Nobel Prize
in Economics to Esther Du
o, Michael Kremer, and Abhijit Banerjee for their experimental
approach to evaluating development initiatives. RCTs in development studies are typically
designed as eld experiments to look at problems in one particular issue area such as health
or education, design an intervention, and compare treatment groups to control groups. In-
deed, they provide a much clearer way to not only measure but correctly measure impacts
of projects. Yet, they remain controversial amongst development experts who question their
external validity and generalizability across projects contexts.
So how do we process these numerous critiques? In order to design better measures, we
need to more carefully consider what we are trying to measure. Table 2.1 oers a clear indi-
cation of how vocabulary alone may cloud our understanding of when a project is eective.
We next turn to an analysis of the terms employed to better identify the dependent variable,
i.e. project eectiveness.
2.1.3 How is Project Evalution Dened?
The OECD's Working Party on Aid Evaluation provides a concise denition of evaluation
in the aid context, which forms the basis of my use of project evaluation: \The systematic
and objective assessment of an on-going or completed project, programme or policy, its
design, implementation and results. The aim is to determine the relevance and fulllment
of objectives, development eciency, eectiveness, impact and sustainability"(OECD, 2004,
18). Thus if a project is poorly designed in its aims, evaluation is not possible. The denition
seems straightforward, but places a heavy burden on the evaluator to assess not only the
project's multiple objectives, but also its eciency, sustainability, and impact. In practice,
42
aid evaluation is broken down into several steps to accommodate the resources and objectives
of project stakeholders (OECD, 2004). The main categories consist of the following three
types of evaluation: monitoring, impact evaluation, and project evaluation.
Monitoring.
Sometimes called monitoring & evaluation or M&E for short, monitoring involves continuous
data collection on designated indicators during a project (OECD, 2004). Those data are
systematically compared to the expected Outputs of the project to ensure that progress is
being made towards achieving its Purpose or Goal. For example, if our Goal is to end polio
in Pakistan, and one of our Outputs is to administer vaccines to all infants, we may use birth
certicates to capture home births and monitor which children are brought in to clinics to
receive the vaccine on a bi-weekly basis. By monitoring who is brought in, we can work
to identify which children of the appropriate age are not brought in for the vaccine, and
potentially contact their families to explain its importance.
Impact Evaluation.
This is the most dicult type of evaluation, as it requires establishing a causal relationship
between the project Inputs and their impact on the project Goal. It typically takes place at
the end of a project's implementation phase, and attempts to compare the project outcomes
to the counterfactual outcomes. That is, what would have happened if the project inter-
vention had not occurred? For example, if we implement an afterschool program aimed at
reducing discrimination against ethnic minority children, an impact evaluation would seek
to determine if there is a change in well-being of such children, and if that change can be
causally attributed to the afterschool program activities. Built in to their denitions of these
terms, organizations like the OECD (2004) and the World Bank (2016) clearly intend for
impact evaluations to be used by policy makers to improve aid programs and allocation.
43
Project Evaluation.
This is the most common form of evaluation when it comes to projects, simply referred to
as evaluation. This type typically focuses on project Outputs, and whether or not the activ-
ities/Inputs of the project are resulting in their expected outcomes on the chosen Indicators
(OECD, 2004). Project evaluation is much easier methodologically and more cost-eective,
as it does not require establishing a causal relationship between Inputs and a project's Goal
or even Purpose. The main focus here is simply about whether the project met its Output
goals.
||
What is the relationship between the three types of evaluations? Imagine your project Goal
was to reduce hunger in Cambodia, and one of your chosen Outputs was to send 200 tons
of rice to ve rural villages with a poor crop yield so you could help feed residents. You
decide one of your Input activities will be to send 20 tons of rice per week to the villages for
10 weeks. You monitor progress each week, to ensure you are sending the proper amount.
Perhaps while monitoring, you discover during Week 5 you have only been sending 15 tons
per week due to the use of the wrong sized bags, so you rectify the situation by changing the
bag size and sending 25 tons the last ve weeks of the project. At the end of the 10 weeks,
your performance evaluation conrms that indeed you sent 200 tons of rice to the village,
and thus you have achieved your desired Output.
At this point you have completed your project evaluation and you could stop and claim
your project a success, which often happens. But if you want to establish that the rice you
sent to that village reduced hunger rates in Cambodia, you need to do an impact evaluation.
Let's say that, as it turns out, hunger was reduced in Cambodia that year, when you count
by your chosen indicator: hunger-related illness cases admitted to medical clinics. But to
determine if your 200 tons of rice in ve villages had an impact on the drop in hunger in
Cambodia, you will need to look into the eating habits of the citizens where you distributed
the food, as well as the clinic admits in those areas. This is time intensive, as it may require
44
more translators and travel to these rural villages. You need to gure out who controlled the
food on the village and the household level, once it was delivered. You need to understand
what people in that village do when they are sick from hunger-related problems, to see if
hunger is indeed being captured through your indicator of new hospital admits. You will
need to gain the trust of residents, in order to inquire about such personal issues. In short,
you and your project team, which must include locals, will need to do a sizable amount
of legwork to map the causal chain linking your rice with hunger patterns in Cambodia.
This is the essence of participatory evaluation methods, which are vital in making impact
assessments.
Yet despite all the resources needed, there is a clear benet to impact evaluations. If we go
by the Outputs found through project evaluation alone, we may miss important information
about how beneciaries interact with the aid project in their daily lives that will help us
better structure projects in the future. This is because there are many other ways that
rice could have been consumed other than simply eating it: maybe the rice was unt for
eating and instead residents fed it to the cows and increased their meat consumption, or
residents received rice from a Swiss relief organization so they sold the rice from your agency
to a neighboring village and used the prot to buy other food, or perhaps the donor did
not consult with the community and the rice immediately spoiled when delivered to a spot
unprotected from moisture or insects. Maybe hunger actually went up in that village, as
the rice was used to ll small play toys for children instead, or the in
ux of rice drove rice
prices down and thus forced local farmers out of business. Yet these increases in hunger were
disguised by the averaging of national-level data. Maybe hunger went up because the donor
had an incorrect village population count, and sent too little rice, and worse still incited
ghts over scarce rice.
The point to remember is that aid projects attempt to solve complex issues, and linear
thinking about simple Outputs cannot always account for the ways these projects actually
play out in practice. Our Inputs and our Goal may have no relationship between them at
45
all at the end of our project, or they may have ve causal paths connecting them all at the
same time. The Inputs may be causing harm in the region, that is disguised by successes in
other regions. We just do not know without an impact evaluation. If we do not understand
what the Cambodian residents value, and the role the rice plays in boosting their wellbeing,
we risk declaring projects a success that in no way improve the lives of those individuals.
We also risk ignoring intended project outcomes that had clear impacts on the community,
even if they were not part of the project design. In the end, all we may know is that 200
tons of rice were delivered, with no measurable impact on those living in the ve villages. It
would be hard to argue that this is development or eective in any sense of the word.
What I argue here is the importance of moving aid eectiveness studies away from con-
sistently con
icting results. Project-level evaluation as a whole provides superior data in
comparison with macro-level data. Unlike country-level data, project-level evaluations ac-
tually have the capability to establish causality between the project intervention Inputs
and the Goal. Projects can be separated out by Goal, but macro-level data cannot make
such distinctions. Even though aid is given for dierent reasons, statistics gathered at the
country-level cannot dierentiate between these reasons. Thus, projects that never intended
to directly boost a country's GDP are still counted in the regressions that seek to establish
increases in economic growth. The aggregated nature of macro-level data, and the infor-
mation lost when standardizing complex measurements across time and countries, simply
cannot tell us whether and which kind of aid is actually causing variation in our desired
outcome. The nuanced information linking individual projects to broad outcomes can reveal
this information, although resource feasibility is still a nontrivial concern.
2.1.4 How are Projects Evaluated?
Micro-level, or project-level, measures can be quantitative or qualitative and are mostly used
by practitioners to inform either practitioners or policymakers. These evaluations focus on
an individual project in order to measure aid eectiveness and they are useful because they
46
can capture the direct eects of aid given by a particular donor, in a particular location, or
for a particular goal. They oer donors a way to evaluate whether a certain type of project
is an eective way to use the aid. The project is treated as a quasi-experiment because it
implicitly tests hypotheses about the eectiveness of aid. Crawford (2004, 178) points out
there are three hypotheses implicitly tested in a typical project evaluation design:
\Management hypothesis: the resources or `inputs' invested by the donor in the project
will be eciently converted to the planned `outputs' by the project implementation
team;
Intervention hypothesis: the `outputs' actually delivered will foster the desired `eects'
in the beneciary community;
Development hypothesis: the `eects' realized will contribute to a desired development
`impact.'"
Conceiving of aid projects in this way allows researchers to better pinpoint the causal mech-
anisms of an outcome.
9
Thus the project planning approach very much in
uences the evalu-
ation and metrics used. More concretely, you can not evaluate the outcomes of your project
if you are not clear on the goals of the project in the rst place. Thus the project planning
process ideally assesses the baseline of the environment where you will be implementing the
project, and evaluates changes it makes to those baselines through its project activities.
Yet the choices stakeholders must make about evaluating project elements come with
trade-os. Typically in the process of enumerating the project's plan, designers also draw
up their project evaluation plan. Evaluation can be performed at each stage, based on
the outcomes you delineate in your LogFrame plan, yet the practitioners involved with the
project are generally held accountable only for achieving the project's Purpose (USAID,
2012). Why not the project Goal? As we have seen, evaluating the ultimate impact of a
project on its Goal is very dicult, both in terms of determining appropriate measurements
and ensuring capacity to measure in the rst place (Savedo et al., 2006). Practitioners are
therefore not typically held accountable to this norm.
9
For an example, see the work of Miguel and Kremer (2004) on the eects of worm treatments on health and
educational outcomes using a randomized control trial (RCT) method to explore causal mechanisms.
47
So how do we use the LogFrame tool beyond listing our objectives? Notice Figure 2.1
also contains columns for Indicators, Data Sources, and Assumptions. It is in these columns
that stakeholders create standards and measure project components. These columns also
give us insight into why project Goals can be so dicult to evaluate. Each column has a
critical function and its own challenges, which include the following:
Indicators.
Indicators allow you to measure progress during each stage of the project. Let's take the
following example:
Project Goal: increase life expectancy in South Africa.
Project Purpose: reduce the prevalence of HIV in South Africa.
There are many dierent measurements or Indicators you could choose from to measure
project Outputs. For example, the total number of HIV cases per year, number of new
cases per year, number of classrooms where safe sexual practices are taught, number of
students who attended safe sex classes, number of condoms distributed, number of HIV tests
performed, number of deaths per year, and the list goes on. There are also measurement
trade-os that come with each of these Indicators, especially when trying to evaluate if a
project is achieving its overall Goal. While it may be easy to measure how many classrooms
oered lessons on safe sex, it is not easy to tell if there is a causal relationship between
teaching students about safe sex (Input) and an increase in life expectancy in South Africa
(Goal). So we may think instead that the total number of HIV cases per year is a sound
way to measure our Purpose, but we may be capturing an increase in patient deaths in that
statistic rather than a decrease in new HIV cases. We need to know which it is; if more
people are dying of HIV, then our project might be having no eect or may even be harmful
to achieving our Goal. The diculty of establishing causal relationships on such complex
issues is one reason why projects tend to rely on Output measurements that directly measure
project activities, rather than its overall Goal or impact.
48
Data Sources.
As any researcher knows, high quality data to measure one's Indicators can be extremely
dicult to obtain, especially in countries where there is little infrastructure to collect data
on a broad scale. This means aid project stakeholders will need to (a) collect their own data,
which requires time and qualied individuals, (b) rely on data collected through existing
means, regardless of the methods used to collect it, or (c) impute information from the
existing data of neighboring regions, past years, etc.
If the stakeholders choose to collect data themselves, they need to factor these costs into
their budget, which may make evaluation unattractive in pure monetary terms. As touched
on previously, typical costs involve properly trained evaluators, translators, investment in a
platform (tablets and software, costs of printing, GIS, etc.), extending the project timeline,
travel between data collection sites, and so on. If these costs are prohibitive, then projects
may have to collect lower quality data or rely simply on Output measures.
If the donor uses existing data, it does not have control over the validity or reliability of
that data. This means that while the Indicators chosen may be good in theory, in practice
the data mean very little because information was not collected in a rigorous way. Morgan
(2013) brings this issue to light when he compares advanced data analysis techniques that
use outdated collection methods, to the Iron Chef cooking a gourmet meal with road kill
(animals unintentionally killed by passing cars) as the main ingredient. Indeed, practitioners
and academics could be using the best-available analysis techniques, but working with poor
data that is not worth analyzing. This would mean we cannot properly evaluate our Outputs,
let alone Purpose or Goal.
The nal Data Source option for donors, to impute the data, faces the same problems
as using someone else's data and compounds those problems with issues of applicability. By
using data from a neighboring country, you risk not only using data that was essentially \road
kill" in the rst place, but applying the inaccurate information in now a second region and
basing project decisions upon that bad information. Additionally, it may be inappropriate
49
to impute data, as the Purpose of the project may be lost during detached mathematical
calculations performed in faraway oces, as opposed to surveying the actual impact of an
agency's project on the ground. If you choose to use infant mortality rates in Uganda to
estimate those for its neighbor D.R. Congo, you will miss the reality of rampant con
ict in
the Congo that greatly impacts its infant mortality numbers.
There are two more considerations to take into account when considering Data Sources:
the level of analysis and the time frame. In terms of level of analysis, data collected on
the national level may be inappropriate for assessing, for example, a literacy intervention in
three towns. Or maybe there are data in the target province, but the data are only on adult
rather than child literacy. You must be sure that the data available are appropriate to use
based on the scope of your project.
Time frame also becomes an issue. If you want to assess literacy improvement, you could
test immediately after the intervention, but maybe it is better to test in a year to see if
students retained the literacy lessons. You must be conscious of the time of year, and which
months students are in school to ensure a maximum sample size. Finally, annual data - the
default in most databases - may be too vague to help when facing a crisis like Ebola. Project
stakeholders must carefully consider each of these elements when choosing Data Sources for
all of their Indicators.
Assumptions.
The Assumptions category is a standard category within all LogFrame analysis. This sec-
tion aims to clarify the external conditions that need to be in place in order for the project
design/plan to work. For example, if we build a sh farm and we assume there will be no ty-
phoons this year, then the community cooperative can sell the harvested sh in a few months.
We cannot plan on a sh harvest in the short-term unless we assume a natural disaster will
not occur. Otherwise we must assume a disaster will occur, and make accommodations in
our project plan up front to address this risk. Additional Assumptions might include that
50
government leadership must remain stable, unemployment levels must stay the same, the
currency will remain stable, current rainfall levels will continue, etc.
Enumerating such Assumptions not only makes the types of uncertainty inherent in the
project clear to project stakeholders so that risks can be prevented or mitigated beforehand,
it allows evaluators to easily see where disappointing results may have gone wrong (USAID,
2012). Practitioners must carefully think through all possible Assumptions ahead of time,
which requires in-depth local knowledge and reinforces the need for beneciary participation.
They must also assess the importance of these Assumptions to overall project success, have
a plan to deal with Assumptions that no longer hold, and consistently monitor Assumptions
during project implementation. Without any one of these elements, a project is likely to fail
in part or in whole, though evaluators would have a clear road map as to why. Assumptions
also reinforce why it can be so dicult to attribute the success of a project to its overall Goals,
as many factors beyond the scope or control of the project implementers aect Indicators
and Outputs.
2.1.5 What do we Measure in Project Evaluation?
Despite the popularity of LFA approaches and our in-depth explanation of LFA vocabulary
here, the vocabulary used by donors is varied and complex in practice. When attempting
to study the eectiveness of aid projects, the dependent variable can be unclear due to
inconsistent terminology, and dicult to compare across donors. What one donor calls a
project's purpose, another calls its objectives, and the distinction between terms such as
output, outcome, and result are fuzzy (see Table 2.2 for a comparison of donor terms). This
problem is exacerbated when attempting to convey these concepts in other languages. For
example, in Spanish there is no equivalent of outcome or output, the closest are translations
for results and impacts. By the same token, there is no dierentiation in English for two
Spanish terms often used in Central American development contexts, manejo (decisions and
actions taken by those who directly manage resources) and gesti on (decisions and actions
51
undertaken by organizations to try and in
uence those who directly manage resources),
which are simply both translated as management in English (Imbach, 2016).
Given the many problems with matching terminology, development evaluator Alejandro
Imbach (2017) encourages individuals to consider the concepts they wish to convey rather
than getting attached to particular words. The most common words to describe project
outcomes and the concepts behind them are:
10
Products or Outputs: The goods and services a project generates or produces as a
result of project activities. For example: workshops, classes, trainings, conservation
projects, etc.
Results or Outcomes: The behavioral changes adopted by organizations or beneciaries
related to project activities. For example, if an organization adopts a manual produced
by the project; if the producers adopt productive practices promoted by the project,
if the local or national government enact laws or regulations using project ideas or
proposals, etc.
Impacts: Changes in the well-being of people and/or environmental quality. The
changes can be positive or negative. They are the ultimate and broadest changes
generated by projects.
Table 2.2: Examples of Aid Project Evaluation Terminology Used by Donors
Adapted from Professor Alejandro Imbach, per Imbach (2017)
General
Concept
NEDA
11
European
Union
World Bank CIDA
12
USAID
13
NORAD
14
Long-Term
Objective
Development
Objective
Global Ob-
jective
Development
Objective
Long-Term
Result
Strategic
Objective
Goal
Project Ob-
jective
Immediate
Objective
Project Pur-
pose
Development
Objective of
Project
Medium-
Term Results
Intermediate
Result
Purpose
Specic Ob-
jective
Product Result Product Short-Term
Result
Project Re-
sult
Product
Activities Activities Activities Activities Activities Activities Activities
10
This list was translated from teaching materials created by CATIE Professor Alejandro Imbach, per Imbach
(2017).
52
Thus when considering how to measure a project's outcomes, we must carefully consider
which concept we are in fact referring to. Using the above list, the most appropriate measure
of a project seems to be its results or outcomes, which I will use from here on out inter-
changeably. We are most interested in whether or not the project created a desired change
in behavior by the project's target population. For example, it may be a social behavior
targeted, such as washing hands before touching food, or a professional behavior targeted,
such as new farming technique. To know if the project was eective, we want to know if
people now wash their hands before touching food, or if farmers now employ the new pro-
duction technique. We also want to know if any laws or regulations were created as a result
of project activities that enforce behavior change.
Yet, most measures of a project's eectiveness are at the product or output level. Goods
or services generated by a project are very easy to measure. Eectiveness is very rarely
measured at the impact level; it is extremely dicult to measure and trace causality back
to a project, it is expensive to carry out these sorts of evaluations, and the appropriate time
to measure a project's impact is unclear. While we may not be able to measure impact
easily, we can certainly make a greater eort at going beyond simple output and measuring
outcomes instead. Outcomes allow us to see if the project brought some sort of sustainable
change, as behavioral adaptation is much more durable then the simple production of goods
during the project period, as products typically stop when the project stops.
2.1.6 Who Evaluates Success and Failure?
How do we interpret project results? What does the result of a regression or an LFA
analysis mean in terms of development progress? Both micro and macro measures change
in substance and relevance depending on the actor measuring, including the assessment of
whether a project was a success or failure. Remember that Heeks and Stanforth (2014, 27)
11
Netherlands Development Assistance Agency
12
Former Canadian International Development Agency, now merged into the Department of Foreign Aairs
13
United States Agency for International Development
14
Norwegian Agency for Development Cooperation
53
point out, \. . . any ongoing development project can never completely succeed or fail because
its designs and its networks are continuously being re-formed." Diallo and Thuillier (2005)
assert each individual denes success per their own interests. Aid eectiveness is a moving
target. Success is not necessarily dened in absolute terms, but rather progress towards goals
that are constantly changing based on that particular moment in time and the individuals
involved.
When designing and choosing project metrics, it is critical to consider who is measuring:
which actors set the objectives and are interested in the outcomes. Evaluation is implicitly
a donor-instigated activity, meanwhile the actual task of evaluation is often delegated by
the donor to beneciaries or other skilled third parties. Applying McCubbins and Schwartz
(1984) classic categorization of monitoring mechanisms is of use here. Police-patrol monitor-
ing requires active oversight of a project by the donor. In development aid, such monitoring
is typically unfeasible in terms of cost and expertise. However, re alarm oversight relies on
agents - in this case, beneciaries or implementers - to raise an alarm if a project element
goes o course, and is a much more ecient use of the donor's limited resources. Project
monitoring during implementation is often performed by beneciaries, while impact and/or
project evaluations at the midterm and end of the project are often performed by third
party implementers or evaluators contracted by the donor. This drastically reduces the cost
and burden on donors, who generally are not living with beneciaries full-time even if a sta
member is assigned to the project, rarely transfer knowledge before they leave a project (sta
turnover is a substantial barrier in aid implementation), and rarely have a way to enforce
changes.
15
The police-patrol vs. re alarm monitoring is an important distinction, because it shows
how a beneciary may sound a project-related re alarm over a potential threat to the success
of the project, and still not share the same notion of what success means. Even when actors
work together to dene project goals, the power dierentials between them may prevent less
15
See OECD (2011).
54
powerful actors from expressing their actual desired outcomes. While donors are the implied
group measuring an ocial project evaluation, other actors certainly have their own measures
of success. Field interviews I performed in the Philippines during May and June 2015 reveal
this tension. The Philippines was chosen as it represents the `typical' aid story of a country
that has received aid for many years, whose economic growth is showing improvement, but
whose human development and governance indicators remain stagnant or worse over time.
16
My 17 interviews were with foreign aid practitioners from bilateral donor agencies and NGO
donors, along with local partners of such aid agencies representing the recipient government
and implementer categories. All participants were asked the question: \Do you know if
locals ever evaluate the success of the project or aid separately, using their own methods or
resources?" Most respondents either said no, or that if locals were evaluating, it was based
more on their feelings than evidence-based assessments. Examples of answers are listed in
Table 2.3 (see Appendix A for the full text of all respondents).
Then again, answers to this question may simply re
ect a dierence in what goals bene-
ciaries were evaluating. Only local partners mentioned that local actors may have dierent
incentives and thus be driven by dierent goals when they participate in the project. Aid
practitioners may believe that the goals indicated in their own LogFrame are the same goals
of the beneciaries. I also asked all respondents: \Do you think locals in the communities
you work in share the same outcomes? Does someone at the [agency] talk about these out-
comes together with them?" All practitioners discussed how they had set up the goals of
the project with the community. Yet naturally, the community members may not reveal
their true incentives for allowing the foreign aid project into their community, for example
as one local partner implied, that it is a source of income for the community and provides
temporary jobs. In a nation like the Philippines that has been accustomed to a fairly steady
rate of foreign assistance since World War II, the goals of the community members may be
to secure their next in
ow, rather than reach a particular Sustainable Development Goals
16
Drawing on the case selection techniques by Gerring (2007).
55
Table 2.3: Selected Responses Concerning Local Actors Evaluation of Projects
Collected in the Philippines, May-June 2015
Aid Practitioner
or Local Partner
Central Oce or
Field Worker
Response to the question: \Do you know
if locals ever evaluate the success of the
project or aid separately, using their own
methods or resources?"
Local Partner Field ``No. Only if there is { if there are people involved
in this project...It's not that they're not interested
in the project, but...they're just less interested be-
cause they're not getting personal gains from the
project."
Local Partner Field ``Yes, separately...You talk with the people who's
{ who are involved in the project...And so I know
what's going on...And of course what's the re-
sult?...But you know, problem is sometimes you
get a good data in [one source]. But it doesn't
match with the [other source's] data...And so it's
a check and balance whether they're telling the
truth or not....because performance is the most
important for them....The government is provid-
ing performance incentive to { to schools that are
performing better..."
Aid Practitioner Field ``No, I don't think they do. They don't do it for
themselves, they don't do it for me, they don't do
it. They tried to just live by the day."
Aid Practitioner Central Oce ``They have their own view of how the projects are
running, no { I don't know if they are really, you
know { one thing that we really train them as part
of our participatory process is teach them how to
evaluate, how to monitor....So did they do a sci-
entic evaluation and monitoring? No, they just
saw these are the impact in their lives, you know,
and so for them that is a successful project, water,
you know, now there is no one getting sick..."
Aid Practitioner Central Oce ``Yes they do. On the level of their { how, how
do you say, on the level of { on their emotions,
so it's like how happy they are like, how satis-
ed they are, your emotional base evaluation...By
really just pure emotional measurement. I have
never heard of a local community or individuals
that has making evaluation like those, you know,
evaluation like competency: 10, goal deliberation:
5, something like that. Not really."
56
(SDGs) as set forth by the UN.
Consider the following statement from a eld aid practitioner during the same set of
interviews in the Philippines. When asked the question: \What sorts of factors in your
opinion determine if local communities accept or decline a project that you're working on?"
the interviewee, a Philippine national working for a foreign aid agency, responded as follows:
\Normally with the communities, any kind of aid they are usually happy to just
accept and sometimes even- there was even a time when- it was kind of really
sad when you hear someone say, `Oh hopefully there is another typhoon so we
get more aid.' I was like, `Oh!' `No no no it's a joke!' But still you know, you
know somewhere there they are kind of thinking once the aid or any kind, any
type of aid will stop, then they'll go back to their- you know, [to] nix." { Field
Aid Practitioner
The goal of this community where the interviewee was working appears to be keeping a
steady income
ow via aid, even if this came at the expense of a natural disaster. In other
words, there are incentives for locals to tell donors essentially what they think they want to
hear in order to achieve that income goal - even if that means agreeing to the donor's goals,
regardless of whether or not the project is something they truly want. Thus the vocabulary
of project success and failure highly depend on whose goals we have in mind. For example,
if there is a project that comes into a community to set up water ltration systems, even
if those ltration systems fail, the project may still be a success to locals if the community
was most concerned about the salaried jobs that project created, even if temporary. This
means that in the common situation where there are multiple donors working separately on
projects with overlapping activities, donors may be upset because they cannot measure their
`success' or return on investment, whereas local beneciaries may be extremely pleased that
there is income entering the community via three dierent donors at one time.
Practitioners are also quick to point out that development success usually means some-
thing very dierent in dierent country and community contexts, and changes in a macro
indicator such as GDP can obscure important group dierences within a country that hide
development setbacks (e.g. Stiglitz et al., 2010, Fioramonti, 2013, Fioramonti, 2017). In fact,
57
a conversation I had with an IGO ocial in charge of compiling and analyzing cross-country
data revealed cases where countries chose to exclusively focus on improving their score on a
standardized wellbeing indicator due to the signal it sends to the international community,
even if the wellbeing of their citizens on the ground does not actually change.
17
Empirical reliance on project evaluation labels such as success, failure, satisfactory, un-
satisfactory, etc., means we are making an assumption that a single static goal exists for
each project, and that each actor in the aid implementation chain shares the same interests:
generally those of the donor. We are also making assumptions that projects operate in vac-
uums and that local actors also consider projects as singular activities leading to a desired
outcome, rather than as one of several eorts to achieve goals in their communities. Such
assumptions obscure the power relations between actors, the political aspects of projects and
their implementation, and realities on the ground aecting development initiatives of inter-
ests. Removing these assumptions as we consider new measurement tools is an important
next step.
2.1.7 What do Project Evaluations Look Like?
Project evaluations are typically documents that are voluntarily published by individual aid
donors on their websites. Evaluations range from downloadable reports with little consis-
tency in how results are reported, to quantied and standardized rating systems that allow
for searches or easy cross-project or cross-country comparisons. For example, Figure 2.2
shows the Table of Contents for two dierent United Kingdom Department for International
Development (DFID) Project Evaluations from 2013-14. You can immediately see that, al-
though there may be some similarities, there is no standardization between these reports
emitted in the same year from the same agency - section titles, content, and even formatting
dier. When exploring the text of actual evaluations, it can be dicult to tell if the agency
believed the project was overall a success, the exact locations, or even the precise dates of
17
Per discussion with a multilateral specialist in February 2017.
58
the project.
18
Take on the other hand Figure 2.3, which shows the landing page of the World
Bank's Independent Evaluation Group website. You can use this interactive tool to compare
project results across regions, and countries, with projects categorized clearly under labels
such as \highly satisfactory," \satisfactory," and so on. Results are standardized and easily
accessible.
19
While the organization of one evaluation may be clearer than another, that
does not necessarily mean the methods used in that evaluation are superior to the other.
But it certainly aects the ease of access to information and ultimately transparency.
In general, evaluations greatly vary between and within donors in terms of their organiza-
tion and the information included. They may be in the language of the donor, the recipient,
or both. Sometimes evaluations are simply summaries of results, and other times specic
data collected or data collection tools are attached as appendices. Evaluations can be per-
formed at various times in a project. There may be a midterm evaluation, nal evaluation,
or special evaluation called for due to some special circumstance within a project. While
creating a single global template for results may not be appropriate across donors, one major
hindrance to transparency goals is the current diculty of accessing and interpreting donor
evaluations.
2.1.8 What Are Project Evaluations Used For?
Based on my research, project evaluations appear to have three main purposes: 1) inform ad-
justments in implementation, 2) form the basis for future policy/programs, and, 3) appease
outside constituents through transparency \performances." In terms of the rst purpose,
informing project adjustments, the idea is that a project can be restructured to accommo-
date problems discovered during midterm evaluations and ensure a project's Purpose is still
achieved. For example, if monitoring reveals an Assumption in the Input phase no longer
holds true (e.g. the rainfall was not sucient to grow tomatoes), stakeholders can trou-
18
For further examples, see: https://dec.usaid.gov/dec/content/search.aspx, or: http://www.adb.org/search/
series/project-performance-evaluation-reports
19
For other examples, see: http://ieg.worldbankgroup.org/ratings
59
Figure 2.2: Table of Contents Examples from UK DFID Project Evaluations
Source: https://www.gov.uk/government/collections/evaluation-reports
60
Figure 2.3: Aggregated Results Example from World Bank Project Evaluations
Source: http://ieg.worldbankgroup.org/ratings
61
bleshoot to deal with the impact of this external condition. Or maybe a certain Data Source
the project depended on is no longer being collected by the local government, thus a new
Indicator should be chosen and/or a new Data Source found.
However, when speaking to a previous bilateral country director, the retired director
revealed that midterm evaluations are rarely used to improve a project.
20
Plans are often
turned into a project agreement between the donor and recipient before the implementation
phase, and changes to that program agreement are dicult and time-consuming to make due
to burdensome bureaucratic processes on both sides. Thus it is often easier for stakeholders
to simply continue down the planned project path than adjust implementation activities,
obviously hampering opportunities to improve project outcomes. More research needs to be
done to nd out what actually happens with information collected from midterm evaluations.
The second principal use of evaluation, providing information for future aid policy and/or
projects, is explicitly mentioned in the OECD's denition of evaluation, yet in practice this
use greatly diers between agencies (OECD, 2004). For example, the United Kingdom's
largest and very well-known aid agency, the Department for International Development or
DFID, explicitly acknowledged the use of evaluation results in their policy and project plan-
ning. In an aid evaluation strategic plan for 2014-2019 they simply write, \Evaluations
generate evidence and information to improve DFID policies and programmes and those of
the wider development community"(DFID, 2014, 6). Yet in the United States, evaluations
traditionally have not been trusted by policymakers. A 2013 Congressional Report tellingly
revealed, \Evaluations are rarely designed or used to inform policy. Lack of faith in the
quality of the evaluation, irregular dissemination practices, and resistance to criticism may
all contribute to this problem, as does lack of time on the part of aid implementers and
policymakers alike to read and digest evaluation reports" (Lawson, 2013, 16). Nevertheless,
the report pointed out that Congress has expressed interest in changing this norm when
introducing new legislation in 2009, and the U.S. Oce of Management and Budget required
20
Per conversation with retired bilateral country director, April 21, 2016.
62
federal agencies demonstrate support for their line items via rigorous evaluation results when
making all budget submissions in 2014.
The nal purpose of evaluations is what I would term \performing" transparency. When
evaluations are carried out despite their costs, yet their results are ignored, I posit that they
are primarily done in an eort to appease those calling for transparency about aid projects
and results. For example, as alluded to previously, USAID rarely integrated lessons learned
though evaluation into their processes before the 2010s, nor did reports circulate beyond
third party agencies contracted to implement the project. Given that evaluations are costly,
there must be some reason to carry them out. USAID provides some clues as to why evaluate
at all in these cases. The agency published a \Decision Tree for Selecting the Evaluation
Design" with 24 questions practitioners should answer to determine just how necessary an
evaluation truly is (see Figure 2.4 for question examples).
21
The questionnaire gives two
reasons to evaluate: stakeholders are demanding information, and there are questions that
an evaluation can help answer. When an evaluation is not leading to changes in the current
project design, and it is not being used as the basis for future projects or policy, than
we might look at this nal reason to evaluate, which stems from international community
pressures to comply with expanding transparency norms.
2.1.9 How Common Are Project Evaluations?
This section has aimed to dene aid projects and project evaluations in order to better
understand what ground micro level measures have already covered and what challenges these
approaches have generated over time. Aid projects are complex in design, in evaluation, and
in impact. Projects involve multiple stakeholders, often far beyond those directly involved
in projects, and their success is highly dependent on their specic context. Project-level
evaluation is important to try and catch these nuances.
Yet despite the fact that project-level evaluation can provide data which compensates
21
For full document, see: https://usaidlearninglab.org/sites/default/les/resource/les/mod7 decision tree for
selecting evaluation design.pdf
63
Figure 2.4: Examples of Questions from the
USAID Decision Tree for Selecting the Evaluation Design
Source: https://usaidlearninglab.org
for many shortcomings of aggregate measures, there is a more serious challenge to their
utility: they are not being performed at all. For example, an analysis of over 700,000
projects reported by USAID from 1997-2012 found only 22% of those projects could be
matched to publicly available evaluations. Three big players in the aid arena - the Center
for Global Development, The Bill and Melinda Gates Foundation, and The William and
Flora Hewlett Foundation - formed the Evaluation Gap Working Group in 2004 to better
understand the perceived \evaluation gap" in aid studies. They charged that all types of
donors fail to require or produce a large enough number of evaluations to be useful, and
those that are performed are often
awed in their methods (Savedo et al., 2006). They
lambasted the lack of incentives to conduct good evaluations, and combined with the many
technical, bureaucratic, and political obstacles in the way, alleged the entire aid community
lacks sucient data to know if projects are eective. They asserted that despite the costs
and institutional challenges, \ignorance is more expensive than impact evaluations" (Savedo
et al., 2006, 23). Despite the scathing nature of the report, the lack of project evaluations
64
continues to limit global development eorts, and to date those that are performed remain
substandard (Waddington et al. 2015; Hoey 2015; Levine and Savedo 2015a; Levine and
Savedo 2015b; Bell and Squire 2016).
There are no international laws or requirements that force donors to evaluate the results
of their aid projects or to publicize those outcomes, only norms pushed by international
organizations. This opens a very important gap in aid data - donors can still choose which
projects to evaluate and which evaluation reports they will publish, meaning some projects
will be left out. This means we do not yet know the parameters of this information gap,
whether the gap can be closed, or even the consequences of closing that gap. Nevertheless,
donors are making greater eorts to increase their transparency, competency for international
transparency awards such as the one given by the International Aid Transparency Initiative
(IATI). As aid evaluation, and the publications of these evaluations, becomes more and
more commonplace, it is critical to re-assess evaluation practices in order to obtain the most
useful data possible from each project, a point I will return to when proposing a new type
of measurement in Section 4.1.1.
2.2 Macro-Level Measures: Aid at a National Level
Macro-level development measures tend to fall into the quantitative camp; large-N regres-
sion analyses of all types are particularly popular. They are frequently seen in high impact
academic journals, and used by practitioners to inform broad policy or implementation strate-
gies. The formula is typically simple: independent variables like aid, number of projects,
etc., are analyzed in terms of their impact on GDP, poverty, inequality, etc., at an aggre-
gate level. Random eects or country xed eects may be used to account for so many
country-specic factors, but for the most part, variables are treated as homogeneous across
country-year observations. Macro studies became particularly popular after a call by the
World Bank (1998) to assess the eects of aid. The call was persuasive and coincided with a
turn in the social sciences toward large-N, quantitative methods and technological improve-
65
ments to facilitate such analyses (Roodman, 2007b). Many of the seminal macro aid studies
on eectiveness came out shortly after, published in the late 1990s and early 2000s in top
economic journals. These regression studies were in line with academic publication incentives
promoting these big data methods, and the topic of aid also enjoyed broad public appeal.
Thus, these macro development studies became popular in both academia and the media,
and their frequently dismal results - showing that aid was by and large ineective - began
to attract much more attention from aid stakeholders. Project evaluations performed by aid
agencies were no longer considered sucient, as peer-reviewed articles showing disappointing
macro development results seemed to imply donors were hiding the truth.
Yet despite well over 100 published studies involving macro measures, researchers now
have discredited many in
uential macro-level quantitative studies which suggest generaliz-
able conditions under which aid works, such as Burnside and Dollar (2000),
22
by adding data
or changing arbitrary choices in model specication (Easterly et al. 2004, Roodman 2007b).
Macro approaches are unable to separate the eects of aid by its varying purposes (e.g.
humanitarian, security, ocial development assistance known as ODA, etc.), locations, or
donors; determine whether growth causes aid or vice versa; or identify intervening variables
like corruption (Bermeo 2007, Bourguignon and Sundberg 2007, Rodrik 2012).
New research points out not only the sensitivity of regression models, but the many hid-
den assumptions that statistical analyses and large-N studies obscure (e.g. Bermeo 2016),
for example the assumption that all aid is fungible. Additional criticisms of macro-level data
include unclear constructs (e.g. development, aid); aid, donors, recipients are heterogeneous
but are treated as homogeneous; unclear results due to ripple eects, spillover eects that
violate causal inference assumptions, problems with measuring short-term vs. long-term
eects, and continued treatment of correlation as causation; the `black box' problem of sta-
tistical analyses can obscure near collinearity and other model sensitivities; causal outcomes
cannot be attributed to the input of aid alone in complex environments; and that academia
22
For others see: Collier and Dollar 2004, Svensson 1999, Collier and Dehn 2001, Guillaumont and Chauvet 2001,
Collier and Hoeer 2004, as cited in Roodman (2007a).
66
favors signicant results and media rewards extreme stances leading to reporting of only
signicant results (Easterly et al. 2004; Roodman 2007a; Roodman 2007b; Bourguignon and
Sundberg 2007; Stiglitz et al. 2010; Moyo 2010; Ramalingam 2013; Bermeo 2016; Dietrich
2016). Since changes in a macro-indicator cannot reliably be attributed to aid as the casual
mechanism, it is dicult for macro studies to establish when aid is working as intended by
donors.
Yet macro measures are alluring to researchers, as one of the few low cost tools available
to measure aid eectiveness. The topic remains a consistent area of study in a number of
elds; for example, at least ten articles in the past two years of top international relations
journal International Organization alone tackle questions of aid eectiveness, most often
taking macro-level or so-called big data approaches to the study of aid.
23
Despite Easterly's
own criticism of macro measures (Easterly et al., 2004), even he falls prey to their dangers
in his well-known book lambasting the aid regime White Man's Burden (Easterly, 2006). In
a review of this work, Amartya Sen illustrates the consequences:
\To arrive at his negative view of economic aid, Easterly draws on large-scale cross-
sectional statistical analysis...Such intercountry comparisons have become fashionable
as a way of isolating solid connections between causes and eects, but they are seriously
compromised by the diculty of comparing diverse experiences...Many such studies
are also impaired by diculties in identifying what is causing what. For example,
a country's economic distress may induce donors to give it more aid - which may,
in terms of associative statistics, suggest a connection between aid and bad economic
performance"(Sen, 2006, 174).
Sen summarizes the widespread skepticism of a one-sized-ts-all macro approach to un-
derstanding development and aid across countries he shares with the many researchers and
practitioners included here. Of course as elaborated here, the alternative of a micro-level
approach is often sold as capturing more nuanced causal mechanisms and development in-
tervention eects but has been criticized in particular for its linear approach to results and
its inability to scale up. So how and why do micro and macro measurements persist, given
their
aws?
23
For several examples, see Volume 70, Issue 1.
67
2.3 The Relationship Between Development Theory and Measure-
ment
The broader theories from which development measurements follow have an important place
in this discussion. Since its beginnings in Modernization theory after World War II, devel-
opment theories have primarily focused on the macro level: theories that attempt to explain
why we see development at a national or international level are also those which drive devel-
opment strategies, policy, practice, and projects at even the local level. As a result, much
of the existing development research centers on macro-macro relationships in its methodol-
ogy as well, such as the eect of total foreign aid received on a country's GDP or Human
Development Index (HDI) score. This focus on the macro-level has had important implica-
tions for the denition of development success in the literature (discussed further in Section
2.1.6). Using macro indicators could lead us to conclude a country is headed towards success
when these indicators exhibit desired changes, i.e. GDP growth, decrease in inequality, etc.
Yet there is a clear consensus among practitioners that success usually means something
very dierent depending on the country and community context, who is measuring, and the
macro indicator chosen. For example, GDP growth is often equated with success but it can
hide important variations within a country's population that instead represent development
setbacks (e.g. Stiglitz et al., 2010; Fioramonti, 2013; Fioramonti, 2017).
This macro-level theoretical focus is attractive in its parsimonious in explanations of
aid eectiveness, but leaves out critical factors that explain variations in outcomes such
as the role of local actors in the implementation of aid projects. Macro-to-macro studies
essentially hold constant that local actors will comply with both the letter and the spirit of
donor designated projects. In this way, macro-macro research assumes that project outcomes
will somehow accumulate and eventually change macro-level indicators, which in turn will
signal a country's development success. But recipients do not consistently comply with donor
wishes, nor do projects occur in a bubble. For example, the World Bank has at least three
68
dierent mechanisms meant to address complaints received by individuals and communities
who believe they have been negatively aected in any way by a project funded by the World
Bank: the Grievance Redress Service, the Inspection Panel, and the Oce of the Compliance
Advisor/Ombudsman (CAO).
24
The existence of multiple services provides evidence that
even an international organization as expansive as the World Bank recognizes donors and
local actors do not always have aligned interests, nor denitions of success. `Local buy-in' is a
well-known ingredient for project success among practitioners, yet our theories do not dene
this buy-in, account for its creation, nor identify how it actually aects project outcomes
(Gibson et al., 2005). While some of the concepts falling under the Alternative Development
theoretical approach do have a micro-level focus, they tend to focus on individual projects
and have trouble scaling up given the complex environments also aecting the projects they
collect data on. As alluded to earlier, recently popularized randomized control trials (RCTs)
also can suer from this empirical problem, as the treatment and control locations chosen
to compare can never truly be random in the real world, even if the individual treatments
within them can be. RCTs are very controversial in the development world, with the 2019
Nobel Prize in Economics going to three economists who specialize in development-related
RCTs (Abhijit Banerjee, Esther Du
o, and Michael Kremer)
25
meanwhile one of the most
vocal critics of these same development-related RCTs won the same prize in 2015 (Angus
Deaton, see Deaton and Cartwright (2016); Deaton and Cartwright (2018)).
Micro-level variables are clearly critical ingredients to explaining real world development
outcomes and the contradictions we see, but our ability to make them into theories has not
yet materialized on a wide scale. Asking how micro-level variables aect micro outcomes,
and how the micro and macro-level interact, is still a gap in development theory and thus
the measurements that follow.
26
Accounting for all possible interactions between the macro
24
See the Grievance Redress Service (http://www.worldbank.org/en/projects-operations/products-and-services/
grievance-redress-service); the Inspection Panel (http://ewebapps.worldbank.org/apps/ip/Pages/Home.aspx), and
the Oce of the Compliance Advisor/Ombudsman (http://www.cao-ombudsman.org/about).
25
See: https://www.theguardian.com/world/2019/oct/14/economics-nobel-prize-abhijit-banerjee-esther-du
o-
michael-kremer
26
See the systemism approach pioneered by Bunge (1996) and emphasized by James (2017) which establishes
69
and micro-levels in our theories, for the context in which these variables function, and for
their interactions with their surrounding environment is an avenue worth exploring in order
to reconcile the contradictions we see within the literature and in practice.
2.4 The Micro-Macro Paradox and Interpreting Results
Knowing all we know about the pros, cons, and origins of each measurement, how do current
macro and micro-level measures t together to inform our understanding of development
progress around the world? Certainly demands for transparency on development aid out-
comes by constituents and international organizations are on the rise. Since aid project
evaluations are clearly underfunded and underperformed, the dearth of project-level data
forces researchers to use broad macro measures based on more readily accessible country-level
data to try and infer the overall eects of aid, despite their serious measurement problems
mentioned thus far. As such, multilateral organizations and funders have placed increased
pressure on donors to perform project-level evaluations both because of their better pre-
cision and to address the sizable knowledge gap regarding micro outcomes. For example,
the OECD - the multilateral agency which serves as the clear global standard-setter in the
management of foreign aid - has been particularly vocal about improving the impact of aid
on development outcomes through donor transparency and aid evaluations. It has spon-
sored four High Level Fora on Aid Eectiveness between 2002-2011 with participants from
over 100 countries in attendance. The Fora brought together donors of all types (bilateral,
multilateral, NGOs, private foundations, and public-private partnerships), implementing or-
ganizations, and recipients to discuss improving development cooperation and aid delivery.
27
Delegates developed the Busan Principles after the Fourth Level Fora with 10 indicators for
assessing progress, including two indicators pushing for transparency and clear measures of
guidelines for evaluating when theories are complete.
27
See: http://www.oecd.org/dac/eectiveness/thehighlevelforaonaideectivenessahistory.htm
70
development co-operation (GPEDC, 2011).
28
Nevertheless, we have seen slow progress on compliance with the Busan principles.
29
This is not surprising, as the literature demonstrates we are still struggling to measure aid
eectiveness, thus there are few results for donors to publish with condence. There is also
a very important gap in aid data - given resource constraints, donors must choose which
projects to evaluate and which evaluation reports they will publish, meaning some projects
will be left out. Scholars are just beginning to ask how agencies decide which projects they
will evaluate and subsequently publish, even though the transnational movement towards
open aid data and transparency has been growing since 2002 (Allcott 2015, Cameron et al.
2016; Sabet and Brown 2018; Rogl a 2020). This means we do not yet know much about
the parameters of this information gap, whether we can realistically close it, or even the
consequences of closing the gap.
Where does this leave us now? Aid eectiveness research ndings are best character-
ized by the micro-macro paradox. Mosley (1986) rst described the paradox: micro-level
data coming from aid projects have been hopeful and shown good returns on investment,
meanwhile macroeconomic data relying on regressions show the eect of aid on gross na-
tional product growth to be disheartening. More specically, large-N work on aid paints an
overall negative view of its eectiveness, whereas positive anecdotes from individual projects
are often drawn upon in practitioner reports to show the need and utility of foreign aid.
Mosley calls the phenomenon \the `pessimistic' macro-based view of aid" and \the `opti-
mistic' micro-based view" (22).
So how can aid projects be succeeding individually, yet as a whole be failing? Mosley
(1986) poses three possible causes of the paradox: 1) data inaccuracies, 2) biases in project
data (e.g. sampling bias, long-term eects not captured, not all aid is in the form of projects),
and, 3) macro eects of projects are unaccounted for (e.g. eects of incoming project re-
28
The term development co-operation for the purposes of this study refers to the relationship between the aid donor
and recipient. See Busan Indicators No. 4 Transparency and No. 7 Mutual Review at: http://eectivecooperation.
org/wordpress/wp-content/uploads/2015/05/GPEDC-Monitoring-Framework-10-Indicators.pdf
29
Although the last progress report occurred back in 2014 (GPEDC, 2014).
71
sources on market values, fungibility of aid). Subsequent research has certainly lent cred-
ibility to his arguments (see Pritchett 2002; Espina and Zinnes 2003; Savedo et al. 2006;
Collier 2007; Roodman 2007b; Allcott 2015; Savedo et al. 2015; Cameron et al. 2016; Law-
son 2016; Sabet and Brown 2018). Mosley's nal assessment: \both the micro and macro
data are seriously defective" (22).
||
I have reviewed traditional measures at the micro and macro levels, introduced the schools of
project planning and evaluation along with their important vocabulary, and discussed where
these measurements have left us today. In the Methodology Chapter 4, I will introduce a
new process-based measure using lessons from past macro and micro measures, approaches,
and terminology that I use to carry out this study. This Institutionalization Measure, as I
call it, attempts to go beyond measuring the before and after state of a problem a project is
employed to x, instead seeking to evaluate whether change, intended or not, has lasted post-
project. Before that however, I will introduce the Theory driving this research on incentives
and project impacts in this next Chapter 3.
72
Chapter 3
A Theory of Incentive Types
I have now laid the groundwork for my theory by establishing two important contexts: 1)
Much of aid eectiveness theories and research center on macro-macro relationships leav-
ing the micro-level relatively unexplored, and these mainstream theories privilege economic
incentives over other incentive types to motivate local actors and long-term project im-
pacts. 2) This privileging of the macro-level has implications for how we dene and measure
development success: we conclude a country is headed towards development success when
national-level indicators show positive changes, even if micro-level wellbeing is not improved.
This has lead to con
icting results and a reliance on `good enough' numbers in development
practice.
We need micro-level theory that can explore how dierent incentive types in
uence aid
eectiveness while still being adaptable to diverse local contexts. Incentivizing `local buy-in'
is a well-known ingredient for project success amongst practitioners, yet our theories do not
account for this process, how to dene buy-in, nor how it actually aects project outcomes.
To ll this gap, I present a theory on the relationship between local actor incentive types
that motivate such buy-in and the duration of foreign-funded development project outcomes.
In this chapter, my approach is as follows: 1) I will begin with some key denitions, 2) I
will outline the pros and cons of previous theories that describe actor dynamics when imple-
menting projects that help us think about the institutions that shape available incentives,
and 3) I present and justify my theory and hypothesis.
73
3.1 Denitions
3.1.1 Local Actors
Rather than just individual project beneciaries within a community, I dene local actors as
any actor who is from or whose interests lie in the recipient country. Aid organizations com-
monly use the strategy of contracting local NGOs or other local implementing organizations
to implement projects, and these organizations often have important relationships and inter-
ests within beneciary communities (Bebbington, 2004). For example, local implementers
contracted to work with local indigenous communities have incentives to maintain those
relationships in order to implement future projects earmarked specically for indigenous
communities.
Figure 3.1: Typical Actors in the Foreign Aid Implementation Chain
Figure 3.1 depicts the links that initially come to mind in the implementation chain for
aid projects: donors, recipients, implementers, and beneciaries. Each actor has been greatly
simplied here for illustrative purposes, however a bit more information about each link in
the chain is in order. Donors are not just restricted to bilateral relations between countries.
74
Multilateral relationships are also included, namely those between intergovernmental orga-
nizations (IGOs) and aid recipient countries, as are relationships between nongovernmental
organizations (NGOs) and recipient countries. I dierentiate donors from implementers in
this study, given that implementers can be a donor state, a recipient state, or a nonstate
actor: an NGO or enterprise native to the recipient state or from the donor state. My
concept of aid recipients is restricted to actors in the state where the aid is implemented,
in particular government ocials who allow aid into the country. Individuals in a recipient
country who benet from aid projects are termed beneciaries.
In practice, these chains can take on myriad forms, where some actors - such as donor
governments - simultaneously serve as donor and implementer. At the same time, recipient
governments always have a hand in regulating the aid environment but may play no role
in disbursing project funds, as these can be emitted directly to a local community group.
Nevertheless, the way these actors interact during implementation shapes the agency and/or
motivation of local actors to participate in a project's implementation, including the recip-
ient government and beneciaries. Later in this chapter, I will elaborate on the main actor
categories and discuss the dynamic interactions between them that shape project implemen-
tation.
3.1.2 Buy-In
Buy-in is often con
ated with the term of ownership. Although the idea of ownership is often
touted as a development agency goals, along with sustainability, it has a murky denition in
the development community (Gibson et al., 2005). Partnership, control, and responsibility
are all some of the concepts invoked by development agencies to describe ownership, but
ownership typically implies some sort of nancial input in projects. I posit that buy-in is
one aspect of ownership, specically an individual or organizational behavior that is in line
with the donor project goals regardless of personal beliefs or nancial support of the project.
If there is a belief in the project itself and its value to the beneciaries, this of course can
75
strengthen buy-in, but there are many situations in which local actors may not believe in
a project but their behaviors are still aligned with project goals due to other motivating
incentives such as peer pressure.
3.2 Theories of Institutions and Incentives
While I have already explored which development theories privilege which incentive types,
there are also theories that frame how we think about shaping actor organization within aid
projects that make implicit predictions about resulting incentives. To talk about institutions
and incentives separately does a disservice to their intertwined relationship when it comes
to all foreign aid projects: \Institutions are the formal and informal rules of behaviour
that constitute incentives for all agents involved in the aid delivery process; they aect the
performance of foreign aid programmes" (Martens et al., 2002, 1). We will now explore
which of these frames is most useful when establishing a new theory regarding the power of
incentive types on the project-level.
3.2.1 Principal-Agent Conceptions of Aid
Since macro-to-macro studies of aid assume other actors' interests align with those of the
donor, these studies also imply a principal-agent relationship: the agent (implementer, recip-
ient, and/or beneciary) typically complies with the principal (the donor), and sometimes
the agent and principal are dicult to distinguish. In the classic principal-agent relation-
ship because these interests align, the incentives align, meaning the principal will nd the
agent easy to motivate as long as they share the same level of information and openness to
risk (Sappington, 1991). Yet, principal-agent theories leave some room for agents to behave
independently (e.g. Hawkins and Jacoby, 2006), meanwhile aid studies tend to ignore the
decisions and behaviors of any actor other than the donor. When we break down the aid
implementation chain into its component parts, it becomes evident that viewing the aid
76
implementation process through a principle-agent lens is wrought with theoretical inconsis-
tencies, as the principals and agents in this story do not have the same interests nor does
the principal necessarily have control over the agent.
The case of bilateral and multilateral donors provides an illustrative example. Taxpayers
are technically the rst principals in the chain (see Figure 3.2), even though they do not
actually participate in the project implementation. In fact, states should be considered
the agents, as they have been delegated the task of spending tax revenues in the most
responsible and ecient way. But instead we see the rst challenge to viewing the aid chain
as a principal-agent arrangement: year after year, across various surveys, there is often low
support and/or a lot of misunderstanding about aid amongst the general taxpaying public in
donor countries. The U.S. public expresses suspicion over foreign aid and generally does not
strongly support it.
1
Along similar lines, the UK public has indicated stable or decreasing
support for development initiatives.
2
Europeans are more supportive of aid, in general, but
this varies based on the wealth of a given country. Wealthier nations, interestingly, show
less support for foreign aid.
3
Nevertheless, ocial aid
ows from all members of the OECD
Development Assistance Committee (DAC) to all recipient countries has steadily increased
since 2000, demonstrating that the agents are going against the principals' wishes.
4
Now, let us assume for a moment that states are the principal, instead of taxpayers. States
promise the aid to a recipient country either directly or through an IGO. The principal-agent
1
Examples include: Kaiser Family Foundation. (2013, November). Survey of Americans on the U.S. Role
in Global Health. Accessed April 15, 2015, at: http://k.org/global-health-policy/poll-nding/2013-survey-of-
americans-on-the-u-s-role-in-global-health. WorldPublicOpinion.org. (2010). American Public Opinion on For-
eign Aid. Accessed April 15, 2015, at: http://www.worldpublicopinion.org/pipa/articles/brunitedstatescanadara/
670.php. PIPA (2005). Americans on Addressing World Poverty. Accessed April 15, 2015, at: http://www.
worldpublicopinion.org/pipa/articles/btdevelopmentaidra/76.php?lb=btda&pnt=76&nid=&id=. Council on Foreign
Relations. (2012). Public Opinion on Global Issues - Chapter 7: World Opinion on Economic Development and Hu-
manitarian Aid. Accessed April 15, 2015, at: www.cfr.org/public opinion.
2
Hogg, M., & Shah, H. (2010). Engaging the public in tackling global poverty (Joint Brieng from DEA and the
Institute of Development Studies). Accessed May 3, 2015, at: http://think-global.org.uk/resources/item/3330
3
Bauhr, M., Charron, N., & Nasiritousi, N. (2013). Dues Corruption Cause Aid Fatigue? Public Opinion and the
Aid-Corruption Paradox (p. 577). International Studies Quarterly, 57: 568-579. EuropeAid. (2015). The European
Year for Development { Citizens' views on development, cooperation and aid (Special Eurobarometer 421 / Wave
EB82.1 { TNS Opinion & Social). Accessed April 15, 2015, at: http://ec.europa.eu/public opinion/index en.htm
4
OECD. 2015. QWIDS: Query Wizard for International Development Statistics. Accessed May 12, 2015, from
http://stats.oecd.org/qwids
77
Figure 3.2: Principal-Agent Relationships in the Aid Implementation Chain
Dotted Lines Indicate Controversial Principal-Agent Relationship
relationship between a state and an IGO is fairly clear in that the IGO is basically an agent
of the state. But there is a very muddled principal-agent relationship between a recipient
state and an NGO, donor state, or IGO. For example, the recipient country may also be a
member state of the IGO, making the IGO an agent of both recipient and donor. Another
common situation is that the donor tacks on conditions to the aid (tied aid) or uses the aid
as an incentive to encourage change.
5
If the aid is tied, it might be easier to imagine the
recipient country as the agent of the principal, given the recipient state must comply with
some donor wishes. However, the recipient does not actually act on behalf of the donor,
but still in its own interest, which is not a clean t with the principal-agent framework.
The donor may tell the recipient to do particular things in order to receive the aid, but the
recipient is not ultimately bound to the directives of the donor, and often receive the aid
before having to meet any donor conditions anyway.
The principal-agent relationship gets even more complex when we consider that donor
states sometimes give the money through a donor state agency, the recipient state directly, a
nonstate actor operating in the recipient state, or to a nonstate actor from one of the donor
5
There have been mixed results on the eectiveness of either strategy. See Ethier (2003), Brown (2005).
78
states. Any of these entities can be the implementer. A donor state agency who implements
the project is a clear agent of the donor state. Nonstate actors might be willing to carry out
actions on behalf of the donor through contract bids tendered by the donor to implement a
project, potentially { though not always { engaging in a more straightforward principal-agent
relationship. However, the norms of sovereignty help shield a recipient state from too much
interference by donor states or NGOs, thus disrupting any clear principal-agent relationship
when the donor gives aid directly to the recipient. Once we get to the project level, no
matter the implementer, it is a stretch to view beneciaries as the agents of implementers
or especially donors. Beneciaries are not tasked with making decisions on behalf of either
actor. Beneciaries can only accept, negotiate, or resist implementer decisions.
Here we reach the crux of the problem: our current understanding of the aid regime
requires that beneciaries comply with donor wishes in order for aid to be `eective,' yet
there is no clear principal-agent relationship between these actors that would facilitate strong
incentives for beneciaries to comply. A donor's `agents' are not really their agents at
all. Realistically, independently-motivated participation by beneciaries is needed to make
positive changes sustainable in the long-term and outcomes feasible, which means that they
will be driven by separate incentives alternative. Thus the way we have conceptualized the
relationship between actors is problematic, and further complicated by the fact that we lack
clear outcome measures of aid eectiveness and ultimately development success. Donors are
the implied actors who measure and determine aid eectiveness, yet they often must rely
on beneciaries to measure outcomes on the ground. All of these concerns are serious and
compounded by the fact that there is often not enough money, time, capacity, or even quality
measures available for most donors to do justice to project-level evaluation or monitoring in
obtain outcome measures. Thus, donors often rely solely on aggregate measures even though
such measures cannot pick up on the specic type of aid or its specic goals. While donors
may realize that these measurements are not helpful in terms of knowing which aid type is
more or less eective, it allows them to capture something to report to their principals, be
79
those taxpayers or individual benefactors.
Even if we accept the principal-agent model as the best way to characterize aid project im-
plementation, Sappington (1991) also recognizes several additional issues with the principal-
agent model that are relevant here. The rst is that the literature puts most emphasis on
principals and agents as intent on maximizing economic gain, thus privileging economic in-
centives above all others. Yet there has been evidence that noneconomic incentives such as
loyalty, pride, altruism, and so on, can be just as eective motivators for actors and that
over-reliance on economic motivators can crowd out more intrinsic noneconomic motivations
(Bowles 2008; Darrington and Howell 2011). Additionally, these models tend to oversimplify
complex relationships, such as those between aid project actors, weakening their explanatory
power when applied to cases like this.
Finally, all of these various principal-agent relationships assume that recipients share the
same concepts or measurements of aid eectiveness as those of the donor. What happens if
these concepts and measurements are dierent? Combined with the fact that donors cannot
completely control implementation or monitoring, to what extent do their implementation
preferences stand? Let us look to alternative theories of aid actor organization to look for
better explanations of actor institutional organization and their motivating incentives.
3.2.2 The Octangle and Aid Delivery
Gibson et al. (2005) take clear issue with the way in which aid delivery is conceptualized as a
principal-agent relationship in most academic work. First, they claim this conceptualization
of aid delivery is incongruent with donor's more recent emphasis on the idea of local buy-in.
If donors are at the top of an aid chain, how can locals take ownership of projects? Second,
they also emphasize the importance of the many nuances within each type of actor. They
identify eight actor categories,
6
though it is important to note that some of these categories
still may overlap depending on the project:
6
See pg. 63 in Gibson et al. (2005).
80
1. Donor government
2. Recipient government
3. Other donors
4. Donor's international development/aid organization
5. Particular organizations within the recipient government
6. Outside implementers (e.g. contractors, NGOs, private consultants)
7. Civil society organizations or other interest groups in the donor and recipient countries
8. Intended beneciaries
Third, they also underscore that Figure 3.2 is rarely how the actual aid delivery chain
works. Why is a top-down, principal-agent, aid chain delivery conceptualization of aid
inappropriate for the study of aid eectiveness? \By conceiving of development cooperation
in this way, donors may infer (incorrectly) that the incentive structures of their internal aid-
disbursing operations are far-distanced, and thus insulated and largely unrelated to problems
of aid sustainability on the recipient side" (Gibson et al., 2005, 62). Essentially by assuming
a principal-agent structure, donors do not take responsibility for the fact that their own
institutions are in
uencing the incentives available to local actor, which in turn are leading
to short over long-term project outcomes.
There are instead dynamic interactions between elements. A linear organizational model
does not capture those relationships, a fact further supported by my various eldwork evi-
dence presented thus far about the power of locals to refuse a project. While we can most
clearly imagine the principal-agent relationship at play within hierarchical donor agencies
and/or other government agencies, \[w]ork in the political economy tradition has found this
simple model of hierarchy|and thus bureaucracy|fundamentally
awed. Numerous schol-
ars have shown that such institutions are riddled with problems of asymmetric information,
generated by their inherent principal{agent relationships" (Gibson et al., 2005, 43).
7
7
Gibson et al. (2005) additionally cite Alchian and Demsetz (1972); Bickers and Williams (2001); and Tirole
(1986).
81
Gibson et al. (2005) posit that in international development,\...it is more valuable to
think about a set of nested situations that may take on any of a variety of productive
or unproductive relationships" (61). They present the Octangle, as shown in Figure 3.3,
where they illustrate the links between actors, showing dyadic and triadic relationships
that characterize international development. They are further able to analyze actions as
part of a linked whole, rather than in isolation, which better captures the way these actors
interact in practice. Within the Octangle, they posit that some of these relationships are
indeed principal-agent ones, for example the outside implementers that are contracted for
projects are agents of donor governments, recipient governments and their agencies, and
beneciaries. Yet others, such as the relationship between the beneciary and recipient
government agencies, are certainly not. Disaggregating aid delivery in this way does allow a
more accurate analysis of who does or could have accountability to whom.
Figure 3.3: Actors in the Development Aid Implementation Octangle (Gibson et al., 2005, 64)
Dotted lines denote indirect relationships.
What other light can the Octangle shed on incentives? When considering how to design
development projects, one must consider the incentives for not only the local actors, but at
all levels of the implementation chain. Gibson et al. (2005) go on to show that highly mo-
82
tivated donor and implementer sta are a necessary but not sucient factor for sustainable
development projects. They nd that even sta who are motivated and dedicated to their
jobs still succumb to perverse economic incentives. Four factors help mitigate these eects:
increasing the length of sta assignments, information about earlier projects and maintaining
follow-up, retaining younger sta members by reducing the number of short-term contracts,
and promotion or other career rewards based on past performance in projects that were con-
sidered very successful to avoid shirking responsibility in multi-sta project environments.
They give additional recommendation on improving the structure if development organiza-
tions that may help them achieve their sustainability goals: mid-term evaluations, involving
local beneciaries in the evaluation process, encouraging learning about key mission concepts
among sta, ensuring evaluations are read by others, and improving evaluations. For ex-
ample, many evaluations are conducted too late to be applied to current projects to correct
course, they ignore previous relevant evaluation reports, and rarely include the key concepts
that agencies claim to be striving for such as local buy-in and sustainability of results. In
their case study of the Swedish International Development Agency known as SIDA, Gibson
et al. (2005) nd there were low levels of agreement even within the agency about what cri-
teria should be used in evaluations, little trust by sta in evaluation results, and few eorts
to aggregate the lessons learned across evaluations especially in regards to mission-related
goals.
Budgetary and time constraints can also create perverse economic incentives for sta to
continue bad habits in agencies. Honig (2018) makes the argument that overly tight controls
by donors on eld implementers prevents those implementers from using their full skill set
to creatively solve problems on the ground. Pressure to show results from these projects in
a world of `good enough' numbers can also decrease aid eectiveness. Institutions that are
carefully structured in the public and private spheres could help mitigate these problems, as
long as they are adapted to their local context and have the trust of locals.
Ostrom (2010), a co-author of Gibson et al. (2005), took the results of the SIDA study
83
and developed further recommendations for the World Bank to not only change the incen-
tive structures to include more monetary incentives for all actors, but do the exploratory
research necessary to understand the intended and unintended incentives at play. Beyond
encouraging the use of evaluations, she argues we must re-examine what is actually meant
by ownership and sustainability. For example, beneciaries must be able to say whether or
not they want the project. Roles and responsibilities of all actors must be clear, account-
ability mechanisms implemented, incentives identied, and all actors' needs and outcomes
contemplated in project design. Any goods being provided by projects must be examined
beforehand for how they may improve or exacerbate public goods or common pool resource
problems amongst the target population, and incentives adjusted accordingly via institu-
tional arrangements to prevent negative impacts. This recommendation also applies to ser-
vices; consultants' and technical assistants' roles must also be examined to see if they are
promoting or inhibiting sustainability. Ostrom also points out that agencies have a decision
to make: do they continue using traditional aid that simply moves money around or do they
commit to institution change and building, which requires long-term investments and puts
their relationship with recipient countries more at risk given the potential impingement on
sovereignty. Additionally the way budgeting is done in most organizations must change to
remove end-of-the-year pressures to spend money on any project that comes their way.
Based on my own interview data and eldwork observations, the Octangle is a much more
accurate depiction in terms of the complicated relationships on the ground in each and every
aid project. It also is a much more visually striking way of showing why actors are rarely
held accountable for the results of aid projects. Each group of actors is interacting with so
many other groups that assigning responsibility to some group or individual for particular
outcomes may be futile. Work such as Ramalingam and Jones (2008) also backs models like
the Octangle over the linear causality thinking that pervades development practice, despite
its \messy realities" in practice (ix). But is it enough to analyze dierences in incentive
types? While incentives play a key role in their analysis, further clarication on which
84
incentives may play a key role in long-term outcomes needs to be claried.
3.2.3 Team Production as a Clarifying Lens
The team production approach, which derives from the eld of economics (Alchian and
Demsetz 1972; Holmstrom 1982; Mitchell 1998), is not a perfect t for application from
rms to aid projects but it can illuminate the role of incentive types and how local actors
seek to overcome typical collective action problems when applied to the development project
process, including lack of trust, asymmetrical information and free riding. Imagine an auto
factory: a car will not run if the engineer erred in designing the engine, or if it was installed
incorrectly by a line worker. Either way, the outcome is failure: the car does not work.
Aid projects require similar teamwork and each actor plays an essential role. Understanding
when actor incentives are aligned allows us to understand when success is more probable.
Viewing the aid implementation chain as a team production process instead of a principal-
agent relationship is more appropriate to the way aid works in practice. A team production
approach recognizes the agency of each actor in the chain, and reveals that there is no
substitutability along the aid chain. Each link in the chain triggers the next actor to perform
their duties. For example, beneciaries cannot perform their critical role unless the donor and
implementer have performed their own roles. Thus we see that project outcomes represent
cumulative multiplier eects, not additive ones. This enables us to look at the relationships
between each actor in the aid implementation chain, rather than assuming other actors will
follow donor wishes and interests. Identifying strengths and weaknesses in these relationships
may better illuminate the conditions under which we should expect aid to be eective.
By imagining aid projects as processes that require cooperation and inputs by each actor
in the chain, we can better identify both the conditions for aid eectiveness and where to
locate dysfunctions that can lead to failure. In our example of an auto factory, the team
production framework shows us that each auto worker has a valuable function, and that
they must perform their role for the car to properly work. Now taking an example to the
85
aid world, projects can fail when each actor does not perform their role in implementation.
An aid practitioner at a multilateral agency whom I interviewed in July 2015 shared a case
where a fraught relationship between locals and donors led to the cancellation of a large
project, stating locals,
\...made a complaint against our support, and they basically were saying that we're
doing all this support but we are not adequately engaging the local stakeholders and
also this was not a good time to develop the sector because the country has such low
capacity that they could never manage it...And so this was extremely frustrating for us
and very depressing because we were working so hard with them. And unfortunately you
know some of the people that raised the concern said that they had never been contacted
and actually that's not true because we did invite them to certain events and we have
their signatures on the attendance list and they still said that they weren't involved.
So it was very frustrating, but the thing is, and I think that's an important point, the
[agency] is very concerned about their image, and so even though in the end the case
was dismissed, we are now no longer pursuing any activities in this country..."
Here we see a lack of aligned donor and local actor interests, manifested by distrust
and frustration between actors. The role of local actors is not substitutable; without their
cooperation, the project failed. So what were local actors' incentives to participate, if any?
What incentives and relationships must be in place to achieve long-lasting project outcomes?
Certainly there is reason to believe in this scenario that economic incentives were not the
only motivator of local actors, who were instead concerned about implementation capacity
and about simply having a say in what was happening. Huck et al. (2012) uses a simple
team production model to analyze social norms and economic incentives in rms. They nd
that the presence of group economic incentives leads to positive externalities, which then
trigger social norms that incentivize further positive externalities. They ultimately nd that
a rm will have increased productivity when social norms - i.e. nonmaterial incentives - are
introduced as motivators.
86
3.3 A Theory of Incentive Types
While the principal-agent frame is the prevailing theoretical model for aid project organiza-
tion in the literature, the Octangle and team production studies presented here gives us a
much more accurate idea of how project roles are organized in practice based on development
institutions and the dierent types of incentives that motivate buy-in to a project. These
various incentive types may all be likely to produce short-term development outcomes, but
for outcomes to continue after the donor and/or implementer leave, I argue that local actors
will be most eectively motivated to continue project activities by nonmaterial incentives
over material incentives (though not mutually exclusive). Economic incentives include money
or items with tangible value such as land, equipment, etc., while nonmaterial incentives are
typically nonmonetary motivators such as status, peer feedback, guilt aversion, or envy (List
and Rasul, 2011).
Five main arguments support my assertion in the case of development projects. Firstly,
economic incentives related to a development project are often temporary. For example, they
may stop when the project stops, they may not be immediately tangible and only appear
in the long-term, or they may not be distributed individually but rather at the community
level, making them a weak motivator. Secondly, change should be institutionalized in the
long-term to increase the chances that they endure (Finnemore and Sikkink, 1998). Whether
changes are institutionalized via informal norms or formal laws, both routes typically require
local buy-in, and both routes may take time. It would be inecient and infeasible for a donor
to provide constant long-term economic incentives in order to create stakeholder buy-in, but
social incentives could feasibly create that buy-in in the long term because actors can expect
to interact again and again with their social group who will judge their actions (Elster, 2010).
Thirdly, if informal norms are strong, they can last longer and/or perform better than formal
institutions (Cardenas and Carpenter 2008; Elster 2010). In many places where governance
capacity is weak, social norms { typically trust and reciprocity { can stand in for formal
institutions and ensure, for example, contracts are fullled. This means project incentives
87
should tap into these social norms to maximize the duration of outcomes. Fourthly, evidence
suggests that creating relationships of accountability aligns actor incentives, which improves
outcomes (Reinikka and Smith, 2004). Although part of creating accountability is nancial
(World Bank, 2003), when material incentives associated with projects disappear, it is logical
that new behaviors must be enforced via nonmaterial means to continue.
Fifthly, there is little reason to believe that coercion works at all in development projects
given that people in states with low governance capacity or ongoing development challenges
have come to expect that there will be few penalties for breaking the law given the low
levels of enforcement, lack of clarity over what is and is not legal, and because the social
stigma (a social incentive in this case) for breaking the law is often also low (Upho and
Langholz, 1998). This appears particularly true in rural areas and/or with new rules; for
example, many studies in Central America show deforestation rates to be higher inside of
national parks/reserves than outside their boundaries. But even if coercion does work to some
extent, studies coming from economic and organizational psychology warns that coercion
can often demotivate people when compared to the motivational power of nonmaterial-
intrinsic incentives due to factors like increased trust, taking innovative risks due to no fear
of judgment/punishment, and the ability to make choices (Stone et al. 2009; Elster 2010;
Kuvaas et al. 2017).
Alam and Wagner (2016) and Tao (2016) provide further evidence for the importance of
nonmaterial incentives. Using data from a Bangladeshi project, Alam and Wagner (2016) ar-
gue that technology interventions meant to increase human development levels in developing
countries only yield a very small material incentive to adopters, even though they decrease
inequality by expanding technology access and improving governance. Thus nonmaterial
incentives in this case, what they term \procedural fairness" (655) and the reduction of un-
certainty - must be leveraged to improve information and communication technology (ICT)
adoption by individuals that will ultimately improve their own skills. Tao (2016) analyzed
why female teachers in Nigeria avoid rural posts. She nds that the high material incentives
88
they receive do not make up for the self-reported signicant toll such posts take on their
well-being as both women and educators in a rural context. This result informed a new Ru-
ral Teacher Incentive Scheme that focused on nonmaterial incentives that would make these
posts more attractive and reduce well-being constraints, demonstrating the importance of
delineating between incentive types when considering development outcomes. Both of these
studies additionally give further evidence that project actor interests do not necessarily align.
For example, Tao (2016) showed how even implementer interests can dier from the donor;
material incentives for implementers (teachers in this case) still were not enough to ensure
actor participation.
Micro-level evidence from six other countries additionally supports the power of nonma-
terial incentives. Using data collected from Uganda, Zambia, and Peru, Reinikka and Smith
(2004) nd that the public provision of goods can be costly because there is no market
competition that would generate relationships of accountability that align actor's incentives
and improve overall performance. Thus, they recommend that social incentives be used to
create accountability relationships in the public sector. The presence of social incentives -
specically altruism and guilt aversion - also help explain why economically inecient xed
wage contract for Philippine rice planters persist (Goto et al., 2011). Meanwhile, absenteeism
among health workers in Laos is reduced when they work in their home communities, suggest-
ing social incentives { like peer monitoring { serve as important motivating forces Yamada
et al. (2013). Sawada et al. (2013) performed an experiment in Sri Lanka on an irrigation
project funded by JICA, Japan's main foreign aid agency. They found that social incentives
such as altruism can reduce the eect of large group size on increasing free-riding behaviors.
Finally, there is also some macro-level evidence supporting the importance of nonmaterial
incentives. For example, previous survey and experimental work has found at the country
level that high levels of trust and strong reciprocity values are correlated with GDP growth,
and decreases in poverty, unemployment, and inequality, contrasting the typical correlation
between GDP and inequality growth (Cardenas and Carpenter, 2008).
89
Finally, evidence reinforcing the power of nonmaterial incentives from two other relevant
elds bears mentioning: micronance and academia. First, micronance is a development
tool used to give high-interest but small loans to the poorest individuals in society that
allow them to access credit and turn their skills into sustainable livelihoods (Ruben, 2007).
They have become hugely popular over the past two decades, as they t in nicely with the
local turn of practitioners towards community-based and bottom-up development processes.
Additionally, they need little nance in terms of start-up costs and are self sucient through
their interest rates. Loans do tend to be paid back, however incentive types have been
found to play a role in terms of increasing or decreasing payment rates. For example, the
literature on group vs. individual accountability nds that when individuals are in group
lending programs - where other members of their social group have also borrowed and keep
each other accountable - and are switched to individual liability, those with weaker social
networks are more likely to default than if they stayed under the group lending scheme
(Gin e and Karlan, 2014). Thus, nonmaterial incentives such as peer pressure are stronger
at motivating individuals to pay back their small loans than solely the economic incentive
of reducing interest payments or obtaining future loans.
Second, academics either study development or implement development projects, and
oftentimes do both. While they can be outsiders brought in, local academics are often called
upon to be a part of projects in their country. Merton (1973) long contended that given the
institutional structure of science and academia, the incentives of academics lie in nonmaterial
incentives such as obtaining social validation of one's work for the purpose of advancing
knowledge (i.e. peer reviewed publications) over material incentives like money or higher
salary. Merton even warns that mixing material and nonmaterial incentives can produce
distractions from advancing knowledge. However, he does acknowledge that the presence of
nonmaterial-extrinsic incentives like awards and recognition can help the researcher feel like
they are on the right track and not so alienated from their community, which can keep their
nonmaterial motivations strong.
90
3.3.1 Hypothesis
Following this body of development-related work on the incentives and social norms driving
local actors, I seek to identify what sort of incentives motivating local actors foment longer-
lasting project outcomes.
8
These incentives could be purposefully generated by the donor,
or be externally-induced but are nevertheless relevant to the project at hand. I attempt
to identify what types of incentives motivated local actors during the project implementa-
tion period, and look to see how those motivations mapped on to the duration of project
outcomes. I posit that while short-term outcomes may be motivated by a number of dier-
ent incentive types, for long-term outcomes to continue after the donor and/or implementer
leaves, local actors must at minimum be motivated to continue to maintain project outcomes
using nonmaterial incentives. In other words:
Hypothesis 1 Incentives: If local actors are motivated by nonmaterial incentives, then
project outcomes are likely to last longer.
I describe the operationalization of all variables in detail in the next Chapter 4 on my
Research Design.
8
Note that I specically avoid use of the term sustainability given that studies of this concept deserve a much
larger discussion than I can provide here. This study focuses on one aspect of the concept, which is duration of
outcomes.
91
Chapter 4
Research Design
Aid as a key topic of inquiry remains vitally important, but we are still struggling to measure
it. Context-specic research of the kind I carry out here is what this situation demands.
In my dissertation, I build a theory about the incentives driving local actors to perform
their implementation roles and the relationship between those incentive types and long-
term project outcomes. In order to evaluate this theory, I conduct archival research, eld
observations and interviews on the implementation and outcomes of four foreign aid projects
executed between 1992-2016 in Costa Rica and Mauritius and developed a new measurement
tool for project outcomes and incentive types.
In this chapter, I will discuss the research process and my methodology for evaluating
my main variables. First, I will outline my general criteria for choosing country and individ-
ual project cases (I will detail the specics concerning my chosen cases of Costa Rica and
Mauritius and the projects within them in the subsequent Chapters 5 and 6). Second, I will
elaborate on my dependent and independent variables and their measurement. Third, I will
discuss my specic research process in the eld for identifying projects that t my criteria,
archival work, coding, site visits and interviews, and data analysis.
4.0.1 Macro Context: Choosing Country Cases
Using the case selection techniques of Gerring (2007), I chose two pathway cases to develop
the theory. Pathway cases are meant to probe causal mechanisms in cases where a hypothesis
has been conrmed. In cases where we already know that aid was `successful' within an
92
overall country context, we still do not know the ingredients that make some individual
project factors likely to have longer-lasting impacts over others. Thus, the pathway case
allows us to explore micro-level data and interactions, and how they are related to macro-
level claims of success. Both cases still need to be commonly classied as developing, have
a history of receiving foreign aid, and should be comparable on key indicators. Holding
these elements constant will help minimize additional confounding variables and to compare
results. The second case will be outside of the geopolitical region of the rst case to examine
the generalizability of my theoretical framework. Of course future research needs to delve
into country cases with variation on macro-level variables to see if those variables are also
playing a role in the micro-level variables selected for this study, but this dissertation is
an important rst step to explore what is happening in contexts where projects have a
supportive macro-environment.
Cases of development `success' at the national level give us more condence that aid at a
project level has the potential to be successful. As we have discussed, success is a problematic
concept in the world of development. My operational denition of success is if a country
has 1) risen in income level to one of the top two World Bank income classications and/or
has a strong HDI score, 2) has evidence of stable institutions, and/or 3) has graduated any
bilateral aid programs. These conditions oer some condence that the country has the
macro-environment conducive to project activities. In countries that have remained low on
economic and well-being indicators, that suer from consistent violent con
ict/war, or have
authoritarian contexts that impede measurement, forces from the macro environment are
likely inhibiting project activities. This prevents us from properly assessing the participation
of local actors.
Given the measurement diculties elaborated on in Chapter 2, my approach that takes
into account the historical trajectory of each case and multiple indicators I believe is a better
strategy for identifying cases then just one or two indicators such as GDP growth. I will
now detail each of these three key characteristics I use for case selection.
93
Economic and Human Development
As elaborated on in Chapter 1, there are any number of ways to dene development. Here I
take a comprehensive approach. I am looking for countries that are still generally considered
by international organizations as \developing" but have risen to the upper middle income
category since independence and currently have a relatively high HDI score or on other well-
being indicators like poverty rates, education, health, etc.
1
I am also open to the inclusion
of other indicators that re
ect a healthy economy and way of life for the recipient country
population. This could include, for example, reverse deforestation, stable health care sys-
tems, members of various regional integration schemes, creation of innovative development
measurements, high levels of FDI/a desirable environment for conducting business, etc.
Institutional Stability
Again, institutions are dened as sets of rules, norms and/or processes, that have become
embedded in the fabric of a given polity, and institutionalization is the creation of such rules
(North, 2009). These can be formal as in written policies and laws, or informal in the case of
mutually understood norms. Institutions pattern and structure actions available to actors,
which means that laws, rights, and protections can be enforced and protected. Institutions
are posited to oer multiple benets to development, including long-term growth (Acemoglu
and Robinson, 2006) and economic resilience to external shocks (Rodrik, 1999b).
I am interested in countries that have relatively strong institutions to bolster the macro
environment such that development projects can survive and
ourish. There are a number
of measures that indicate strong national-level institutions. These include:
Scores concerning governance
2
Strength and age of democracy
3
1
I do not use inequality indicators as a standard here because a rise in inequality is often associated with increases
in GDP, although I do include it in narrative descriptions about cases.
2
See World Governance Indicators (WGIs): https://info.worldbank.org/governance/wgi/
3
See Polity Data: http://www.systemicpeace.org/polityproject.html
94
Freedom of expression protections and strong rule of law
4
Membership in international governmental organizations (IGOs)
Existence of national-level development institutions
Development Cooperation
In conjunction with these indicators, I look for cases that have received large amounts of ODA
since independence compared to other countries, but with amounts that have decreased over
the years. I am interested in NGO, multilateral, and bilateral donors who are members of the
OECD's Development Assistance Committee (DAC), as this group represents longstanding
donors. Donors may use the language of \graduating" aid programs or simply lower the
amounts allocated over time. My interpretation is that the international community views
these as cases where development is occurring, rather than places where aid is being rescinded
as a punishment for not meeting pre-established conditions.
4.0.2 Micro Context: Choosing Development Aid Projects Within
Country Cases
By its nature, large-N research treats aid as if it is identical across all instances. However,
aid is distributed in various ways, and aid projects are diverse and complex. Projects involve
dierent donors both in terms of countries and with respect to public vs. private aid. Projects
are in dierent sectors, they follow dierent implementation patterns, they are awarded in
varied ways, to various public and private beneciaries, across dierent time periods, in
dierent governing environments, in extremely diverse cultural and geographical locations,
and often overlap with other projects sponsored by completely dierent donors. Much of the
macro-level measurement that has been done on the eectiveness of foreign aid thus misses
the mark in terms of understanding these complex projects. How can we take all the aid
awarded by all the donors in one country and compare it to another, when almost every
4
See Freedom House Scores: https://freedomhouse.org/report/freedom-world/freedom-world-2018
95
aspect of the way that aid is awarded and used operates dierently across contexts? Macro
measures seek to convert the nonstandardized practice of aid into a single generalizable
theory, which does not make sense when looking at the actual practice of aid projects in the
real world.
We have not yet come up with a useful generalizable theory at the macro level about aid
eectiveness. Is micro-level theory possible, given the diversity of aid projects on which I
have just elaborated? What part of the practice of foreign aid might we be able to understand
across projects? We can take the observable outcomes and behavior of local actors in aid
projects as our starting point and work backwards. First, to see if targeted outcomes indeed
changed and second, to identify incentives the project oered to motivate change. Even if
people vary in the types of incentives they ultimately value in these situations, variables like
location or sector may help us identify the social, cultural, and professional norms that drive
individuals to make those value judgments. This may indeed lead to some generalizability
at the micro level and towards theory-building that allows for important variation between
projects. We can also take the information about behavioral change and incentives learned
from various projects and see if this can tell us something about the overall state of devel-
opment of a particular country. If only monetary incentives are used in a country that has
experienced disappointing development outcomes, and we nd that nonmonetary incentives
are more eective in other places, we will have discovered a new variable to test in explaining
such a macro outcome.
As elaborated in Section 1.4.2, there are three main ways that aid is delivered: bud-
getary support, projects, and programs. My focus is on aid projects for two reasons: they
can be discretely measured, and they constitute the main way aid is delivered by bilateral
and private donors, as well as some multilateral donors.
5
Projects need to have begun im-
plementation at least ve years prior to the site visit to allow time to see some longer-term
5
Programs are notoriously dicult to measure because there is not a clear delineation in aid databases for this
category, and given that these include both budget support and project funds, there is currently no way to identify
which budget support and project observations were coordinated under a program approach.
96
impacts once the project has ended.
While I will focus on one delivery method - project aid - due to its common use and
possibility to be identied in datasets, delivery method is not the only way to delineate
types of aid. Aid is also commonly categorized in major aid databases by the variables
donor type, sector, funding type, approach in terms of who initiated the project, and how it
was implemented. A graphical summary of these characteristics and how they vary across
projects is summarized in Figure 4.1. While there is typically data available identifying the
donor and sector of a particular aid commitment, the funding type is generally not readily
identiable in aid datasets, nor is the approach used - or rather, who initiated the aid
transaction. This information needs to be researched in individual project documents. My
goal is to identify a set of projects that illustrate some variation in these dierent categories
to see if my theory is widely applicable.
Figure 4.1: Main Characteristics of Foreign Aid
However, it is worth noting that there are many more project variables that provide useful
information beyond just those outlined in Figure 4.1, and that represent under-researched
97
variables in aid studies. In an ideal world, we would see far more data collected on each
project, such as the following variables:
6
Anticipated Start Date
Actual Start Date
Anticipated End Date
Actual End Date
Other Organizations Involved
Role of Each Organization
Type of Each Organization (NGO, En-
terprise, Government, etc.)
Planning Approach (e.g. LFA, RBM)
Evaluation Done/Info on Evaluation
Standardized eectiveness categories
Sector ID (per AidData.org)
Donor Project ID
Amount Committed
Amount Disbursed and/or Total Cost
Project Website or Document
Geographic Location / Geocode
Urban or Rural
If Current Violence/Con
ict in Project
Location
Many of these variables are either not included in standard donor reporting mechanisms, or
the variables are included but data not recorded. Thus, existing studies of aid projects are
limited at the moment by the information available. Nevertheless, the information we do
have can help us move forward in developing project-level theories that might encourage more
funding and resources to properly collect more observations on these variables to improve
our theories in the future.
This variety of aid project characteristics could easily confound my hypothesized rela-
tionships if not carefully considered. Comparing projects with the same implementer but
dierent donors can help us better identify the inner workings of disparate bureaucratic cul-
tures across donors. I will focus on projects implemented in just one to two sectors, given
that results can be dicult to compare across sectors especially in terms of duration of re-
sults (i.e. building a bridge vs. an afterschool program vs. increasing civic participation).
Importantly, I will steer clear of projects based on donor or implementer reported `success'
6
These are in part taken from variables oered by AidData.org, and supplemented with additional variables I
am proposing here.
98
on outcomes, both because the outcomes are my dependent variable and to avoid the lan-
guage of `success' and `failure.' As I have argued, this subjective designation depends on
which actor is doing the measuring.
4.1 Variables
4.1.1 Dependent Variable: Project Outcomes
The concepts measured by traditional development project indicators (i.e. economic growth,
poverty levels, life expectancy) are still subject to local, domestic, and international factors
that may have nothing to do with the project, making it dicult to separate out project
eects from other contextual forces at play. Measures rely on donor interpretations of success
or failure, which may not represent the desired outcomes for all actors involved. As a result,
\good enough" numbers are the norm in development policy even if severely
awed, given
international organizations' scarce evaluation resources (Rocha de Siqueira, 2016, 1). The
literature published since Mosley (1986) put forth the Micro-Macro Paradox still does not
refute his original assessment; it is time to consider a new way to think about measuring
project outcomes.
There is already a clear awareness in the development and participatory governance
literature that project changes must be institutionalized, or rendered stable to have a long-
term impact. One way to innovate is to consider focusing on the paths and processes put
into place by projects as our outcomes of interest. We can analyze whether project changes
become stable within a given target population, and trace the causal mechanisms that aect
stability. Thus, I am proposing a new tool for operationalizing outcomes that focuses on
the institutionalization of project outcomes. I ask to what extent have the changes brought
about by the project been embedded within the target population?
Institutions are a useful way to conceptualize development outcomes, since terms like
policies and norms appear frequently in the development literature as salient to the be-
99
havioral change most development initiatives seek. For example, Friedmann (1998) argues
that for any long-term change to take hold, people must actively participate in political
processes and have a role in articulating the policies and norms that shape their own com-
munities. However, their creation or change is not typically measured;Ostrom (2010) cites
the need for more academic research on building institutions through international devel-
opment work and foreign aid to inform better policy. This is especially relevant given that
development IGOs often talk about their initiatives in terms of institutions. The World Bank
has long referred to its development work as promoting norm changes (Martinsson, 2011).
The UN Millennium Declaration changed the norms surrounding development objectives by
establishing the measurable Millennium Development Goals (MDGs), and by reorienting the
development mindset toward a human development focus instead of solely economic growth
(Fukuda-Parr and Hulme 2011; Poku and Whitman 2011; Fukuda-Parr 2012). Changes in
global development priorities have also been called norm changes, for example, the rapid
growth in projects promoting gender equality over the last couple of decades (Swiss, 2012).
Yet despite this use of rhetoric, the MDGs and now SDGs do not focus on norm changes.
Instead they rely on traditional indicators, which again are subject to too many external
forces. For example, one of the targets established by the UN for MDG 7: Ensure Environ-
mental Sustainability was to decrease the number of individuals without sustainable access
to potable water by 50%. But water access is certainly subject to forces outside the control
of individual projects as natural disasters or the nearby construction of a new livestock farm
by a private company could easily interfere. More informative measures would focus on the
processes put into place. For instance, how many people now seek safe water through ltra-
tion systems or nding alternative sources? Or, how much research has been done into root
causes? Ostrom (2010) cites the need for more academic research on building institutions
through international development work and foreign aid to inform better policy decisions.
Measuring the institutionalization of project outcomes { the changes observed during and
after project activities { could avoid some of the above measurement pitfalls and the current
100
trap of \good enough" numbers accepted by many international organizations. For example,
if a project is dedicated to agriculture, but a drought has set in, we might miss the potential
gains of a project in terms of mitigating outcomes that could have been much worse when we
only consider traditional indicators. What are `traditional' indicators? These are sometimes
called `counting' indicators because they are variables that can easily be counted: X number
of people attended a meeting, X number of people were treated for malaria, etc. These
indicators are common in evaluations and easy to generate without needing costly training,
or skilled evaluators. But they do not oer information as to whether such numerical changes
will last over time. A measure of institutionalization evaluates if change occurred among the
target population, and if so, to what degree and how. Focusing on the degree to which a new
rule is institutionalized allows us to identify bottlenecks and also allows for adaptation to
local contexts. As so many in the development community have argued, there is no one-size-
ts-all solution, and traditional indicators often ignore context-specic institutions, goals,
barriers, and comparative advantages. An institutionalization measure at the micro-level
focuses on analyzing processes instead of counting indicators.
I will use a real world example from Armenia to oer a more in-depth breakdown of the
dierences between traditional indicators and the more process-based indicators that I am
proposing here (see a summary of this comparison in Table 4.1). The Millennium Challenge
Corporation (MCC) is a prominent U.S. aid agency begun under the G.W. Bush administra-
tion. It approves aid for a few countries that meet MCC pre-requisites and works with those
countries to design and implement development projects. MCC worked with the government
of Armenia from September 2006 to September 2011 (MCC, 2012). The overall project objec-
tives were \to reduce rural poverty through better economic performance in the agricultural
sector as a result of rural road rehabilitation and improved possibilities for irrigated agricul-
ture" (MCC, 2011, 4). By May 2012, MCC made its nal report on project activities. Using
a country-led implementation strategy to create local buy-in, it worked with various public
and private actors in Armenia to rehabilitate rural roads and improve the water supply.
101
Indicators used to measure road improvements included the average annual daily trac on
roads, the sections of roads rehabilitated, and road roughness scores. Water-related indica-
tors included canals cleaned, pumping stations remodeled, and the number of farmers and
specically female farmers now using these new water management techniques. While impact
assessments are mentioned, they were not slated to be conducted after 2014, so long-term
impact data are unavailable. This evaluation strategy simply does not tell us if sustainable
processes were set up three years post-project. While MCC mentions anticipated eventual
beneciary income increases due to its investment, it does not address whether training or
nancial investments led to such changes or the establishment of processes concerning the
way roads and water supply were used and managed.
Table 4.1: Dierences in Traditional vs. Process-Based Development Project Measurements
Illustrated via the MCC Project in Armenia (2006-2011)
Sector Traditional Indicators Process Indicators
Roads Average annual daily trac on
roads
Sections of roads rehabilitated
Road roughness scores
Number of road sections completed by
the local partner after training
Number of future projects done by the
local partner using the new skills
Annual road maintenance by the local
agency post project
Annual rehabilitation/maintenance
level of new roads
Water Canals cleaned
Pumping stations remodeled
Number of farmers and female
farmers using these new water
management techniques
Number of new farmers later trained by
original farmers
Continued use of new practices by
trained farmers
Improvements of techniques by trained
farmers/further adaptation for local con-
text
Can we say that the MCC project in Armenia was successful, at least per MCC standards?
It certainly reported meeting its targets on its indicators. But if these one-time interventions
on roads and irrigation infrastructure only brought short-term (e.g. 3-5 year) gains that did
not endure, would the donor still deem the project successful? Would locals? If the project
102
did not bring road or water improvements, should we attribute it to external forces such
as the diminishing democratic governance occurring in the country around 2012, which the
MCC nal report mentions, or poor project design/implementation? We are hard pressed
to say. It is additionally not clear what the partial accomplishment of a goal might mean.
The MCC target was to rehabilitate 24.4 roads, which is exactly how many they counted
as rehabilitated. If they had only done 14, would it still have been assessed as successful
by their standards, or would it not be considered worth the MCC investment? Would it
make a dierence if those 14 roads accounted for 90% of the vehicles on the road or 10%?
Typical outcome indicators cannot provide us with this information, thus, the utility of the
evaluation is limited. We need to understand the processes the project put in place in order
to fully understand its impacts.
To this end, we could instead measure whether or not professional norms in the road
and water industries changed. We could observe how many road sections were completed
by the local partner after training, and whether future projects done by the local partner
used the new skills that were attained. We could measure whether there was a change in
processes or policies instituted in the local department of transportation. This could be
monitored in subsequent years to determine 1) whether these roads were maintained by
the local agency, and 2) the rehabilitation/maintenance level of new roads. We could nd
out how many new farmers were later trained by locals on water management, and whether
those farmers continued to use the practices they were taught. We could explore whether the
farmers improved on those techniques, or further adapted them to local context. Traditional
indicators can still be helpful by revealing disconnects between practices and outcomes, but
they should not be the only measurement used. We will have far more telling results when
we look at processes instituted rather than indicators alone, as these can easily be aected
by other confounding macro variables, including market prices, natural disasters, private
investments, and so on.
103
A New Tool
The analysis of institutionalization is not exclusive to any one discipline. Subjects from pop
culture to educational curriculum involve the institutionalization of new rules and behaviors.
As Polski and Ostrom (1999) argue, institutional analysis must be multi-disciplinary, given
the complexity of human interaction on the collective level. Thus, this tool I am proposing
has been created with an eye toward attempts to characterize, analyze and/or measure the
phenomenon of institutionalization in a variety of disciplines. I draw on the work of politi-
cal scientists March and Olsen (2011); economist North (2009); political economist Ostrom
(2010) and her earlier authored work with Gibson et al. (2005); educational experts Miles
(1987), Miles and Louis (1987), Curry (1992), Fullan (2007); and sociologists Hedstr om
(2005), Colyvas and Powell (2006), and Pawson and Tilley (2010). Additionally, I fortied
the tool via interviews with 11 development evaluation specialists based in Costa Rica, who
represent ve nationalities and experience working on development projects in almost every
Latin American nation along with complementary work on several other continents around
the world. These individuals have decades of experience teaching development theory and
implementing/evaluating aid projects. The thresholds of rankings/rubric is being established
through applied eldwork in Costa Rica, evaluating three dierent foreign-funded develop-
ment projects implemented between 1990-2016 in dierent sectors and representing dierent
donors and geographic locations within the country.
Based on the combination of literature review, expert interviews, and eldwork, I posit
that we can divide institutionalization into ve types of change processes: two at the group
or collective level (rules, resources), and three at the individual level (behavior, knowledge,
preferences). While development projects often target tangible changes to rules, behavior,
and knowledge, the literature demonstrates that consistent resources and preferences are
important factors in sustaining change. All ve types are complementary; new institutions
will be more stable the greater degree of change we see across categories because of the
extensive interaction between levels, wherein group variables reinforce individual-level vari-
104
ables. However, changes in these ve categories may occur at dierent rates, not at all, or
in dierent directions. Changes in the macro environment of a project also aect the micro,
and are accounted for in the measurement tool by examining target populations involved in
a project on a national and subnational level.
An overview of these main categories can be seen in Figure 4.2, and a complete version
of the tool can be found in Appendix C. The goal is to assign a level of Institutionalization
to project outcomes, regardless of whether project consequences were intended. As such, I
have created the measure with an eye toward several components:
Level of Analysis. Given that new sets of rules are formed at both the collective and
individual level, this component breaks down observable changes into the Group and
Individual levels.
Variable. This captures the ve types of change processes just mentioned: two at the
Group level (rules, resources), and three at the Individual level (behavior, knowledge,
preferences).
Characteristics of Change. This component breaks down the ve change Variables into
more discrete categories, in order to give guidance on where to look for changes in each
of the categories, and to organize complex interactions in the environment into groups
of related activities so that the Variables can be operationalized and measured in more
meaningful ways. They have also been arranged by Capital, per the seven community
capital classications identied by Flora et al. (2004), to ensure that changes occurring
across community assets are captured. The seven capitals are: human, social, cultural,
political, nancial, natural, and built.
Observable Attributes. This category elaborates specic observable activities that are
related to each Characteristic of Change. A Scale has been established for each Charac-
teristic in order to assign a level of change, where `0' equals no change, and `3' signies
that the change has been institutionalized. Additionally, the tool asks questions about
105
Figure 4.2: Basics of an Institutionalization Measure to Assess Project Outcomes
A complete version of the tool can be found in Appendix C.
106
whether the changes are Still Active (to identify if changes are ongoing/one-time/or oc-
curred but were later reversed), Who Made Them (to identify loci of power/leadership),
and Why the Change Did/Did Not Occur (to identify bottlenecks in the process).
Observable Populations. Change often occurs within multiple groups that function on
national and subnational levels. Changes to organizations or people at the national
and local levels must be observed in order to better capture macro-level changes that
are also aecting project outcomes.
A New Collection Method
In the case where the project baseline has not already been established by a researcher,
the tool is intended to be used as follows. All documents related to the project should
be collected through online searches, library and archival databases, and project actors.
This includes grant proposals, evaluations, publicity pieces, theses and dissertations, peer-
reviewed articles, books, media stories, government reports and other primary and secondary
project materials that shed light on project implementation and outcomes. These documents
are used for two purposes:
1. To establish basic project parameters such as geographic location(s), budget, initial
goals/objectives, years, donors and donor organization types, implementers and im-
plementing organization types, other organizational partners and their roles, sector,
related publications, named individual actors, and references to additional related doc-
uments.
2. To assign values on an ordinal institutionalization scale for each Observed Attribute.
Values will be assigned using the extensive guidelines established for each Observable
Attribute that is included in the evaluation tool.
7
7
A scale from [-1] to [3] was used for coding, however despite training, the codes still appeared to be too disparate.
Additionally, many of the outcomes coded needed to be double checked in the eld or were dierent in the eld, thus
the scale needs to be further polished for future use. For the purposes of this study, the researcher conrmed whether
or not objectives were met during site visits and interviews with a simple scale: yes, no, somewhat.
107
The documents collected help construct narratives concerning the project, such as the
project initiation or design process, actor interactions, and challenges/achievements. Given
that there are individual incentives to hide embarrassing truths, and the risk of collecting
fallible accounts from memory when relying solely on interviews, these documents provide
a clearer window into the past when attempting to assign institutionalization levels to each
Observed Attribute.
8
To increase reliability, the documents should be coded using this
instrument by several individuals on the research team, and several of the variables should
be randomly picked to check for agreement among coders.
9
Any questions on the project
that cannot be resolved by the documents should be resolved jointly by the team.
Once the research team has established a value for each Observed Attribute, the values
can be conrmed through on-site observations and interviews with key actors in the project
that represent dierent organizations - the donor, implementer, beneciaries, etc. Any re-
maining questions from the research team can also be posed to these actors for answers.
Final values on each Observed Attribute should then be assigned.
Finally, to determine the current level of Institutionalization of the outcome of interest,
the researcher will use a qualitative aggregation method, which takes the mode of the ordinal
score assigned for each Attribute (to avoid arbitrarily taking a mean of these ordinal scores).
The mode will rst be calculated for each Characteristic and secondly for each Variable,
to increase the amount of disaggregated data available to researchers. The researchers will
also produce a complementary narrative that explains aberrations, aggregations that do not
accurately capture the real situation, etc., to not lose important qualitative data gathered
in the process.
In the case that the baseline has already been established, and/or researchers are able
to use this tool from the beginning of the project, then data should be gathered using par-
ticipatory methods (essentially gathered in consultation with local actors).
10
Participatory
8
See descriptions of recall biases in Bradburn et al. (1987); Grimes and Schulz (2002).
9
I trained four dierent English/Spanish bilingual research assistants to code this particular project, with a
minimum of two researchers looking at a given document.
10
For example, the OECD uses the denition of the former Canadian International Development Agency (CIDA),
108
evaluations have been a buzzword in the aid eld since the 1990s. These are often associated
with the move towards the human development paradigm, which emphasizes participatory
approaches as its preferred method.
11
The benet to my approach is that the power dynam-
ics and motivational incentives of local actors will be more clear by discussing them directly
with local actors, and new information may surface that is left out when only discussing
with donors and/or implementers.
Advantages
The proposed tool oers a number of advantages over traditional micro and macro measures.
The rst is that it attempts to include all project consequences, even if they were not
directly caused or intended by the project, emphasized through employment of the term
relevant, addressing the concern of Pawson and Tilley (2010) about capturing unintended
consequences in evaluations. A second is that we can avoid the terms success and failure.
Instead, the focus is on outcomes in a target population, so that all outcomes are included,
regardless of who desired them or whether they were intended. This allows us to better
analyze why some outcomes may have occurred and not others, and in whose interest each
outcome may have been.
Third, it can be adapted to local contexts by allowing categories that can accommodate
for context-specic goals, barriers, and comparative advantages. Fourth, it can be used to
evaluate domestic development projects. Fifth, we can better identify causal chains. For
example, we can observe bottlenecks in the micro or macro environment that are aecting
the institutionalization of changes, such as changes in only one area rather than all ve.
to explain the dierence between traditional and participatory evaluations: \Traditionally, evaluation tended to be
managed with an outsider perspective, often giving little recognition to local expectations and the potential for
stakeholder contributions. In eect, stakeholders were the objects of evaluations { rather than key participants.
Beneciaries, local organizations and governments in recipient countries were left without substantive roles. With
participatory evaluations, key stakeholders can become integrally involved in: [s]etting up frameworks for measuring
and reporting on results...[r]e
ecting on progress, proposing solutions/directions to respond to issues/challenges,
and...[h]elping with the implementation and sharing of evaluation results" (CIDA, 2001, 3).
11
See references to participatory approaches in documents from the Austrian Development Agency in
2008 (https://www.oecd.org/development/evaluation/dcdndep/47069197.pdf), JICA in 2001 (https://www.jica.go.
jp/jica-ri/IFIC and JBICI-Studies/english/publications/reports/study/topical/participatory/pdf/par.pdf), and US-
AID in 1996 (https://pdf.usaid.gov/pdf docs/PNABS539.pdf).
109
Last, and potentially the most innovative advantage, is that the approach can be scaled
up. By analyzing the processes occurring at the project-level, we can rst compare project
experiences in a way that acknowledges variables particular to the local context, and then
scale up to understand a country's development trajectory. For example, we can assess the
number of projects aimed at a particular behavior change, and then evaluate whether that
new behavior is observed on a regional or national level. We have little idea about how
project results might aggregate up to macro level outcomes, but this will give us some idea
whether projects have ripple eects, and whether changes have diused to non-project areas.
We can also then compare institutionalization levels with other project-related variables to
assess the overall development aid strategy in a particular country or to explain its macro
outcomes.
Authors like Sen (1999) and Fioramonti (2013), organizations like the UNDP, and world-
wide multilateral initiatives like the Sustainable Development Goals have all made wide-
reaching arguments that the purpose of development is to increase individual-level freedoms,
yet our metrics of development aid eectiveness continue to focus on macro inputs to macro
outcomes, leaving the micro in
uences undertheorized and omitting the politics of develop-
ment cooperation. Analyzing the impact of local actors on development initiatives is now
essential for understanding outcomes at all levels. An institutionalization-based measure-
ment for aid project outcomes combines approaches within the elds of economics, political
science, international relations, and human development together constituting a vital inter-
disciplinary approach to the study of global problems.
4.1.2 Independent Variable: Incentive Types for Local Actors
Local actors - again dened as any actor who is from or whose interests lie in the recipi-
ent country, rather than just individual project beneciaries - are often open about having
their own disparate interests that do not always align with those of the donor. This obvi-
ously in
uences the eectiveness of incentives and their behavior during and after project
110
implementation. Yet, little is understood about how these incentives function within devel-
opment projects. There are several existing incentive classication schemes in the economics
and organizational behavior literature that can be adapted. For example, economists List
and Rasul (2011) categorize incentives as material (i.e. economic), nonmaterial (status
goods/feedback/social preferences), and other (gifts, shirking). Information scientist Dalkir
(2007) uses a similar classication: remunerative (material rewards), moral (right/admirable
thing to do, individually or in society), and coercive (punishment for failures).
The dierences essentially boil down to three main groupings: material, nonmaterial,
and punitive actions. This creates three types of incentives:
Coercive: some sort of punishment was enacted for failure to comply with a directive,
e.g.:
{ Fine, additional tax
{ Jail time
{ Forced community service
Material: economic/monetary or other tangible reward that saves money, e.g.:
{ Money, tax breaks
{ Land, equipment
{ Food
{ Health services
Nonmaterial:
{ Extrinsic: right/admirable thing to do per one's society/culture (imposed by
others), e.g.:
Status goods, like a new title (but not a raise)
Peer pressure/accountability to others
Praise/positive feedback from others in the community
{ Intrinsic: right/admirable thing to do for an individual (self-imposed), e.g.:
One's moral beliefs about the issue at hand, i.e. a belief in supporting projects
that support rural farmers, gender equality, children's health, etc.
12
Belief in universal principles/human rights, that all are entitled to the same
rights or treatment
12
Moral beliefs include religion, but in such cases of religion, it is important to dierentiate here if the individual
does something because they believe it is the right thing to do (nonmaterial-intrinsic), or because of how they think
others in their faith may view them (nonmaterial-extrinsic).
111
A New Tool
As such, I have created a new tool to classify these incentives, which can be found in Appendix
D. The tool contains the following data columns:
13
Incentive. This captures the three types of incentives previously mentioned: coercive,
material, nonmaterial.
Observations. This category elaborates specic activities to observe that are related
to each Incentive.
Source. This identies the source of the incentive, e.g. the donor, the local or national
government, if it was already present in the community before the project, attributed to
some other phenomenon in the environment such as a natural disaster, from a dierent
donor, from a dierent civil service organization, etc.
Planned. This identies if the the incentive was a planned/intentional part of the
project design.
A New Collection Method
This tool uses the same data sources as the Institutionalization Measure of project impacts.
First, all documents related to the project should be collected through online searches, library
and archival databases, and project actors: grant proposals, evaluations, publicity pieces,
theses and dissertations, peer-reviewed articles, books, media articles, government reports
and other primary and secondary project materials that shed light on project implementation
and outcomes. These documents are used to identify each mention of an incentive present in
the project environment, using an attached Code Book that denes incentives and capitals.
14
13
The tool contains an additional column labeled: Capital. This captures which of the seven capitals from the
community capitals framework (CCF) developed by sociologists Flora et al. (2016) the incentives falls into: natural,
cultural, human, social, political, nancial, and built. While not used in the analysis for this dissertation, it provides
additional information for future research that could provide further nuance to the incentive classication. Further
description to identify each capital is provided within the Code Book for the tool, found in Appendix D.
14
Note: Although combined here, the Code Book does dierentiate between nonmaterial-intrinsic and nonmaterial-
extrinsic categories to make way for future research that separates these categories.
112
The basic process for dierentiating between each incentive is to look for benets present
in the environment that meet the main criteria of each category: coercive: punishment,
material: monetary or money saving benet, nonmaterial: social benet or an expression of
one's personal beliefs.
Again, given that there are individual incentives to hide embarrassing truths, and the
risk of collecting fallible accounts from memory when relying solely on interviews, these doc-
uments provide a clearer window into the past when attempting to identify Observations
of incentives. To increase reliability, the documents should be coded using this instrument
by several individuals on the research team, and several of the variables should be ran-
domly picked to check inter-coder agreement.
15
Any questions on the project that cannot
be resolved by the documents should be resolved together by the team.
Once the research team has established each Observation, the presence of the incentive
can be conrmed through on-site observations and interviews with key actors in the project
that represent dierent organizations - the donor, implementer, beneciaries, etc. Any re-
maining questions from the research team can also be posed to these actors for answers.
4.2 Method
I collect three main types of data from each aid project. 1) Observational data at the previous
project sites to evaluate whether outputs and/or impacts are intact today. 2) Interviews
with any actor I can locate that was involved with the project. As there are nevertheless
incentives to hide embarrassing truths in interviews, 3) I attempt to minimize this issue
by looking for archived dissertations/theses, media articles, government reports and other
secondary research materials that shed light on project implementation and outcomes online
and in available archives within each country.
My methodology for project evaluation has been adapted from previous aid project eval-
15
I used the same trained research assistants to code this particular project, with a minimum of two researchers
looking at a given document.
113
uations.Throughout the process I attempted to use participatory evaluation methods - es-
sentially conrming that my research question was relevant to and gathering data in consul-
tation with local actors.
16
My general process to collect data in my initial country case was
as follows:
1. Creation and adaptation of two data collection instruments to guide interviews, archival
research, and eld site observations: a) an institutionalization tool to measure project
outcomes, and b) an incentive classication tool.
17
July 2017-May 2019.
2. Finalization of an interview guide and eld visit protocol including appropriate IRB
approvals for working with human subjects. March-June 2018.
3. Collection and coding of all existing documentation that could be accessed online,
through local libraries, and through contacts. December 2018-May 2019; October-
November 2019. A complete list can be found in Appendix B.
4. Hiring of four bilingual research assistants to code project documents based on the two
data collection instruments. June-July 2019.
5. Field site visits to previous project sites to collect additional data using the following
methods: observations, interviews, focus groups, and/or informal meetings. June-July
2019; October-November 2019. The overview of interviewees can be found in Appendix
B.
Since I have elaborated on the data collection instruments, here I will focus on the steps
I have taken to identify and evaluate projects once in-country.
4.2.1 Project Identication
My strategy is to rst speak to any relevant contacts in academic institutions or other
development-related organizations to get a sense of how the relationships between actors in
the aid implementation chain have played out in practice in each country. This gives me
a good idea of what projects existed, what sort of data I would be able to collect, how far
back I can look into projects given various constraints, where I might be able to nd archival
16
E.g. the UNDP guide to evaluation (see UNDP 2009), and the midterm and nal evaluations of the fth
operational phase of the GEF Small Grants Program (SGP) in Costa Rica (see Imbach 2014 and Imbach 2015).
17
Described in more detail in Sections 4.1.1 and 4.1.2.
114
documents, and what sort of information people may be willing to share at project locations.
I then select projects to analyze.
4.2.2 Archival Work
Given that there are incentives to skim over project or personal weaknesses during interviews,
I attempt to minimize this issue by triangulating sources. Once the projects are identied
I move to archival work. I look for dissertations, theses, media articles, government reports
and any other secondary research materials that shed light on project implementation and
outcomes through neutral sources such as online and library databases, as well as through
recommendations and personal documents of informants.
18
I use these documents to establish basic project parameters such as geographic loca-
tion(s), budget, initial goals/objectives, years, donors and donor organization types, imple-
menters and implementing organization types, other organizational partners and their roles,
sector, related publications, named individual actors, and references to additional related
documents. These parameters are conrmed with key actors such a previous project lead-
ers during in-person and written communications. I additionally use these documents to
re-construct narratives about the project implementation, such as the project initiation or
design process, actor interactions, and challenges/achievements. Such narratives not only
give me more information about my key variables of interest, but help establish whether
other important dynamics were at play during the project implementation. For example,
documents may reveal that a key actor was replaced in the middle of a project, that a con
ict
between organization arose, or that a natural disaster hit the project site that re-oriented
project activities.
18
A Spanish/English bilingual Master's student in Development Practice was hired to help with this document
collection in Costa Rica.
115
4.2.3 Coding
Four additional research assistants were hired to code documents. All research assistants
spoke at minimum English and Spanish (and a native speaker in one of the other) and had
Master's degrees in either Development Practice or Agroforestry & Sustainable Agriculture.
I trained all four in the use of the instruments and worked with them to address any initial
questions when rst using the instruments.
4.2.4 Site Visit and Interviews
Once the projects had been identied, project documents collected and coded to the fullest
extent possible, then I planned in-person site visits. My contacts typically helped to leverage
initial interviews, and I used the snowball method to gain additional contacts. I rst got
in touch with any key informants identied by my initial institutional contacts and/or who
were identied in project documents. If their contact info was not personally given to me,
I searched for this information on the Internet or contacted the broader organizations with
whom they were aliated (when relevant). One of two outcomes occur upon contact:
1. The contact responds and we set up a date and time for an initial meeting. I travel
to the site for those meetings. If they oer additional contacts, I also get in touch
with those individuals to set up meetings. If the project involved entire communities
at the household level rather than a specic group, I also knocked on doors in selected
neighborhoods.
2. The contact does not respond. If I have no other contacts who I am able to get in touch
with, I go directly to the project site and look for the headquarters of any relevant
NGOs, CSOs, or other organizations that had worked on the project to present myself
in person. If the project involved entire communities at the household level rather than
a specic group, I also knocked on doors in selected neighborhoods.
Once on site, the data is gathered using participatory methods (in consultation with
local actors). The rst way is through observations. Where possible, actors can take me to
see tangible project results such as infrastructure, equipment, crops, etc. The second way is
through interviews.
116
For the purposes of this study, interviews oer a number of benets over other research
methods. Although randomized control trials (RCTs) have become increasingly popular in
aid eectiveness studies, these typically focus on short-term eects. Such experiments only
allow us to see immediate impacts and gloss over one of the biggest threats to aid eec-
tiveness: the uncertainty and complexity of daily life. Additionally, the project beneciary
population can be dicult to reach in developing countries due to language barriers, rural
locations, lack of education or literacy, and long working hours. For various reasons, local
actors are sometimes distrustful of outsiders (Mosley, 2013). This means that electronic
methods or big surveys will not work; the only way to access reliable information is through
face-to-face interviews.
As such, I perform interviews with any actors that I can access who had some role
in the project. I speak to people representing dierent roles related to designated aid
projects: A) government ocials involved in design/implementation decisions (e.g. na-
tional, provincial, local); B) community leaders (e.g. of churches, civil society organizations,
co-operatives); and, C) populations targeted by the project intervention (e.g. farmers in
agricultural projects, parents in child nutrition or education projects).
Interviews occur during site visits in the eld, but are also performed by standalone
appointments or by Skype as needed. Overall, I aim to determine the following:
1. What motivated local actors to participate in the project during implementation?
2. Are there dierences in people's motivation during and after the project?
3. Did the creation of institutions or other sustainable structure during the project change
people's incentives?
To conduct the needed interviews in Costa Rica, I leveraged my
uent Spanish skills and
the access and trust aorded by my aliation with a local scholarly institution, which has
cultivated deep relationships with locals across the country. CATIE is well-connected with
project beneciaries, local government, other local universities, and other aid implementers
working in the country. Interviews in Mauritius were conducted in English, an ocial lan-
117
guage of the country, and I worked with contacts at the University of Mauritius to help
identify projects that had been previously independently evaluated by someone at the uni-
versity. I was careful to dress similarly across all eld interview days in both countries in
plain outdoors type clothing used for eldwork in rural areas in Latin America.
I relied on the snowball method to increase my interview sample, and a semi-structured
format where I asked open-ended questions of interviewees, as they are appropriate for
testing theory given they maximize neutrality and limit leading questions (Mosley, 2013). I
also included questions that the research assistants felt needed to be claried about project
activities when coding project documents for the institutionalization and incentives tools.
Additionally, given the nature of the study and the projects selected, most respondents
come from communities that are rural, impoverished, and/or indigenous, and one project
took place in a prison. Given these demographics, I elected not to audio record interviews,
so that interviewees felt more comfortable and assured of their anonymity. Instead I kept
hand-written notes in the same eld journal across all cases.
4.2.5 Compiling and Analysis
Upon return from the eld, I reviewed and typed up when possible (or asked a research
assistant to type up) eld notes. I then compiled the codes by the research assistants
to look for patterns across variables as described in the sections on my dependent and
independent variables, and adjusted them given my follow-ups at eld sites. I additionally
looked for patterns and inconsistencies in the narratives across documents and interviewees
when analyzing the notes. I present results in each country case chapter.
4.3 Summary
I have presented here three important aspects of my methodology for this project: criteria
for my case selection at the macro (country) and micro (project) level; a detailed outline of
118
my main variables and how they will be measured. This included two new instruments for
measuring project impacts and identifying incentives, as well as a step-by-step guide to my
research process. In the next chapter, I will introduce my two country cases, Costa Rica and
Mauritius, along with the four project cases that I evaluated.
119
Chapter 5
The Case of Costa Rica
In this chapter, I will start by discussing my selection of Costa Rica as my initial case
based on the three aspects I discussed in Chapter 4: economic and human development,
institutional stability, and development cooperation. Next, I will discuss the selection of the
aid projects I analyze within Costa Rica. For each project, I discuss a brief summary of
activities, the project objectives, and the main actors. I end the chapter with a discussion of
the results for each project I found based on archival research, eld site visits, and interviews.
5.1 Case Selection: Costa Rica
Using the case selection techniques of Gerring (2007), I have chosen to study Costa Rica
because it is a pathway case. Pathway cases are meant to probe causal mechanisms. Aid
has been widely considered as `successful' in Costa Rica across donors, although we do not
know how individual projects may have contributed to the dependent variable - development
- within this case. This pathway case allows us to explore micro-level data and interactions,
and how they are related to macro-level claims of success. My goal is to oer a more
complete explanation of those factors that have fostered Costa Rica's unique outcomes as
an aid recipient by including micro-level data of past aid projects.
Aid as a key topic of inquiry remains vitally important, but we are still struggling to
measure it. Context-specic research is what this situation demands. A detailed study
of Costa Rica's development culture allows me to identify variables that can be used and
compared within other contexts. To justify my selection of Costa Rica as a pathway case,
120
I will rst elaborate more on its macro context, and the variables that may have led to its
noted development `success.' Then I will discuss what I am looking for at the micro level
in terms of project case selection, and elaborate on the aid projects in Costa Rica I have
chosen to explore in-depth in order to propose a theory of how local actor incentives aect
the outcomes of projects.
5.1.1 Macro Context: Costa Rica as a Case of Development `Suc-
cess'
Costa Rica has deed the odds. Rottenberg (1993) predicted that a decline in aid
ows
to the country would lead to economic stagnation (as cited in Fox 1998). That prediction
was wrong: when aid
ows fell in the 1990s, Costa Rica's development performance actually
improved on a variety of economic and well-being indicators, such as an increase in per capita
income and a decrease in poverty levels. Changes in such macro indicators have traditionally
been how the academic and practitioner community dene development success. Costa Rica's
successes, then, have often been attributed to factors such as export-led growth and sound
state institutions capable of sustaining stable growth (Shafer, 1994). Looking to Costa
Rica's colonial history, another prevalent development narrative highlights the emergence
over time of a large auent middle class and strong values that support democracy and
social inclusion (Mahoney 2010; Acemoglu and Robinson 2006). Although these macro-level
interpretations oer important insights, they do not fully account for the country's success.
Costa Rica received USD 4.6 billion in foreign aid - specically ODA - from the end of World
War II through 1995 (Fox 1998; Paus 2005). Overall, it has received USD 10.9 billion in
international funding from 1952-2013 (AidData, 2019). This USD 10.9 billion in aid invested
at the micro-level by NGO, multilateral, and bilateral donors has played a vital role in the
country's achievements, meaning micro-level explanations also need to be explored.
So, what lessons can we learn from projects in the Costa Rican case to better answer
the question posed in much of the development aid literature: under what conditions is aid
121
eective? First, I will discuss Costa Rica's development trajectory to understand the context
in which these projects operate.
1
Economic and Human Development
While Costa Rica's per capita average income was low in the 1960s along with its Central
American neighbors, by 2002 the World Bank had classied it as an upper middle-income
country (World Bank, 2019e). It is a world leader in terms of environmental policies, the only
tropical country to reverse deforestation, and it has tripled its GDP per capita since 1960
(World Bank, 2019a). Regionally, Costa Rica has one of the lowest poverty rates, highest
growth rates, top scores on human development measures, and one of the lowest corruption
scores. Its high numbers on human development indicators such as literacy, infant and
maternal mortality, and access to clean water and sanitation have been better than other
countries with similar a GDP per capita in Latin American or even the world average for
all middle income countries (Deneulin, 2005). These achievements have been the result of
a complicated path originating with its inherited colonial structure and winding through
dependence on two agricultural crops, several nancial crises, controversial economic and
environmental practices, and drastic changes in governmental policy and strategy around
the 1980s that precipitated its high economic and human development scores today.
A natural place to start the narrative is with coee, an export of particular importance
to Costa Rica's development history and a crop often associated with the very name of
the country internationally. Nugent and Robinson (2010) posit that Costa Rica's focus on
small scale coee plantations over large scale ones have brought numerous economic benets,
despite having relatively equal factor endowments as its regional counterparts like El Salvador
and Guatemala. Costa Rica has counted on high human capital (education levels, literacy
rates, etc.) and more equal distribution of land since independence, which has allowed the
benets of coee to be distributed more equally across the population. They argue this is
1
For specic scores on a number of development indicators, see 6.2 in the following chapter where Costa Rica
and Mauritius scores are listed for comparison.
122
in part because politics were less militarized in Costa Rica due to its low importance to
colonial Spain and geopolitical factors that favored peace with its neighbors. I would posit
that the small indigenous population and small number of slaves brought to the country were
additional key factors in more equal land distribution. Instead, power has been continuously
re-allocated through democratic elections, and money invested in social services over military
ventures. Nugent and Robinson (2010) additionally argue this occurred because political
elites did not own the majority of land and thus could not control the organization of the
agricultural sector, so political elites instead focused on maximizing the commercialization
of coee globally.
Certainly these commercialization eorts can be seen to this day; coee has brought a
lot of notoriety to the small Central American country and in particular their emphasis on
organic and fair trade coee, even though it does boast several other major agricultural
exports. Coee (grown in seven regions of the country) along with bananas (mostly con-
centrated on the Eastern Caribbean side) made up 90% of exports as of 1962 (Mesa-Lago,
2003). Beef and sugar were large exports starting in the 1960s (Fox, 2003). Later, pineapple
(mostly concentrated in the north), and palm oil (mostly found in the Western Pacic side
of the country) were added to their portfolio. However the high use of pesticides on these
other three crops has been widely criticized in the international community due to their
controversial environmental and health impacts; research carried out at Costa Rica's Na-
tional University showed an alarming average of 18.2kgs of pesticides is used per hectare
2
of
cultivated land, placing the country as rst in the world for its pesticide use (Araya, 2015).
3
The banana industry has been a particularly high user of pesticides over the decades,
however there have been strong incentives for the government and private industry to keep
the banana industry strong throughout national history. After the two oil shocks that rocked
world economies in the 1970s paired with an abrupt drop in coee prices, the so-called Banana
Wars began where Central American countries joined together to try and increase banana
2
For reference, 1 hectare = 2.47 acres.
3
To compare, China ranks second at 17kgs/hectare, while the U.S. averages 2.5kgs/hectare (Araya, 2015).
123
revenue to oset the economic shocks. They created export cartels, however, large companies
like United Fruit Company and Standard Fruit retaliated by reducing export demands by
30% in Costa Rica, creating further pressure on the small nation and a lot of tension in
the region (Bucheli, 2008). The country simultaneously received quite a bit of negative
international attention for its high rates of deforestation, despite its creation of a national
park system meant to conserve forestlands in 1970. More than 75%of the country was covered
in indigenous forest in the 1940s but by 1983, uncontrolled logging for agricultural purposes
left just 26% of the country's forests, resulting in a deforestation rate of 50,000 hectares per
year (Blasiak, 2011).
Costa Rica needed to nd rapid solutions to their economic crises and the hits to their
international reputation, but the government's initial attempts in the 1980s did not have the
results they hoped for. The government attempted to borrow money from the IMF in 1980
and 1981, but their loans were both rescinded due to noncompliance with required austerity
measures that would reduce their social welfare spending which was unacceptable to the
identity of Costa Rican citizens (Deneulin, 2005). Thus they turned away from their previ-
ously favored development model of import substitution industrialization (ISI) and towards
the promise of export orientation and FDI to diversify their economy and create a high-value
manufacturing sector. They created economic free trade zones - known as zonas francas -
to attract multinational companies, although benets from the initial companies who took
up residence in these spaces had little spillover eect for most Costa Rican citizens (Ferreira
and Harrison 2012; Cordero and Paus 2012). After a period of government-implemented
austerity during the Monge administration, the country did take World Bank structural ad-
justment loans in 1985 and 1989. However, these loans ended up increasing the country's
scal decit in the end by weakening the local currency and overspending to attract FDI,
making the country fairly dependent on foreign investors (Rankin, 2012).
So how did the country nd their way out of these nancial crises? Firstly, aid - par-
ticularly from the U.S. - rose substantilly in the 1980s to assist Costa Rica with its debt
124
(Deneulin, 2005). Secondly, Costa Rica's high levels of education, high rates of bilingualism
(Spanish and English), and investment in other social services due to disbanding its military
in the 1940s have long provided a strong competitive advantage (and the reason the govern-
ment found themselves in debt at that point) that was more fully utilized by the country's
private sector in the 1980s and eventually paid o in the 1990s and beyond. USAID funded
the Coalicion Costacrricense Iniciativas de Desarrollo (CINDE) in 1982, a non-partisan pri-
vate organization created by 76 businessmen meant to promote FDI in order to diversify
exports and help with economic recovery during the Latin American debt crisis (Paus 2005;
Rankin 2012). Leveraging the 1985 Tourism Investment Incentives Law that gave special
economic privileges to tourism-related business as well as former President

Oscar Arias'
award of the Nobel Peace Prize in 1987 that established Costa Rica as a peaceful destina-
tion, CINDE worked with the Costa Rican Tourism Institute (ICT) to aggressively promote
tourism infrastructure development (Matarrita-Cascante, 2010). CINDE's principal focus
was on attracting FDI in the clothing sector, however, USAID's funding was conditional
on decreasing public sector spending on welfare programs while also liberalizing monetary
policies, leading to a devaluation in their currency that reduced their competitive advantage
internationally in the apparel sector. When USAID stopped funding in the 1990s as part
of Costa Rica's \graduation" from U.S. aid programs, CINDE re-focused their eorts on
two more industries where their human capital and low wages had a competitive advantage:
information technology and medical device manufacturing.
CINDE's decisions marked a key turning point in the Costa Rican economy. Their eorts
made tourism the third largest source of FDI by the early 1990s, and the country eventually
began promoting their environment as its biggest draw, providing one more incentive to stop
deforestation and promote organic farming for agrotourism purposes (Matarrita-Cascante
2010; Zumbado-Morales 2010). CINDE landed information technology giant Intel's interest
in 1995 who built a USD 300 million microchip manufacturing plant to employ 3,500 locals
- by 2000, exports of microchips reached USD 1 billion, dwarng all other products in gross
125
export value (Paus 2005; Oviedo et al. 2015). CINDE also landed Abbot Laboratories,
a medical device and technologies manufacturer, around the same time. The decisions of
these two multinational companies to set up factories in Costa Rica attracted major national
growth in FDI outside of the apparel industry (Onugha 2000; Paus 2005). Education in the
country improved, turning out a new generation of a highly educated workforce for these
industries. While Intel shut its plant down in 2013, companies like Amazon have recently
lled some of the gap, rapidly building new warehouses and providing many local jobs with
more future plans for expansion. From 1940 to 1990, the country showed huge gains in
indicators such as literacy, infant mortality, health care coverage, and water and sanitation
access (Deneulin, 2005).
Costa Rica remains to this day a model for economic and human development in Central
America and all of Latin America. Nevertheless, Costa Rica is still considered a develop-
ing country in the international community and it continues to struggle scally and with
inequality. In 2016, Costa Rica began implementing a ve-year plan in conjunction with a
USD 420 million World Bank loan, which aims to strengthen health care, scal sustainabil-
ity, and rural development. This loan is eight times the amount of net ODA and ocial
aid the country received in 2014. The austerity measures and debt crisis of the 1980s intro-
duced new factions of the population more interested in a market orientation than a social
welfare one at the national level, creating new cultural and governance tensions to this day
(Deneulin, 2005). Recent corruption scandals have rocked the country.
4
Costa Rica is also
receiving renewed national and international attention over their treatment of Nicaraguan
political refugees despite their dependence on Nicaraguan migrant workers in the agriculture
industry, and also concerning their pesticide use in pineapple plantations, with a number of
local protests and international research looking at the health impacts particularly of drink-
ing water contamination.
5
However, its strong institutional base appears to be keeping the
4
For example, see the 2017 cementazo scandal, reaching up to former President Luis Guillermo Sol s: https:
//www.americasquarterly.org/content/costa-ricas-and-after-corruption-scandal-casts-long-shadow
5
For examples of current news, see: The Guardian: https://www.theguardian.com/business/2010/oct/02/truth-
about-pineapple-production#maincontent;
126
country in tact, providing checks and balances to the governance system at all levels.
Institutional Stability
Institutions are dened as sets of rules and/or processes, and institutionalization is the
creation of new rules (North, 2009). They can be formal as in written policies and laws, or
informal in the case of norms. Institutions pattern and structure actions available to actors,
which means that laws, rights, and protections can be enforced and protected more easily.
Costa Rica is known for its relatively strong institutions particularly compared to its regional
neighbors. It is particularly known for its signicant investments in educational and social
policy, in particular the abolition of the death penalty in 1882, the establishment of free and
compulsory primary education for all children in 1886, the introduction of social security
reforms and workers' rights protections in 1941, and the abolition of their army in 1948.
They are often referred to as the oldest democracy in Central America, with their demo-
cratic institutions often credited for their social development gains. The country has consis-
tently been rated a 10 out of 10 - Full Democracy - since the Polity Dataset began collecting
Costa Rican data in 1946 (Marshall and Jaggers 2011a; Marshall et al. 2019). They have
held elections since their civil war ended in 1949, when they also secured voting rights for
women and the black population, and established a Supreme Electoral Tribunal to oversee
and ensure fair and transparent elections (Booth, 1987). They have a high freedom rat-
ing according to Freedom House at 91 out of 100; the highest ranking of freedom, political
rights, and civil liberties; strong freedom of expression protections; and strong rule of law
(Freedom House, 2019a). The 1987 Central American Peace Accords - signed by Costa
Rica, El Salvador, Guatemala, Honduras, and Nicaragua to ended decades of military vi-
olence and civil con
ict - were led by later Nobel Prize Winner Oscar Arias, President of
Costa Rica at the time. The Accord gave rise to a strong civil society in Costa Rica with
The Costa Rica Star: https://news.co.cr/un-calls-for-sustainability-in-costa-ricas-pineapple-industry/45310;
The Tico Times: https://ticotimes.net/2017/05/15/pineapple-expansion-costa-rica;
Delno.cr: https://delno.cr/2019/11/deuda-pinera-costa-rica-continua-sin-medir-impacto-en-la-salud-de-las-
comunidades-contaminadas
127
many community-level organizations, i.e. producers, women's cooperatives, and tourism for
development associations (Pearce, 1998). These groups tend to interact with national and
international organizations to implement foreign funding on development projects.
The state has been considerably involved in the economic development eorts of the
country since the end of their civil war, providing signicant institutional restraints on the
private sector while still working in partnership with such companies (Booth 1987; Rankin
2012; OECD 2015). They rst began with the eventually failed Costa Rican Development
Corporation (CODESA) in the early 1970s that promoted economic development by acting
as a majority partner in joint ventures with private corporations with the goals of promoting
growth in industries essential for economic development, like attracting higher value added
manufacturing. CODESA was an attempt to decrease dependence on foreign companies
who were taking much of the prots generated from investment and provide the needed
support and growth for national businesses. The government holding company invested in
railways, cement and aluminum production, and agricultural technologies like fertilizers and
new equipment, although many of their investments were never paid back and CODESA
was discontinued. Nowadays, Costa Rica's most important development governance struc-
tures include the Ministry of Planning and Economic Policy (MIDEPLAN) who oversee
the National Development Plan (NDP) prepared with each new presidential term every four
years. Human Development institutions were created in the in the 1970s and continue today:
the Mixed Institute for Social Aid (IMAS), and the Social Development Fund and Family
Allocations (FODESAF) (Deneulin, 2005). The country additionally has created the 2030
National Development Goals and a National Planning System that does all monitoring and
evaluation of essentially every government action.
Internationally, Costa Rica is a member of the Central American Common Market or
Mercado Com un Centroamericano (MCCA), part of the broader regional integration initia-
tive known as the Central American Integration System or Sistema de la Integraci on Cen-
troamericana (SICA). They have signed a number of bilateral and multilateral free trade
128
agreements, and the country began the process of joining the Organisation for Economic
Co-operation and Development (OECD), an organization of essentially the most economi-
cally powerful countries in the world.
6
The Inter-American Court of Human Rights - one of
the two branches protecting human rights on behalf of the Organization of American States
(OAS) - sits in the country's capital, San Jos e. Additionally, Costa Rica is a member of all
major IGOs, smaller regional development banks such as the Andean Development Corpo-
ration, has helped create the UN Children's Fund, the Optional Protocol to the Covenant
Against Torture, the UN Commissioner for Human Rights, and the Inter-American Human
Rights System (Brysk, 2005). All these actions set precedents within their country and in-
ternationally to model what is appropriate behavior, and also to make appropriate behavior
habitualized.
Development Cooperation
In 1961, Costa Rica ranked 71st out of 119 nations reporting in terms of net ocial develop-
ment assistance and ocial aid received
7
(60th percentile) (World Bank, 2019d). However
by its lowest point in 2013, Costa Rica's ODA in
ows had dropped to a ranking of 124th
out of 142 nations reporting (87th percentile). Costa Rica's aid in
ows peaked in 1985 at
USD 591.4 million and ODA at 7.7% of GNI, but rapidly dropped o in the 1990s as its
development successes became apparent. By the 1990s Costa Rica was paying back more
than it was receiving in soft loans and aid, and at its lowest point in 2013, ODA accounted
for only 0.06% of its GNI. A recent World Bank loan has bumped this up, but the amount
still remains low, with most recent data reporting ODA accounts for just 0.18% of its GNI
in 2017 (World Bank, 2019d).
Of the aid to Costa Rica reported to AidData.org spanning entries back to 1958, one
percent of entries are for general budget support to the country, accounting for 27% of donor
6
See: https://www.oecd.org/latin-america/countries/costarica/
7
In constant 2015 USD.
129
committed funds.
8
Budget support to Costa Rica has been exclusively given by bilateral and
multilateral donors, whereas projects are supported by both government and private donors,
consistent with global aid practices. The data is far from perfect, but given the information
available at the time of project selection, the majority of funds entering Costa Rica since
the 1950s have been for projects.
9
Costa Rica's strong commitment to democratic principles since World War II and abo-
lition of its army in 1948 after the Costa Rican Civil War, made it an increasingly safe bet
across time for donors to make aid investments. So much so that donors were willing to
take a participatory and community-initiated approach to aid. A report marking 70 years of
aid collaboration between the U.S. and Costa Rica emphasizes that the two actors worked
together to determine aid projects: \[f]rom the beginning the two countries were careful
that activities supported by the U.S. would respect Costa Rica's independence and national
identity, while responding directly to local needs. Throughout the years, the U.S. supported
initiatives either begun or requested by Costa Rica" (Mack and McNeil, 2012, 1). The report
goes on to quote former U.S. Ambassador to Costa Rica, Frank McNeil, who emphasized
the importance of a productive donor-recipient country relationship: \To the extent U.S. aid
worked well, it was due primarily to a collaborative relationship between Costa Rican and
U.S. authorities. . . In my experience elsewhere, in Southeast Asia and Latin America, the
recipient's relationship to donor nations was often less collegial, hence less eective" (Mack
and McNeil, 2012, 4).
The idea of a participatory approach to development projects continues to be a major
8
This percentage does not change when only looking at projects listed as starting in 2000 or after, after Costa
Rica is considered to have graduated many donors' aid programs.
9
AidData.org is considered as the most expansive aid database to date, incorporating data from a number of
sources including the OECD, which previously was considered the most expansive aid dataset. Nevertheless, data
on aid from around the world is notoriously incomplete, most likely from a combination of a lack of human and
technological resources to keep complete data over time, to political factors that incentivize hiding information. In
the data analyzed to come up with these percentages (downloaded on October 1, 2017), it is important to note that
66% of entries for Costa Rica have no date attached to the projects, making it very dicult to be condent in whether
aid delivery has changed since 2000. Additionally, 5% of the projects have no budget commitment amount listed, which
includes 21% of the listing for budget support, meaning my conclusion that most aid to Costa Rica comes in the form
of projects may be inaccurate. Nevertheless, the data used here represents the most complete information to date.
To search aid given to Costa Rica through AidData.org, see: http://aiddata.org/dashboard#/advanced/analytic-
dashboard.
130
buzzword used in Costa Rica amongst international and domestic development entities. An
annual summit by the international NGO Future Without Poverty held in Turrialba in 2017
continuously pushed the idea of needing local buy-in, and ownership from local communities.
The 6th People, Planet, Peace Conference on International Sustainable Tourism held in San
Jos e in 2017 hosted by the Costa Rican government entities in cooperation with a number of
NGOs and academic institutions continuously focused on the idea of sustainable tourism as a
development strategy for developing countries, again pushing that locals must be involved in
planning, implementation, and maintenance for tourism to serve as a successful development
strategy. Given the importance of the tourism sector to Costa Rica's development goals, the
Costa Rican government recently revised their criteria for their Sustainable Tourism Certi-
cation (CST) Program because participatory processes involving local community members
are the new norm. Costa Rica is certainly a strong example of this new development trend
that privileges local ownership and participation over the past two decades.
Costa Rica is additionally a global leader in terms of implementing new and more inclu-
sive development measurements. Such eorts are aligned with their international relations
strategy, as Costa Rica has long sought to in
uence other states when it comes to well-being.
Brysk (2005) has termed the in
uencer states as Global Good Samaritans (GGSs), identi-
fying six who frequently act in this way including Costa Rica.
10
States are likely to act like
GGSs when they want to advance their reputation within the international community as
one with high moral currency; essentially they want to be a leader in moral or human rights.
Brysk nds that states are likely to become GGSs when they have the following four char-
acteristics: globalized (economic, politically, and culturally), domestic democracy, adequate
development and security, and a middle power in the international community.
Former Vice President H.E. Ana-Helena Chac on particularly pushed Costa Rica towards
10
The six countries are Canada, Sweden, the Netherlands, Costa Rica, South Africa, and Japan - see also Brysk
(2009). Brysk (2005) states GGSs can provide powerful in
uence on framing or ratication for treaties, helping
victims, monitoring oenders, sanctioning oenders, creating institutions and recruiting personnel, gaining resources,
encouraging peace processes, encouraging transnational action on new issues, and contributing dierent understand-
ings of what rights are.
131
new indicators and indices that show the state of well-being in Costa Rica beyond economic
growth. For example, the country is one of the rst to implement the UN's Multidimensional
Poverty Index (MPI) as a more encompassing measure of poverty at the national level,
launching its use in October of 2015.
11
In a recent UN General Assembly High-Level Side
Event in September 2017 on the MPI as a tracker of Sustinable Development Goals (SDG)
progress, Chac on spoke as one of a handful of selected bilateral representatives on this
achievement:
\We are the rst country in the world to have a public-private partnership that has
helped us to implement the Multidimensional Poverty Index. This alliance with Hori-
zonte Positivo, which brings the best of the Costa Rican private sector, has facilitated
to a great extent this process. In addition, we are integrating the Multidimensional
Poverty Index to the national development plan and for this reason we are the rst
country to base the national budget on the MPI. This means that, with the same re-
sources, we can identify where the deprivations are so that we focus our budget on
them." - H.E. Ana-Helena Chac on, Former Vice President of Costa Rica
12
Additionally, they are the rst country to adapt and implement a Business MPI measure-
ment, launched in 2017, which focuses on poverty levels of employees.
13
They began to use
the Social Progress Index, which focuses on non-economic indicators of well-being, and in
2016 and 2017 showed levels of social progress that were similar to those of Western Europe,
despite its relatively small GDP.
14
More recently, President Carlos Alvarado has sought to
position the country as the global authority on national-level eorts to mitigate climate
change as a development tool, attempting to be the rst country to ban all single-use plas-
tics and become carbon neutral by 2021.
15
Costa Rica continues to push for its role as a
Global Good Samaritan, despite its other continued development challenges mentioned here.
11
See: http://www.ophi.org.uk/multidimensional-poverty-indices-launched-in-costa-rica-and-el-salvador/
12
For a summary of the meeting on September 19. 2017, and the transcription of this text, see: http://www.ophi.
org.uk/ophi stories/unga-2017/
13
See: http://www.ophi.org.uk/ophi stories/costa-rica-launches-an-innovative-tool-business-mpi/
14
See reporting in the Tico Times http://www.ticotimes.net/2016/07/04/costa-rica-social-progress-index and the
2017 raw data https://www.socialprogressindex.com/assets/downloads/resources/en/English-2017-Social-Progress-
Index-Findings-Report embargo-d-until-June-21-2017.pdf.
15
See news articles, for example: https://www.outdoorjournal.com/featured/environment/plastic-free-carbon-
neutral-can-costa-rica-keep-promises/
132
5.1.2 Micro Context: Project Selection
Costa Rica clearly provides a macro environment conducive to project success. To explore
how individual project dynamics operate within its context, I chose three past projects
each implemented approximately 10 years apart. I focus on aid projects with the same
implementer: CATIE (in English, the Tropical Agricultural Research and Higher Education
Center), a graduate school and research center whose headquarters are located in Turrialba.
CATIE's focus is on natural resource management, rural development, and sustainability in
Costa Rica and throughout Latin America. CATIE itself has been a direct aid beneciary
(many of their buildings and programs were donated by USAID) as well as an implementer
of many foreign-funded development projects. CATIE's portfolio of foreign aid projects
includes, for example, the war on drugs, child emigration crises, CAFTA trade-opportunities,
and hurricane relief eorts. We know from the prolic research on international organizations
and state institutions that bureaucracy plays a huge role in an organization's behavior (e.g.
Barnett and Finnemore 2004, Schuller 2012). Again, comparing projects with the same
implementer but dierent donors can help us better identify how the confounding variable
of disparate bureaucratic cultures across donors manifests. Another benet as a higher
education institution is that CATIE has quite a bit of data from this period collected and
archived in the form of reports and student master's and doctoral theses that were not
funded by the donor.
16
Again, I will not be choosing projects based on donor or implementer
reported `success' on outcomes, both because the outcomes are my dependent variable.
In the following sections, I will introduce the three Costa Rican projects I have selected
for in-depth study in order to build my theory concerning local actor incentives in foreign
aid development projects. The main variables of all projects studied have been summarized
in Table 5.1. The table shows that while variation exists within each category, this variation
16
An initial analysis of the main sectors for which Costa Rica has received aid are dicult to determine given
somewhat vague classication using data reported to the OECD Credit Reporting System, in turn reported by
AidData (see: http://aiddata.org/aiddata-research-datasets). Nevertheless, it appears that the social sector (health,
education, and governance), environmental protections, and agricultural projects are particularly popular among
reporting donors.
133
has been strategically divided to still ensure comparability between the four total projects,
yet re
ect external validity in terms of the vast variety of project variability. For example,
I have looked at two bilateral and two multilateral donors, projects in two sectors, in two
countries, three time frames, university implementers partnering with local CSOs, and a
variety of geographical locations and initiators.
Table 5.1: Summary of Aid Projects Chosen
Project Years Site Recipient
Country
Donor
Country
Donor
Type
Sector
17
Imple-
menter
Initiator
Proyecto
Manglares
1990-
2002
T erraba-
Sierpe
National
Wetland
Costa
Rica
Sweden,
Norway,
Denmark
Bilateral Agriculture,
Forestry &
Fishing
CATIE,
IUCN
CATIE,
IUCN
Proyecto
Carbono
2004-
2006
Indigenous
territories
of Tala-
manca
Costa
Rica
World
Bank
Multi-
lateral
Agriculture,
Forestry &
Fishing
CATIE CATIE,
Local
CSOs
AC3 2014-
2016
Nicoya
Peninsula
Costa
Rica
Canada Bilateral Water Sup-
ply & Sani-
tation
CATIE CATIE
SWMP
18
2014-
2016
Petit
Verger
Prison
Mauritius GEF Multi-
lateral
Water Sup-
ply & Sani-
tation
Association
Kinou et e,
UoM
Beneciary
Proyecto Manglares - Wise Use of the Mangrove Resources in T erraba-Sierpe
Project Summary The Conservation for Sustainable Development in Central America
project, known as OLAFO, was carried out in several countries. In Costa Rica, it was a two-
part project taking place from 1990-2002 and funded initially by Sweden (agency: SIDA),
with Norway (agency: NORAD) soon joining and Denmark (agency: DANIDA) following
closely after. Part One concerned the use of non-timber products in the Talamanca region,
as lumber extraction was illegal at the time in the conservation-designated region, thus,
it focused on how locals could still establish sustainable livelihoods. Part Two, known as
Proyecto Manglares (English: Project Mangrove), took place from 1992-95 in the T erraba-
Sierpe National Wetlands in the Puntarenas province on the south Pacic coast. There, the
17
Per OECD Classications.
18
Project discussed in Chapter 6 on the comparison country case study of Mauritius.
134
focus was on the sustainable use of red mangroves in charcoal production through a part-
nership with a local mangrove harvesting cooperative known as Coopemangle R.L. OLAFO
had seven locations in total in Central America but for the purposes of the present research,
I focus on Proyecto Manglares in Costa Rica.
Objectives Proyecto Manglares documents list a number of smaller goals carried out in
the areas in and surrounding the T erraba-Sierpe National Wetlands, including:
Research and creation of databases on the status of Wetlands resources, including
biophysical, social, economic, and production information.
Consolidate Coopemangle as a base organization capable of sustainably managing man-
grove resources through organizational, administrative, management and equipment
strengthening.
Technical assistance to improve harvest yields in a sustainable way.
Promotion across local communities of the appropriate and sustainable use of natural
resources.
Environmental training for teachers and students of schools in communities around the
Wetlands.
Coordination between institutions that operate in the Wetlands.
Formulation of a management plan for the Wetlands and creation of a local organization
to implement the new management plan.
Develop administrative alternatives for mangroves located in humid areas.
Disseminate within the community - as well as amongst technical and decision-making
levels - the potential of mangrove resources for rural development.
Institutional strengthening of the (former) Forest Directorate as the principal govern-
ment institution in charge of all protected mangroves at the national level.
These were all meant to build up to the main objective of the project, which was to
validate and demonstrate that communities can sustainably manage mangrove production
activities, and later scale the model to other Central American communities. The objec-
tives appeared feasible based on implementers expertise, donor resources, and the fact that
135
economic activities were already occurring in the area but needed to be adjusted for more
sustainable management.
Proyecto Manglares was considered a participatory project. The IUCN/CATIE went to
the regions that were chosen, told local actors what sort of project they wanted to work on,
and asked the community what they wanted to do within those parameters. Nevertheless,
it was initiated from the outside by CATIE { not the donor per se in this case, but an
implementing organization with similar power to an outside donor, that catalyzed the aid
from several bilateral sources.
Actors OLAFO's inception is very much tied to its principal actors. The International
Union for Conservation of Nature (IUCN) was founded in 1948 at the behest of UNESCO, at
rst a small organization located in Europe dedicated to coordinated the work of conservation
scientists to protect natural resources.
19
The organization - a public-private partnership that
brings together IGOs, governments, and NGOs - did not itself fund conservation research,
but instead formed special working groups of externally-funded experts to focus in particular
on protected areas and endangered species (Imbach, 2017). They later co-founded the World
Wildlife Fund in the early 1960s to help fund their activities. As of 2019, the IUCN currently
claims to have over 1,100 member organizations and access to more than 15,000 experts in
environmental issues.
20
Their model functioned well up through the Stockholm Conference in 1972, which marked
the beginning of `green' thinking at an international level, later birthing the concept of sus-
tainable development that today frames many multilateral agency activities (Imbach, 2017).
The IUCN actively participated in this conference, but as their work was primarily dedi-
cated to conservation up through this time, they were challenged to consider how to combine
conservation and development, to both empower people and ensure conservation eorts were
sustainable. They launched the famous document, the World Conservation Strategy, in 1980
19
For more information on IUCN, see their main website at: https://www.iucn.org/about
20
See: https://www.iucn.org/
136
where they embraced the idea of people as a part of conservation eorts, rather than people
at the expense of conservation. At that time, the IUCN decided to branch out and open
more oces in other countries.
In 1986, the IUCN approached CATIE about creating operations in Central American
(Imbach, 2017). They sent a researcher from Sweden, Johan Ashuvud, to work full-time
at CATIE in 1987, funded by SIDA. As he began to put together small projects, he con-
tracted Alejandro Imbach at CATIE as a regional specialist, and gave him the initial task of
determining how to combine conservation and development to benet locals. He was then
asked to nd locations within the country for projects that spurred local development using
natural resources. However, Ashuvud passed away in 1988, leaving Imbach to maintain the
connection with SIDA and the IUCN to CATIE. Imbach developed various projects, includ-
ing Proyecto Manglares in T erraba-Sierpe, negotiating the details of these projects with the
communities in 1989 and pulling in funding from SIDA and NORAD in 1990, and DANIDA
in 1993 when Imbach also became Project Coordinator. Figure 5.1 represents a network map
of the involved actors.
Figure 5.1: Network Map of Actors Involved in OLAFO
137
Proyecto Carbono - Carbon Sequestration Project in Talamanca
Project Summary Costa Rica had one of the highest rates in the world of deforestation
in the 1970s and 80s, due in part to inexpensive credit available to cattle farmers and land
title laws that incentivized deforestation for agricultural production and the rapid expansion
of highways in the country (Somarriba et al., 2008). Shortly thereafter, Costa Rica changed
many of its laws to instead incentivize reforestation and sustainable forest management,
switching to a more long-term view of the environmental benets of preserving forests from
the more short-term focus on economic benets. The `Carbon Sequestration Project and
development of environmental markets in cacao plantations and other indigenous agroforestry
systems in Talamanca', or simply Proyecto Carbono as it is known, was implemented between
2004-2006 nanced through the World Bank using Japanese funds. It was considered by the
implementers and project beneciaries as a continuation of the World Bank-implemented
GEF MSP Indigenous Agroforestry Cocoa Production project, although the funds were
specically earmarked for carbon capture projects rather than the biodiversity focus of the
previous project. Such funds were a sign of the times, as carbon capture was a trending
area of research and foreign funding for development projects (Segura Madrigal, 1999). The
project was locally initiated, but the various aid actors most likely had dierent project goals
in line. Ocially, it was intended to help farmers from the Bribri and Cab ecar indigenous
territories in the Talamanca region capitalize on the nascent trend of selling carbon capture
environmental services in order to generate additional income without additional negative
environmental impacts.
Objectives Specically, the project sought to develop a `green market' for their carbon
sequestration capacity that could fund production improvements to current cacao, fruit, and
timber crops. It was considered a pilot project by CATIE to test out the strategy and
feasibility of the concept before implementing it in other areas of Central America with the
following project activities in mind (CATIE, 2006):
138
Building a strategy to increase carbon sequestration capacity on the indigenous-run
farms
Building a mechanism to market and administer the carbon sequestration program
Monitoring and certication of carbon sequestration
Assessing impacts of the program in the socio-cultural and environmental arenas
Developing eective local participatory management skills and structures
Like most of the other projects, Proyecto Carbono had a number of subgoals to further to
the overall well being of indigenous producer households. For example, the project strength-
ened the local community organizations, both in terms of interactions within organizations
and and their interactions with the other organizations (Villalobos and Soto, 2006). This
encouraged, for example, the sharing of seeds and grafted buds to grow better quality prod-
ucts, the sharing of experiences between producers to improve advice about best practices to
producers, and strengthening other socio-cultural bonds within and across the communities
that contribute to community resiliency and well-being. Project documents claim it led to
a new strategy and mechanism for con
ict resolutions between the territories, a resurrect-
ing of their forest development plan, a revitalization of their cabinetmaking industry, plan
to diversify current products such as chocolate, links to other indigenous groups in other
countries, and links to academics from other countries.
Five hundred households from these communities were originally targeted by Proyecto
Carbono (CATIE, 2006). By the end of 2005, 700 producers had been introduced to the
project, 486 had demonstrated interest, and 400 cacao farms had been restructured to in-
crease their production and thus capture more carbon (Villalobos and Soto, 2006).
Actors The actors involved in the project represented a diverse set of international, na-
tional, and local organizations (Somarriba et al., 2008). Figure 5.2 represents a network
map of the involved actors, which shows a complex web of government, CSO, and university
actors working together from the local to the multilateral level. Given the complex nature
of governance of most indigenous territories around the world, this arrangement was meant
139
Figure 5.2: Network Map of Actors Involved in Proyecto Carbono
Translated and Adapted from Somarriba et al. (2008)
to empower voices from various often underrepresented sectors of society, although the pres-
ence of so many stakeholders also meant more time and resources needed to be spent by
each actor in order to reach agreements with so man voices at play.
So how did it work in practice? CATIE worked along with the Costa Rican Ministry of
the Environment and Energy (MINAE) and four local civil society organizations in the area,
listed with their Spanish acronyms: Indigenous Bribri Integral Development Association
of Talamanca (ADITIBRI), Indigenous Cab ecar Integral Development Association of Tala-
manca (ADITICA), Indigenous Bribri Women's Association of Talamanca (ACOMUITA),
Indigenous People and Farmer Coordinating Association for Community-based Agroforestry
(ACICAFOC). ACOMUITA was selected as the organization that would lead all the ad-
ministrative responsibilities of charging for these environmental services, and with getting
the political buy-in from the local community members and producers. ACICAFOC was in
charge of liaising between the three other indigenous organizations and CATIE, creating an
140
indigenous guidance committee for the project, evaluating the opportunities to expand the
project to other parts of Central America and creating a network of indigenous organizations.
CATIE in turn provided the link between the local organizations and the Costa Rican
government and World Bank, and the technical and scientic skills needed to achieve project
goals. For example, it subcontracted the Rainforest Alliance, an international NGO, to
evaluate the market and certication options best suited to the Talamanca region to market
this service on a global level. Meanwhile MINAE organized the other relevant arms of the
Costa Rican government - the FONAFIFO (National Fund for Forest Financing), Ministry
of Agriculture, and the provincial government - and put together an Assessment Committee
that specically dealt with the political issues of the Proyecto Carbono implementation at
the international and national level. At the local level, a Project Guidance Committee was
formed that oversaw overall implementation of the program.
AC3 - Water for Human Consumption, Communities and Climate Change
Project Summary The AC3 project was initiated by a group of environmental economists
at CATIE to the International Development Research Centre (IDRC), and organization es-
tablished by the Canadian parliament in 1970.
21
Given the role of access to water in poverty
and human development, it was proposed as primarily a research project to understand so-
lutions based on behavioral incentives to this issue in areas vulnerable to changing climates,
again to scale solutions to other areas of the world. The project focused on the struggles
of volunteer-run community organizations - of which there are around 24,000 across Central
America - overseeing water distribution in rural areas in several countries in Guatemala,
Nicaragua and Costa Rica given extreme climate conditions in particular geographical re-
gions that are further deteriorating under climate change. In Costa Rica, the focus was on
areas around the northwestern Nicoya Peninsula that are particularly dry, served mainly by
surface and underground water sources, and vulnerable to increasing droughts and demands
21
For more information on the IDRC, see: https://www.idrc.ca/en/about-idrc
141
on limited water systems by a strong tourism sector in the area. The community water
management boards there - known by their Spanish acronym of ASADAs - are charged with
proving supplying water to households and business that are served by the water pumps
and/or wells under their purview.
Objectives Overall, the project had, \the aim of improving the decision-making processes
related to the design of plans and policies for climate change adaptation of [community-based
drinking water organizations] in Central America" (EfD, 2013, para. 4). The individual
objectives for the project were as follows:
1. To analyze, identify and map the impact of climate change on the users of specied
water resources in order to dene priority areas of action at the regional level.
2. Identify and evaluate from a multidisciplinary perspective the factors that facilitate
the successful adaptation of community water administration boards that are facing
intense droughts.
3. Analyze the incentives and constraints, beyond the nancial aspects, aecting the
decisions of communities and households to invest in adaptation to climate change,
including risk aversion, short-term framing, and prosocial preferences.
4. Develop guidelines for the design of multi-tiered and multi-stakeholder (local, national,
NGOs) local adaptation strategies.
5. Contribute to the building of local capacities for adaptation to climate change.
6. Disseminate the results at multiple scales, from the local level up to scientic forums.
The project worked in 81 communities in Costa Rica, and had three phases. The rst
was to create a model predicting the eects of climate change across time in the region.
The second was to do a comprehensive household survey on water consumption, habits,
and attitudes that included an experiment that oered pre-information to some households
about consumption before the survey began. The third was an experiment in just nine of
those 81 communities where researchers and partners installed water reduction technology
in household faucets to see what incentivized people to keep them installed. This nal phase
included a control group, a group where the technology was installed in the faucets, and
142
a group where the technology was installed and they were oered a nancial incentive of
approximately USD 50 if the technology was still installed during a surprise visit six months
later. For the purposes of this research, I focused on communities where the nal phase of
the project was implemented but that were not part of a follow-up project that is currently
being implemented by CATIE, leaving me with three possible communities. One community
declined to be part of any future research,
22
leaving two communities in which I evaluated
AC3 impacts.
Actors The AC3 project was very large, working with numerous communities and in sev-
eral countries, all funded by the Canadian IDRC. Figure 5.3 represents a network map of
the involved actors. While the number of people involved was large, the categories of actors
on all Costa Rican projects were just threefold: CATIE, AyA, and ASADAs. CATIE was
the main implementer, specically their Education for Development (EfD) arm, and worked
primarily with the elected members of each individual water board known as ASADAs within
each community. The Costa Rican Institute of Aqueducts and Sewers (AyA) manages the
main water supplies in Costa Rica which are primarily in urban areas; ASADAs are Costa
Rica's answer to managing the water supply in rural areas, although some areas still have no
coverage at all by any water regulation body and ASADAs are not always able to overcome
the challenges of providing water in some rural areas on their own.
23
ASADAs generally have
four to six volunteer members who are elected on terms of one to three years. Members have
no special training in water management, although the AyA is still in charge of providing all
infrastructure in rural areas and it provides ASADAs technical assistance. ASADAs operate
o of water usage and/or monthly fees. The AyA did have some involvement in the project
22
This was communicated to me by the head of eld work on the CATIE follow-up project. When they contacted
this community, the current ASADA leadership expressed that they felt saturated by research requests with no real
incentive to participate, as results were rarely shared with them or were at aggregate levels, meaning they did not
see any benet to these sorts of projects. Neither the CATIE implementer nor the current ASADA leader had any
prior contact or relationship in order to negotiate this point, so CATIE dropped them from the sample. This is an
interesting outcome in and of itself, as it shows the power of incentive types in motivating local actors to participate
in line with my theory.
23
See: https://www.aya.go.cr/asadas/Paginas/default.aspx
143
by working with CATIE to develop nationwide capacity building workshops on adaptation
to climate change for rural ASADAs.
Figure 5.3: Network Map of Actors Involved in AC3
5.2 Results
Project documents have been coded for all three projects implemented in Costa Rica as
well as Mauritius (discussed further in the next chapter), along with observational data and
interviews - for a complete list of data collected, see Appendix B. As mentioned in Footnote
7, the Institutionalization tool in practice was more complicated to implement then intended.
A scale from [-1] to [3] was used for coding as explained in the Code Book (see Appendix
C), however despite training, the codes still appeared to be too disparate across coders.
Additionally, many of the outcomes coded needed to be double checked in the eld or were
dierent in the eld, thus the tool needs to be further elaborated for easier use in the future.
However a considerable amount of information was still gleaned from the coding process, as
coders were also requested to provide a robust amount of qualitative data during the coding
process that helped the researcher identify consistencies and inconsistencies across codes and
eventually project outcomes. For the purposes of this study, the researcher then conrmed
whether or not objectives were met during site visits and interviews with a simple scale: yes,
144
no, somewhat.
The considerable amount of qualitative data collected does show support for my theory:
nonmaterial incentives for local actors is related to longer-lasting project outcomes. Table 5.2
represents a summary of long-term impacts and incentives present for each donor objective
across all four projects. Photos have also been included of each eld site, however photos
have only been included here that show information that is already part of the public domain
in order to protect the identity of research participants as well as beneciaries in general.
Table 5.2: Summary of Results: Project Impacts and Incentives
Legend for Target Population of Each Objective: I = Implementer, B = Beneciary
Legend for Incentives: C = Coercive, M = Material, NM = Nonmaterial
Project Donor Objective Long-
Term
Impact
Target
Pop.
Incentives
Manglares Validate and demonstrate that communities can sustainably manage
mangrove production activities.
No B M
Manglares Research and creation of databases on the status of Wetlands resources,
including biophysical, social, economic, and production information.
Yes I NM
Manglares Consolidate Coopemangle as a base organization capable of sustainably
managing mangrove resources through organizational, administrative,
management and equipment strengthening.
No B M
Manglares Technical assistance to improve harvest yields in a sustainable way. No B M to C
Manglares Promotion across local communities of the appropriate and sustainable
use of natural resources.
No B -
Manglares Environmental training for teachers and students of schools in commu-
nities around the Wetlands.
No B NM to C
Manglares Coordination between institutions that operate in the Wetlands. Yes I NM
Manglares Formulation of a management plan for the Wetlands and creation of a
local organization to implement the new management plan.
Some I M, NM
Manglares Develop administrative alternatives for mangroves in humid areas. No I -
Manglares Disseminate within the community - as well as among technical and
decision-making levels - the potential of mangrove resources for rural
development.
No B -
Manglares Institutional strengthening of the (former) Forest Directorate as the
principal government institution in charge of all protected mangroves
at the national level.
Yes I M, NM
Manglares Scale the model to other Central American communities. No I -
Carbono Build a strategy to increase carbon sequestration capacity on
indigenous-run farms.
No B M
Carbono Build a mechanism to market and administer the carbon sequestration
program.
No B M
Carbono Monitor and certication of carbon sequestration. No B M
Carbono Assess impacts of the program in the socio-cultural and environmental
arenas.
Some I, B NM
Carbono Develop eective local participatory management skills and structures. Yes B NM
AC3 To analyze, identify and map the impact of climate change on the users
of specied water resources in order to dene priority areas of action
at the regional level.
Yes I NM
145
Table 5.2: Summary of Results: Project Impacts and Incentives - Continued
Project Donor Objective Long-
Term
Impact
Target
Pop.
Incentives
AC3 Identify and evaluate from a multidisciplinary perspective the factors
that facilitate the successful adaptation of community water adminis-
tration boards that are facing intense droughts.
? I NM
AC3 Analyze the incentives and constraints, beyond the nancial aspects,
aecting the decisions of communities and households to invest in adap-
tation to climate change, including risk aversion, short-term framing,
and prosocial preferences.
? I NM
AC3 Develop guidelines for the design of multi-tiered and multi-stakeholder
(local, national, NGOs) local adaptation strategies.
No I NM
AC3 Contribute to the building of local capacities for adaptation to climate
change.
Some B M, NM
AC3 Disseminate the results at multiple scales, from the local level up to
scientic forums.
Some I NM
SWMP
24
To develop a sensitization campaign on wastes [sic] segregation at the
prison and at the community level for green wastes.
Yes I NM
SWMP To carry out a waste characterization study at the prison so as to
identify potential waste treatment options in view of decreasing the
amount of wastes going to the landll.
Yes I M, NM
SWMP To create meaningful activities for detainees in view of their rehabili-
tation and resettlement.
No B M
SWMP To train detainees and ocers on the making of compost on a large-
scale basis, for example, in windrow and to outline the various design
requirements for a large-scale composting facility.
Yes B C
SWMP To generate income through sale of compost. Yes I M
SWMP Uses of compost produced to the prison garden. Yes I M
SWMP Participation of the community in the vicinity of Petit Verger Prison
in the project will motivate the community to adopt a better attitude
towards detainees which will eventually help released detainees in their
resettlement and rehabilitation in the community as honest citizens.
25
No B NM to -
While material incentives may be more self-explanatory, what are examples of nonmate-
rial incentives that appeared when coding projects? Proyecto Carbono saw clear instances
where the indigenous worldview concerning the importance of cacao took precedence over
all other motivating factors. Proyecto Manglares found that recognizing land as nationally
protected areas brought in outside interest and opportunities to promote conservation ef-
forts, including academic studies, children's education programs, etc. that brought the local
community together on conservation eorts as well. In AC3, a simple certicate and stamp
by a national body certifying the quality of water was a strong incentive for better ASADA
management by local actors over any sort of material reward.
24
Project discussed in Chapter 6 on the comparison country case study of Mauritius.
25
Listed in project description though left out of list of project objectives in nal report.
146
A very interesting pattern emerges in this data: when the target actor for a particular
donor objective - that is, the actor who is meant to carry out the objective - is a local
implementer, material and nonmaterial incentives appear to result in long-term impacts at
more or less the same rate. However, when the target is a community-level beneciary, long-
term impacts are associated with nonmaterial incentives at a higher rate than any other type
of incentive. The data clearly has many nuances, and each project will be described in more
detail below.
A few limitations should be mentioned right away. First, in this data, it is hard to nd
evidence of examples of nonmaterial-intrinsic incentives. While the analysis here does not
dierentiate between extrinsic and intrinsic examples, it is an important note for future
research to consider how to better identify these types of incentives. One reason intrin-
sic motivations were not detected could be due to how project documents and/or people
in interviews speak about incentives, or the fact that this sort of incentive is pre-existing
within individuals and not necessarily something that can be motivated by others or one's
environment. This is an issue that will be explored at a future date.
Second, the fact that the implementers were mostly from academic institutions has
brought the issue of who each objective intends to target to the forefront, as material and
nonmaterial incentives appear to be functioning in distinct ways for local implementers ver-
sus local beneciaries. Local academic implementers may be motivated in dierent ways
than for example local NGO implementers.
Third, the results do not take into account how dicult the objective may be to achieve.
While it certainly seems intuitive that it is much easier to disseminate information or results
within a community than it is to scale a project across Central America, this dierentiation
still represents a profound measurement challenge. Diculty to implement would need to be
another column added to a future iteration of the project outcome evaluation tool presented
here, so that it is coded by multiple researchers given its subjectivity.
Fourth, the objectives presented here are clearly listed as donor objectives in this ta-
147
ble. I have attempted to elaborate on unlisted local actor/beneciary objectives as well as
unintended project consequences discovered in the process of data gathering within the dis-
cussions on each individual project below. However, future research needs to consider how
to combine the analysis of these objectives in a clear an rigorous way.
5.2.1 Proyecto Manglares
Archival Research
Coding for outcomes in 38 project documents revealed that the main achievements of the
project were training the local charcoal cooperative called CoopeMangle on how to more
sustainably harvest mangrove wood for charcoal production, and and how to derive multi-
ple benets from those harvests. For example, they helped start-up a working bakery for
the wives of the all-male mangrove harvesters that used leftover wood to power the ovens
needed to make bread products, generating additional diversied income for the community.
Additional benets were reported at the national level. Firstly, an oce was established
by the national government specically to manage and protect the Terrab a-Sierpe National
Wetland. Secondly, it planted the seed for CATIE to develop an eventual Master's degree
in Bioconservation and one in Development Practice.
Site Visit Observations and Interviews
Photos from the site visit can be found in Figure 5.4. Interview participants were initially
generated by contacting members of the government oce established to oversee the National
Wetland. Given the project ended in 1995, it was more dicult to nd relevant project actors,
sometimes due to outside forces such as the passing of some key actors. Nevertheless, eight
local actors were interviewed, half on the implementation side and half on the beneciary
side. The key piece of information that emerged across all these interviews was that the year
after the project ended, a new law was passed in Costa Rica preventing mangrove extraction
of any kind for commercial purposes. One local implementer led a petition for the members
148
of CoopeMangle to continue their livelihood for six to twelve months more, but after that,
all harvesting activities had to cease. The bakery survived approximately two more years,
but since it had originally relied upon leftover mangrove wood from the charcoal production,
it also closed down.
Despite what appears to have been very strong social relationships between local bene-
ciaries and implementers during the project, once the new law was passed and the petition
was led, there was essentially a total loss of communication between the two parties. One
interviewee on the implementer side expressed happiness that the law was passed because
it was better for conservation. All actors interviewed on the beneciary side expressed feel-
ings of total abandonment by implementers, with no sense of accountability. Beneciaries
expressed they suer to this day from the repercussions of the law and the abandonment by
implementers and particularly the national government post-project. They stated that most
of the charcoal workers were too old to be hired by other corporations, meaning they had
no way to tap into any retirement/pension funds, and had to continue to work in random
short jobs until they no longer were physically able or passed away. Other individuals had to
return to former careers usually in other parts of the country. Women had no other source
of secure income. The expensive charcoal ovens bought by project funds disintegrated away
over the years, and bakery equipment remains largely unused in the house of the former head
of Coopemangle. Beneciaries did have fond memories of one implementer, who they kept
in touch with every few years, however they had not been in contact for several years at that
point and the contact seems to have been only social in nature.
Incentives and Project Outcomes
Objectives where long-term impact was generated were those where local implementers were
targeted, particularly with material incentives - new management arrangements that pro-
vided oces, new jobs, existing job security, etc. However, implementers were also motivated
by nonmaterial incentives such as publishable research opportunities and developing a social
149
Figure 5.4: Photos from Proyecto Manglares eld site observation, reading by row from top left to
right: (1) and (2) Entrance to the Coopemangle neighborhood where most former members of the
cooperative still live. (3) Beginning of the Terraba-Sierpe National Wetlands and the mangroves
within the Coopemangle neighborhood. (4) The Terraba River. (5) Endangered
ora and fauna
call the National Wetlands home, such as this pair of red macaws. (6) One of the many palm oil
plantations bordering the majority of the National Wetlands borders, causing major biodiversity
concerns. (7) A red mangrove on the left in the Sierpe River within the National Wetlands, and
(8) a close-up on its seed. This was the species harvested by Coopemangle.
150
network. When beneciaries were targeted, material incentives did not lead to long-lasting
impacts, and given the law prohibiting all mangrove extraction passed right after the project
ended, previous material and nonmaterial incentives were soon converted into coercive in-
centives as all mangrove harvesting became illegal. While the market activities ceased to
exist and economic incentives were indeed lost, other potential benets like organizational
strengthening or changes in individual leader skillset still were not observed from the eld
site visit or mentioned in interviews. The bottom line: local implementers beneted in
the long-term from the project, while local beneciaries were harmed by the post-project
conversion of material to coercive incentives due to the new national law.
5.2.2 Proyecto Carbono
Archival Research
Coding for outcomes in 79 project documents revealed that the project was meant to provide
the scientic and business basis to develop an indigenous-run carbon capture product, but
project activities were clearly expanded beyond its original scope after the implementation
began based on negotiations by local beneciary CSOs to include their interests in the
project. For example, ACOMUITA fought for women to take a larger role in the project
and to be the main leader of the indigenous CSOs involved along with consultants to help
with establishing their own chocolate business, ADITIBRI requested consulting to improve
their cabinet-making company, and ADITICA requested soil studies. Based on the data
collected, no carbon product was ever sold by any of the indigenous CSOs, although one
Italian company does continue to pay for the conservation of a section of indigenous land.
However, all CSO requests do appear to have been fullled as an extensive disc was developed
post-project with all the consulting reports compiled in one place.
151
Site Visit Observations and Interviews
Photos from the site visit can be found in Figure 5.5. Site observations and interview
participants were generated by contacting members of the local organizations that originally
worked with CATIE: ACOMUITA, ADITICA, and ADITIBRI. Both the Bribri and Cab ecar
indigenous territories were visited for interviews, 20 interviews were performed with past
and present leaders of all three organizations, along with two more interviews o-site for a
totall of 22 informants. Two seperate trips were made to these territories to accommodate
the schedule of all informants. All agreed that there was no carbon capture project ever
sold under the CATIE project framework, though they do receive some carbon capture
income through a national scheme under the Costa Rican National Fund for Forest Financing
(FONAFIFO). Both indigenous communities and all three organizations did receive CATIE
consultants who worked on special projects requested by each respective CSO.
Dierent experiences emerged between the CSOs via interviews. Members of the Bribri
group maintained that it was important to keep up relationships with CATIE for access to
future projects and benets, and agreed that they are in better shape politically and orga-
nizationally than they were before, in particular thanks to one of the CATIE Coordinators.
ACOMUITA, a women's group that contains members of both indigenous groups, worked
with the same coordinator and expressed the same sentiment. Given they had to ght for
the inclusion of women and for consulting projects based on their current needs, they said
the project certainly fomented their negotiation skills to better ght for their interests in
future projects.
A representative of the Cab ecar, however, worked with a dierent CATIE Coordinator.
They expressed they felt CATIE took advantage of sharing their ancestral knowledge, using
it to receive academic accolades, equipment through the World Bank, and funded travel
to international conferences, while leaving very few benets for them. They expressed to
CATIE they would no longer work with them on future projects because they did not respect
their intellectual property rights, and left few material gains behind. Other Cab ecar repre-
152
Figure 5.5: Photos from Proyecto Carbono eld site observation, reading by row from top left to
right: (1) and (2) ACOMUITA facilities. (3) and (4) ADITICA facilities. (5) Banner hung
outside a school stating \We are a territory, not a reserve." (6) ADITIBRI facilities. (7) and (8)
ACOMUITA chocolate sold in the nearby popular tourist town of Puerto Viejo.
153
sentatives felt neutral about CATIE or had no idea who they were. All parties involved in
the original project, however, appeared to be content with the World Bank, as all reported
positive experiences with the World Bank liaison who frequented the indigenous territories
during the project period.
One common experience was that the CSOs were generally not successful in passing down
the information garnered through these additional studies to subsequent administrations, or
had trouble understand the technical jargon used in these documents. So although the
information was helpful in the short-term, its long-term impacts were not felt.
Incentives and Project Outcomes
We see that when beneciaries were targeted with material incentives, no long-lasting im-
pacts were found, but when they were targeted with nonmaterial incentives, long-lasting
impacts did exist. Implementers were targeted by nonmaterial incentives - namely pub-
lishable research - which did show evidence of long-lasting impacts. The bottom line: the
only long-term achievements of the project were generating new academic research for im-
plementers and improving local organizational capacity amongst beneciaries.
5.2.3 AC3
Archival Research
Coding for outcomes in 44 project documents revealed that the main achievements of the
project beneted implementers: generating new academic research and predictive climate
models. Results were not reported on the community-level, and results of the information
and water reduction technology experiments in 2015 still have not been published by the
researchers (hence the \?" in Table 5.2). Results were mentioned of trainings and skill
building for local ASADA leaders to improve services and cope with future droughts in
project documents, though I found little evidence for this through interviews and site visits.
154
Figure 5.6: Photos from AC3 eld site observation, reading by row from top left to right: (1)
Abandoned shrimp farm due to unsafe water. (2) Small scale cattle operations replace large dairy
operations that were shut down due to cattle illness. (3), (4), and (5) AC3-painted water
connections mostly remain though most are missing their number plaques in Community A. (6)
and (7) A water pump and an anti-litter reminder stating \Garbage in the end reaches streams,
rivers, and seas." in Community B.
155
Site Visit Observations and Interviews
Two of the original nine communities that received water reduction technology were chosen
for site observations and interviews.
26
Photos from the site visit can be found in Figure
5.6. In the rst community (Community A), I interviewed seven leaders and 11 households,
including members of the local ASADA. In the second community (Community B), I inter-
viewed two community leaders and 42 households. The technology was still being used in
some houses or was only recently removed due to calcication or other changes or repairs
needed in the houses, though Community A had more users in my neighborhood sample
than Community B. Several still gave the technology to another family member in the com-
munity to use if they changed water systems that could no longer accommodate it. Several
households expressed that they bought the technology on their own after they ended up in
the control group or if their original water saver broke.
However, in Community A, the ASADA appeared to be more organized and households
in general better remembered CATIE's time there and the lessons learned from them, and
more water savers were still installed. The ASADA members, past and present, very fondly
remembered CATIE's visit - in particular their eld coordinator - and stressed the importance
of maintaining that relationship, which in part was why they made time to meet with me
for interviews. In Community B, the ASADA was less organized with accusations of past
corruption during the time the project was implemented, and fewer households remembered
CATIE's visit or had the technology still implemented. There was confusion on the part of
the past ASADA board over the CATIE project, whether they came at all or what they did
while there. One interviewee mentioned CATIE did not follow protocol and there was some
resentment against them, implying their sense of entitlement for doing a project without
obtaining proper ASADA permissions. Several previous ASADA leaders avoided all contact
26
As mentioned in Footnote 22, six of the other original communities had another follow-up project implemented
by a similar team right after the project, and a seventh community refused participation in that project due to a
feeling of project saturation without any real benet for their community. I use letters A and B here to refer to the
two communities studied to protect their privacy, as their names are not in any existing published documents unlike
the other projects studied, and photos presented are only of the general region on the Nicoya Peninsula.
156
with me, or arranged meetings and never appeared.
Incentives and Project Outcomes
Nonmaterial incentives were the incentive type of choice for all objectives of the project
except the installation and potential payment for using the water savings technology. How-
ever, the implementers were the main target actor in this project, who were motivated by
nonmaterial incentives such as publishable research that has led to long-lasting impacts on
those objectives. The only place where beneciaries were targeted was with the water saving
technology, the strengthening of the ASADAs, and activities not mentioned in the ocial
objectives like setting up a household numbering system on each water plaque. Material
incentives were used to encourage longer use of the water reduction technology, though the
results of this choice have not yet been released by researchers so it is hard to know the
outcomes outside of the random sample of household surveys I performed. The household
numbering system no longer exists; one interviewee said that kids had stolen the numbers
from the pump connections, and it appears that there was no other incentive in place to
keep or replace numbers.
Meanwhile when nonmaterial incentives were successfully employed, they did lead to an
ASADA board in Community A that was more organized and better trained to serve their
communities in the long-term. They developed a better building system, moved from analog
to digital records, and set up a payment system with a local bank to simplify customer
payments. The bottom line: implementers saw long-term benets motivated by nonmate-
rial incentives, while beneciary benets are still largely unknown although there is some
evidence that nonmaterial incentives were associated with long-term impacts.
|||{
Results for the rst country case have been presented here along with initial impressions
and ndings. In the next chapter, I will introduce my second country case, Mauritius, and
the nal aid project I evaluate following the same format as this chapter. Analysis of project
157
results will additionally mention any dierences with the Costa Rican projects.
158
Chapter 6
The Case of Mauritius
In this chapter, I will discuss my selection of Mauritius as my second case, just as in the
previous Chapter 5. First I begin with a macro-analysis of my case selection based on the
three aspects I discussed in Chapter 4: economic and human development, institutional
stability, and development cooperation. I then include a comparison of the Costa Rican and
Mauritian cases at the macro-level to show where they coincide and dier on a number of
characteristics: their development narratives, global indicators, and the role of aid in their
national development trajectories. Next, I will discuss the selection of the aid project I
analyze within Mauritius. I discuss a brief summary of activities, the project objectives, and
the main actors. I end the chapter with a discussion of the project results I found based on
archival research, eld site visits, and interviews.
6.1 Case Selection: Mauritius
A small island nation o the eastern coast of Madagascar, Mauritius is now often touted
as the nancial gateway to Sub-Saharan Africa. Like Costa Rica, development has largely
been considered `successful' in Mauritius by the international community, thus, the country
is also considered a pathway case per the case selection techniques of Gerring (2007). Again,
this allows us to probe causal mechanisms including the micro-level data and interactions
that operate in this favorable macro environment. But rst I will justify my selection of this
country by looking at the same macro variables used to chose Costa Rica, and then detail
the aid project chosen for in-depth study.
159
6.1.1 Macro Context: The \Mauritian Miracle" of Development
Like Costa Rica, Mauritius has deed its naysayers concerning its development prospects.
James Meade, a Nobel-winning economist, authored a 1961 government report for Mauritius
as they prepared for independence from the U.K. He stated the country had weak poten-
tial for peaceful development, blaming overpopulation and resource competition between the
ethnically diverse Mauritian population (Meade, 2011). Although no indigenous populations
in Mauritius, French and British colonizers imported many slaves from the African continent
- known as creoles - to work in the sugar industry, and the British relied on indentured ser-
vants mostly from South Asian countries like India and Pakistan after slavery was outlawed,
although the French maintained an elite minority on the island. Meade's concerns were
understandable given the country's rapidly growing population in a very conned space, not
to mention the island nation is isolated from the African continent, depended on a mono-
culture export strategy (sugar), and does not have any natural resources to extract; in fact,
the rst visits recorded by Arab sailors in the Middle Ages and later the Portuguese in the
early 1500s did not mention even a single mammal living on the uninhabited island. Ten
years later and just a few years after independence, another Nobel-prize winner, Vidiadhar
Surajprasad Naipaul, predicted the same dark fate for Mauritius (Naipaul, 1973).
Yet Mauritius' economic and human development successes are now often referred to
as the \Mauritian Miracle," receiving attention over the years from a number of famous
economists and political economists such as Brautigam (1997), Sachs and Warner (1997),
Rodrik (1999a), Subramanian and Roy (2001), and Stiglitz (2011) who have all attempted to
explain the country's success. Since 1992, it has been classied as an upper middle income
country by the World Bank (World Bank, 2019e)
1
. Explanatory factors oered include taking
into account the initial conditions of the former colony, its openness strategy to exports
and FDI (particularly its preferential agreement with the EU for sugar exports), strong
1
They were actually further raised to the level of high income in 2020, see: https://blogs.worldbank.org/
opendata/new-world-bank-country-classications-income-level-2020-2021
160
participatory institutions, and its ethnic diversity that connected it to diasporas around the
world and separated the political and economic elite which avoided killing its lucrative sugar
industry (as summarized in Subramanian and Roy 2001).
Again, these macro-level factors certainly played a role in the Mauritian Miracle, however,
Mauritius has received more than USD 7.2 billion in aid from just bilateral and multilateral
donors since 1988 alone (AidData, 2019). I argue this amount of project aid certainly has
played a role in the development of the country. But rst I will turn to more specic
indicators for Mauritius concerning the three main characteristics I am looking for in my
case studies.
Economic and Human Development
Real GDP in Mauritius increased on average 5.3% between 1970-2009, while its regional
neighbors in Sub-Saharan Africa only experienced an average GDP growth of less than 2.5%
in the same period (Padayachy, 2010).
2
Mauritius ranks 49th globally in the World Eco-
nomic Forum's Global Competitiveness Report - the best performer in Sub-Saharan Africa.
This enormous growth and global business competitiveness is most surprising given that
various economists have argued that the country's initial conditions in fact put Mauritius at
a development disadvantage. Geographically it is 25% farther away from the world's major
economic centers than other African countries, and 30% farther away than most developing
countries (Subramanian and Roy, 2001). A major population boom was predicted in the
1960s, precipitating labor surpluses and high unemployment (Meade, 2011). Their popula-
tion density remains one of the highest in the world, at 623 people per sq.km. (Padayachy
2010; World Bank 2019b). The country depended on one agricultural crop - sugar - intro-
duced by French colonizers, with no other natural resources available to extract from the
two main islands and several outlying islands that make up the nation.
Like Costa Rica, Mauritius also presented a winding path through a monocrop agri-
2
The source (Padayachy, 2010) was provided by the Mauritian Economic Development Board for this research.
161
cultural sector, failed economic policies, nancial crises, and an eventual turn to exports
and FDI. The development history as told by the Mauritian Economic Development Board
economist (Seechurn, 2019), sugar accounted for 85-90% of exports in the 1960s, precipitat-
ing danger given the dependence of the economy on just one crop (Seechurn, 2019). The
sugar industry had reached its labor limit, but the population was growing quickly, and
unemployment was around 15% at the end of the 1950s (Padayachy, 2010). Diversication
was a necessity to avoid a national economic and social collapse.
In the late 1960s, the country turned to an ISI strategy with a focus on manufacturing
(Padayachy, 2010). However, ISI ultimately failed - unemployment levels did not decrease,
and the country continued to teeter on the brink of economic crisis. Ultimately they had
no capital to develop their own industries, there was a lack of qualied laborers, and local
entrepreneurs were risk averse despite government-provided incentives to encourage local
manufacturing expansion.
Government leaders looked to Asian counterparts, small nations like Hong Kong, Puerto
Rico, and Taiwan, and saw the range of incentives they oered to encourage FDI, and
switched to an export-orientation instead, creating export processing zones like Costa Rica's
zonas francas in 1970. Because of the lack of local capital, foreign companies were aggres-
sively courted through incentives to come to Mauritius and hire locals in the manufacturing
sector. Like Costa Rica, they had a highly educated and multilingual population that was
attractive for foreign companies, along with their status as part of the British Common-
wealth post-colonialism gave them access to European markets that many Asian companies
did not have easy access to (Padayachy, 2010).
In the beginning of the 1970s, this strategy appeared fruitful, along with investments
in tourism by the private sector. GDP increased threefold and per capita income grew by
112% (Padayachy, 2010). They received further benets under trade preferences provided
by the Multi Fibre Arrangement and Cotonou Agreement. At the same time, Mauritius
were given preferential access to EU sugar markets through the Sugar Protocol, agreeing to
162
a guaranteed market price despite world market prices - though the price was on average 90
percent above the market price. This proved to be a key decision, as sugar prices eventually
dropped but Mauritian industries were protected while they built their export strategy.
Although the money from sugar was reinvested in the country, they were hit by an
economic recession in the late 1970s/early 1980s. The sugar boom from the rst half of
the decade drove up salaries not only within the industry but across industries, leading
to high rates of in
ation (Padayachy, 2010). The two global oil shocks hit and the price
of sugar dropped for those outside the EU agreement, and the country was hit with two
cyclones,
oods, and a drought each year from 1975-1979. Mauritius was again on the brink
of economic collapse. They were forced to turn to the IMF and World Bank, who required
structural adjustment policies and a devaluation of their currency in both 1979 and 1981.
While the rate of in
ation was cut dramatically by 1983, the country's unemployment rate
sat at an all-time high of 17%.
In the mid-1980s, oil and commodity prices began to level out again and Mauritius began
to recover. They focused on the export of textiles and further diversifying their economy
by building up the tourism sector. They created a national airlines - Air Mauritius - and
encouraging hotels (Seechurn, 2019). In 1992, the Mauritius Oshore Business Activities
Act
3
was passed to encourage and regulate all oshore business activities in an eort to
diversify the economy and to compete with other global oshore nancial centers. Companies
register in Mauritius, while channeling their investments to other countries in particular to
India (Padayachy, 2010). Throughout the 1990s, GDP rose on average almost 6% a year
and in
ation rates fell.
Preferences provided through the Multi Fibre Agreement and the Sugar Protocol were
dismantled due to new international trade rules so Mauritius continued a sectoral diversica-
tion strategy in an eort to prevent any future crises and recover jobs. In the late 1990s, the
country turned to information technology (ICT), given their high levels of multilingualism
3
Text of the MOBA Act can be found at: http://www.assetprotection.cz/knihovna/mauritius/mauritius-
oshore business activities act.pdf
163
(English, French, Creole, and a number of Indian languages), and their time zone that fa-
cilitated call centers globally. They used a government-owned private company to facilitate
and construct what would become Cybercity in the southern suburb of Ebene of the capital
city in the 2000s. Cybercity provides a home to a number of companies in outsourcing ser-
vices, software development, and call centers (Padayachy, 2010). While preferences provided
through the Multi Fibre Agreement and the Sugar Protocol were dismantled due to new
international trade rules. In the mid-2000s, the country implemented a number of structural
reforms to harmonize all the new sectors, instituting a uniform income tax of 15%, making
permits easier to obtain, and simplifying tax collection. The reforms proved benecial, as
the government tax base grew to better provide services and companies expanded. Real
estate schemes also improved around this time, bringing an increase in FDI to the islands.
Int he 2010s, the Mauritian government has looked to their oceans given the country's small
land area. Aquaculture, marinas, and deep ocean exploration have all been part of their
more recent economic expansion plans, though their capacity remains low in this sector to
carry out activities and thus plans are moving more slowly. Their seafood exports are strong,
sending a large volume of tuna to the EU (Padayachy, 2010).
Mauritius now ranks 66th in the Human Development Index - the highest in the African
region and even two spots higher than Costa Rica (UNDP, 2019). Mauritius is also known
for its abundant social protections. The country has trade unions that are large, active,
and engaged in centralized wage bargaining; prices are controlled on a number of items in-
cluding socially important ones; and generous social security particularly for civil servants -
all without needing to increase taxes due to the country's economic growth and age demo-
graphics (Subramanian and Roy, 2001). Local economists cite education as one of the main
motivating factors for increases in productivity (Padayachy, 2010). Mauritius has converted
Meade's feared labor scarcity into a labor surplus, focusing on job creation in the 1980s and
1990s with accompanying investments in education and training through the present to ll
those jobs, along with government-provided incentives for foreigners to come study, work
164
and stay in the country (Subramanian 2013; Seechurn 2019).
Mauritius still faces important development challenges (Seechurn, 2019). It is now suf-
fering from a labor shortage particularly in manufacturing and ICT, and universities cannot
produce enough graduates. Unemployment sits around 7% - approximately 40,000 individ-
uals - but these residents are not currently qualied to do the work or adaptable to such
jobs and women and youth are overrepresented in this number World Bank (2019c). While
government incentives have been put in place to attract more workers, it is hard to convince
many working age people to come and live in the more remote island nation, and it will
bring its own set of problems given the high population density. A government scheme to
attract those from the diaspora abroad - particularly if well-educated - back to the island
has only attracted about 250 Mauritians back to the islands. It is also hard to convince more
outside FDI due to this geographic challenge, especially given the islands' vulnerability to
climate change. Their market (population) size is just under 1.3 million, which is not a large
enough captive market for many companies and their distance from the continent and rest
of the world means that the cost of shipping exports remains one of their weaknesses in
attracting FDI. The end of their preferential trade agreement for sugar with the EU meant
a drastic decrease in the country's main export and the closing of many companies, and for
the producers that remain, prices are now subject to the
uctuations of the world market.
Sugar dominance also led to extreme deforestation across the island. The country has not
been untouched by corruption; in 2018, the rst elected female leader of the nation Ameenah
Gurib-Fakim resigned after allegations of spending USD 26,000 on personal purchases using
a NGO-issued credit card, as did the minister of gender equality and the deputy assembly
speaker after apparent ties to drug trackers (Freedom House, 2019b).
New development aid has been pouring in from the EU to promote sustainable livelihoods
in new industries. The World Bank has invested in knowledge production, looking at the
drivers of rising inequality there especially of women, and labor mobility. Aid via programs
like the GEF's Small Grants Program is also coming in to deal with the uncertain eects
165
of climate change and livelihood adaptation in a post-sugar economy, especially for coastal
residents and those on the smaller inhabited island of Rodrigues. As an island nation with
high population density, they are especially vulnerable to the adverse eects of climate change
including natural disasters, and will need to adjust production and consumption practices
in the coming years (Padayachy 2010; World Bank 2019c).
Institutional Stability
Despite its distance from mainland Africa and other challenges previously mentioned, Mau-
ritius has been promoted in the international community both by its own government and
the international community as the gateway to the African continent because of its stable
institutions. It outperforms most of its Sub-Saharan regional neighbors on governance in-
dicators, and its strong education institutions and low military spending set it apart. It
has built its previous growth on institutions such as property right protections and highly
developed infrastructure (Von Uexkull et al., 2019). This stability and predictability has
been key in elevating the certainty of foreign investors in their investment.
Like Mauritius' economic development prospects at the time of independence, its prospects
for political stability were also slim and for the same reasons: economic stagnation and ethnic
tensions (Marshall and Jaggers, 2011b). But also like its development prospects, it was able
to draw on the benets of a diverse population and nd common interests to instead because
known over the past few decades as one of the most democratic and stable countries in the
entire African region (Marshall and Jaggers, 2011b). The Ibrahim Index of African Gover-
nance (IIAG) rates Mauritius as number one in overall governance among all 54 countries
rates (Mo Ibrahim Foundation, 2018).
4
The country moved from a 9 out of 10 up to a full
10 - Full Democracy - in 1982 (Marshall and Jaggers, 2011b). A parliamentary democracy,
4
The IIAG is a widely used index providing a quantitative evaluation of governance across African countries since
2007 in the following categories: Safety and Rule of Law, Participation and Human Rights, Sustainable Economic
Opportunity, and Human Development. It is published annually and used by a variety of government, NGO, and
academic researchers for monitoring purposes. It was originally developed in partnership with Harvard University,
though is now mainly an African initiative with a public-private partnership advisory council. For more information,
see: https://mo.ibrahim.foundation/iiag
166
they have held competitive multiparty elections for their prime minister since their indepen-
dence in 1968 with the exception of 1972. They have a high freedom rating according to
Freedom House at 89 out of 100 (free); a freedom rating at 1.5/7, political rights at 1/7,
and civil liberties at 2/7 (all high); and strong civil liberties such as freedom of expression,
associational and organizational rights, rule of law, and individual rights (Freedom House,
2019b).
Like Costa Rica, the Mauritian state has taken a strong hand in the economic develop-
ment trajectory of the country, consistently passing business-friendly policies year after year
to open their economy to the world no matter what party is in place. Given their attention
to a new sector almost every decade, numerous new branches and regulatroy commissions
have been created to oversee each of these new endeavors, leading to a number of overlap-
ping organizations that the government is working to streamline each year. For example, the
The Mauritius Industrial Development Authority was previously known as Mauritius Export
Development and Investment Authority - created at the same time as Costa Rica's CINDE
- until the government created the Board of Investment and all investment-related duties
were taken away in 2001.
5
Then the Mauritian Economic Development Board was creating
from the 2018 merging three bodies: the Board of Investment, Enterprise Mauritius, and
the Financial Services Promotion Agency.
6
Despite constant re-organizations, Mauritius still
has numerous domestic government bodies and government-created independent regulatory
bodies dedicated to development, selected bodies are listed in Table 6.1 just to name a few
of the many relevant bodies.
7
On an international level, Mauritius remains a member of the Commonwealth of Nations
for ex-territories of the British Nation, but is also a part of many regional multilateral orga-
nizations such as the Indian Ocean Commission (IOC), the Southern African Development
Community (SADC), Common Market for Eastern and Southern Africa (COMESA), they
5
See: http://prb.pmo.govmu.org/English/Pages/prb%202003/Parastatal%20and%20Other%20Statutory%
20Bodies/Mauritius-Industrial-Development-Authority.aspx
6
See: http://edbmauritius.org/about-us/
7
A full list can be found at: http://www.govmu.org/English/GovernmentBodies/Pages/default.aspx
167
Table 6.1: Examples of Development Bodies in the Mauritian Government
Economic Development Human Development
Ministry of Finance, Economic Planning and
Development
Ministry of Social Security, National Solidarity, and
Environment and Sustainable Development
Ministry of Foreign Aairs, Regional Integration and
International Trade
Ministry of Justice, Human Rights and Institutional
Reforms
Ministry of Industrial Development, SMEs and
Cooperatives
Ministry of Social Integration and Economic
Empowerment
Ministry of National Infrastructure and Community
Development
Ministry of Gender Equality and Family Welfare
Mauritius Board of Investment Ministry of Agro Industry and Food Security
Economic Development Board Maurice
^
Ile Durable Fund
Development Bank of Mauritius Equal Opportunity Commission
Small and Medium Enterprises Development Authority National Human Rights Commission
Financial Services Commission Law Reform Commission
Competition Commission of Mauritius National Women Entrepreneur Council
are part of an Economic Partnership Agreement between the EU and Eastern and Southern
African states, and signed the newly created African Continental Free Trade Agreement,
in addition to bilateral trade treaties with partners like China and India. They are part
of all major IGOs including the WTO and are continuously looking for ways to grow their
presence in the international community.
Development Cooperation
In 1961, Mauritius was 64th out of 119 nations reporting in terms of net ocial development
assistance and ocial aid received
8
(54th percentile) (World Bank, 2019d). However by its
lowest point in 2017, Mauritius' ODA in
ows had dropped to a ranking of 131st out of 141
nations reporting (93rd percentile). Mauritian aid in
ows peaked in 1990 at USD 118.3
million and ODA at 3.4% of GNI, but rapidly dropped o in the 1990s as its development
successes became apparent. At its lowest point in 2017, ODA accounted for only 0.08% of
its GNI (World Bank, 2019d).
Also like Costa Rica, aid returned in the late 2000s with greater
ows from 2008-2013
to assist with education reform, banking oversight, climate change impacts and disaster
management, and technical assistance on understanding its current development challenges.
8
In constant 2015 USD.
168
However, Mauritius is no longer receiving the aid it once was; for example, there are no active
World Bank lending operations in the country at the moment. One interview respondent in
the present study who has been working in Mauritian NGOs for a decade said that in 2019
it is, \hard to get money from international organizations because people think everyone in
Mauritius is rich....[that] everyone has a boat, drives a BMW...[that] Mauritius is an African
country but not `Africa'." They estimated that as many as 99% of Mauritian NGOs presently
get their funding from local sources.
Of the aid to Mauritius reported to AidData.org including entries back to only 1988,
1.9% of entries are for general budget support to the country, accounting for approximately
14% of donor committed funds. Budget support to Mauritius has been exclusively given
by bilateral and multilateral donors, whereas projects are supported by both government
and private donors, consistent with global aid practices and with Costa Rica. The data is,
again, far from perfect, but given the information available, the majority of funds entering
Mauritius since at least 1988 have been for projects.
9
Unlike Costa Rica, Mauritius is not known as a global good Samaritan or an innovator
on development measurements. Participatory approaches to projects within the country
have been dependent on general donor policies to giving aid, rather than recipient driven.
They are not known for their sustainability approaches in any particular sector. Mauritius'
popularity in the international community comes from their saavy approach to capitalist
market forces and their reputation for peace. In fact in conversations with scholars and
practitioners on the island, it seems clear the development community is small, and actors
all generally know one another. This practitioner network appears to be the driving force in
9
As mentioned in regards to Costa Rica in the previous chapter, AidData.org is considered as the most expansive
aid database to date, incorporating data from a number of sources including the OECD, which previously was
considered the most expansive aid dataset. Nevertheless, data on aid from around the world is notoriously incomplete,
most likely from a combination of a lack of human and technological resources to keep complete data over time, to
political factors that incentivize hiding information. In the data analyzed to come up with these percentages (most
recently downloaded on December 20, 2019), it is important to note that a majority 82% of entries for Mauritius
have no date attached to the projects. Additionally, 16% of the projects have no budget commitment amount
listed, which includes 26% of the listing for budget support, meaning my conclusion that most aid to Mauritius
comes in the form of projects may be inaccurate. Nevertheless, the data used here represents the most complete
information available at the time this was written. To search aid given to Mauritius through AidData.org, see:
http://aiddata.org/dashboard#/advanced/analytic-dashboard.
169
obtaining projects for the island that are externally funded, given the small amount of ODA
currently
owing into the country.
6.1.2 Comparing Costa Rica and Mauritius
Now that I have introduced the two macro-levels contexts I have used for this research, Costa
Rica and Mauritius, it is worth providing a comparison summary of the two contexts before
moving on to the micro-level project I selected in Mauritius. In this section, I have included
a brief look at how the historical development narrative compares across both countries, in
addition to similarities and dierences in their rankings across dierent global indicators of
development. Despite two very dierent cultural contexts, these two countries have a variety
in common and play similar roles as representatives of their distinct geopolitical regions.
Development Narrative
Fox (2003) points out that both countries overall have worked to implement stable macroeco-
nomic policies such as monitoring exchange rates, and make constant adjustments to compete
for FDI globally by reducing obstacles, increasing eciency, appropriate pricing, and oer-
ing special services and market access. However there are some notable dierences: Costa
Rica is a fairly homogeneous population with little cultural or religious diversity, whereas
Mauritius is known for its ethnic and religious diversity which is particularly notable in their
multiparty political system that demands coalitions across ethnic-based parties. While each
path was unarguably distinct, the development narratives presented for both countries show
a number of initial conditions and strategies in common:
Former colonies who both had small to no indigenous populations and were peripheral
to their colonizers
Small countries whose geographies favored peace in their regions:
Mauritius: no army after Independence
Costa Rica: abolished army in 1948
170
Instituted and maintained strong democratic institutions generally respected by the
population since independence.
Prioritize social welfare policies such as education, fostering multilingualism, and health
care for their populace since independence.
Huge drops in fertility:
Mauritius: 1950-70s fell from 6 to 3 children/woman
Costa Rica: 1950-70s fell from 7 to 4 children/woman
Agro-based industries reliant on one or two exports:
Mauritius: until 1970 sugar 95% of exports
Costa Rica: until 1962 bananas + coee 90% of exports
Pursuit and failure of ISI strategy in manufacturing:
Mauritius: 1964-1970
Costa Rica: 1962-1979
Pursuit of export-oriented strategy and FDI via export processing zones:
Mauritius: 1970 to present
Costa Rica: 1981 to present
Financial crises and turn to IMF and/or World Bank:
Mauritius: 1979-1986
Costa Rica: 1980-1990
Large in
ows of foreign aid:
Mauritius: peak in 1990, rapid drop o over next decade
Costa Rica: peak in 1985, rapid drop o over next decade
Focus on similar industries for diversication:
Mauritius: Tourism early 1980s, ICT late 1990s
Costa Rica: Tourism mid 1980s; ICT mid 1990s
In an interview with Seechurn (2019), discussions have occurred in late 2019 in Mauritian
government circles about a possible turn to the Costa Rican model, where they focus on just
a few sectors like manufacturing and ICT rather than the more scattered sectoral approach
that has occurred over the past few decades. This point is backed up by the most recent
World Bank in Mauritius report, which talks about the prospects of Mauritius nding an
171
`eciency-seeking' investor as Costa Rica did with Intel - one that is not there to exploit the
host country or exclusively send benets back to the home country, but simply contribute
to the global market (Von Uexkull et al., 2019). This sort of FDI brings many jobs and
international notoriety that encourages further FDI. They state this can be achieved by
focused governments that know how to provide an enabling environment for rms to compete
globally.
In a technical research report on factors supporting successful integration of countries
into the world's economy, submitted to USAID, Fox (2003) supports what I have presented
here thus far:
\Both countries have received signicant amounts of foreign aid, each using its demo-
cratic tradition and other tools to encourage donor largesse. Costa Rica received mas-
sive assistance from the United States during the 1980s. Mauritius had more diverse
sources. The county marketed itself for both aid and market access as a former colony
of France as well as England, and as an African and Indian Ocean country. It is the
largest recipient of foreign aid from India" (20).
Thus, there is additionally evidence that aid at the micro-level has played a role in the
achievements of both Costa Rica and Mauritius that macro-level explanations alone cannot
account for, reinforcing that the current exploration of project-level dynamics in these two
country cases is a worthwhile endeavor.
Development Indicators
The World Bank's Worldwide Governance Indicators rank countries on six dimensions of
governance: voice and accountability; political stability and absence of violence; government
eectiveness; regulatory quality; rule of law; and control of corruption. While not without
their critics in terms of methodology, they are considered generally reliable in the study of
politics, and are one indicator that can be considered when measuring development in terms
of institutions. Scores for the Costa Rica and Mauritius have been summarized in Figure
6.1. The scores for both countries almost entirely fall between the 60th and 80th percentiles,
often interchanging high and low scores between indicators. However, notice that all the
172
condence intervals overlap, meaning the dierence in scores between these two countries is
not statistically signicant.
Figure 6.1: Worldwide Governance Indicator Scores for Costa Rica and Mauritius (2018)
Source: World Bank (2019f)
Additional selected indicators at the country-level are presented here in Table 6.2. This
table is meant to show that the two countries tend to be rated similarly across many types
of measurements - either their scores are similar, or they rank rst in their respective world
regions. Given the problems with macro-level measurement described in Chapter 2, I at-
tempted to include a wide variety of quantitative and qualitative indicators to increase the
amount of information available to the reader on each country.
6.1.3 Micro Context: Project Selection
The purpose of including a second country case is to have an idea about whether or not
the results learned through studying the three Costa Rican projects are applicable in other
10
All data from 2019 with the exception of Climate which is from 2018.
11
By percentile rank, all data from 2018.
12
All data from 2019 with the exception of Global Competitiveness Rank and the Corruption Perceptions Index
Rank which are both from 2018.
173
Table 6.2: Summary of Selected Development Indicators for Costa Rica and Mauritius
Sources: World Bank (2019b); World Bank (2019f); Freedom House (2019a); Sundberg and
Melander (2013); Beck et al. (2018); Transparency International (2018); UNDP (2019); World
Economic Forum (2019a); World Economic Forum (2019b)
Indicator Costa Rica Mauritius
Basic Statistics
10
Surface area (km
2
) 51,100 2,040
Population (millions) 4,999,441 1,265,303
Populations Density (pax/km
2
) 97.9 623.3
Government Presidential republic Parliamentary republic
Aggregate Freedom Score 91/100 (Free) 89/100 (Free)
Income Class Upper middle Upper middle
Open Economy Yes Yes
Civil War/Con
ict Stable since 1948 civil war Stable since 1968 independence
Colonized Spanish until 1821 Dutch/French/British until 1968
Climate Tropical: rainforest/monsoon Tropical:rainforest/monsoon
Worldwide Governance Indicators
11
Control of Corruption 70.7% 63.0%
Government Eectiveness 67.8% 77.4%
Political Stability/Absence of Violence 62.5% 79.0%
Regulatory Quality 69.2% 81.7%
Rule of Law 69.2% 76.9%
Voice and Accountability 84.7% 72.4%
Various Development Measures
12
HDI Rank 68 66
Life Expectancy at Birth (Years) 80.1 74.9
Mean Years of Schooling 8.7 9.4
GNI per Capita (PPP$) 14,790 22,724
Global Competitiveness Rank 55/140 49/140
Travel & Tourism Competitiveness Rank 41/140 54/140
Corruption Perceptions Index Rank 48/180 56/180
country contexts. Thus, one more project was selected in Mauritius for comparison sake. I
specically sought to identify a project 1) in one of the same sectors studied in Costa Rica,
and 2) where another higher education institute was involved as a principal implementer,
both so it could be better compared with the projects studied in Costa Rica that were
implemented by CATIE, but also in the case that documents evaluating the project produced
by impartial third parties - e.g. professors or students - might again be available. I will now
outline the details of the project selected (the main variables of all projects studied have
been summarized in Table 5.1).
174
SWMP - Sustainable Waste Management Practices Project
Project Summary The Sustainable Waste Management Practices Project at Petit Verger
Prison (SWMP) was implemented from 2014-2016 and funded by the Small Grants Program
(SGP). The project came about based on a need for better waste management practices
at the all-male prison due to the sheer volume of waste being produced and the challenges
this geographically small island nation is facing with the increasing repercussions of global
climate change. As 90% of the waste produced there was determined to be organic in nature,
the idea was to teach current inmates at this medium-security prison composting techniques
they would use to not only better manage the waste produced at the prison but also in the
surrounding community of Pointe aux Sables (Vieillesse, 2016). The benets expected were
to improve waste management, to give prisoners new skills they could turn into livelihood
strategies after prison, and to improve the relationship between the community and the
prison/prisoners. Several of the promotional materials for the project promoted the slogan,
\From Waste to Wealth" (Mauritius Prison Service 2018; GEF Small Grants Programme
2015). Additionally, the project included the retailing of compost produced, which went
back into government funds meant for prison and community projects (Vieillesse, 2016).
The project received the maximum SGP award of USD 50,000, with addition in-kind co-
nancing of USD 107,733 (GEF Small Grants Programme, 2015).
Prison ocers and detainees alike were training by technical teams from the University of
Mauritius and composting infrastructure and equipment such as drums and a wood chipper
were installed at the prison (Vieillesse 2016; Le Mauricien 2016). Work was done in the sur-
rounding community to sensitize them to the importance of sustainable waste management
in the Mauritian context and introduce how they could gather their own waste in order to
participate in the project. The project additional spun o the construction of a composting
facility at a nearby women's prison, with hopes of scaling it to all Mauritian prisons on both
of the country's two islands in the future and to form a community composting network. The
nal report stated that the composting initiation was successful, participation was high, and
175
that those trained in composting skills were satised with these trainings (Vieillesse, 2016).
The project won the national Public Service Excellence Award 2013 under the category of
"Maurice Ile Durable" (Le Mauricien, 2016). Known as MID, this governmental strategic
vision was announced in 2008 and the SWMP project clears falls under its aim: \to make
Mauritius a world model of sustainable development, particularly in the context of SIDS
(Small Island Developing States)" (Maurice Ile Durable, para. 3).
Again, the main variables of all projects studied have been summarized in Table 5.1.
Objectives The project was pitched through the SGP as a way to rehabilitate inmates and
allow them to make reparations for their oenses, all in an eort to decrease recidivism (GEF
Small Grants Programme, nd). It was seen as a win-win eort to reach this goal while also
addressing climate change issues in Mauritius, educating the surrounding community, and
increasing sustainable practices within the national prison system. The project objectives
as listed by the implementers are (Vieillesse, 2016, 3-4):
\To develop a sensitization campaign on wastes [sic] segregation at the prison and at
the community level for green wastes.
To carry out a waste characterization study at the prison so as to identify potential
waste treatment options in view of decreasing the amount of wastes going to the landll.
To create meaningful activities for detainees in view of their rehabilitation and reset-
tlement.
To train detainees and ocers on the making of compost on a large-scale basis, for
example, in windrow
13
and to outline the various design requirements for a large-scale
composting facility.
To generate income through sale of compost.
Uses of compost produced to the prison garden."
It is important to note that the objectives did not make clear where the income from sales
would go, or how inmate outcomes would be measured in terms of future job opportunities
or rates of recidivism.
13
Large-scale composting technique where organic waste is piled in long rows and consistently turned to introduce
air into the mixture.
176
Actors The GEF and specically their SGP have a very large footprint in the develop-
ment community, but are nevertheless a lesser known multilateral development donor in the
international community. The SGP is a global grant program created by the Global En-
vironment Facility (GEF), a multilateral agency, meant to be used for community-driven
projects. GEF was initiated from the 1992 Rio Earth Summit proceedings to deal with
urgent global environmental issues by coordinating donor funds. GEF states they currently
partner with 183 countries, have initiated close to 5000 projects in 170 countries since its
inception, and given more than USD 17 billion in grants and while catalyzing USD 88 billion
in project nancing.
14
Specically under the auspices of the SGP, they have worked in 128
countries on about 24,000 projects. Its reach is vast and powerful, and its network includes
organizations in the public, private, and civil society sectors. The GEF touts that social
and political empowerment of local people is central to their mission, as environmental and
economic sustainability will not be possible without strong and fair governance at all levels.
Thus they created the SGP globally in order to cater to the needs of NGOs and commu-
nity organizations (CSOs); grants through the SGP are up to USD 50,000 and are typically
awarded through a request for proposal (RFP) process.
15
Because of the large reach of the SGP and the organizational structure of the GEF,
implementing the projects across 170 countries quickly became inecient. Originally the
money was given to three organizations: the UNDP, UNEP, and the World Bank. However,
it was the UNDP that ended up becoming the main implementer of this program across
the world. Unlike the UNEP, the UNDP has staed oces in every country, and the World
Bank usually only works with much larger grants, making it outside their organizational
scope (Imbach, 2017). Thus the SGP is managed by the UNDP oce within each country.
Mauritius began participating in the SGP program in 1995, just three years after the in-
ception of GEF.
16
To date, the SGP has supported 168 projects within the country, investing
14
See the main GEF website: https://www.thegef.org/about-us
15
See: https://sgp.undp.org/about-us-157.html
16
For this information on the SGP in Mauritius, see their Country Page on the SGP website: https://sgp.undp.
org/component/countrypages/?view=countrypage&country=75&Itemid=204
177
in grants of over USD 5 million, and even more in co-nancing and in-kind co-nancing. SGP
projects have seen quite a bit of recognition at the national and international levels, win-
ning prestigious awards including the Mauritian Public Service Excellence Award, Whitley
Foundation Award, Energy Globe Award, Green Africa Award, the Golden Jubilee Marine
Environment Award, the Arab Gulf Programme for Development (AGFUND) Prize, and the
Island Bright Spot Award (Somaroo and Koodoruth, 2016). Mauritian project themes fall
into the following arenas:
Biodiversity
Chemicals
Climate change mitigation
Community based adaptation
International waters
Land degradation
Figure 6.2 represents a network map of the involved actors. The SWMP had two local
implementers: the University of Mauritius (UoM) and the Association Kinou et e, a Mauritian
NGO. The UoM was tasked with leading the technical implementation of the composting
project, given their vast experience implementing projects concerned with waste manage-
ment around Mauritius and on the African continent. Kinou et e was meant to manage the
participation of the beneciaries: the prisoners and the community surrounding the prison.
The mission of Kinou et e is focused on current and former prisoners: \To accompany, help
and guide their beneciaries through their rehabilitation and reintegration process in a pro-
fessional manner" (ACTogether.mu).
The governmental agency Mauritius Prison Service (MPS) was a key partner to the two
implementers as the organization who runs the Petit Verger prison. Their sta participated in
trainings to implement the new program with the prisoners, and allowed for the construction
of waste management facilities on-site. Given their own agency initiatives on food security,
the program t in nicely with their own goals and they also planted additional trees on-site
178
Figure 6.2: Network Map of Actors Involved in SWMP
using the compost generated (GEF, 2016). Additionally, the MPS has employed a 10-year
plan (2013-2013) to reduce the amount of recidivism in Mauritius, and this project aims
to contribute to this plan by providing previous oenders with new job skills to keep them
employed and away from illegal activities (GEF Small Grants Programme, nd).
The two implementers partnered with several other organizations during and after project
implementation. The Mouvement Civique de Pointe aux Sable (MCPAS) is a civil society
organization formed by community members in the area immediately surround the prison,
that helped organize informational sessions and pamphlet distribution to neighbors in order
to coordinate waste collection from the community members (GEF, 2016). ATICS Ltd.,
a private waste management company, then picked up and transported that waste to the
composting facilities built at the prison.
179
6.2 Results
Project documents have been coded for the SWMP project implemented in Mauritius, along
with observational data and interviews. The Mauritian data also shows support for my
theory: nonmaterial incentives for local actors is related to longer-lasting project outcomes,
however coercive incentives do play a larger role in this case, as will be described. Again,
Table 5.2 represents a summary of long-term impacts and incentives present for each objective
across all four projects. The patterns observed in the Costa Rican data were for the most
part repeated in this Mauritian project, with the caveat that coercive incentives were better
explored in this case, as described below.
6.2.1 SWMP
Archival Research
Coding for outcomes in 19 project documents revealed that the main achievements of the
project beneted implementers: the Petit Verger Prison used detainees to carry out cost-
savings and cost-generation measures that beneted the prison itself. Given the SWMP
used prison labor to carry out its project activities, it is very dicult to distinguish between
coercive and other incentives present in this project, given that all project involvement of
detainees was arguably coercive in nature.
An additional point of importance is that the way the objectives for the project are writ-
ten makes them very easy to achieve and shows little impact: they are essentially measured
by counting indicators. Creating an awareness campaign, doing a research study, and per-
forming trainings are simple activities that can be tallied - as opposed to making the desired
impact of these activities the actual project objective. This means that making inferences
about long-term impacts for these objectives may not be possible.
180
Site Visit Observations and Interviews
I faced two important challenges when collecting data on this project because it was located
at a prison: my original IRB approval did not allow for interviews with detainees, and I
would additionally face a long process of seeking permission from the Mauritian authorities
to enter the prison that went beyond my visa allowance. Thus, my focus for interviews was
with any implementing parties, especially those who work(ed) directly with detainees, and
community households who were also targeted by the composting campaigns and collection
services. I was able to speak to three individuals on the implementing side, representing three
dierent agencies, including one that worked directly with detainees and released detainees.
I additionally knocked on hundreds of doors in the neighborhood that were targeted by the
project (as indicated on a map by implementer interviewees) and spoke to individuals in 34
households.
The project has been viewed as a great success from the implementer point-of-view. It
won awards, SGP recognitions, and although it appears that it did not expand to other
prisons in the end, Petit Verger did become a hub for composting for all prisons on the main
island of Mauritius. This has greatly reduced costs for the prison system by generating some
additional income through compost sales to the public and other government agencies which
has been reinvested in the prison itself, and by using the compost to fertilize the prison's
garden which provides meals for detainees. This has also reduced waste, the need for waste
transportation, and use of landlls, which ultimately benets the entire island nation and
taxpayers.
On the community side, only three out of 34 households had any recollection whatsoever
of the composting project, and did not associate it with any sort of a campaign to improve
the community attitude towards the prison. Although interviewees indicated a neighborhood
of about 12 streets where the project took place, all three of these houses were located on
the only street where composting bins were still visible. Nevertheless, interviewees had a
hard time explaining when the compost bins were installed, how to use them, or if they were
181
still in use. They did not know who picked up trash from those bins and ATICS Ltd. no
longer services that area. An unmarked garbage truck passed by during eld interviews and
a household subject mentioned they believe it was owned by the prison itself, however it did
not stop to pick up any trash.
It is not actually clear if the bins even belonged to this project. A sticker on the outside
says \Sponsored by Petit Verger Prison & Cernol Ltd | Design by: Alladeen." Cernol
and Alladeen appear nowhere in project documentation; Cernol appears to manufacture
detergents, hygiene products, and industrial chemicals, and Alladeen only appears as a
common last name in Internet searches. The street where the remaning bins are located is
the main street to the prison, and composting infrastructure observed was as follows: two
(2) single bins, twelve (12) double bins, one (1) single bracket with a missing bin, and ve
(5) double brackets with missing bins. Most were missing the lids, or were cracked, through
some remained intact. Visual inspection showed quite a bit of non-compostable trash in the
remaining bins, if used at all. Photos from the site visit can be found in Figure 6.3.
Incentives and Project Outcomes
In general, local beneciaries do not appear to have experienced any long-term impacts
whether generated by material, coercive, or no incentives. Teaching detainees this particular
composting skill set requires those individuals to invest in equipment without any start-
up capital unavailable to prisoners upon release. Compost from community members was
collected during the project, but the program fell apart afterwards with little collective
memory about the project and only the abandoned remnants of compost bins in front of
only one street in the neighborhood. On the other hand, local implementers saw important
long-term impacts generated by material and nonmaterial incentives, with the Mauritian
government being the main winner in this project. They have reduced prison costs, improved
their reputation via news coverage, and generated some additional income. However, these
benets have been controversially built on the availability of prison labor, while oering few
182
long-term benets to this workforce. The bottom line: local beneciaries had no impacts
and no nonmaterial incentives, while local implementers had abundant impacts associated
with material and nonmaterial incentives.
183
Figure 6.3: Photos from SWMP eld site observation, reading by row from top left to right: (1) A
double compost bin still intact in the surrounding neighborhood, though unused. (2) Petit Verger
Prison walls. (3) Entrance to Petit Verger Prison with the composting center visible in the
background. (4) Close up on the prison's composting center, with a clear reference to its inclusion
in the national climate change initiative Maurice Ile Durable'. (5) Broken double composting bin
in the neighborhood. (6) A surviving compost bin being used for organic waste.
184
Chapter 7
Conclusion: The Power of Nonmaterial Incentives in
Aid Eectiveness
Now that I have presented results for all four projects studied, I will oer a nal discussion
of the data and conclusion in this nal chapter. I will focus on the implication for my
arguments concerning development theory, measurement, institutional and incentive-related
theories, then move to the limitations of the study, and oer recommendations for future
research along the way.
7.1 Discussion
Results show support for my theory that nonmaterial incentives lead to longer-lasting project
impacts, but only when local beneciaries are targeted by a particular project objective.
Instead, material incentives alone appear to lead to long-lasting impacts at the same rate as
nonmaterial incentives when the local implementer is the objective's target population. This
may be explained by the fact that implementers are often under contract with the donor, thus
their very involvement in the project is predicated on material incentives, making them just
as powerful of a long-term motivator. It could be explained by the fact that implementers
have greater material and nonmaterial resources than beneciaries and thus have greater
capacity to institutionalize project changes and/or solve challenges over the long-term.
The role of peer accountability as a particularly important type of nonmaterial incentive
came up again and again. Where there was a well-liked implementer among beneciaries,
long-lasting impacts were seen. Poorly-liked or nondescript implementers appear to be as-
185
sociated with no evidence for long-term impacts. The same goes on the beneciary side; the
importance of charismatic leaders came up again and again in interviews for creating the
institutionalization tool to measure the dependent variable. Most implementers believe that
when there is an in
uential member of the beneciary community participating - whether
that person holds formal or informal power - we are more likely to see long-term results. The
presence of peer accountability may even combat the frustration by beneciaries over project
saturation - when multiple donors come to the same beneciaries over time, promising results
that often never come at the expense of the individual's time and eort. Looking forward,
these ideas should be
ushed out in a rigorous way to look at the impact specically of peer
accountability between implementer and beneciary on project outcomes.
Narrative descriptions here of beneciary-led objectives are harder to incorporate into
the same analysis as they are usually only discovered through in-person interactions, if
discovered at all. Some of the donor objectives listed here were created due to beneciary
requests, however many were unspoken or unmeasured outcomes that can be picked up by
the institutionalization tool but whose incentives are more dicult to identify. While I
have tried to thoroughly include the beneciary desired outcomes I discovered in my results
narratives, future research needs to contemplate how to better identify and incorporate these
objectives into a single rigorous analysis.
7.2 Further Implications
7.2.1 Development Theory
I have shown support for the non-mainstream development theories that make space for
nonmaterial incentives as key motivating factors for aid actors, over the mainstream theories
that privilege material incentives as principal motivators. Economic incentives alone seem
to be largely insucient for motivating long-term development project outcomes. While
communities certainly may lack access to capital in order to carry out development objectives,
186
nonmaterial incentives like peer relationships, social status and recognition, and concerns
with social stigma are associated with longer-lasting project objectives in the four projects
studies here. This may re
ect a broader community-level belief in the project, which certainly
would have an in
uence on the duration of project impacts. Coercive incentives also played
a role as demonstrated in the SWMP project, however more due to actors who had no choice
but to participate rather than deliberate decisions by actors choosing to break a rule by not
participating.
I have additionally suggested here that the typical narrative in development theory con-
sidered through a political lens is that local actors have little in
uence in the realm of
development. While some development theories make a small space for the power of local
actors, these theories remain on the outside of mainstream IR theory. I posit here that
although their power is most often exerted at the micro-level in development projects, lo-
cal actor choices still have consequences at the macro-level, and I have provided evidence
through a wide variety of evidence for this phenomenon. As power dynamics shift in the
international system, the development theories used in IR should also show why power dif-
fuses across actor types. A number of current world leaders and issues are demonstrating
the problems with our many theoretical assumptions about the state as a black box; indi-
viduals can indeed change the system. My research joins a growing body of IR scholarship
that recognizes the independent causal power of low-prole actors like aid recipients, even
in interactions with huge power asymmetries. Rosenau (1990), for example, stimulated an
ongoing argument that the role of individual actors will increasingly matter within the world
politics arena. A micro-level analysis of the impact of everyday people on development is
now key for understanding even macro-level outcomes.
7.2.2 Measurement
The two tools created and presented here have great potential for measuring both incentive
types and the institutionalization of outcomes, but after this pilot implementation, certainly
187
need to be further developed to better capture project outcomes across disparate coders. I
believe in particular that the institutionalization tool has the possibility of better measuring
development outcomes than our [past attempts at the micro-level, but understanding how to
implement it retrospectively - after a project has already occurred - is a challenge I will need
to take up in future research. Project documents may certainly give the impression something
happened, but site visits tend to make clear that no matter the codes and modes determined
ahead of time, results may have all but disappeared by the time eld observations occur.
The incentives tool also provides for easy coding of the three types of incentives here across
coders, however its possibility to capture more types of incentives and nuances in terms of
what community capital (natural, social, political, etc.) was targeted by the incentive still
needs further development to ensure not only that coders are on the same page, but that
the tool is generating helpful information for practitioners and researchers alike.
7.2.3 Theories of Institutions and Incentives
Data collected reveals a few other important points about previous theories and results in
the development eld. As expected, the principal-agent framework does not map well on to
aid projects. For example, while detainee beneciaries in the SWMP project may be clear
agents, this is due to their literal captivity. Community member beneciaries in this same
project clearly have felt no obligation towards the implementers during or post-project and
certainly do not behave like agents. In Proyecto Carbono, indigenous groups are clearly the
principals and have exercised their power through objective negotiation and decisions over
whether they will work with implementers in the future. In AC3's Community B, there was
clear resentment by the community beneciaries over whether the implementers overstepped
their bounds in implementing the project in their community without proper permissions,
placing the beneciaries in the principal position and the implementers as a rogue agent.
Principal-agent conceptions simply do not appear to be fruitful frameworks for analyzing aid
actors or incentives at the project-level.
188
We did however see support for the Octangle framework, at least in interpreting the
complex relationships present in aid projects. Of all the projects, the complicated web of
stakeholders present in Proyecto Carbono (see Figure 5.2) nicely illustrates the many actors
and relationships between them. SWMP shows how one donor in reality is made up of three
dierent bodies, whereas a project like AC3 shows how recipients and beneciaries can in
fact play the same role. The Octangle framework also reminds us that there are incentives
playing at all actor levels. Given that the implementers in this research were all academic
institutions, their incentives - as we know from Merton (1973) - come from institutional
structures that promote nonmaterial incentives such as peer validation through conference
presentations and publications. This most likely has aected the result here that nonmaterial
incentives were most important when beneciaries were the target population for a particular
objective, but not when implementers were - because implementers in this case would be
motivated no matter the incentive given their own career interests.
We additionally see some utility in using the team production framework to analyze
results. Certainly when local actors did not play their part in the process - for example, the
Cab ecar in Proyecto Carbono refused to continue working with CATIE in the future, which
has stopped most CATIE initiatives all together in the territories. The SWMP has had
great success with in-prison composting because prisoners have no choice but to play their
role, whereas all composting in the nearby community has not only stopped but is barely
remembered by community members just a few years later, who appear to never really have
had a clear incentive to perform their roles in the project.
7.3 Limitations
7.3.1 Country Cases
No matter the study, it is notoriously diculty to compare across countries as no two coun-
tries are the same. The two country contexts I chose are widely considered development
189
successes, which was a deliberate choice as a rst step to
eshing out this new micro-level
development theory. Both of these countries are very concerned with their international
reputation, given its deep links to their priority economic sectors like tourism and interna-
tional nance. This could in fact be a motivator for countries to put more resources into
promoting development success at the micro-level. On the other hand, institutional stability
could be another determining factor in producing buy-in at the local level. Further research
should diversify country cases along more variables, including cases where development has
not been so successful.
7.3.2 Sectors
I chose here two projects in the Agriculture, Forestry and Fishing sector, and two in the
Water Supply and Sanitation sector. This was meant to provide some diversity across sectors
while still maintaining some comparability given vast dierences between sectors. Indeed,
not all sectors show clear, measurable changes. Humanitarian and Governance purposes are
two of the most notoriously dicult sectors to measure. While the tools are designed to
apply to any project, future research needs to address whether sectors with greater outcome
measurement challenges can also be used in such research.
7.3.3 Time and Resource Constraints
As argued earlier, project evaluation is a time consuming and costly venture, and this research
is no dierent. It often took months to plan visits, only for individuals to cancel or reschedule
at the last minute. Travel was often long, given locations were rural and often not on maps
or had no cell phone signal to nd one's way via maps or calling contacts. Some permissions
could not be obtained in time given travel constraints, for example, access to the prison in
Mauritius. This limitation is in line with the realities of aid project evaluation, however the
hope is with continued development of the evaluation tools presented here, one may be able
to obtain higher quality information when evaluating projects, in particular the possibility
190
of capturing unintended project consequences.
7.3.4 Implementers
I have chosen to use academic institutions as implementers in order to minimize the eect
of implementer culture in this study and focus instead on incentive types. However, future
research very clearly needs to look into other types of implementers - especially those that
represent local actors in projects - who clearly would have dierent motivations than say
presenting project discoveries at academic conferences or in peer-reviewed journals. We can
imagine that implementers coming from private companies, NGOs, bilateral agencies, etc.,
all have very dierent incentives driving their performance in a project, and the tools here
do still have the ability to capture those distinct motivators.
7.4 Conclusion
Given the evidence presented here, I am condent to say that incentive types are related to
the duration of project outcomes, though their eect may be mediated by the actor targeted
by each project objective. This study could provide a basis for eventually understanding the
link between micro-level projects and macro-level outcomes on the aggregate national level.
This endeavor allows us to better understand the conditions under which aid is eective in the
long-term, and gives us a way to conceptualize and operationalize aid eectiveness beyond
macro-measures. It also has a clear policy implication that could benet practitioners when
designing and monitoring aid projects.
191
Appendices
192
Appendix A
Field Interview Responses on Aid Evaluation
Field interviews were performed on four islands in the Philippines during May and June 2015
with foreign aid practitioners from bilateral donor agencies and NGO donors, along with
local partners of such aid agencies representing the recipient government and implementer
categories. All participants were asked the question: \Do you know if locals ever evaluate
the success of the project or aid separately, using their own methods or resources?"
Answers were as follows:
A.1 Local Actors: Recipient Government and Implementers
\No. Only if there is { if there are people involved in this project...It's not that
they're not interested in the project, but...they're just less interested because they're
not getting personal gains from the project." - Field Partner
\Yes, separately. Separately. Oh oh [yes]. Separate for every project...You talk with
the people who's { who are involved in the project...And so I know what's going on.
Yeah. Later I want to ask so many more questions...And of course what's the result?
I [nd out] through the focal persons that are in charge...But you know, problem
is sometimes you get a good data in [one source]. But it doesn't match with the
[other source's] data...And so it's a check and balance whether they're telling the truth
or not....because performance is the most important for them....The government is
providing performance incentive to { to schools that are performing better..." - Field
193
Partner
\We are forcing them to need to do that, but yeah. But some of them, I know, most of
them are not really doing it....We would be happy if there would be one organization
who would tell us that this is what the contribution of the [project details] are for us,
but right now, hmm. As I've said, they are only forced to do that if we asked them to
submit a report to support a new request." - Central Oce Partner
A.2 Donors
\No, I don't think they do. They don't do it for themselves, they don't do it for me,
they don't do it. They tried to just live by the day." - Field Aid Practitioner
\Yeah...but I'm not sure if they give feedback to the head of the local oces to give
to the employees...there probably is some communication, it's just not given to me
directly. There's less involvement of her giving me feedback. I think she expects that
[bilateral agency] would be doing that more or less." - Field Aid Practitioner
\I don't think so." - Field Aid Practitioner
\I would say the only instance I've had where that occurred was after the mangrove
planting last year. We were { I went on one of the inspecting trips after they were
planted. And it was a [person] that had been doing this for a while, just on their own.
And the reason they said they felt was shing better afterwards. Um { they felt like it
had improved. So, which for a sherman, that's hey, you know, that is denitely, you
know, money in your pocket, food in your mouth kind of thing. Um, but other than
that I haven't really { I haven't really come across it too much. But I wouldn't say
that I've had ample opportunity to either." - Field Aid Practitioner
\...they do, but it's in their best interest - they've got a con
ict of interest, they can't
turn around and say that the clique that they've just given too many dollars for we
194
haven't actually utilized....it's sort of in their best interest to say their equipment's
being used because they've received the funds so they can get more funds. But it's
actually not true." - Field Aid Practitioner
\We ask them to participate { they have their own view of how the projects are running,
no { I don't know if they are really, you know { one thing that we really train them as
part of our participatory process is teach them how to evaluate, how to monitor....So
did they do a scientic evaluation and monitoring? No, they just saw these are the
impact in their lives, you know, and so for them that is a successful project, water, you
know, now there is no one getting sick....they were the next, the newest, the newest
ODF community, so open defecation free community. So, again, did they did they
come up with scientic? No, they know they had been awarded because no one is
defecating anymore in public, you know? And that each of the household has toilet.
So yeah, I am indirectly answering your question, but where we do have structured
evaluation, we ask them to participate and we train them how to do it." - Central
Oce Aid Practitioner
\I doubt it. But if you're talking about organizations { those who have the means will
do it...So there's a few who do it, but I don't see { I don't think like a civil society would
start like until like { do an assessment because that takes like I heard it's expensive
and takes time and money. And it would be great, but I don't think { I doubt it." -
Central Oce Aid Practitioner
\When you have an organized beneciary, they can actually do their own monitoring
{ very informal though { sometimes they don't have the { they don't document these
kinds of processes among them. But I think they are also providing us self-monitoring
system, they are also in their own ways monitoring projects...although they might not
really label it as monitoring system, like what we do as aid funding agencies, but they
have this kind of looking at things among them that allows them to see if this project
195
is eective or not, or maybe we have to shift our focus. Yeah. So you know, we cannot
really underestimate the knowledge, the local knowledge, and how they come about
with decisions of that level." - Central Oce Aid Practitioner
\Yes they do. On the level of their { how, how'd you say, on the level of { on their
emotions, so it's like how happy they are like, how satised they are, your emotional
base evaluation, we've noticed that whenever you go to a, like a community, you will
sort of gauge how happy they are when you arrive, and you will be welcome in a
certain area. When some expectations from the individual is not met you will sense
it, so those sorts of things. By really just pure emotional measurement. I have never
heard of a local community or individuals that has making evaluation like those, you
know, evaluation like competency: 10, goal deliberation: 5, something like that. Not
really." - Central Oce Aid Practitioner
\Not to my knowledge, no. Like I said, I mean, as much as possible, we do try to do
the evaluation together or leave it, try as much as possible to have our partners do it
in the rst place so then, you know, it's not a separate evaluation in the sense that {
what am I trying to say { I think that the thing is that is there is this joint evaluation
of our activities with our partner organizations." - Central Oce Aid Practitioner
196
Appendix B
List of Project Documents Collected
B.1 Costa Rica: Proyecto Manglares
B.1.1 Documents and Academic Articles
1. Acosta, J. G. (2013). Celebraci on D a Mundial de los Humedales

Area Silvestre
Protegida Humedal Nacional T erraba-Sierpe Sitio Ramsar, Costa Rica, 2013. Accessed
at: http://osaconservation.org/wp-content/uploads/2013/06/Celebraci%C3%B3n-D%C3%
ADa-Mundial-de-los-Humedales2.pdf
2. Aguilar R., E. (1990). Regeneraci on en el manglar en areas intervenidas en la Reserva
Forestal de T erraba-Sierpe, Costa Rica (Magister Scientiae). CATIE, Turrialba, Costa Rica.
3. Ammour, T. (1993). Conservaci on y desarrollo sotenible en Am erica Central: Manejo
y aprovechamiento de la biodiversidad. Revista Forestal centroamericana, 2(5), 20{26.
4. Ammour, T., Ocampo, R., & Robles, G. (1994). Caracterizaci on de los sectores aso-
ciados a la producci on, comercializacio on y transformaci on de plantas medicinales en Costa
Rica. (No. 3; p. 36). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/bitstream/
handle/11554/659/Caracterizacion de los sectores asociados.pdf?sequence=1&isAllowed=y
5. Ammour, Tania. (1998a). Proyecto uso sostenible del manglar:

Area demostrativa
manglares de terraba-sierpe, Cant onde Osa, Puntarenas, Costa Rica.
6. Ammour, Tania. (1998b). Semestral Report July-December 1998: Main Achievements
{ Olafo-Mangroves Impact Assessment 1989-1998 (No. 18; p. 19). Turrialba, Costa Rica:
CATIE.
197
7. Ammour, Tania. (1999). Evaluaci on parcial de impactos de los proyectos Olafo y
Manglares: Dise~ no y demostraci on del desarrollo rural basado en el manejo de ecosistemas
naturales.
8. Ammour, Tania, Imbach, A., Suman, D., & Windevoxhel Lora, N. J. (1999). Manejo
productivo de manglares en Am erica central. Accessed at: https://portals.iucn.org/library/
node/8411
9. Asch Quir os, C. (1993). Informe nal, estudio de fotointerpretaci on y confecci on de
mapas uso de la tierra y unidades vegetativas de la Reserva Forestal T erraba-Sierpe, Proyecto
Danida Manglares (p. 32). Turrialba, Costa Rica.
10. Asch Quir os, C., & Solano Brenes, O. (1991). Clasicaci on y perspectivas de manejo
de los humedales Terraba-Sierpe (p. 228). Accessed at: http://opac.bibliotecaorton.catie.
ac.cr/cgi-bin/koha/opac-detail.pl?biblionumber=406725
11. Barton, D. N. (1995). Valoraci on Economica Parcial de Alternativas de Manejo
para los Humedales de T erraba Sierpe Costa Rica (Magister en Pol tica econ omica
para Centro Am erica y el Caribe con enfasis en Econom a Ec ol ogica y Desarrollo
Sostenible, Universidad Nacional). Accessed at: https://www.researchgate.net/publication/
305692200 Partial Economic Valuation and Complementary Decision Making Criteria in
the Management of the Terraba-Sierpe Mangroves Costa Rica
12. Casta~ neda Molina, K., Castro Gonz ales, M., Olivares, I., & D az Hern andez, L.
(2005). Humedales de Terraba Sierpre|Segundo estudio de caso. Accessed at: https://
www.academia.edu/8753537/8caso manglares terrabasierpre
13. CATIE. (1994). Annual Report 1994 (Annual Report No. 14; p. 253).
Accessed at: https://books.google.co.cr/books?id=iRcOAQAAIAAJ&printsec=frontcover&
source=gbs ge summary r&cad=0#v=onepage&q&f=true
14. CATIE. (1995). Annual Report 1995 (Annual Report No. 15; p. 203). Bib. Orton
IICA / CATIE. Accessed at: https://books.google.co.cr/books?id=4RgOAQAAIAAJ
15. CATIE (Ed.). (1998). CATIE's contribution to sustainable rural development
198
in Central America: An overview of the impacts of CATIE/OLAFO and Mangroves
projects 1989-1998. Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/bitstream/
handle/11554/732/CATIE s contribution to sustainable rural development.pdf?sequence=
1&isAllowed=y
16. CATIE, & IUCN. (1996). Wise use of the mangrove resources in Estero Real
(Nicaragua) and Terraba-Sierpe (Costa Rica)|Mangrove Project: Progress report April-
June 1996 (No. 3; p. 17). Accessed at: http://201.207.189.89/bitstream/handle/11554/
4324/Wise use of the mangrove April-June.pdf?sequence=1&isAllowed=y
17. CATIE, & UICN. (1990). Wise use of the mangrove resources in Estero Real,
Nicaragua and Terraba-Sierpe, Costa Rica: A proposal submitted by CATIE/UICN to
DANIDA (No. 1; p. 61). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/
bitstream/handle/11554/4325/Wise use of the mangrove.pdf?sequence=1&isAllowed=y
18. Cordero P erez, P. (2000). El manglar m as grande de Costa Rica: Experiencias de la
UICN en el proyecto DANIDA-MANGLARES de T erraba-Sierpe (p. 40). San Jos e, Costa
Rica: IUCN.
19. Espinosa Sep ulveda, A. R. (1992). Evaluaci on de la estructura y composici on
del bosque de manglar y lineamientos para su manejo s lvicola en la reserva forestal de
T erraba-Sierpe, Costa Rica (Magister Scientiae, CATIE). Accessed at: http://repositorio.
bibliotecaorton.catie.ac.cr:80/handle/11554/1942
20. Jimenez B., G. (2018). Humedal Terraba-Sierpe [Archivo jpg]. In Area de Con-
servaci on Osa (Guillermo Jimenez B). Accessed at: http://www.sinac.go.cr/ES/ac/acosa/
Mapas/H04 Terraba-Sierpe.jpg
21. Jim enez, J. A. (2003, November 11). Alternativas de manejo de los manglares en
el contexto del pacico de centroamerica. Accessed at: http://www.mag.go.cr/congreso
agronomico ix/A01-1277-21.pdf
22. Jim enez-Ram on, J. A., Arguedas-Campos, C. L., & Coto, J. (1991). Es-
tudio de factibilidad de producci on de carb on y taninos en polvo a partir del
199
mangle, en Sierpe-T erraba. Organizaci on para Estudios Tropicales. Accessed
at: http://www.sidalc.net/cgi-bin/wxis.exe/?IsisScript=oet.xis&method=post&formato=
2&cantidad=1&expresion=mfn=011271
23. Kent, J., Marmillod, D., Ammour, T., & Villalobos Soto, R. (1998). Investigaci on y
desarrollo: Hacia el manejo sostenible de un insecticida natural en Costa Rica. Accessed at:
http://repositorio.bibliotecaorton.catie.ac.cr:80/handle/11554/2621
24. Kent, Justine, & Bianco, L. (1994, December 12). Market-related constraints to non-
timber forest product development in Central America: Experiences from the CATIE/Olafo
Project. 18. Annapolis, MD.
25. Lahmann, E. J. (1999). La Reserva Forestal de T erraba-Sierpe, Costa Rica: Un
ejemplo de Uso Adecuado del Manglar. Ecosistemas de Manglar en Am erica Tropical, 1,
291{298.
26. Mainardi, V. (1996). El manglar de Terraba-Sierpe en Costa Rica (Sandra
Ram rez Rivera). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/bitstream/
handle/11554/1576/El manglar de Terraba Sierpe.pdf?sequence=1&isAllowed=y
27. Marin Melendez, M. E. (1991). Estudio de caso sobre el uso actual de la Reserva
Forestal de T erraba-Sierpe y evaluaci on de la rentabilidad de un proyecto de maricultura
y silvicultura para Coopemangle (Magister Scientiae, CATIE). Accessed at: http://orton.
catie.ac.cr/repdoc/A1576e/A1576e00.pdf
28. Marmillod, D. (1993, December 4). Diagnostico del Estado del Conocimiento en el
Area Demonstrativa (AD) Terrab a-Sierpe (minitaller del 25.10.1993).
29. Matute Ir as, L. (2016). An alisis de situaci on de las comunidades del humedal
de Sierpe, cant on de Osa, Costa Rica, como insumo para la toma de decisiones sobre
aprovechamiento y control de la tilapia silvestre introducida (Master en pr actica de la con-
servaci on de la biodiversidad, CATIE). Accessed at: http://repositorio.bibliotecaorton.catie.
ac.cr:80/handle/11554/8684
30. McCarthy R., R., & N., W. L. (1995). Plan de Manejo Humedal Nacional T erraba-
200
Sierpe (p. 119). Accessed at: CATIE, Turrialba (Costa Rica). Programa Uso Sostenible
de los Recursos del Manglar; IUCN, San Jos e (Costa Rica). Programa de Humedales para
Mesoam erica. Accessed at: http://www.sidalc.net/cgi-bin/wxis.exe/?IsisScript=orton.xis&
method=post&formato=2&cantidad=1&expresion=mfn=057850
31. McCarthy Ram rez, R. G. (1993). El sistema de areas protegidas de Costa Rica y su
aporte a la conservaci on y manejo de los humedales (Magister Scientiae, CATIE). Accessed
at: http://repositorio.bibliotecaorton.catie.ac.cr:80/handle/11554/1604
32. Mory Villase~ nor, S. (2017). Programa de educaci on ambiental en el agua, Humedal
Terralba-Sierpe. Sitio Ramsar Costa Rica. Accessed at: http://www.sinac.go.cr/ES/
partciudygober/Documents/PROGRAMA%20EDUCACI%C3%93N%20AMBIENTAL%
20EN%20EL%20AGUA%20HUMEDAL%20TERRABA%20SIERPE.pdf
33. Pizarro B., F., & Angulo Z., H. (1993). Diagn ostico de los manglares de la costa
pac ca de Costa Rica (p. 67). Accessed at: Universidad Nacional, Heredia (Costa Rica).
Escuela de Ciencias Biol ogicas. Laboratorio de Ecolog a y Manejo de Ecosistemas Estuarinos;
CATIE, Turrialba (Costa Rica). Proyecto Uso Adecuado de los Recursos del Manglar en
T erraba-Sierpe CATIE-UICN-DANIDA. Accessed at: http://www.sidalc.net/cgi-bin/wxis.
exe/?IsisScript=orton.xis&cantidad=50&expresion=AVICENNIA%20BICOLOR
34. Reyes, V., Miranda, M., Monge, C., & Salas, F. (2007). Valoraci on econ omica del
ecosistema Humedal Nacional T erraba-Sierpe y propuesta de mecanismos para su sosteni-
bilidad, Costa Rica. In An alisis de casos en Iberoam erica: Vol. 1. Sustento del uso
justo de materiales protegidos por Derechos de autor para nes educativos (Alexandra
Mel endez C., p. 604). Accessed at: http://www.ucipfg.com/Repositorio/MLGA/MLGA-
09/Unidades academicas/Unidad3/001.pdf
35. S anchez Pe~ na, R. O. (1986). Metodolog a descriptiva para determinar los posibles
usos de las areas de manglares, y su aplicaci on en Coronado-Sierpe, Costa Rica (Magis-
ter Scientiae, CATIE). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/bitstream/
handle/11554/3104/Metodologia descriptiva.pdf?sequence=1&isAllowed=y
201
36. SGP. (2004). Desarrollo de un servicio eco tur stico novedoso en los manglares de
Terraba Sierpe por parte de COOPEMANGLE R.L., para apoyar su conservaci onThe GEF
Small Grants Programme. Accessed May 11, 2019, at: https://sgp.undp.org/index.php?
option=com sgpprojects&view=projectdetail&id=7478&Itemid=205
37. Stilma, A. (1994). Problems and prospectives of a mangrove exploiting cooperative
in Costa Rica. Accessed at: http://www.sidalc.net/cgi-bin/wxis.exe/?IsisScript=oet.xis&
method=post&formato=2&cantidad=1&expresion=mfn=011513
38. Suazo Suazo, A. M. (1997). Caracterizaci on de los enfoques metodol ogicos para la
participaci on comunitaria en el manejo integrado de los recursos naturales (Magister Scien-
tiae, CATIE). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/bitstream/handle/
11554/4611/Caracterizacion de los enfoques metodologicos.pdf?sequence=1
B.1.2 Interviews and Meetings
Please note that interviewees are protected by my IRB protocol, thus limited information
has been given about each individual to avoid identication.
TOTAL: 8
1. Implementers: 4 (from CATIE, the IUCN, and several Costa Rican government enti-
ties)
2: Beneciaries: 4 (from CoopeMangle)
B.2 Costa Rica: Proyecto Carbono
B.2.1 Documents and Academic Articles
1. Altieri, M. A. (2011). Agroecology versus Ecoagriculture: Balancing food production
and biodiversityconservation in the midst of social inequity. Accessed at: http://agroeco.
org/wp-content/uploads/2011/02/ECOAG.pdf
2. Andrade Casta~ neda, H. J., & Segura Madrigal, M. A. (2008). >C omo estimar
202
r apidamente el carbono almacenado en la biomasa a erea de los sistemas agroforestales
ind genas de Talamanca, Costa Rica. Agroforester a en las Am erica, 6(46), 7.
3. Andrade, H. J., Segura, M., Somarriba, E., & Villalobos, M. (2008). Valoraci on
biof sica y nanciera de la jaci on de carbono por uso del suelo en ncas cacaoteras ind genas
de Talamanca, Costa Rica. Agroforester a en las Am erica, (46), 45{49.
4. Arce, N. (2006). Informe nal de consultoria.
5. Arce, N., Ortiz, E., Villalobos, M., & Cordero, S. (2008). Existencias de carbono en
charrales y sistemas agroforestales de cacao y banano de ncas ind genas bribri y cab ecar de
Talamanca, Costa Rica. Agroforester a en las Am erica, (46), 30{33.
6. Borge, C. (2006). Estrategia de fortalecimiento institucional acomuita (p. 106) [Con-
sultoria]. Talamanca, Costa Rica: CATIE.
7. Borge, C., & Castillo, R. (1997). Cultura y conservaci on en la Talamanca ind gena.
San Jos e, Costa Rica: EUNED.
8. Borrero, G. (2005). Resumen ejecutivo del diagn ostico de grupos productores de cacao
en centroamerica con potencial para incorporarse al mercado regional de carbono (p. 109)
[Resumen Ejecutivo]. Talamanca, Costa Rica: CATIE.
9. Bowyer-Walker, N. L. (2004, February 19). Prop ositos, T erminos y Condiciones de
la Donaci on Japonesa para la Captura de Carbono y Desarrollo de Mercados Ambientales
en Sistemas Agroforestales Ind genas con cacao en Costa Rica, Acuerdo TF052118 [Pdf
Document]. (GrantTF052118.Firma.Espa~ nol 25 mayo.doc).
10. Calero Borge, W. A. (2008). Producci on e incrementos de madera y carbono de laurel
(Cordia alliodora) y cedro amargo (Cedrela odorata L.) de regeneraci on natural en cacao-
tales y bananales ind genas de Talamanca, Costa Rica (Magister Scientiae en Agroforester a
Tropical, CATIE). Accessed at: http://orton.catie.ac.cr/repdoc/A2386e/A2386e.pdf
11. Calero Borge, W. A. (2009). Producci on de Madera y Carbono en la Regeneraci on
de Sistemas Agroforestales en al Reserva Ind gena de Talamanca, Costa Rica. Ciencia e
Interculturalidad, 4(2), 108{150.
203
12. Candela Restrepo, S. (2007). Convergencias y Divergencias de las Organizaciones
Ind genas y Externas Involucradas en la Gesti on de los Recursos Naturales de la Alta
Talamanca/Caribe (Magister Scientiae en Agroforester a Tropical, CATIE). Accessed at:
http://www.sidalc.net/REPDOC/A1426E/A1426E.PDF
13. Candela, S. E., Stoian, D., Somarriba, E., & Villalobos, M. (2013). Formas y efectos
de la gobernanza forestal en los territorios ind genas bribri y cab ecar de Alta Talamanca,
Costa Rica. Agroforester a en las Am erica, 49(49), 85{92.
14. Carnap, M. (2005). Informe t ecnico, proyecto captura de carbono y desarrollo de
mercados ambientales en sistemas agroforestales ind genas con cacao en Costa Rica (p. 54)
[Consultoria]. Talamanca, Costa Rica: CATIE.
15. Carvajal Cascante, E. (2006). Integraci on t ecnica de resultados de cosnultorias del
componente II dle proyecto de captura de carbono (p. 54) [Consultoria]. Talamanca, Costa
Rica: CATIE.
16. CATIE. (2006a). Captura de carbono y desarrollo de mercados ambientales en
cacaotales y otros sistemas agroforestales ind genas en Talamanca. Accessed at: http://
orton.catie.ac.cr/repdoc/A0928e/A0928e.pdf
17. CATIE. (2006b). Proyecto jaci on de carbono en sistemas agroforestales con cacao
de ncas ind genas en Talamanca, Lim on, Costa Rica.
18. Cerda Bustillos, R., Espin Chion, T., & Cifuentes, M. (2013). Carbono en sistemas
agroforestales de cacao de la Reserva Ind gena Bribri de Talamanca, Costa Rica. Agro-
forester a en las Am ericas, (49), 9.
19. Cerda, R., Deheuvels, O., Calvache, D., Niehaus, L., Saenz, Y., Kent, J., Vilchez,
S., Villota, A., Martinez, C., & Somarriba, E. (2014). Contribution of cocoa agroforestry
systems to family income and domestic consumption: Looking toward intensication. Agro-
forestry Systems, 88(6), 957{981.
20. Chaverri, D. (2005a). Analisis de la legislaci on forestal y marco legal referente a
la participaci on de ocmunidades ind genas en el pago de servicios ambientales. (p. 40)
204
[Consultoria]. Talamanca, Costa Rica: CATIE.
21. Chaverri, D. (2005b). Propuesta de mecanismos necesarios para que funcione el pago
por servicios ambientales (an alisis legal y contratos), Memoria de talleres.
22. Chaverri, D. (2007). Estrategia legal para el establecimiento y operaci on del mercado
de captura de carbono en los territorios ind genas Bribri y Cab ecar (Consultoria No. TF-
052118; p. 82). Talamanca, Costa Rica: CATIE.
23. Cole, J. (2005). Sobre lo plural de las ideas y los discursos de desarrollo econ omico
local. Estudio de caso del pueblo ind gena bribri de Talamanca, Costa Rica. Revista Cen-
troamericana de Ciencias Sociales, 2(1), 135{168.
24. Dahlquist, R. M., Whelan, M. P., Winowiecki, L., Polidoro, B., Candela, S., Harvey,
C. A., Wulfhorst, J. D., McDaniel, P. A., & Bosque-P erez, N. A. (2007). Incorporating liveli-
hoods in biodiversity conservation: A case study of cacao agroforestry systems in Talamanca,
Costa Rica. Biodiversity and Conservation, 16(8), 2311{2333.
25. Deheuvels, O., Rousseau, G. X., Soto Quiroga, G., Decker Franco, M., Cerda, R.,
V lchez Mendoza, S. J., & Somarriba, E. (2014). Biodiversity is aected by changes in
management intensity of cocoa-based agroforests. Agroforestry Systems, 88(6), 1081{1099.
26. Gaete, M. (2006). Evaluaci on de los resultados e impactos sociales y culturales
del proyecto captura de carbono (p. 59) [Informe de Evaluaci on]. Talamanca, Costa Rica:
CATIE.
27. Garz on, A. M. B., & Cuaspud, D. F. C. (2006). Determinaci on de los Destinos de
Madera y Carbono de Laurel (Cordia alliodora) y Cedar (Cedrela odorata) de Regeneraci on
Natural en Cacaotales y Bananales en os Territorios Ind genas Bribr y Cab ecar de Tala-
manca, Costa Rica (Magister Scientiae en Ingeniero Agroforestal, Universidad de Nari~ no).
Accessed at: http://www.sidalc.net/repdoc/A11219e/A11219e.pdf
28. Guzm an, J. (2006). Evaluaci on de resultados ambietales, evaluaci on de resultados e
impactos econ omicos (p. 140) [Consultoria]. Talamanca, Costa Rica: CATIE.
29. Herrera, A. C., Hip olito-Romero, E., Junkin, R., & Aguilar, A. E. (2013). El rol
205
de los sistemas cacaoteros en los medios de vida de los hogares productores del Cant on de
Talamanca, Costa Rica (No. Bolet n T ecnico No 64; p. 58). Turrialba, Costa Rica: CATIE.
30. Hicks, F., Rivera, L., Alp zar, E., & Alvarez, H. (2006). Componente Dise~ no de
una estrategia de comercializaci on, certicaci on y monitoreo de carbono jado en ncas
agroforestales de los territorio ind genas Bribri y Cab ecar (p. 45) [Consultoria]. Talamanca,
Costa Rica: Rainforest Alliance.
31. Imbach, M. (2005). Consultoria para rehabilitaci on de ncas y fomento de la pro-
ducci on sostenible en las areas afectadas por las inundaciones.
32. Imbach, M. (2006a). Los pueblos ind genas deben asumir la autoridad de gobierno
local y de propietarios privados de sus tierras.
33. Imbach, M. (2006b). Perl organizativo ACOMUITA.
34. Islas, A. S. (2008). >Cu antos y cu ales arboles de laurel (Cordia alliodora) se pueden
cosechar en los cacaotales y bananales ind genas de Talamanca, Costa Rica? Agroforester a
en las Am erica, 46, 6.
35. Lansing, D. M. (2009). The Spaces of Carbon: Calculation,Technology, and Dis-
course in the Production of Carbon Forestry Osets in Costa Rica (Graduate Program in
Geography, The Ohio State University). Accessed at: https://etd.ohiolink.edu/pg 10?0::
NO:10:P10 ACCESSION NUM:osu1250637819
36. Lansing, D. M. (2013). Not all baselines are created equal: A Q methodology analysis
of stakeholder perspectives of additionality in a carbon forestry oset project in Costa Rica.
Global Environmental Change, 23(3), 654{663.
37. Morales, F. (2005). Consultoria elaboraci on de contenidos de programas radiales,
materiales impresos y murales, producci on de materiales y un centro de informaci on. (p. 16)
[Consultoria]. Talamanca, Costa Rica: CATIE.
38. Murillo Araya, M. E., & Salazar Morales, B. (2005). Informe nal de la consultoria
para la coordinaci on de la cooperaci on externa en el marco del proyecto captura de carbono
en los territorios ind genas Bribri y Cabecar (p. 160) [Consultoria]. Talamanca, Costa Rica:
206
CATIE.
39. Orozco, L. (2006). Sistematizaci on y an alisis de diagn osticos de ncas cacaoteras de
Talamanca (p. 33) [Consultoria]. Talamanca, Costa Rica.
40. Orozco, L., Deheuvels, O., Villalobos, M., & Somarriba, E. (2015). El Sector Cacao
en Centroamerica Estado de desarrollo en el a~ no 2007 (T enico No. 401; p. 82). Turrialba,
Costa Rica: CATIE.
41. Ortiz Guerrero, A. M., & Riascos Chalapud, L. D. (2006). Almacenamiento y
Fijaci on de Carbono del Sistema Agroforestal Cacao (Theobroma cacao) y Laurel (Cordia
alliodora|Ruiz & Pav on) Oken en la Reserva Ind gena de Talamanca, Costa Rica (Ingeniero
Agroforestal, Universidad de Nari~ no). Accessed at: http://orton.catie.ac.cr/repdoc/A3706e/
A3706e.pdf
42. P erez Bonilla, M. E. (2006, May). Evaluaci on Participativa de la Coejecuci on del
Proyecto con las Organizaciones Involucradas.
43. Project Coordination Council. (2006). Informe de cierre Captura de Carbono y
Desarrollo de Mercados ambientales en sistemas Agroforestales Ind genas con cacao en Costa
Rica. / Implementation completion report|Costa Rica: Carbon capture and development
of environmental markets in indigenous cocoa agroforestry systems in Costa Rica. GRANT
TF 52118 (p. 63) [Consultoria]. Talamanca, Costa Rica: CATIE.
44. Ramos, J. R. A. (2011). Generaci on y aplicaci on de conocimientos de tesis gener-
adas por estudiantes del CATIE entre 2002-2009 en los territorios ind genas Bribr de Tala-
manca: Rastreando sus rutas, resultados y usos (Magister Scientiae en Agricultura Ecol ogica,
CATIE). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/handle/11554/2212
45. Reyes Fern andez, V. (2006, April). Consultoria para apoyar el seguimiento de ADIT-
ICA y los consejos de vecinos.
46. Rivera, A. (2006). Asistencia adminsitrativa y nanciera del monitoreo y seguimiento
de la evaluaci on del impacto sociocultural de carbono por parte de las organizaci on indigenas
ACOMUITA, ADITIBRI, ADITICA.
207
47. Rodriguez, C. (2005). Asesor a t ecnica, pol tica y organizativa a ADITIBRI, ADIT-
ICA y ACOMUITA en el monitoreo y seguimiento de la evaluaci on dle impacto sociocultural
dle proyecto captura de carbono (p. 16) [Consultoria]. Talamanca, Costa Rica: CATIE.
48. Rodr guez, M., & Davidson-Hunt, I. J. (2018). Resilience and the Dynamic Use of
Biodiversity in a Bribri Community of Costa Rica. Human Ecology, 46(6), 923{931.
49. Rodr guez Romero, G. (2006a). An alisis de la problem atica tendencias y perspectivas
que tienen los ancianos, J ovenes y educadores y lideres in
uyentes en el desarrollo y la
conservaci on de los recursos naturales.
50. Rodr guez Romero, G. (2006b). Consultor a para actualizar mecanismos de admin-
istraci on de los territorios ind genas Bribri y Cab ecar.
51. Rodr guez Romero, G. (2006c). Un nuevo enfoque para la educaci on ind gena Bribriss
y Cab ecares.
52. Rojas, V., Martinuz, A., Candela, S., & Cheverry, D. (2007). Caracterizaci on
de una experiencia existosa de Pago por Servicios Ambientales (PSA), en los teritorios
ind genas Bribr y Cab ecar de Costa Rica (Sistematizaci on No. 1; p. 77). Accessed
at: https://www.conafor.gob.mx/innovacion forestal/wp-content/uploads/2018/04/Rojas-
V.-et-al.Caso-PSA zona-indigena Costa-Rica.pdf
53. Sampson, A. L. (2006). Edici on del n umero especial de la Revista Agroforester a en
las Am ericas sobre el Proyecto Carbono. (p. 156) [Consultoria]. Talamanca, Costa Rica:
CATIE.
54. Segura, M. (2005). Estimaci on del carbono almacenado y jado en sistemas ageofore-
stales ind genas con cacao en la zona de Talamanca, Costa Rica (Consultoria No. TF-052118;
p. 139). Talamanca, Costa Rica: CATIE.
55. Segura, M. (2006). Estimaci on de carbono y dise~ no de l nea base para proyectos
MDL (p. 10) [Consultoria]. Accessed at: CATIE website.
56. Segura Madrigal, M. (1999). Valoraci on del servicio de jaci on y almacenamiento de
carbono en bosques privados en el Area de Conservaci on Cordillera Volc anica Central, Costa
208
Rica (Magister Scientiae, CATIE). Accessed at: http://repositorio.bibliotecaorton.catie.ac.
cr:80/handle/11554/4298
57. Selles, C. S. (2005). Informe nal de la consultoria Suwa Pakok realizado en los
territorios ind genas Bri Bri|Cabecar de Talamanca (p. 20) [Consultoria]. Talamanca,
Costa Rica: CATIE.
58. Somarriba, E., Andrade, H. J., Segura, M., & Villalobos, M. (2008). >C omo jar
carbono atmosf erico, certicarlo y venderlo para complementar los ingresos de productores
ind genas en Costa Rica? Agroforester a en las Am erica, 6(46), 8.
59. Somarriba, E., Cerda, R., Orozco, L., Cifuentes, M., D avila, H., Espin, T., Mavisoy,
H.,

Avila, G., Alvarado, E., Poveda, V., Astorga, C., Say, E., & Deheuvels, O. (2013).
Carbon stocks and cocoa yields in agroforestry systems of Central America. Agriculture,
Ecosystems & Environment, 173, 46{57.
60. Somarriba, E., & Harvey, C. A. (2003). >C omo integrar la producci on sostenible
y conservaci on de la biodiversidad en cacaotales org anicos ind genas? Agroforester a en las
Am erica, 10(37{38), 6.
61. Somarriba, E., Quesada, F., & Villalobos, M. (2008). La captura de carbono: Un
servicio ambiental en ncas cacaoteras ind genas (No. Manual T ecnico No. 64; p. 28).
Accessed at: http://www.sidalc.net/repdoc/a0870e/a0870e.pdf
62. Somarriba, E., Su arez-Islas, A., Calero-Borge, W., Villota, A., Castillo, C., V lchez,
S., Deheuvels, O., & Cerda, R. (2014). Cocoa{timber agroforestry systems: Theobroma
cacao{Cordia alliodora in Central America. Agroforestry Systems, 88(6), 1001{1019.
63. Somarriba, E., Trivelato, M., Villalobos, M., Su arez, A., Benavides, P., Moran, K.,
Orozco, L., & L opez, A. (2003). Diagn ostico agroforestal de peque~ nas ncas cacaoteras
org anicas de ind genas Bribri y Cab ecar de Talamanca, Costa Rica. Agroforester a en las
Am ericas, 10(37{38), 7.
64. Somarriba, E., Villalobos, M., Sucre, L., L opez, M., Torres, F., Torres, A., & Ro-
driguez, K. (2008). El proyecto Captura de carbono y desarrollo de mercados ambientales
209
en cacaotales y otros sistemas agroforestales ind genas en Talamanca, Costa Rica. Agro-
forester a en las Am erica, (46), 8{13.
65. Steiner, W. (2005). Dise~ no de un plan de desarrollo forestal en los territorios Bribrio
y Cabecar (p. 61) [Consultoria]. Talamanca, Costa Rica: CATIE.
66. Su arez Islas, A. (2001). Aprovechamiento sostenible de madera de Cordia al-
liodora y Cedrela odorata de regeneraci on natural en cacaotales y bananales de ind genas
de Talamanca, Costa Rica (PDF Document, CATIE). Accessed at: http://repositorio.
bibliotecaorton.catie.ac.cr:80/handle/11554/4156
67. Suatunce, P., Somarriba, E., Harvey, C., & Finegan, B. (2003). Composici on
or stica
y estructura de bosques y cacaotales en los Territorios Ind genas de Talamanca, Costa Rica.
Agroforester a en las Am ericas, 10(37{38), 6.
68. Talavera, E. (2006, February). Intercambios de experiencias de agricultores indigenas
productores de cacao.
69. Ulate, R. (2004, June 3). Carta de entendimiento MINAE-CATIE para la im-
plementaci on del proyecto \Carbon capture and development of environmental markets in
Indigenous cocoa agroforestry systems in Costa Rica" Grant TF052118 [Word document].
70. Uma~ na, L. (2005). Consultoria, para el establecimeinto de 10 parcelas demostrativas
de cacao con dise~ nos agroforestales mejorados, capacitaci on de 25 promotores y la producci on
de 5200 plantas de cacao injertadas (p. 7) [Consultoria]. Talamanca, Costa Rica: CATIE.
71. Vargas, M. (2006). Sistematizaci on y Evaluaci on Participativa de la Experiencia de
ACOMUITA en el Proyecto Estudio Captura de Carbono y desarrollo de Mercados Am-
bientales en Sistemas Agroforestales Ind genas con cacao en Costa Rica [Sistematizaci on].
Talamanca, Costa Rica.
72. Villalobos, M. (2005). Calendario 2005: Captura de carbono y desarrollo de mercados
ambientales en cacaotales y otros sistemas agroforestales ind genas en Talamanca, Costa
Rica. Accessed at: http://orton.catie.ac.cr/repdoc/A0863e/A0863e.pdf
73. Villalobos, M., & Acu~ na, K. (2007, May). Una experiencia de manejo colaborativo
210
de proyectos en los territorios ind genas Bribri y Cab ecar de Talamanca, Costa Rica: El caso
del Proyecto Captura de Carbono.
74. Villalobos, M., & Acu~ na, K. (2009). Una experiencia de manejo ambiental en cacao
en los territorios ind genas Bribri y Cab ecar de Talamanca-Costa Rica: El caso del Proyecto
Captura de Carbono. Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr/bitstream/
handle/11554/7746/103.pdf?sequence=2&isAllowed=y
75. Villalobos, M., & Soto, C. (2006). Calendario 2006: Captura de carbono en cacaotales
y otros sistemas agroforestales ind genas y desarrollo de mercados ambientales en Talamanca,
Costa Rica. Accessed at: http://orton.catie.ac.cr/repdoc/A0864e/A0864e.pdf
76. Whelan, M. P. (2005). Reading the Talamanca landscape: Land use and livelihoods
in the Bribri and Cab ecar indigenous territories (Mag ster Scientiae in Tropical Agroecology,
CATIE). Accessed at: http://repositorio.bibliotecaorton.catie.ac.cr:80/handle/11554/3743
77. World Bank. (2003). Japan Trust Funds Proposal|Special Program FY03.
78. Ying, S. (2006). Plan empresarial ACOMUITA.
79. Zamora, Ing. H. V. (2006, March). Elaboraci on de mapas de parcelas demostrativas.
B.2.2 Interviews and Meetings
Please note that interviewees are protected by my IRB protocol, thus limited information
has been given about each individual to avoid identication.
TOTAL: 22
1. Implementers: 2 (from CATIE)
2: Beneciaries: 20 (from ACOMUITA, ADITICA, ADITIBRI, and other involved com-
munity members)
211
B.3 Costa Rica: AC3
B.3.1 Documents and Academic Articles
1. Alp zar, F. (2015, September). Eciencia en el uso de agua y energ a en hogares:
Instrumentos de pol tica econ omica. Presented at the EfD Annual Policy Interaction Work-
shop, Heredia, Costa Rica. Accessed at: http://www.efdinitiative.org/sites/default/les/
2015 09 11 policy day fa 1.pdf
2. Alpizar, F., Bernedo, M., & Ferraro, P. (2016, June 23). Household time and risk
preferences and adaptation to climate. Presented at the European Association of Environ-
mental and Resource Economists 22nd Annual Conference, Zurich, Switzerland. Accessed
at: https://web.archive.org/web/20170929035056/http://www.webmeets.com:80/EAERE/
2016/prog/viewpaper.asp?pid=722
3. Carpio, M. N. B. D. (2016). Essays on Behavioral Economics in Climate Change
Adaptation (Ph.D.). Georgia State University.
4. CATIE. (2012). T tulo de la propuesta: Agua para consumo humano, comunidades y
cambio clim atico: Impactos esperados y adaptaci on en Am erica Central (p. 26) [Propuesta
de Investigaci on]. Turrialba, Costa Rica: IDRC.
5. CATIE. (2016). Informe Final: Adaptaci on al cambio clim atico de las organizaciones
rurales proveedoras de agua en Am erica Central [Informe T ecnico Final]. Turrialba, Costa
Rica: IDRC.
6. CATIE - AC3. (2016, March 4). Home|AC3. Accessed December 5, 2018, via Agua
para consumo humano, comunidades y cambio clim atico website at: https://web.archive.
org/web/20160304175533/http://aguaycomunidades.org/
7. DICYT. (2016, October 27). Am erica Central enfrentar a problemas de disponi-
bilidad de agua debido al cambio clim atico. Accessed December 6, 2018, via Sci-
ence Guatemala website at: http://www.dicyt.com/noticias/america-central-enfrentara-
problemas-de-disponibilidad-de-agua-debido-al-cambio-climatico
212
8. Dobbin, K. B., & Sarathy, B. (2015). Solving Rural Water Exclusion: Challenges and
Limits to Co-Management in Costa Rica. Society & Natural Resources, 28(4), 388{404.
9. EfD. (2013, July 18). AC3 Project Water for human consumption, com-
munities and climate change [Text]. Accessed May 20, 2019, via the EfD - Ini-
tiative website at: http://efdinitiative.org//our-work/projects/ac3-project-water-human-
consumption-communities-and-climate-change-expected
10. Gentes, I., & Madrigal, R. (2010). Sostenibilidad para los acueductos comunales en
Costa Rica: Desaf os pendientes en la gobernabilidad h drica. VertigO, (Hors s erie 7).
11. Hannah, L., Donatti, C. I., Harvey, C. A., Alfaro, E., Rodriguez, D. A., Bouroncle,
C., Castellanos, E., Diaz, F., Fung, E., Hidalgo, H. G., Imbach, P., L aderach, P., Landrum,
J. P., & Solano, A. L. (2017). Regional modeling of climate change impacts on smallholder
agriculture and ecosystems in Central America. Climatic Change, 141(1), 29{45.
12. Imbach, P. A., Georgiou, S., Calderer, L., Coto, A., Nakaegawa, T., Chou, S. C., Lyra,
A. A., Hidalgo, H. G., & Ciais, P. (2016). Hydrological Dynamics of Central America: Time-
of-Emergence of the Global Warming Signal. AGU Fall Meeting Abstracts, 43. Accessed at:
http://adsabs.harvard.edu/abs/2016AGUFM.H43K1632I
13. Imbach, P., Chou, S. C., Lyra, A., Rodrigues, D., Rodriguez, D., Latinovic, D.,
Siqueira, G., Silva, A., Garofolo, L., & Georgiou, S. (2018). Future climate change scenarios
in Central America at high spatial resolution. PLOS ONE, 13(4).
14. Kuzdas, C., Warner, B. P., Wiek, A., Vignola, R., Yglesias, M., & Childers, D. L.
(2016). Sustainability assessment of water governance alternatives: The case of Guanacaste
Costa Rica. Sustainability Science, 11(2), 231{247.
15. Kuzdas, C., & Wiek, A. (2014). Governance scenarios for addressing water con
icts
and climate change impacts. Environmental Science & Policy, 42, 181{196.
16. Kuzdas, C., Wiek, A., Warner, B., Vignola, R., & Morataya, R. (2014). Sus-
tainability Appraisal of Water Governance Regimes: The Case of Guanacaste, Costa Rica.
Environmental Management, 54(2), 205{222.
213
17. Kuzdas, C., Wiek, A., Warner, B., Vignola, R., & Morataya, R. (2015). Integrated
and Participatory Analysis of Water Governance Regimes: The Case of the Costa Rican Dry
Tropics. World Development, 66, 254{268.
18. LMA-CATIE. (2017, February). Mesoamerica/Mapas/Apps. Accessed December 5,
2018, at: http://mesomapps.info/
19. Lyra, A., Imbach, P., Rodriguez, D., Chou, S. C., Georgiou, S., & Garofolo, L.
(2017). Projections of climate change impacts on central America tropical rainforest. Cli-
matic Change, 141(1), 93{105.
20. Madrigal, R oger. (2009). Rendici on de cuentas y participaci on co-
munitaria en la provisi on de agua potable [Policy Brief]. Accessed via CATIE
website at: https://www.researchgate.net/publication/315447767 Rendicion de cuentas y
participacion comunitaria en la provision de agua potable
21. Madrigal, Roger. (2014, September). Agua para consumo humano, co-
munidades y cambio clim atico. PDF document presented at the V Encuentro Lati-
noamericano de Gesti on Comunitaria del Agua, CTEC, San Carlos, Costa Rica. Ac-
cessed at: http://www.efdinitiative.org/sites/default/les/presentacion roger madrigal ac3-
v encuentro clocsas san carlos 2014 0.pdf
22. Madrigal, R., Alp zar, F., Imbach, P., Le on, E., Moraes, C., Capit an, T., Fer-
raro, P., & Bernedo, M. (2015, September 16). Acueductos comunales y la adaptaci on al
cambio clim atico. PDF document presented at the EfD Annual Policy Interaction Work-
shop, Heredia, Costa Rica. Accessed at: http://www.efdinitiative.org/sites/default/les/
ac3 ppt setiembre 2015 0.pdf
23. Madrigal, R oger, Alp zar, F., & Schl uter, A. (2011a). Determinants of Performance
of Community-Based Drinking Water Organizations. World Development, 39(9), 1663{1675.
24. Madrigal, R oger, Alp zar, F., & Schl uter, A. (2011b, January 10). Individual Percep-
tions and the Performance of Community Based Drinking Water Organizations. Presented
at the Sustaining Commons: Sustaining Our Future, the Thirteenth Biennial Conference of
214
the International Association for the Study of the Commons, Hyderabad, India. Accessed
at: http://dlc.dlib.indiana.edu/dlc/handle/10535/7345
25. Madrigal, R., & Capit an, T. (2014). Household and community-level adaptation to
droughts in rural water systems of Costa Rica [EfD Research Brief]. CATIE.
26. Madrigal-Ballestero, Roger. (2011). Estimating the Cost and Benets of Supplying
Hydrological Ecosystem Services: An Application for Small-Scale Rural Drinking Water
Organizations. In F. DeClerk, J. F. L. Coq, B. Rapidel, & J. Beer (Eds.), Ecosystem
Services from Agriculture and Agroforestry: Measurement and Payment (pp. 161{177). US
& UK: Earthscan from Routledge.
27. Madrigal-Ballestero, R oger, Alp zar, F., & Schl uter, A. (2013). Public perceptions
of the performance of community-based drinking water organizations in Costa Rica. Water
Resources and Rural Development, 1{2, 43{56.
28. Madrigal-Ballestero, R oger, & Naranjo, M. A. (2015). Adaptive capacity, drought
and the performance of community-based drinking water organizations in Costa Rica. Jour-
nal of Water and Climate Change, 6(4), 831{847.
29. Rivera, J. E. (2016). Technical Evaluation of IDRC's Project: Water, Communities,
and Climate Change in Central America [Technical Evaluation]. Washington, D.C: IDRC.
30. Robalino, J., Jimenez, J., & Chac on, A. (2015). The Eect of Hydro-Meteorological
Emergencies on Internal Migration. World Development, 67, 438{448.
B.3.2 Media
1. Astorga, Yamileth. (2014, October 3). AC3 Project interview with Yamileth Astorga,
CEO of the Costa Rican Institute of Aqueducts and Sewers (AYA). (Efdinitiative, Inter-
viewer) [EfD News]. Accessed at: http://www.efdinitiative.org/news/archive/ac3-project-
interview-yamileth-astorga-ceo-costa-rican-institute-aqueducts-and-sewers
2. Backteman, K. (2015, April 9). Rural people must deal with threats to drinking water.
EfD - Initiative. Accessed at: http://www.efdinitiative.org/our-work/policy-interactions/
215
rural-people-must-deal-threats-drinking-water
3. EfD. (2013, April 23). Field visits kicko for AC3 project on water and climate change
adaptation. EfD - Initiative. Accessed at: http://www.efdinitiative.org/news/archive/eld-
visits-kicko-ac3-project-water-and-climate-change-adaptation
4. EfD. (2014, September 24). EfD Fellow R oger Madrigal presents at the 5th
Latin American Meeting of Community Water Management. EfD - Initiative. Ac-
cessed at: http://www.efdinitiative.org/news/archive/ac3-project-involvement-5th-latin-
american-meeting-community-water-management
5. EfD. (2015a). Water, communities and climate change: AC3. Accessed at: https:
//www.youtube.com/watch?v=YCFRoJllO1Q
6. EfD. (2015b, March 11). EfD-CA senior researcher Roger Madrigal participates
in IWRA World Water Congress 2015 and the ICARUS IV Global Meeting. EfD - Ini-
tiative. Accessed at: http://www.efdinitiative.org/news/archive/efd-ca-senior-researcher-
roger-madrigal-participates-iwra-world-water-congress-2015-and
7. EfD. (2015c, May 22). IDRC brings researchers to the World Water Congress. EfD -
Initiative. Accessed at: http://www.efdinitiative.org/news/archive/idrc-brings-researchers-
world-water-congress
8. EfD. (2015d, August 24). EfD Central America Projects AC3 and CAS-
CADE are in the top 500 best projects of Latin America. EfD - Initiative.
Accessed at: http://www.efdinitiative.org/news/archive/efd-central-america-projects-ac3-
and-cascade-are-top-500-best-projects-latin-america
9. EfD. (2015e, September 21). EfD Central America hosted the EfD Annual Policy
Interaction Workshop. EfD - Initiative. Accessed at: http://www.efdinitiative.org/news/
archive/efd-central-america-hosted-efd-annual-policy-interaction-workshop
10. EfD. (2016a). Impactos del cambio clim atico y disponibilidad de agua- Proyecto
AC3-CATIE. Accessed at: https://www.youtube.com/watch?v=F3Tkke7x2TM
11. EfD. (2016b, July 15). EfD-CA researchers Francisco Alpizar and
216
Laura Villalobos presented @ EAERE 2016. EfD - Initiative. Accessed at:
http://www.efdinitiative.org/news/archive/efd-ca-researchers-francisco-alpizar-and-laura-
villalobos-presented-eaere-2016
12. EfD. (2016c, August 30). EfD Central America is present during World Wa-
ter Week. EfD - Initiative. Accessed at: http://www.efdinitiative.org/news/archive/efd-
central-america-present-during-world-water-week
13. Madrigal, R. (2012, June 8). Managing Rural Water Systems: Keys to Success.
Resources, Spring/Summer(180), 20.
14. Soto, M. (2013, October 1). Cambio clim atico: Centroam erica ten-
der a a ser m as `caliente' hacia el 2025. La Naci on, Grupo Naci on. Accessed
at: https://www.nacion.com/ciencia/medio-ambiente/cambio-climatico-centroamerica-
tendera-a-ser-mas-caliente-hacia-el-2025/4B5W5UAMTZHGPPRJLXJB2I5ZBE/story/
B.3.3 Interviews and Meetings
Please note that interviewees are protected by my IRB protocol, thus limited information
has been given about each individual to avoid identication.
TOTAL: 65
1. Implementers: 3 (from CATIE)
2. Beneciaries in Community A: 7 (from local ASADA)
3. Household Interviews in Community A: 11
2. Beneciaries in Community B: 2 (from local ASADA)
3. Household Interviews in Community B: 42
217
B.4 Mauritius: SWMP
B.4.1 Documents and Academic Articles
1. Chen, S., & Huus-Hansen, I. (2017). Community-Based Chemicals and Waste Manage-
ment: Experiences from the GEF Small Grants Programme. United Nations Development
Programme.
2. GEF Small Grants Programme. (2016). Case Study on Sustainable Waste Manage-
ment Practices.
3. Imrit, C. (2018). CO2 Emissions Reduction in Community Based Projects in Mauritius
[BEng (Honors) Chemical and Environmental Engineering]. University of Mauritius.
4. Ministry of Environment, Sustainable Development, and Disas-
ter & Beach Management, Mauritius Prison Service, & Indian Ocean Cen-
tre for Education in Human Values. (2016). Report on Earth Day 2016.
http://environment.govmu.org/english//DOCUMENTS/REPORT%20EARTH%20DAY%202016.PDF
5. Somaroo, G., & Koodoruth, I. (2016). Country Programme Strategy 2015-
2018 Republic of Mauritius (Operational Phase 6) (Country Programme Strategy Op-
erational Phase 6; p. 54). Global Environmental Facility Small Grants Pro-
gramme. https://sgp.undp.org/all-documents/country-documents/600-op6-sgp-mauritius-
country-programme-strategy/le.html
6. Vieillesse, M. (2016). Final Report: Sustainable Waste Management Practices at Petit
Verger Prison.
B.4.2 Media
1. ACTogether.mu. (n.d.). Kinou et e Association.
2. Bruneau, J. P. (n.d.). United Nations Public Service Awards Case Studies: Sustainable
Waste Management Practices at Petit Verger Prison. Knowledge Base of UN Public Service
Awards Initiatives.
218
3. GEF. (2016, June 5). Less Waste in the Landlls: Compost Production in Mauritius.
Global Environment Facility.
4. GEF Small Grants Programme. (n.d.). Sustainable Waste Management Practices at
Petit Verger Prison.
5. GEF Small Grants Programme. (2015, July 28). SGP Mauritius|Public Service
Excellence Award 2013. https://www.youtube.com/watch?v=OJHisJTXzxE
6. Government Information Service, Prime Minister's Oce. (2016, April 23). Journ ee
mondiale de la Terre 2016: Mise en terre de 200
amboyants par des d etenus et le personnel
p enitentiaire. Republic of Mauritius, 17.
7. Le Mauricien. (2016, February 27). GESTION DES D

ECHETS:
La prison de Petit Verger fait acquisition d'un broyeur de bois. Actualit es.
https://www.lemauricien.com/article/gestion-des-dechets-la-prison-petit-verger-fait-
acquisition-d-broyeur-bois/
8. Maurice Ile Durable. (n.d.-a). About MID.
9. Maurice Ile Durable. (n.d.-b). MID Consultative Process.
10. Mauritius Prison Service. (n.d.). Green Waste Composting at PVP.
11. Mauritius Prison Service. (2014, December 17). Mauritius Prison Ser-
vice|Sustainable Waste Management Practices at Petit Verger Prison.
12. Mauritius Prison Service. (2018, March 1). Green Waste Composting Video.
http://prisons.govmu.org/English/Media-Library/Pages/default.aspx#
13. University of Mauritius. (n.d.). Faculty of Engineering CEE - Research areas &
expertise in the Department.
B.4.3 Interviews and Meetings
Please note that interviewees are protected by my IRB protocol, thus limited information
has been given about each individual to avoid identication.
TOTAL: 37
219
1. Implementers: 3 (from Association Kinoute, UNDP, and ATICS Ltd.)
2. Household Interviews in Community: 34
220
Appendix C
Tool to Measure Institutionalization of Project
Outcomes
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
Appendix D
Tool to Measure Incentive Types
243
244
245
Bibliography
Abtew, M. S. (2017). Sustainable Development Goals, Governance, and the Private Sector. In
Management Challenges in Dierent Types of African Firms, Frontiers in African Business
Research, pages 73{93. Springer, Singapore.
Acemoglu, D. and Robinson, J. A. (2006). Economic Origins of Dictatorship and Democracy.
Cambridge University Press, Cambridge ; New York.
ACTogether.mu. Kinou et e Association.
AidData (2019). Project Level Aid Dashboard.
Alam, M. M. and Wagner, C. (2016). The Relative Importance of Monetary and Non-
Monetary Drivers for Information and Communication Technology Acceptance in Rural
Agribusiness. Information Technology for Development, 22(4):654{671.
Alchian, A. A. and Demsetz, H. (1972). Production, Information Costs, and Economic
Organization. The American Economic Review, 62(5):777{795.
Alkire, S. (2005). Capability and Functionings: Denition & Justication. HDCA Introduc-
tory Brieng Note., Human Development and Capability Association.
Allcott, H. (2015). Site Selection Bias in Program Evaluation. The Quarterly Journal of
Economics, 130(3):1117{1165.
Araya, J. (2015). Costa Rica es el consumidor m as voraz de plaguicidas en el mundo.
Semanario Universidad.
Asian Development Bank (2004). Eectiveness of Participatory Approaches: Do the New
Approaches Oer an Eective Solution to the Conventional Problems in Rural Develop-
ment Projects? Special Evaluation Study Special Evaluation Study SST: REG 2005-01
(Final), Asian Development Bank.
Barnett, M. N. and Finnemore, M. (2004). Rules for the World: International Organizations
in Global Politics. Cornell University Press, Ithaca, N.Y.
Bebbington, A. (2004). NGOs and uneven development: Geographies of development inter-
vention. Progress in Human Geography, 28(6):725{745.
Beck, H. E., Zimmermann, N. E., McVicar, T. R., Vergopolan, N., Berg, A., and Wood, E. F.
(2018). Present and future K oppen-Geiger climate classication maps at 1-km resolution.
Scientic Data, 5.
246
Bell, C. and Squire, L. (2016). Providing Policy Makers with Timely Advice: The Timeliness-
Rigor Trade-o. The World Bank Economic Review, pages 1{23.
Bell, S. A. and Aggleton, P. (2012). Integrating Ethnographic Principles in NGO Mon-
itoring and Impact Evaluations: NGO Monitoring and Impact Evaluation. Journal of
International Development, 24(6):795{807.
Bermeo, S. B. (2007). Foreign Aid, Foreign Policy, and Development: Sector Allocation in
Bilateral Aid. Paper Presented at the International Studies Association Annual Conven-
tion. In International Studies Association Annual Convention, Chicago, IL.
Bermeo, S. B. (2016). Aid Is Not Oil: Donor Utility, Heterogeneous Aid, and the Aid-
Democratization Relationship. International Organization, 70(01):1{32.
Bickers, K. N. and Williams, J. T. (2001). Public Policy Analysis: A Political Economy
Approach. Houghton Miin, Boston ; New York.
Binnendijk, A. (2000). Results Based Management in the Development Co-Operation Agen-
cies: A Review of Experience. Background Report, Development Assistance Committee
(DAC) Working Party on Aid Evaluation, OECD.
Blasiak, R. (2011). Ethics and Environmentalism: Costa Rica's Lesson. United Nations
University OurWorld 2.0.
Booth, J. A. (1987). Costa Rican Democracy. World Aairs, 150(1):43{53.
Bourguignon, F. and Sundberg, M. (2007). Aid Eectiveness: Opening the Black Box. The
American Economic Review, 97(2):316{321.
Bowles, S. (2008). Policies Designed for Self-Interested Citizens May Undermine "The Moral
Sentiments": Evidence from Economic Experiments. Science, 320(5883):1605{1609.
Bradburn, N. M., Rips, L. J., and Shevell, S. K. (1987). Answering autobiographical ques-
tions: The impact of memory and inference on surveys. Science, 236(4798):157{161.
Brautigam, D. (1997). Institutions, Economic Reform, and Democratic Consolidation in
Mauritius. Comparative Politics, 30(1):45{62.
Brown, A., Foster, M., Norton, A., and Naschold, F. (2001). The status of sector wide
approaches. Technical report, Overseas Development Institute, London.
Brown, S. (2005). Foreign Aid and Democracy Promotion: Lessons from Africa. The Euro-
pean Journal of Development Research, 17(2):179{198.
Brysk, A. (2005). Global Good Samaritans? Human Rights Foreign Policy in Costa
Rica. Global Governance: A Review of Multilateralism and International Organizations,
11(4):445{466.
Brysk, A. (2009). Global Good Samaritans: Human Rights as Foreign Oolicy. Oxford
University Press, Oxford ; New York.
247
Bucheli, M. (2008). Multinational corporations, totalitarian regimes and economic national-
ism: United Fruit Company in Central America, 1899{1975. Business History, 50(4):433{
454.
Bunge, M. (1996). Finding Philosophy in Social Science. Yale University Press, New Haven,
CT.
Burnside, C. and Dollar, D. (2000). Aid, Policies, and Growth. The American Economic
Review, 90(4):847{868.
Cameron, D. B., Mishra, A., and Brown, A. N. (2016). The growth of impact evaluation
for international development: How much have we learned? Journal of Development
Eectiveness, 8(1):1{21.
Cardenas, J. C. and Carpenter, J. (2008). Behavioural Development Economics: Lessons
from Field Labs in the Developing World. The Journal of Development Studies, 44(3):311{
338.
Cardoso, F. H. and Faletto, E. (1979). Dependency and development in Latin America.
University of California Press, Berkeley.
Carlsson, J., Kohlin, G., and Ekbom, A. (1994). The Political Economy of Evaluation -
International Aid Agencies and the Eectiveness of Aid | Jerker Carlsson | Palgrave
Macmillan. St. Martin's Press, New York, NY.
CATIE (2006). Captura de carbono y desarrollo de mercados ambientales en cacaotales y
otros sistemas agroforestales ind genas en Talamanca. Technical report, CATIE, Turrialba,
Costa Rica.
Chambers, R. and Conway, G. (1992). Sustainable Rural Livelihoods: Practical Concepts
for the 21st Century. Number 296 in DP. Institute of Development Studies, Brighton,
England.
CIDA (2000). RBM Handbook on Developing Results Chains: The Basics of RBM as
Applied to 100 Project Examples. Technical report, Results-Based Management Division,
Canadian International Development Agency.
CIDA (2001). Participatory Evaluations. Technical Report No. 3, Canadian International
Development Agency, Gatineau, Quebec, Canada.
Clemens, M. A. (2012). The Collision of Development Goals and Impact Evaluation. Revue
d' economie du d eveloppement, Vol. 20(4):175{205.
Collier, P. (2007). The Bottom Billion: Why the Poorest Countries Are Failing and What
Can Be Done About It. Oxford University Press, Oxford ; New York.
Colyvas, J. A. and Powell, W. W. (2006). Roads to Institutionalization: The Remaking
of Boundaries between Public and Private Science. Research in Organizational Behavior,
27:305{353.
248
Comeliau, C. (2000). The Limitless Growth Assumption. International Social Science Jour-
nal, 52(166):457{465.
Cordella, T. and Dell'Ariccia, G. (2007). Budget Support Versus Project Aid: A Theoretical
Appraisal. The Economic Journal, 117(523):1260{1279.
Cordero, J. and Paus, E. (2012). Foreign Investment and Economic Development in Costa
Rica: The Unrealized Potential. In Gallagher, K. P., Chudnovsky, D., and Ocampo,
J. A., editors, Rethinking Foreign Investment for Sustainable Development, pages 97{124.
Anthem Press, London.
Crawford, P. (2004). Synthetic Indices: A Method for Evaluating Aid Project Eectiveness.
Evaluation, 10(2):175{192.
Crosswell, M. (2015). Country Selectivity and Extreme Poverty. Technical report, USAID.
Curry, B. K. (1992). Instituting Enduring Innovations. Achieving Continuity of Change
in Higher Education. ASHE-ERIC Higher Education Report No. 7, 1992. Publications
Department, ASHE-ERIC Higher Education Reports, The George Washington University,
Washington, D.C.
Dahl, R. A. (1957). The concept of power. Behavioral Science, 2(3):201{215.
Dalkir, K. (2007). Knowledge Management in Theory and Practice. Elsevier Butterworth-
Heinemann, Amsterdam, 3rd edition.
Darrington, J. W. and Howell, G. A. (2011). Motivation and incentives in relational contracts.
Journal of Financial Management of Property and Construction; Bingley, 16(1):42{51.
Davies, R. (2004). Scale, Complexity and the Representation of Theories of Change. Evalu-
ation, 10(1):101{121.
Deaton, A. and Cartwright, N. (2016). The limitations of randomised controlled trials.
Deaton, A. and Cartwright, N. (2018). Understanding and misunderstanding randomized
controlled trials. Social Science & Medicine, 210:2{21.
Deneulin, S. (2005). Development as Freedom and the Costa Rican Human Development
Story. Oxford Development Studies, 33(3-4):493{510.
DFID (2014). DFID Evaluation Strategy: June 2014 to 2019. Technical report, Department
for International Development.
Diallo, A. and Thuillier, D. (2005). The success of international development projects, trust
and communication: An African perspective. International Journal of Project Manage-
ment, 23(3):237{252.
Dietrich, S. (2016). Donor Political Economies and the Pursuit of Aid Eectiveness. Inter-
national Organization, 70(01):65{102.
249
Easterly, W., Levine, R., and Roodman, D. (2004). Aid, Policies, and Growth: Comment.
American Economic Review, 94(3):774{780.
Easterly, W. R. (2006). The White Man's Burden: Why the West's Eorts to Aid the Rest
Have Done So Much Ill and So Little Good. Penguin Press, New York, NY.
EfD (2013). AC3 Project Water for human consumption, communities and cli-
mate change. http://efdinitiative.org//our-work/projects/ac3-project-water-human-
consumption-communities-and-climate-change-expected.
Elster, J. (2010). Explaining Social Behavior: More Nuts and Bolts for the Social Sciences.
Cambridge Univ. Press, Cambridge, reprinted edition.
Escobar, A. (2012). Encountering Development: The Making and Unmaking of the Third
World. Princeton University Press, Princeton, N.J.
Espina, C. and Zinnes, C. (2003). Institutional Incentives Within USAID: How do They
Aect Projects Outcomes? NIE-Based Toolkits for USAID Applications, Session 2 Forum
5, Center for Institutional Reform and the Informal Sector (IRIS), Washington, D.C.
Ethier, D. (2003). Is Democracy Promotion Eective? Comparing Conditionality and In-
centives. Democratization, 10(1):99{120.
Evans, P. B. (1979). Dependent Development: The Alliance of Multinational, State, and
Local Capital in Brazil. Princeton University Press, Princeton, N.J.
Ferreira, G. F. and Harrison, R. W. (2012). From Coee Beans to Microchips: Export
diversication and economic growth in Costa Rica. Journal of Agricultural and Applied
Economics, 44(4):517{531.
Finnemore, M. and Sikkink, K. (1998). International Norm Dynamics and Political Change.
International Organization, 52(4):887{917.
Fioramonti, L. (2013). Gross Domestic Problem: The Politics behind the World's Most
Powerful Number. Economic Controversies. Zed Books, London ; New York.
Fioramonti, L. (2017). The World after GDP: Politics, Business and Society in the Post
Growth Era. Polity, Malden, MA, 1st edition.
Flora, C. B., Flora, J. L., and Fey, S. (2004). Rural Communities: Legacy and Change.
Westview Press, Boulder, Colo., second edition.
Flora, C. B., Flora, J. L., and Gasteyer, S. P. (2016). Rural Communities: Legacy and
Change. Westview Press, Boulder, Colo, fth edition.
Fox, J. (1998). Real Progress: Fifty Years Of USAID in Costa Rica. USAID Program and
Operations Assessment Report 23, USAID.
Fox, J. W. (2003). Successful integration into the global economy: Costa Rica and Mauritius.
Nathan Associates Inc., TCB Project, available at www. tcb-project. com.
250
Frediani, A. A. (2006). Participatory Methods and the Capability Approach. Technical
report, Human Development and Capability Association.
Freedom House (2019a). Costa Rica.
Freedom House (2019b). Mauritius.
Friedman, M. (1962). Capitalism and Freedom. University of Chicago Press, Chicago.
Friedmann, J. (1998). Empowerment: The Politics of Alternative Development. Blackwell,
Cambridge, Mass., reprint edition.
Fukuda-Parr, S. (2012). Recapturing the Narrative of International Development. Tech-
nical Report Research Paper No. 2012{5, United Nations Research Institute for Social
Development (UNRISD).
Fukuda-Parr, S. and Hulme, D. (2011). International Norm Dynamics and the `End
of Poverty': Understanding the Millennium Development Goals. Global Governance,
17(1):17{36.
Fullan, M. (2007). The New Meaning of Educational Change. Teachers College Press, New
York, fourth edition.
GEF (2016). Less Waste in the Landlls: Compost Production in Mauritius.
GEF Small Grants Programme (2015). SGP Mauritius - Public Service Excellence Award
2013.
GEF Small Grants Programme (n.d.). Sustainable Waste Management Practices at Petit
Verger Prison.
Gerring, J. (2007). Case Study Research: Principles and Practices. Cambridge University
Press, New York.
Gibson, C. C., Andersson, K., Ostrom, E., and Shivakumar, S. (2005). The Samaritan's
Dilemma: The Political Economy of Development Aid. Oxford University Press, New
York.
Gin e, X. and Karlan, D. S. (2014). Group versus individual liability: Short and long term
evidence from Philippine microcredit lending groups. Journal of Development Economics,
107:65{83.
Goto, J., Aida, T., Aoyagi, K., and Sawada, Y. (2011). Incentives and Social Preferences in a
Traditional Labor Contract: Evidence from Rice Planting Experiments in the Philippines.
Journal of Behavioral Economics and Finance, 4:94{96.
GPEDC (2011). The Monitoring Framework of the Global Partnership for Eective Develop-
ment Co-operation. Technical report, Global Partnership for Eective Development Co-
operation. Accessed November 29, 2015: http://eectivecooperation.org/wordpress/wp-
content/uploads/2015/05/GPEDC-Monitoring-Framework-10-Indicators.pdf.
251
GPEDC (2014). Making Development Co-operation More Eective: 2014 Progress
Report. Technical report, Global Partnership for Eective Development Co-
operation. Accessed November 29, 2015: http://www.keepeek.com/Digital-
Asset-Management/oecd/development/making-development-co-operation-more-
eective 9789264209305-en#page79.
Grimes, D. A. and Schulz, K. F. (2002). Bias and causal associations in observational
research. The Lancet, 359(9302):248{252.
Gunder Frank, A. (1967). Capitalism and Underdevelopment in Latin America: Historical
Studies of Chile and Brasil. Monthly Review, New York; London, two thousand, ninth
edition.
Gutierrez-Montes, I., Emery, M., and Fernandez-Baca, E. (2009). The Sustainable Liveli-
hoods Approach and the Community Capitals Framework: The Importance of System-
Level Approaches to Community Change Eorts. Community Development, 40(2):106{
113.
Hardin, G. (1968). The Tragedy of the Commons. Science, 162(3859):1243{1248.
Hawkins, D. G. and Jacoby, W. (2006). How agents matter. In Hawkins, D. G., Lake,
D. A., Nielson, D. L., and Tierney, M. J., editors, Delegation and Agency in International
Organizations, pages 199{228. Cambridge University Press, Cambridge.
Hayek, F. A. (1960). The Constitution of Liberty. University of Chicago Press, Chicago.
Hedstr om, P. (2005). Dissecting the Social: On the Principles of Analytical Sociology. Cam-
bridge University Press, Cambridge.
Heeks, R. and Stanforth, C. (2014). Understanding Development Project Implementation:
An Actor-Network Perspective. Public Administration and Development, 34(1):14{31.
Hirschman, A. O. (2015). Development Projects Observed. Brookings Institution Press,
Washington, D.C.
Hoey, L. (2015). \Show me the Numbers": Examining the Dynamics Between Evaluation
and Government Performance in Developing Countries. World Development, 70:1{12.
Holmstrom, B. (1982). Moral Hazard in Teams. The Bell Journal of Economics, 13(2):324{
340.
Honig, D. (2018). Navigation by Judgment: Why and When Top Down Management of
Foreign Aid Doesn't Work. Oxford University Press, Oxford, New York.
Honig, D. and Gulrajani, N. (2018). Making good on donors' desire to Do Development
Dierently. Third World Quarterly, 39(1):68{84.
Hoskin, K. (1996). The `awful idea of accountability': Inscribing people into the measurement
of objects. Accountability: Power, Ethos and the Technologies of Managing, 265.
252
Huck, S., K ubler, D., and Weibull, J. (2012). Social norms and economic incentives in rms.
Journal of Economic Behavior & Organization, 83(2):173{185.
Imbach, A. (2017). Personal Communication, September - October.
Imbach, A. C. (2014). Midterm Review (MTR) of the Fifth Operational Phase of the GEF
Small Grants Program in Costa Rica. Midterm Project Evaluation GEF PIMS: 4560 /
UNDP Project: 00079305, GEF and the UN Development Program, Costa Rica.
Imbach, A. C. (2015). Terminal Evaluation (TE) of the Fifth Operational Phase of the GEF
Small Grants Program in Costa Rica. Final Project Evaluation GEF PIMS: 4560 / UNDP
Project: 00079305, GEF and the UN Development Program, Costa Rica.
Imbach, A. C. (2016). Planicacion Estrategica Territorial Participativa, volume 1: Concep-
tos y metodolog a. Geolatina Ediciones, Turrialba, Costa Rica.
Independent Evaluation Group (2016). Learning from IDA Experience : Lessons from IEG
Evaluations. Technical report, The World Bank, Washington, D.C.
International Commission on Education for Sustainable Development Practice (2008). Final
Report from the International Commission on Education for Sustainable Development
Practice. Technical report, Earth Institute at Columbia University; John D. and Catherine
T. MacArthur Foundation, Brookyln, NY.
IUCN (1980). World Conservation Strategy: Living Resource Conservation for Sustainable
Development. International Union for Conservation of Nature and Natural Resources,
Gland.
Jain, S. (2007). Project Assistance versus Budget Support: An Incentive-Theoretic Analysis
of Aid Conditionality. Review of World Economics, 143(4):694{719.
James, P. (2017). Clarifying Causal Mechanisms in International Relations. In Oxford
Research Encyclopedia of Politics. Oxford University Press, Online.
Jerve, A. M. (2002). Ownership and Partnership: Does the New Rhetoric Solve the Incentive
Problems in Aid? Forum for Development Studies, 29(2):389{407.
Johnson, D., Goldstein-Plesser, G., and Zajonc, T. (2014). Can Foreign Aid Create an
Incentive for Good Governance? Evidence from the Millennium Challenge Corporation.
Technical report, Social Science Research Network.
Johnson, D. and Zajonc, T. (2006). Can Foreign Aid Create an Incentive for Good Gover-
nance? Evidence from the Millennium Challenge Corporation. SSRN Scholarly Paper ID
896293, Social Science Research Network, Rochester, NY.
Keohane, R. O. and Nye, J. S. (1973). Power and interdependence. Survival, 15(4):158{165.
Kothari, U. (2016). A Radical History of Development Studies: Individuals, Institutions and
Ideologies. Zed Books Ltd.
253
Kuvaas, B., Buch, R., Weibel, A., Dysvik, A., and Nerstad, C. G. L. (2017). Do intrinsic
and extrinsic motivation relate dierently to employee outcomes? Journal of Economic
Psychology, 61:244{258.
Lawson, M. L. (2013). Does Foreign Aid Work? Eorts to Evaluate U.S. Foreign Assistance.
CRS Report for Congress R42827, Congressional Research Service.
Lawson, M. L. (2016). Does Foreign Aid Work? Eorts to Evaluate U.S. Foreign Assistance.
CRS Report for Congress R42827, Congressional Research Service.
Le Mauricien (2016). GESTION DES D

ECHETS: La prison de Petit Verger fait acquisition
d'un broyeur de bois.
Levine, R. and Savedo, W. (2015a). Aid at the frontier: Building knowledge collectively.
Journal of Development Eectiveness, 7(3):275{289.
Levine, R. and Savedo, W. (2015b). The Future of Aid: Building Knowledge Collectively.
Technical Report CGD Policy Paper 050, Center for Global Development.
Lipset, S. M. (1959). Some Social Requisites of Democracy: Economic Development and
Political Legitimacy. American Political Science Review, 53(01):69{105.
List, J. A. and Rasul, I. (2011). Field Experiments in Labor Economics. In Handbook of
Labor Economics, volume 4, pages 103{228. Elsevier, Great Britain.
Mack, S. and McNeil, F. (2012). 70 Years of U.S. Development Assistance. Special Informa-
tive Section, USAID.
Mahoney, J. (2010). Colonialism and Postcolonial Development: Spanish America in Com-
parative Perspective. Cambridge University Press, New York.
Mansuri, G. and Rao, V. (2003). Evaluating Community-Based and Community-Driven
Development: A Critical Review of the Evidence. Technical report, Development Research
Group, The World Bank.
March, J. G. and Olsen, J. P. (2011). The Logic of Appropriateness. In The Oxford Handbook
of Political Science, Oxford Handbooks Online. Oxford University Press, Online.
Marshall, M. G., Gurr, T. R., and Jaggers, K. (2019). Polity IV Project: Dataset Users'
Manual. Technical report, Center for Systemic Peace.
Marshall, M. G. and Jaggers, K. (2011a). Polity IV Country Report 2010: Costa Rica.
Technical report, Center for Systemic Peace.
Marshall, M. G. and Jaggers, K. (2011b). Polity IV Country Report 2010: Mauritius.
Technical report, Center for Systemic Peace.
Martens, B., Mummert, U., Murrell, P., and Seabright, P., editors (2002). The Institutional
Economics of Foreign Aid. Cambridge University Press, New York.
254
Martinsson, J. (2011). Global Norms: Creation, Diusion, and Limits. Technical report,
The International Bank for Reconstruction and Development / The World Bank.
Matarrita-Cascante, D. (2010). Tourism Development in Costa Rica: History and Trends.
e-Review of Tourism Research, 8(6):136{156.
Maurice Ile Durable. About MID.
Mauritius Prison Service (2018). Green Waste Composting Video.
MCC (2011). Armenia Compact Monitoring and Evaluation Plan.
https://www.mcc.gov/resources/doc/armenia-compact-me-plan.
MCC (2012). Monitoring and Evaluation.
McCubbins, M. D. and Schwartz, T. (1984). Congressional Oversight Overlooked: Police
Patrols versus Fire Alarms. American Journal of Political Science, 28(1):165.
Meade, J. E. (2011). The Economic and Social Structure of Mauritius. Routledge, Oxon,
UK; New York.
Merton, R. K. (1973). The Sociology of Science: Theoretical and Empirical Investigations.
University of Chicago Press, Chicago ; London.
Mesa-Lago, C. (2003). Market, Socialist, and Mixed Economies: Comparative Policy and
Performance-Chile, Cuba ... Johns Hopkins Univ Press, Baltimore.
Miguel, E. and Kremer, M. (2004). Worms: Identifying Impacts on Education and Health
in the Presence of Treatment Externalities. Econometrica, 72(1):159{217.
Miles, M. (1987). Practical Guide for School Administrators: How to Get There. In Sym-
posium on Eective Schools Programs and the Urban High School: The Management of
Large-Scale Change, Annual Meeting of the American Educational Research Association,
Washington, DC. Center for Policy Research.
Miles, M. and Louis, K. (1987). Research on Instititonalization: A Re
ective Review. In
Miles, M., Ekholm, M., and Vandenberghe, R., editors, Lasting School Improvement:
Exploring Process of Instititonalization, pages 25{44. Acco, Leuven, Belgium.
Ministry of Foreign Aairs of Denmark (2014). Review of Budget Support Evaluations.
Technical report, Ministry of Foreign Aairs of Denmark.
Mitchell, L. E. (1998). Trust and Team Production in Post-Capitalist Society. Journal of
Corporation Law, 24:869.
Mo Ibrahim Foundation (2018). Ibrahim Index of African Governance (IIAG) Data Portal.
Moore, B. (1966). Social Origins of Dictatorship and Democracy: Lord and Peasant in the
Making of the Modern World. Beacon Press, Boston.
255
Morgan, T. C. (2013). The Secret Ingredient on Iron Chef|Road Kill! Con
ict Management
and Peace Science, 30(1):3{10.
Mosley, L., editor (2013). Interview Research in Political Science. Cornell University Press,
Ithaca.
Mosley, P. (1986). Aid-Eectiveness : The Micro-Macro Paradox. IDS Bulletin, 17(2):22{27.
Moyo, D. (2010). Dead Aid: Why Aid Is Not Working and How There Is a Better Way for
Africa. Farrar, Straus and Giroux, New York.
Naipaul, V. S. (1973). The Overcrowded Barracoon. Knopf, New York, 1st american ed.
edition.
North, D. C. (2009). Institutions, Institutional Change and Economic Performance. Political
Economy of Institutions and Decisions. Cambridge University Press, Cambridge.
Nugent, J. B. and Robinson, J. A. (2010). Are factor endowments fate? Revista de Historia
Economica - Journal of Iberian and Latin American Economic History, 28(1):45{82.
Nussbaum, M. C. (2000). Women and Human Development: The Capabilities Approach.
The John Robert Seeley Lectures. Cambridge University Press, Cambridge ; New York.
OECD (2004). Glossary of Key Terms in Evaluation and Results Based Management. Tech-
nical Report Reprinted in 2010, DAC Working Party on Aid Evaluation (WP-EV), OECD.
OECD (2006). Evaluation of General Budget Support: Synthesis Report. Technical report,
OECD.
OECD (2011). Aid Eectiveness 2005{10: Progress in Implementing the Paris Declaration.
Better Aid. OECD Publishing, Paris.
OECD (2013). The DAC Network on Development Evaluation - 30 Years of Strengthen-
ing Learning in Development. Technical report, OECD DAC Network on Development
Evaluation, Paris.
OECD (2015). Public Governance Reviews: Costa Rica Highlights. Technical report, OECD.
OECD (2016). Ocial development assistance (ODA) - Net ODA - OECD Data.
Onugha, I. (2000). Case 11: Costa Rica Medical Devices. Technical report, World Economic
Forum Global Agenda Council on the Future of Manufacturing.
Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective
Action. The Political Economy of Institutions and Decisions. Cambridge University Press,
Cambridge ; New York.
Ostrom, E. (2010). Overcoming the Samaritan's Dilemma in Development Aid.
256
Oviedo, A. M., Sanchez, S. M., Lindert, K. A., and Lopez, J. H. (2015). Costa Rica's
development: From good to better - systematic country diagnostic. Technical Report
97489, The World Bank.
Padayachy, R. (2010). Le D eveloppement

Economique de Maurice: Un Parcours Atypique.
The Mauritius Chamber of Commerce and Industry CCI Info, 46:14{22.
Paus, E. (2005). Foreign Investment, Development, and Globalization: Can Costa Rica
Become Ireland? Palgrave Macmillan, New York.
Pawson, R. and Tilley, N. (2010). Realistic Evaluation. SAGE, London, reprinted edition.
Pearce, J. (1998). From Civil War to `Civil Society': Has the End of the Cold War Brought
Peace to Central America? International Aairs, 74(3):587{615.
Poku, N. K. and Whitman, J. (2011). The Millennium Development Goals and Development
after 2015. Third World Quarterly, 32(1):181{198.
Polski, M. M. and Ostrom, E. (1999). An Institutional Framework for Policy Analysis and
Design. Technical Report W98-27, Workshop in Political Theory and Policy Analysis,
Indiana University.
Pritchett, L. (2002). It pays to be ignorant: A simple political economy of rigorous program
evaluation. The Journal of Policy Reform, 5(4):251{269.
Ramalingam, B. (2013). Aid on the Edge of Chaos: Rethinking International Cooperation
in a Complex World. Oxford University Press, Oxford, 1st edition.
Ramalingam, B. and Jones, H. (2008). Exploring the science of complexity: Ideas and
implications for development and humanitarian eorts. Technical Report 285, Overseas
Development Institute, London.
Rankin, M. A. (2012). The History of Costa Rica. The Greenwood Histories of the Modern
Nations. Greenwood, an imprint of ABC-CLIO, LLC, Santa Barbara, California.
Reinikka, R. and Smith, N. (2004). Public Expenditure Tracking Surveys in Education.
UNESCO : International Institute for Educational Planning, Paris.
Rocha de Siqueira, I. (2016). Development by Trial and Error: The Authority of Good
Enough Numbers. International Political Sociology, 0:1{19.
Rodrik, D. (1999a). The New Global Economic and Developing Countries: Making Openness
Worl. Policy Essay 24, Overseas Development Council, Washington, D.C.
Rodrik, D. (1999b). Where Did All the Growth Go? External Shocks, Social Con
ict, and
Growth Collapses. Journal of Economic Growth, 4(4):385{412.
Rodrik, D. (2012). Why We Learn Nothing from Regressing Economic Growth on Policies.
Seoul Journal of Economics, 25(2):137{151.
257
Rogl a, J. (2020). Leveraging Aid Transparency for Political Gain: Project Evaluations as
Donor Bargaining Chips in International Cooperation Negotiations.
Roodman, D. (2007a). The Anarchy of Numbers: Aid, Development, and Cross-Country
Empirics. The World Bank Economic Review, 21(2):255{277.
Roodman, D. (2007b). Macro Aid Eectiveness Research: A Guide for the Perplexed.
Working Paper 134, Center for Global Development, Washington, D.C.
Rosenau, J. N. (1990). Turbulence in World Politics: A Theory of Change and Continuity.
Princeton University Press, Princeton, N.J.
Rostow, W. W. (1959). The Stages of Economic Growth. The Economic History Review,
12(1):1{16.
Rottenberg, S., editor (1993). The Political Economy of Poverty, Equity, and Growth: Costa
Rica and Uruguay. Oxford University Press, published for the World Bank, New York and
Oxford.
Ruben, M. (2007). The promise of micronance for poverty relief in the developing world.
CSA LLC: Proquest.
Sabet, S. M. and Brown, A. N. (2018). Is impact evaluation still on the rise? The new trends
in 2010{2015. Journal of Development Eectiveness, 10(3):291{304.
Sachs, J. D. and Warner, A. M. (1997). Sources of Slow Growth in African Economies.
Journal of African Economies, 6(3):335{376.
Sappington, D. E. M. (1991). Incentives in Principal-Agent Relationships. Journal of Eco-
nomic Perspectives, 5(2):45{66.
Savedo, W., Perakis, R., and Schwanke, B. (2015). Shifting the Foreign Aid Paradigm -
Paying for Outcomes. Part of the White House and the World 2016 Brieng Book, Center
for Global Development.
Savedo, W. D., Levine, R., and Birdsall, N. (2006). When Will We Ever Learn? Improving
Lives Through Impact Evaluation. Report of the Evaluation Gap Working Group, Center
for Global Development, Washington, D.C.
Sawada, Y., Kasahara, R., Aoyagi, K., Shoji, M., and Ueyama, M. (2013). Modes of collective
action in village economies: Evidence from natural and artefactual eld experiments in a
developing country. Asian Development Review, 30(1):31{51.
Schuller, M. (2012). Killing with Kindness: Haiti, International Aid, and NGOs. Rutgers
University Press.
Scott, J. C. (1985). Weapons of the Weak: Everyday Forms of Peasant Resistance. Yale
University Press.
Seechurn, S. (2019). Personal Communication, October.
258
Segura Madrigal, M. (1999). Valoraci on del servicio de jaci on y almacenamiento de carbono
en bosques privados en el Area de Conservaci on Cordillera Volc anica Central, Costa Rica.
Magister Scientiae, CATIE, Turrialba, Costa Rica.
Sen, A. (1970). The Impossibility of a Paretian Liberal. Journal of Political Economy,
78(1):152{157.
Sen, A. (1999). Development as Freedom. Oxford University Press, New York, NY.
Sen, A. (2006). Man without a Plan - Can Foreign Aid Work, The Review Essay. Foreign
Aairs, 85:171{178.
Shafer, D. M. (1994). Winners and Losers: How Sectors Shape the Developmental Prospects
of States. Cornell Studies in Political Economy. Cornell University Press, Ithaca.
Shimomura, Y., Page, J., and Kato, H. (2016). Japan's Development Assistance: Foreign
Aid and the Post-2015 Agenda. Springer.
Smith, T. (1981). The Logic of Dependency Theory Revisited. International Organization,
35(4):755{761.
Solow, R. M. (1956). A Contribution to the Theory of Economic Growth. The Quarterly
Journal of Economics, 70(1):65{94.
Somaroo, G. and Koodoruth, I. (2016). Country Programme Strategy 2015-2018 Republic
of Mauritius (Operational Phase 6). Country Programme Strategy Operational Phase 6,
Global Environmental Facility Small Grants Programme.
Somarriba, E., Villalobos, M., Sucre, L., L opez, M., Torres, F., Torres, A., and Rodriguez, K.
(2008). El proyecto Captura de carbono y desarrollo de mercados ambientales en cacaotales
y otros sistemas agroforestales ind genas en Talamanca, Costa Rica. Agroforester a en las
Am erica, 46:8{13.
Soto, H. D. (2000). The Mystery of Capital: Why Capitalism Triumphs in the West and
Fails Everywhere Else. Basic Books.
Stem, C., Margoluis, R., Salafsky, N., and Brown, M. (2005). Monitoring and Evaluation in
Conservation: A Review of Trends and Approaches. Conservation Biology, 19(2):295{309.
Stiglitz, J. (2011). The Mauritius miracle, or how to make a big success of a small economy
| Joseph Stiglitz. The Guardian.
Stiglitz, J. E. (2007). Making Globalization Work. W.W. Norton & Co, New York.
Stiglitz, J. E., Sen, A., and Fitoussi, J.-P. (2010). Mismeasuring Our Lives: Why GDP
Doesn't Add Up. New Press, New York.
Stone, D. N., Deci, E. L., and Ryan, R. M. (2009). Beyond Talk: Creating Autonomous
Motivation through Self-Determination Theory - Dan N. Stone, Edward L Deci, Richard
M. Ryan, 2009. Journal of General Management.
259
Subramanian, A. (2013). The Mauritian Success Story and its Lessons. In Fosu, A. K.,
editor, Achieving Development Success, pages 204{231. Oxford University Press.
Subramanian, A. and Roy, D. (2001). Who Can Explain the Mauritian Miracle: Meade,
Romer, Sachs, or Rodrik? IMF Working Papers, 01(116):1.
Sundberg, R. and Melander, E. (2013). Introducing the UCDP Georeferenced Event Dataset.
Journal of Peace Research, 50(4):523{532.
Swiss, L. (2012). The adoption of women and gender as development assistance priorities:
An event history analysis of world polity eects. International Sociology, 27(1):96{119.
Tao, S. (2016). Improving Female Teacher Deployment to Rural Schools. In Transforming
Teacher Quality in the Global South, pages 125{144. Palgrave Macmillan, London.
Tirole, J. (1986). Hierarchies and Bureaucracies: On the Role of Collusion in Organizations.
Journal of Law, Economics & Organization, 2:181.
Transparency International (2018). Corruption Perceptions Index 2018.
Tryzna, T., editor (1995). A Sustainable World: Dening and Measuring Sustainable Devel-
opment. California Institute of Public Aairs and Earthscan for IUCN, Sacramento and
London.
ul Haq, M. (1999). Re
ections on Human Development. Oxford India Paperbacks. Oxford
University Press, Delhi ; New York, expanded ed edition.
UNDP (2009). Handbook on Planning Monitoring and Evaluating for Development Result.
Technical report, United Nations Development Programme, New York, NY.
UNDP (2019). 2019 Human Development Index Ranking. Human Development Reports,
UNDP.
Upho, N. and Langholz, J. (1998). Incentives for avoiding the Tragedy of the Commons.
Environmental Conservation, 25(3):251{261.
USAID (2012). The Logical Framework. Technical Report Number 2, Version 1, USAID.
Vieillesse, M. (2016). Final Report: Sustainable Waste Management Practices at Petit
Verger Prison.
Villalobos, M. and Soto, C. (2006). Calendario 2006: Captura de carbono en cacaotales y
otros sistemas agroforestales ind genas y desarrollo de mercados ambientales en Talamanca,
Costa Rica.
Von Uexkull, J. E., Echandi, R., and Ranzani, M. (2019). World Bank in Mauritius. Technical
Report 141916, The World Bank.
Waddington, H., Davies, P., and Puri, J. (2015). Improving lives through better evidence.
Journal of Development Eectiveness, 7(3):267{274.
260
Wallerstein, W. and Duran, B. (2003). The conceptual, historical and practice roots of
community based participatory research and related participatory traditions. In Minkler,
M. and Wallerstein, N., editors, Community-Based Participatory Research for Health:
From Process to Outcomes, pages 27{52. John Wiley & Sons, San Francisco, CA.
Williamson, J. (2008). A Short History of the Washington Consensus. In Serra, N. and
Stiglitz, J. E., editors, The Washington Consensus Reconsidered: Towards a New Global
Governance, The Initiative for Policy Dialogue Series, pages 14{40. Oxford University
Press, Oxford ; New York.
World Bank (1998). Assessing Aid: What Works, What Doesn't, and Why. A World Bank
Policy Research Report. Oxford University Press, Washington, D.C.
World Bank (2003). Making Services Work for Poor People. World Development Report
2004. World Bank Publications.
World Bank (2016). Impact Evaluation - Overview.
World Bank (2019a). Costa Rica Home.
World Bank (2019b). Data for Mauritius, Costa Rica.
World Bank (2019c). Mauritius: Overview.
World Bank (2019d). Net ocial development assistance and ocial aid received (constant
2015 US$).
World Bank (2019e). World Bank Country and Lending Groups: Country Classications.
World Bank (2019f). Worldwide Governance Indicators.
World Commission on Environment and Development (1987). Our Common Future (Brundt-
land Report). Technical report.
World Economic Forum (2019a). Competitiveness Rankings: The Global Competitiveness
Report 2018. Technical report, World Economic Forum.
World Economic Forum (2019b). The Travel & Tourism Competitiveness Report 2019:
Travel and Tourism at a Tipping Point. Technical report, World Economic Forum.
Yamada, H., Sawada, Y., and Luo, X. (2013). Why is Absenteeism Low among Public Health
Workers in Lao PDR? Journal of Development Studies, 49(1):125{133.
Zumbado-Morales, F. (2010). Agrotourism and Agro-Ecotourism in Costa Rica. e-Review
of Tourism Research, 8(6).
261

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Economic policy-making in a changing environment: the politics of success in Peru vis-à-vis China

PDF

Developing countries in the digital era: state and business interactions for industrial upgrading

PDF

Political determinants and economic effects of HIV/AIDS: a push for the multisectoral approach

PDF

Politics behind appointments: policy, patronage, and party systems in the Andes

PDF

Identity salience and political violence: centering social identity in the study of intrastate conflict

PDF

Habits and policy: the socio-cognitive foundations of foreign policy stability

PDF

The political economy of oil, terrorism and institutional development in Iran and its impact on the Middle East

PDF

The murky risk of trade protectionism in an interconnected and uncertain global economy: a state, industrial, and regional analysis

PDF

Power to the masses? The rise of direct democracy in Latin America

PDF

The allocation and accessibility of aid: the role of recipient government agency in dictating donor decisions

PDF

The effects of interlocal collaboration on local economic performance: investigation of Korean cases

PDF

Social policy and presidential ideology in Latin America: the political economy of social spending and anti-poverty programs

PDF

The erosion of democracy in the post–Cold War world: understanding the phenomenon and its causes

PDF

Policies for genetically modified crops in developing countries: the role of non-state actors

PDF

Budget institutions and the positive theory of fiscal policy

PDF

Flinging the boomerang: locating instability and the threat potential of identity-bias in US national security policy

PDF

Three essays on economics of early life health in developing countries

PDF

The effects of health-related spousal influence on couples coping with chronic heart failure: an application of the actor-partner interdependence model

PDF

Global health diplomacy: a new era of health in U.S. foreign policy

PDF

Effects of economic incentives on creative project-based networks: communication, collaboration and change in the American film industry, 1998-2010

Asset Metadata

Creator Roglà, Jennifer (author)

Core Title The role of local actors in international development: how incentive types affect foreign aid effectiveness

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Publication Date 11/22/2020

Defense Date 03/10/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag AID,aid effectiveness,assessment/evaluation,communities,Costa Rica,developing countries,Development,development projects,foreign aid,grassroots,incentives,institutions,international development,international political economy,local actors,localization,Mauritius,monitoring,OAI-PMH Harvest,Political Economy

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wise, Carol (committee chair), Munck, Gerardo (committee member), Nugent, Jeff (committee member)

Creator Email jrogla@gmail.com,rogla@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-396017

Unique identifier UC11667176

Identifier etd-RoglaJenni-9144.pdf (filename),usctheses-c89-396017 (legacy record id)

Legacy Identifier etd-RoglaJenni-9144-0.pdf

Dmrecord 396017

Document Type Dissertation

Rights Roglà, Jennifer; Rogla, Jennifer

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA