Emerging: Challenges

Emerging
Challenges
and Innovations
in Microfinance
and Financial
Inclusion
Edited by
Michael O’Connor
& Joana Silva Afonso
Emerging Challenges and Innovations
in Microfinance and Financial Inclusion
Michael O’Connor • Joana Silva Afonso
Editors
Emerging Challenges
and Innovations in
Microfinance and
Financial Inclusion
Editors
Michael O’Connor Joana Silva Afonso
Portsmouth Business School Portsmouth Business School
University of Portsmouth University of Portsmouth
Portsmouth, UK Portsmouth, UK
ISBN 978-3-030-05260-7 ISBN 978-3-030-05261-4 (eBook)

https://doi.org/10.1007/978-3-030-05261-4
© The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland
AG 2019
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and trans-
mission or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
Cover illustration: Witthaya Prasongsin / Getty Images

Cover design by Tjasa Krivec
This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Foreword
Microfinance is an unusual discipline, with one foot in the world of busi-

ness and finance and another in the world of sociology and development
economics. This means that the investors and practitioners from the busi-
ness and finance side get to enjoy the wealth of research that accompanies
the developmental side of the field, while academics gain partners and an
audience beyond scholars and policymakers.
This publication gathers some of the best work from the fifth European
Research Conference on Microfinance, organised by the University of
Portsmouth, UK, in June 2017. Appropriately for the sector, it spans a
wide range of subjects—some directly applicable to practitioners, such as
the effectiveness and viability of Islamic finance products, or those aimed
at strengthening climate change adaptation. For those actors with a more
developmental or social orientation, there is an assessment of lending
practice disparity between women and men and a “natural” control study
of the market effects following the withdrawal of microfinance lending in
Uzbekistan. And for those focused on regulation and policy, there is a
review of the impact of interest rate regulations on the quality of finan-
cial services.
All this is bookended by a wonderful Microfinance Alphabet high-
lighting the key issues in the sector from the team that started the Research
Conference series in 2009, as well as a conclusion that helps set the book’s
v
vi Foreword
chapters into the context of broader research in the sector—thus setting

the scene for future research.
The European Research Conference on Microfinance, a bi-annual
event that has been going strong for a decade, is a key event in the micro-
finance academic calendar. But it is more than that—it is a vehicle to
bring together the sector’s best researchers with practitioners, investors,
and others, creating a unique opportunity to cross the boundaries of the
discipline.
As founding partners and sponsors of the European Research
Conference on Microfinance, we congratulate the authors and editors of
this book for sharing the best of this Conference with the broader com-
munity and look forward to the next decade of success!
European Microfinance Platform Daniel Rozas

Luxembourg, Luxembourg
Acknowledgements
This publication results from a selection of outputs from researchers who

took part in the fifth European Research Conference on Microfinance
that was hosted at the University of Portsmouth in June 2017. It goes
without saying that without those contributing chapters, this publication
would not have been possible. As editors, we are deeply appreciative of
the professionalism of those involved in contributing to the publication.
The editors wish to acknowledge the many organisations and individu-
als without whom the conference itself would not have been a success—
these include Daniel Rozas, Christoph Pausch, Niamh Watters, Camille
Dassy and Gabriela Erice from the European Microfinance Platform, the
primary promoter and funder of the conference; Professors Ariane
Szafarz, Marc Labie and Marek Hudon from the Centre of European
Research in Microfinance (CERMi); as well as members of the Scientific
Committee including organisers of previous iterations of the event.
We also wish to thank our many conference sponsors and supporters,
including ADA Microfinance, Lendwithcare, the Microfinance Gateway,
the European Microfinance Network and Microfinance Centre; our
many moderators, speakers and presenters together with Michael Buchan,
Jinx Prowse and Music Fusion; and our many supportive colleagues at
Portsmouth Business School, University of Portsmouth.
vii
viii Acknowledgements
We would like to thank the Afonso and O’Connor families for not
asking too many questions together with Tula Weis and the Palgrave
Macmillan team for nudging us gently towards the finishing line in com-
pleting this work.
Contents
1
The Microfinance Alphabet 1
Marek Hudon, Marc Labie, and Ariane Szafarz
2 Addressing Climate Change with Microfinance Plus:

Experiences in Cattle and Coffee Regions of Nicaragua 13
Johan Bastiaensen, Milagros Romero, and Frédéric Huybrechs
3 Remittances: Loan Funds for a Rural Economy? Evidence

from the Kayes District (Western Mali) 39
Michel Namé and Philippe Lebailly
4 Islamic Microfinance: Exploring the Experience of

Akhuwat in Promoting Qard Hasan in Pakistan 61
Joana Silva Afonso and Ajaz Ahmed Khan
5 Relationship Lending in Microfinance: Do Women Benefit

As Much As Men? 85
Mathilde Bauwin
ix
x Contents
6 What Happens When Microfinance Programmes Are

Withdrawn? A Conceptual Framework for Analysing
Causal Effects123
Nargiza Alimukhamedova
7 Good Customer Service Versus Bad Regulation149

Mark Hannam
8 Reframing Microfinance and Financial Inclusion Research:

Case Studies and Synthesis163
James Copestake
Index175
Notes on Contributors
Nargiza Alimukhamedova received a PhD in economics in 2014 from

CERGE-EI (the Center for Economic Research and Graduate Education –
Economics Institute), Charles University in Prague. She holds a researcher
position in CERGE-EI and also teaching positions in the University of
Economics in Prague and Westminster International University in
Tashkent. Her main research areas include microfinance, development eco-
nomics, economics of transition, and institutional economics.
Johan Bastiaensen is professor at the Institute of Development Policy
(IOB), University of Antwerp, Belgium, and associate researcher at the
Institute of Research and Development Nitlapan, Universidad
Centroamericana (UCA), Managua, Nicaragua. His main research inter-
est lies with the transformative role of microfinance for a more socially
inclusive and environmentally less destructive rural-agricultural develop-
ment. This research is inspired by critical institutional theory and is often
done in cooperation with the Instituto Nitlapan in Nicaragua, an aca-
demic partner of the IOB since 1988. Much of the applied research is
related to the policy debates in the context of the Fondo de Desarrollo
Local (FDL), today one of the largest and most agricultural microfinance
institutions in Nicaragua.
xi
xii Notes on Contributors
Mathilde Bauwin graduated with a PhD in development economics in

2017 from the French Institute of Demographic Studies (INED) and
Paris Dauphine University. She is now working as a research project
officer at Ada Microfinance, an NGO promoting responsible financial
inclusion and supporting microfinance institutions all over the world.
James Copestake is Professor of International Development at the
University of Bath. His research interests embrace agrarian change, devel-
opment finance, impact evaluation, wellbeing assessment, and the global
political economy of development. Ongoing activities include promoting
qualitative impact evaluation through a non-profit company, Bath Social
and Development Research, and being Director of Studies for Bath’s
Professional Doctorate in Policy Research and Practice.
Mark Hannam is an associate research fellow at the Institute of
Philosophy, School of Advanced Study, University of London. From 2007
to 2016, he was chair of Fair Finance, the London-based microfinance
company.
Marek Hudon is professor at the Solvay Brussels School of Economics
and Management (SBS-EM), Université Libre de Bruxelles (ULB),
Belgium. He is also co-director of the Centre for European Research in
Microfinance (CERMi), and co-director of the Centre d’Etudes
Economiques et Sociales de l’Environnement (CEESE). He has initiated
and is still scientific coordinator of the European Microfinance Programme
(EMP). He has conducted research in India, Kenya, Mali, Morocco,
Vietnam, and the Democratic Republic of Congo.
His research mainly focuses on social and non-profit enterprises/entre-
preneurship (microfinance and complementary currencies, particularly
in the context of developing economies), sustainable development, and
business ethics.
Frédéric Huybrechs is a post-doc researcher at the Institute of
Development Policy (IOB), University of Antwerp, Belgium. He cur-
rently studies the role of microfinance in the advancement of the cattle-
driven agricultural frontier in Nicaragua. In his PhD thesis, he analysed
the topic of Green Microfinance from a political ecology perspective and
in the context of rural development. He conducted fieldwork in Nicaragua
Notes on Contributors xiii
in order to analyse a particular Green Microfinance project which com-

bined green credits with payments for ecosystem services and technical
assistance. He is also a member of the action group “Green Inclusive and
Climate Smart Finance” of the European Microfinance Platform and has
been involved in a consultancy project on Green Microfinance in Latin
America for the Inter-American Development Bank.
Ajaz Ahmed Khan is senior microfinance adviser with CARE
International, an international development agency with operations in
more than 90 countries. He holds a PhD in development economics and
has more than two decades of practical field experience from working as
an agricultural economist in Latin America, Eastern Europe, Africa, and
Asia. He has helped to establish microfinance organisations in various
countries which now provide financial services to tens of thousands of
low-income people. He has written a number of articles and books on
microfinance in general and Islamic microfinance in particular.
Marc Labie is a professor in the Department of Management of the
Warocqué School of Business and Economics at the University of Mons
(UMONS), and for the European Microfinance Programme (EMP). He is
co-director of the Centre for European Research in Microfinance (CERMi)
and co-director of the European Microfinance Programme (EMP).
Labie holds a bachelor’s degree in economics and social science, and a
PhD in business administration. He has also studied for brief periods at
the London School of Economics and Political Science and at the
Universidad de Salamanca. He is an alumnus of the Financial Institutions
for Private Enterprise Development (FIPED) programme at Harvard
University in Cambridge, Massachusetts, US, in which he has also been
lecturing for 11 years. His field experience includes researches and case
studies in Colombia, Bolivia, Democratic Republic of Congo, Indonesia,
Kenya, Madagascar, Morocco, Mexico, and Peru.
His articles have appeared in journals such as Ethics and Economics,
Gestion, Mondes en Développement, Management & Avenir, Management
Decision, and Quarterly Review of Economics and Finance. He has co-edited
various books including The Handbook of Microfinance in 2011 and The
Crises of Microcredit in 2015.
xiv Notes on Contributors
Philippe Lebailly has a PhD in agricultural sciences. He is Professor of

Economics and the research director and head of the unit of economy
and rural development at Gembloux Agro-Bio Tech, University of Liege.
Michel Namé is a researcher (PhD candidate in rural economics) at
Gembloux Agro-Bio Tech, University of Liege. He is an engineer in farm-
ing techniques, specialising in agricultural mechanics and rural develop-
ment (High School of Agronomy—Institut National Polytechnique Félix
Houphouët-Boigny). He also has a master’s degree in economic analysis
and international development, specialising in sustainable development in
developing countries (Centre d’Etudes et de Recherches sur le
Développement International—University of Auvergne). His research
relates to remittances, food security, climate change, and rural develop-
ment, with his current research focussing on the ability of migrant remit-
tances to contribute to food security and poverty alleviation.
Michael O’Connor is a Senior Lecturer in the Accounting and Financial
Management subject group at the Faculty of Business and Law, University
of Portsmouth. As a law graduate and finance professional, he has worked
with Andersen and KPMG in Ireland as well as the European Commission
in Brussels. He has significant overseas experience in Russia and China,
particularly in financial training. In this regard, he has also worked as a
contractor for international consultancy firms on behalf of public and
private sector clients in Western Europe, West Africa, and Central Asia.
He is a graduate of the European Microfinance Programme (EMP) at
Université Libre de Bruxelles (Belgium) and was a co-organiser of the
fifth European Research Conference on Microfinance.
Milagros Romero is a researcher at the Institute of Research and
Development Nitlapan of the Universidad Centroamericana (UCA),
Managua, Nicaragua, and a PhD student at the Institute of Development
Policy (IOB), University of Antwerp, Belgium. Her research focuses on
analysing the insertion of microfinance programmes in rural territorial
dynamics in Nicaragua, addressing the opportunities and constraints in
improving the social and ecological consequences of these programmes,
in a setting with wide inequalities, high poverty levels, and accelerated
Notes on Contributors xv
environmental degradation. Her research is mainly based on the experi-

ences of the development programmes of Nitlapan and the financial
institution Fondo de Desarrollo Local (FDL), and their Microfinance
Plus model combining microcredit and technical assistance services.
Joana Silva Afonso is a senior research associate in the Economics and
Finance Department at the University of Portsmouth. She has experience
as both practitioner and researcher in the microfinance sector. She first
became involved with microfinance in Portugal where she was a micro-
credit officer at Associação Nacional de Direito ao Crédito (ANDC) (non-
governmental organisation—NGO) between 2005 and 2012. In 2013,
she completed the European Microfinance Programme at the Université
Libre de Bruxelles (Belgium), which included fieldwork in the Dominican
Republic. Between 2015 and 2018, as part of her PhD at the University of
Portsmouth, she conducted research on the challenges associated with
impact evaluation in microfinance with fieldwork in Pakistan and
Zimbabwe. She was a co-organiser of the fifth European Research
Conference on Microfinance which took place in June 2017.
Ariane Szafarz is Professor of Finance at the Solvay Brussels School of
Economics and Management (SBS-EM), Université Libre de Bruxelles
(ULB). She holds a PhD in mathematics and an MD in philosophy. Her
research topics include microfinance, gender discrimination, and financial
markets. She is co-director of the Centre for European Research in
Microfinance (CERMi), co-director of the European Microfinance
Programme (EMP), co-director of the SBS-EM Doctoral Programme in
Management Sciences, and president of the Marie-Christine Adam Fund.
She was a visiting professor at Université de Lille II, Université Catholique
de Louvain, and the Luxembourg School of Finance. She sits in the editorial
boards of the Brussels Economic Review and Bankers, Markets and Investors.
Szafarz has authored several books and her scientific articles have
appeared in, for example, Econometric Theory, Economics Letters, European
Economic Review, Journal of Asset Management, Journal of Banking and
Finance, Journal of Business Ethics, International Review of Law and
Economics, Journal of Development Studies, Journal of Empirical Finance,
Journal of International Money and Finance, Labour Economics, Nonprofit
xvi Notes on Contributors
and Voluntary Sector Quarterly, Oxford Economic Papers, Quarterly Review

of Economics and Finance, Small Business Economics, Review of Finance,
and World Development. Together with co-authors Simon Cornée and
Panu Kalmi, she received the 2016 Warren Samuels Prize awarded by the
Association for Social Economics for the article “Selectivity and
Transparency in Social Banking: Evidence from Europe” presented at the
Allied Social Science Associations (ASSA) Meetings, San Francisco, USA,
2016, and now published in the Journal of Economic Issues.
List of Figures
Fig. 5.1 Evolution of financial indicators, in value and ratio (medians) 95

Fig. 5.2 Average amounts granted over credit cycles by gender 98
Fig. 5.3 Evolution of loan amounts from one credit cycle to another by
gender (in ratios over the previous amount) 98
Fig. 5.4 Predictions of amounts granted from cycle 2 108
Fig. 5.5 Predictions of growth rates of loan amounts from a cycle to
another (in ratio) 115
Fig. 6.1 Summary of the multi-dimensional framework 129
Fig. 6.2 Cumulative growth of non-bank MFIs in Uzbekistan, 1998–
2011139
xvii
List of Tables
Table 4.1 AIM main activity indicators 67

Table 5.1 Average loan amount by credit cycle and by gender 97
Table 5.2 Estimation of amounts granted from cycle 2 (all loans) 102
Table 5.3 Estimation of amounts granted from cycle 2 (non-agri. and
agri. loans separately) 105
Table 5.4 Growth rate of loan amount from one credit cycle to another
(in ratio) 111
Table 6.1 Understanding the reasons for MFI(s) termination 130
Table 6.2 Respondents and objectives for in-depth interviews 131
Table 6.3 Difference-in-differences estimator of the impact of
microfinance sector closure 137
Table 6.4 Microcredit and microdeposit services in Uzbekistan 141
Table 8.1 The scope of microfinance and financial inclusion research 165
Table 8.2 Mainstream, alternative and plural disciplinary perspectives
on microfinance 166
xix
1
The Microfinance Alphabet
Marek Hudon, Marc Labie, and Ariane Szafarz
The fifth European Research Conference on Microfinance was organized

in 2017 at the University of Portsmouth, UK. Previously organized in
Belgium (Université Libre de Bruxelles—ULB, 2009), the Netherlands
(University of Groningen, 2011), Norway (Agder University, 20,013),
and Switzerland (University of Geneva, 2015), the conference is a great
opportunity for researchers in microfinance to exchange ideas on their
common field of interest.
Much has been learnt in microfinance over the last ten years. But there
is yet so much to discover on how to improve financial inclusion and
development. A child’s fifth birthday is a great moment to celebrate. It is
also the typical moment where the child discovers the world of reading.
M. Hudon (*)
Solvay Brussels School of Economics and Management (SBS-EM), Université
Libre de Bruxelles (ULB), Brussels, Belgium
Centre for European Research in Microfinance (CERMi), Brussels, Belgium
Centre d’Etudes Economiques et Sociales de l’Environnement (CEESE),
Brussels, Belgium
e-mail: mhudon@ulb.ac.be
© The Author(s) 2019 1

M. O’Connor, J. Silva Afonso (eds.), Emerging Challenges and Innovations
in Microfinance and Financial Inclusion, https://doi.org/10.1007/978-3-030-05261-4_1
2 M. Hudon et al.
Likewise, we are happy to celebrate the fifth edition of the European

Research Conference on Microfinance by contributing to this book with
a—evidently subjective—microfinance alphabet, hoping to so provide
the microfinance scientific community with an opportunity to “read
together” both where we stand and where we are heading. We hope you
will enjoy it!
A: Asset and Liability Management (ALM)
ALM refers to the way in which banks address and manage the risks
attributable to potential mismatches between the assets and liabilities.
Even though ALM is barely considered in the microfinance literature, it
is a key challenge for microfinance institutions (MFIs) willing to develop
products that meet the needs of the financially excluded. Most MFIs are
reluctant to supply long-term loans, and microfinance start-ups pay
much more attention to the asset side of their balance sheet—notably to
the quality of their portfolio—than to their liabilities. Indeed, recently
created MFIs are mostly financed by seed capital and subsidies, so that
their liabilities are often not considered a priority. Later developments
make, however, liabilities management necessary to address MFIs’ long-
term needs for capital. When subsidies dry out, MFIs typically take loans
from regular banks, at least in countries where commercial loans are
accessible to them. In this way, MFIs can sustain their growth for a cer-
tain time, provided that the quality of their portfolio remains under con-
trol. Otherwise, commercial banks will consider MFIs too risky and cut
their financing. But if an MFI’s portfolio keeps growing, its debt/equity
M. Labie
Warocqué School of Business and Economics at the University of Mons
(UMONS), Brussels, Belgium
e-mail: marc.labie@umons.ac.be
A. Szafarz
Solvay Brussels School of Economics and Management (SBS-EM), Université
Libre de Bruxelles (ULB), Brussels, Belgium
e-mail: aszafarz@ulb.ac.be
1 The Microfinance Alphabet 3
ratio will at some point reach a level such that bank loans will become
difficult to obtain. It is at this point that most MFIs start thinking seri-
ously about collecting savings, which are the only long-term source of
funds that allow a financial institution to grow steadily without liquid-
ity issues.
B: Bottom Lines
MFIs are usually understood as double-bottom-line institutions.
While economists tend to use the social-versus-financial trade-off to
describe the objective function of MFIs, management scholars desig-
nate these dual concerns by referring to welfare and market institu-
tional logics. Yet, the actual existence of a “simple” trade-off is challenged
on both sides. On the one hand, commercialization, which typically
downplays the social concern, pushes towards a single business-like
objective (see “C”). On the other, some argue in favour of adding a
third bottom line of environmental performance, leading to the notion
of green microfinance.
C: Commercialization
The microfinance commercialization trend is now well recognized: The
Consultative Group to Assist the Poor (CGAP), a consortium of donors
active in microfinance, mentioned it already in its 1990s’ five-year plans.
Yet, the bulk of the commercialization wave developed during the 2000s.
Cases in point include the initial public offerings of two leading MFIs:
Compartamos and SKS. Some scholars consider commercialization as a
pre-requisite for MFIs to address financial inclusion efficiently. By con-
trast, others see commercialization as a major source of mission drift
observed when organizations focus on profits and move away from their
social goal. The co-existence of such contrasting views probably explains
why we are still lacking a broadly accepted definition of mission drift.
D: Diversification of Products
Microfinance started as microcredit, as illustrated by the fact that the
largest meeting in the field was named “The Microcredit Summit
Campaign.” Later, financial services for the disadvantaged populations
were extended to savings, insurance, and money transfers, leading the
industry to move from microcredit to microfinance, and more recently to
financial inclusion (see “F”). Meanwhile, the conversation evolved around
the advantages and drawbacks of providing financial services only—the
4 M. Hudon et al.
so-called minimalist approach—or combining them with financial liter-

acy, professional training, marketing support, or even business-unrelated
services—the “integral” approach.
E: Ethics of Randomized Control Trials (RCTs)
RCTs are often considered as the gold standard of impact evaluation in
microfinance. RCT results are known for their high internal validity in
terms of causal linkages, associated with limited external validity due to
their poor degree of generalizability. Nowadays, prestigious economic
journals refrain from publishing non-RCT empirical studies due to endo-
geneity concerns. Strikingly, the RCT conversation focuses on methodol-
ogy and the ethical issues are left aside. Yet, RCT ethics committees
would gain from internationally binding guidelines based on strict, uni-
versal ethical rules as is the case in medicine when it comes to, for exam-
ple, patient consent and providing ex ante equally attractive options to
the treated and control groups, a rule known as equipoise. The serious
history of mistreatments of third world populations by past medical
experiments is a strong incentive to increase our vigilance in that regard.
F: Financial Inclusion
Financial inclusion is increasingly replacing microfinance in the alpha-
bet of financial services to the poor and the unbanked. In fact, this change
in language broadens the scope of the sector since it goes beyond MFIs
and includes new service providers, such as fintech, and new technolo-
gies, such as mobile banking. It might, however, obscure the original
objectives of poverty alleviation and empowerment of the disadvantaged
populations.
G: Governance
Any microfinance alphabet must address MFI governance. This topic
encompasses the role of the boards, which is increasingly investigated
quantitatively with international datasets. Governance studies also check
whether the decisions of MFI executives match the mission of the orga-
nization they serve. Top executives gain both increased legitimacy from
stakeholders, such as donors and public authorities, as well as extra finan-
cial manoeuvre room from commercial fund providers. Some relevant
governance mechanisms are internal, such as boards, organization charts,
incentive structures, and corporate culture. Others are driven by the envi-
ronment, such as accounting standards, regulation, and supervision
requirements, as well as the due diligence of debt providers. Corporate

governance in microfinance points to major risks as pinpointed by the
Banana Skins reports.1
H: Human Rights and Microfinance
Muhammad Yunus considers that microcredit should be a human
right. He states that access to credit is instrumental to economic develop-
ment, poverty alleviation, and the improved welfare of all citizens. This
view is challenged by those arguing that credit is not a universal need.
Moreover, credit that is too easily obtainable can have perverse effects like
over-indebtedness. Meanwhile, other financial services, such as savings
and money transfers, are less risky and sometimes even more useful
socially than credit.
I: Interest Rates
The interest rates charged by MFIs have been criticized since the early
days of microcredit in the 1970s. Outsiders find it hard to understand
why lending to the poor is costly. MFIs must cover three types of inevi-
table costs: the cost of funds, the operational costs, and loan loss provi-
sioning. Well-managed MFIs can limit both their costs of funds and loan
loss provisioning, but reductions in operational costs are difficult to
achieve since numerous small proximity loans imply costly features, such
as multiple branch and frequent home visits. But this evidence is not a
reason for charging abusive interest rates, and more so from MFIs benefit-
ting from a monopoly/oligopoly power over financially vulnerable cli-
ents. Excessive interest rates have been exposed since the emergence of
microcredit, but criticisms have grown widespread of lending practices in
leading MFIs in certain countries, such as Mexico.
J: Juggling with Financial Products
Microfinance clients juggle with formal and informal sources of funds
and financial services. They use a variety of financial tools, linked to
informal networks, local merchants, family and friends, and various pro-
viders of financial services. The dark side of this resourcefulness is the risk
of over-indebtedness and subsequently an increased frequency of credit
defaults that can undermine the microfinance sector.
K: Keeping an Open Mind
Microfinance is not a panacea against poverty. And microfinance alone
cannot tackle all the complex issues associated with poverty reduction.
6 M. Hudon et al.
MFIs do, however, respond to a real demand from the poor and unbanked
populations for reliable financial services. At the very least, like banks,
MFIs help smooth income. The question raised by Richard Patten—“If
microfinance is the answer, what is the question?”—is still relevant.
L: Loan Officers
Loan officers are key players in the microfinance arena. Not only are
they the visible side of the organization, but they are also those who
screen clients and enforce loan reimbursements. The typical day of a
microfinance loan officer is made up of many house visits so that they
become an integral part of the community they work in. Loan officers
face conflicting challenges: They are the partners of their clients analysing
their needs and providing them counselling, while also having to pro-
mote their employer’s agenda including portfolio growth—which implies
pushing credits—and timely repayments—with various forms of pres-
sure. Even though the role of loan officers is fundamental, it is only
recently that scholars have started studying them specifically.
M: Microfinance Promise
Researchers in microfinance typically start with reading the influential
and most-quoted article on the sector, “The Microfinance Promise” by
Jonathan Morduch. This paper was published in 1999 in the Journal of
Economic Literature, at the same time as Jonathan’s other seminal paper
“The Microfinance Schism” was published in World Development. Twenty
years later, these two papers are still a wonderful introduction to under-
standing the very essence of microfinance.
N: Non-Governmental Organizations (NGOs) and Savings and
Credit Cooperative Societies (SACCOs)
Historically, the vast majority of MFIs started their operations as
NGOs or SACCOs. NGOs have played a major role in the microfinance
industry as their basic characteristics allowed them to be particularly
innovative. These characteristics include being non-regulated institutions
as they were not collecting savings and being largely financed by subsidy
providers. Key methodological principles of microfinance, such as group
lending and progressive lending, were originally tested by NGOs.
Nowadays, NGOs still represent a significant segment of microfinance,
but their overarching dominance has faded.
Originally created in the nineteenth century in Germany and Britain,

SACCOs have spread all over the world and are now part of the financial
landscape in general, and microfinance in particular. The comparative
advantages of SACCOs are twofold: They are imbedded locally, and they
benefit from members’ voluntary workforce. On the flip side, their
growth can be detrimental to the interactions between members and
executives, resulting in threats to their governance (see “G”).
O: Ohio-State-University Pioneers
The pioneering contributions of Dale Adams, Claudio Gonzalez-Vega,
and Richard Meyer had an influential role in the field of microfinance
and financial inclusion. These scholars have left their mark on a variety of
topics. They argued that “cheap credits” from donors and subsidized
credit lines and grants can prevent the entry of private actors, and there-
fore be detrimental to the development of rural financial systems. These
pioneers stressed the relevance of micro-savings and proved that some
informal markets are less irrational and exploitative than origi-
nally thought.
P: Portsmouth
Following Brussels, Groningen, Kristiansand, and Geneva, Portsmouth
was the fifth city to host the European Research Conference on
Microfinance. The event was perfectly organized, and all the participants
enjoyed the enthusiasm and dedication of the host institution. Portsmouth
will remain present in the collective memory of the European microfi-
nance community. We extend our special gratitude to Joana and Michael,
the editors of the current volume.
Q: Quantitative Models
The role of models in economic theory is much debated. Regardless,
the common wisdom equating mathematical modelling with the tenets
of neoclassical economics is misleading. Mathematical models can be
adapted to various objective functions, provided that they are formally
well-defined. Microfinance models have addressed relevant issues, such as
the consequences of the double-bottom-line approach (see “B”) to finan-
cial services in the context of informational asymmetries. By making
trade-offs explicit, mathematical models can bring to light the real chal-
lenges facing MFIs and help go beyond the often-simplistic rhetoric of
unconstrained “good-doing.” Even though most—if not all—arguments
8 M. Hudon et al.
can be made without any equation, the role of mathematics stems from
its capacity of holding the formal strictness of reasoning. Importantly, the
purpose of mathematical modelling is to address a theoretical research
question with minimal working assumptions rather than reflect the full
complexity of the real world.
R: Regulation and Supervision
Originally, public authorities viewed reluctantly the development of
MFIs. They thought that, if lending to the poor was a feasible and sustain-
able economic activity, it would already have existed for a long time. This
attitude evolved with the success of major MFIs worldwide. Nowadays,
authorities are contemplating ways of nurturing microfinance, notably
thanks to specific non-banking financial institutions (NBFIs). Typical
regulatory frameworks permit various types of MFIs: Simple institutions
cannot collect public savings and are therefore only lightly controlled,
while those providing multiple financial services are subject to rules simi-
lar to those imposed on banks. Altogether, many countries have invested
significant effort in microfinance regulations. But the same cannot be said
about supervision: Most authorities lack properly trained manpower. As a
result, officially supervised MFIs do sometimes look safer than they are.
S: Subsidies
Subsidies have always played an important role in the funding of MFIs.
Most microfinance pioneers were supported by grants or concessionary
loans. After decades of microfinance practice, many MFIs still receive
significant subsidies and are even subsidy-dependent. Only a minority of
MFIs are true commercial actors and totally un-subsidized. While grants
and concessionary loans still exist, subsidized equity—equity holders
with below market expectations—represents the largest type of subsidy.
T: Time-Inconsistency
Procrastination, under-saving, and over-consuming are typical features
of human behaviour, grouped by economists under the label of time-
inconsistency. These features contrast with the textbook characteristics of
the so-called homo economicus who makes time-consistent decisions. Even
though time-inconsistency affects agents with any income level, its con-
sequences are far more harmful to cash-constrained individuals than to
those who are better off. MFIs address the hypothetical time-inconsistency
of their clients by supplying illiquid financial products meant to disci-
pline financial impulses. While discipline can be a useful commitment

device to counteract temptations, they are counter-productive for poor
people confronted by adverse shocks. Poor households are coping not
only with income variability and unpredictability but also with uninsured
idiosyncratic risks, such as sickness and robbery, as well as systemic risks,
including the consequences of climate change and political unrest. To
address these tough issues, the poor badly need access to flexible products.
U: Unfair Discrimination in Lending
Discrimination in lending against vulnerable groups—such as women
and minorities identified by their ethnicity, religion, caste, disability, and
sexual orientation—refers to unfair stereotypical criteria used in loan
allocation. The most common definition of discriminatory loan alloca-
tion relates to a lender’s practice of deviating from profit-maximization
logic. According to this view, a biased lender is identified by the fact that
it denies loans to the targeted group members more frequently than to
the other applicants, all else being equal. The problem with this approach
arises from its implicit assumption that both the group members and the
other loan applicants are equally creditworthy. In some cases, however,
this assumption can be wrong. For instance, evidence in microfinance
suggests that women tend to reimburse their loans more swiftly than
men, for reasons that are still obscure and not captured by the usual con-
trol variables used to stick to the “all else equal” statistical requirement. If
the assumption of equal creditworthiness does not hold, then even equal
denial to both groups, with all else being equal, can end up being unfair.
V: Village Banking
Village banks come under various forms. Some are old and have hardly
changed in decades. It is the case of the Indonesian BKD (Badan Kredit
Desa—literally, village credit organizations), which is among the oldest
worldwide. Other village banks, such as Finca, were initiated more
recently by international players. All village banks share the ideas of start-
ing small, at the village level, and being collectively managed. In some
cases, the bylaws impose that the leader of the bank is a locally elected
person, such as the village’s mayor. Village banks are stable institutions,
able to provide financial services to customers typically disregarded by
other MFIs. The main challenge facing village banks founded with inter-
national support is surviving the sponsor’s exit.
10 M. Hudon et al.
W: Women’s Empowerment
Many MFIs claim to target women. Nevertheless, recent evidence sug-
gests that microfinance commercialization is associated with a decline in
the share of female clients. A possible explanation is that women are
poorer than men on average and are therefore less profitable clients.
Accordingly, female borrowers and savers would be less attractive to
profit-oriented financial institutions. Another, and potentially comple-
mentary, rationale is that commercialization brings along the typical
biases plaguing the banking industry, such as unfair discrimination (see
“U”) of some groups of loan applicants. When women’s empowerment
stops being viewed as a moral priority, the common stereotype that links
business capacities and masculine traits makes it more difficult for female
entrepreneurs to raise capital and business loans. Further developments
on this issue could pay more attention to the characteristics, including
gender, of microfinance leaders, managers, and loan officers (see “L”)
who are willing to empower women. Hypothetically, gender affinity—
and, more generally, homophily—is a promising theoretical avenue to
understand the determinants of the segments served by MFIs.
X: The Unknown Author Who Will Answer All Our Questions
and Doubts
What is the future of microfinance? Are most MFIs doomed to become
either another type of bank or philanthropic institution? Will microfi-
nance institutions still exist in twenty years or will mobile banking and
traditional financial providers overtake the microfinance market? Will
subsidies in microfinance dry up soon? Will RCTs adopt generally agreed
ethical rules? Is green microfinance going to scale up and have a strong
environmental impact (see “B”)? Will anybody be able to be Yunus’s (see
“Y”) successor? Will commercialized microfinance actors totally integrate
the mainstream financial sector? Should MFIs focus on remittances? Will
the double-bottom-line narrative survive the move towards financial
inclusion? Will pro-social financial institutions learn from corporate-
governance crises? How should governments supervise MFIs properly?
These questions and many others are awaiting X!
Y: Yunus, Muhammad
Muhammad Yunus is the iconic figure of microfinance. Born in
1940 in Bangladesh, he was awarded the 2006 Nobel Peace Prize, together
with the Grameen Bank, for promoting the concept of microcredit. His
notoriety allowed the microfinance community to become a priority for
donors and enter business circles. Yunus is not only the central figure of
microfinance, but is also seen as the poster child for social entrepreneurship
and one of the most well-known among the Fellows of the famous Ashoka
network of social entrepreneurs.
Z: Zero Default
A striking figure of microcredit is, crises aside, its overall low probabil-
ity of default. Many leading MFIs have experienced a low portfolio at risk
compared to traditional commercial banks active in the same country.
Does this mean that zero default is the ultimate goal? Certainly not, for
two reasons: First, lending money is a risky business, and a zero-default
score would signal an overly restrictive lending technology. Targeting zero
default means taking very low risks, which can eventually harm the social
bottom line of MFIs. Second, a loan portfolio with zero default sounds
too good to be true, and probably is.
Note
1. The Banana Skins reports are published by the Centre for the Study of
Financial Innovation (CSFI), a non-profit think-tank, established in 1993.
They provide a barometer of the risks facing the banking, insurance, and
microfinance industries (CSFI website, consulted on 3 December 2018).
2
Addressing Climate Change
with Microfinance Plus: Experiences
in Cattle and Coffee Regions
of Nicaragua
Johan Bastiaensen, Milagros Romero,
and Frédéric Huybrechs
This chapter reflects on the role of rural microfinance in the context of

climate change. It develops a conceptual framework to inform strategies
of microfinance institutions (MFIs). This framework avoids too simple
standard panaceas and acknowledges these strategies’ inevitable taking
part in complex context-specific institutional bricolage, thus highlighting
J. Bastiaensen (*) • M. Romero

Institute of Development Policy (IOB), University of Antwerp,
Antwerp, Belgium
Institute of Research and Development Nitlapan, Universidad
Centroamericana (UCA), Managua, Nicaragua
e-mail: johan.bastiaensen@uantwerpen.be; mromero@nitlapan.org.ni
F. Huybrechs
Institute of Development Policy (IOB), University of Antwerp,
Antwerp, Belgium
e-mail: frederic.huybrechs@uantwerpen.be

14 J. Bastiaensen et al.
the importance of path-dependent local embeddedness. We illustrate the

framework with an analysis of the ‘green Microfinance Plus’ approach of
the Fondo de Desarrollo Local (FDL) and the institute Nitlapan-UCA
(Universidad Centroamericana) in Nicaragua. Their approach consists of
the combination of short term and/or investment credit with compulsory
technical assistance (which is partially or entirely paid by the clients,
depending on their wealth level). In some instances, it is also combined
with Payments for Ecosystem Services (PES). We first introduce our ana-
lytical framework and then present two cases in cattle and coffee regions.
The chapter concludes with insights about the role of microfinance in
the context of climate change.
2.1 C
hallenges of Climate Change Policies
in Rural Territories
Given the complexity and uncertainties involved in climate change, miti-
gation and adaptation cannot be translated into straightforward policies
or top-down social engineering of socio-institutional arrangements.
Attempts to promote either mitigation or adaptation will unavoidably
give rise to complex and interactive processes of ‘institutional bricolage’,
reflecting that different stakeholders will piece together justifying princi-
ples and other authoritative resources, informing evolving rules-in-use as
well as local social and individual practices (Cleaver, 2001). Path-
dependent and interactive institutional bricolage takes place in many dif-
ferent ‘political arenas’1 involving a variegated set of local and outside
stakeholders with unequal resources, knowledge and power. The out-
comes of institutional bricolage will condition the social ecological sys-
tem, that is the way in which nature and human livelihoods recursively
interact and co-shape the societal and ecological processes in a specific
territory (Ostrom & Cox, 2010), and thus contribute or not to mitiga-
tion or adaptation.
In the face of adaptation and mitigation, we are “dealing with the com-
plex dynamics of social and ecological processes, and the high degrees of
2 Addressing Climate Change with Microfinance Plus… 15
uncertainty in planning for sustainability in the current era” (Wise et al.,

2014, p. 327). This implies that we should avoid “to ‘close down’ too rapidly
to a small set of decision alternatives by reconfiguring uncertainty into more
manageable, but inappropriately narrow, calculations of risk and cost-bene-
fit equations” and rather open up “policy processes to wider participation,
thus increasing the diversity of values and ideas, as well as equity in decision-
making (Stirling, 2006)” (Wise et al., 2014, p. 327). In this perspective,
Leach, Scoones, and Stirling (2010) speak of sustainabilities (plural) and the
need for different pathways to sustainability, that is “alternative possible tra-
jectories for knowledge, intervention and change which prioritise different
goals, values and functions” (Leach et al., 2010, p. 5).
It is important to acknowledge that institutions, social structure and
perceptions are not exogenous but endogenous to the evolution of the
social ecological system (Bromley, 2012). One cannot identify any
exogenous ‘nature’, nor objective scientific knowledge on the ‘optimal
pathway of change’ to keep the social ecological system on a sustainable
path. There is significant room for inputs from scientific knowledge, as
“the human mind is indeed capable of constructing many different
models of nature, which it may then use to act upon nature, but this
does not preclude that ‘nature will talk back’ to our conceptions of it”
(Van Hecken, Bastiaensen, & Windey, 2015, p. 120). This ‘talking
back’ will lead human beings to ‘update their beliefs about nature’
(Bromley, 2012) and come up with adjusted models to address human-
nature interrelations.
We can thus envisage a socio-culturally embedded and dynamic-
evolutionary process of adaptive governance taking place to address the
perceived challenges of climate change. Whereas Leach et al. (2010) use
the concept ‘pathways to sustainability’ metaphorically, opening up a set
of imaginaries informing the multiplicity of ‘sustainabilities’, we give it
a more operational content. Inspired by De Haan and Zoomers (2005)
and ideas from the ‘Territorial Rural Development’ paradigm (Berdegué
& Schejtman, 2008), we adopt a complexity perspective to the meso-
level rural ‘territory’ (Ambrosio-Albalá & Bastiaensen, 2010), where
development pathways emerge as dominant over others as the outcome
of dynamic interactions of global and local actors with different ideas

and knowledge, (perceived) interests, resources and power within local
‘political arenas’ and geographical spaces characterised by concrete bio-
physical conditions.
The interactions between actors, and between actors and their physical
environments, take place within the earlier-mentioned process of power-
laden institutional bricolage (Cleaver, 2012; Van Hecken, Bastiaensen, &
Huybrechs, 2015) conditioning pathways grounded in “a set of shared
ideas that inspire the actions of the actors, their organisations and their
social networks, and rules of the game that govern the interactions among
the actors around certain economic activities” (Bastiaensen, Merlet, &
Flores, 2015). These pathways generate opportunities for certain house-
hold trajectories while limiting others and dynamically incorporate feed-
backs from those trajectories reproducing or changing prevailing ideas,
networks, organisations and rules of the game (Bastiaensen, Merlet, &
Flores, 2015; De Haan & Zoomers, 2005).
The emerging pathways obviously also condition the ecological out-
comes of the aggregated human activities in the rural territory. The ‘ter-
ritorial pathway’ concept contextualises individual decision-making and
opportunities as an at least partially collective and not always fully con-
scious process, where individual strategies are connected to shared liveli-
hood styles, that is “a specific cultural repertoire composed of shared
experiences, knowledge, insights, interests, prospects and interpretations
of the context; an integrated set of practices and artefacts (crop varieties,
instruments, cattle); a specific ordering of the interrelations with the mar-
kets, technology and institutions, and responses to policies” (De Haan &
Zoomers, 2005, p. 40). This approach explains why and how behaviour
of producers and households is dependent upon territorial structural
conditions beyond mere individual-household knowledge and access to
resources. Promoting transformations that address the challenges of cli-
mate change thus requires a thorough understanding of the dynamics of
these pathways processes to identify relevant inroads for feasible socially
embedded (path-dependent) change processes. This insight is crucial for
our reflection about the role of microfinance in addressing the challenges
of climate change in rural territories.
2.1.1 T
he Role of Rural Microfinance in a Climate
Change Context
The relatively recent debate about the role of microfinance in climate

change action is part of the broader argument about a third environmen-
tal bottom-line and ‘green microfinance’ (Fenton, Paavola, & Tallontire,
2015; Hammill, Matthew, & McCarter, 2008; Huybrechs, Bastiaensen,
& Forcella, 2015; Schuite & Pater, 2008). The initial models about the
environmental impact of microfinance (e.g. Agrawala & Carraro, 2010;
Fenton, Paavola, & Tallontire, 2017) distinguish a public and a private
policy realm. Speaking about adaptation, Fenton et al. (2017) introduce
the distinction between ‘planned’ and ‘autonomous’ adaptation, with the
former pertaining to the sphere of national states and its bi- and multilat-
eral allies (mobilising the needed funding) and the latter to the sphere of
households and enterprises. The role of microfinance is connected to
autonomous adaptation, as “microfinance is one key way of mobilizing
private finance (and channelling public finance) for autonomous house-
hold adaptation” (Fenton et al., 2015, quoted from Fenton et al., 2017,
p. 192). Credit and other financial products are to strengthen autono-
mous adaptation capacity of (poorer) households and enterprises,
enabling them to invest in adjusted livelihood activities and enhancing
resilience in the face of vulnerability and (survival) risk. Regarding miti-
gation, conditional (subsidised) credit is expected to increase the capacity
of households/enterprises to invest in environmentally sound productive
adjustments and can be used to push activities in a desired direction
(Cranford & Mourato, 2014).
Microfinance can be combined with other ‘plus’ services and incen-
tives such as awareness raising, technical assistance and PES. Such
‘Microfinance Plus’ configurations are set up to strengthen or guide
households and enterprises in the ‘right’ direction, contributing to moti-
vations, perceptions and knowledge, while modifying incentives.
Mainstream thinking about climate-related behavioural change builds
upon an individualistic economic-psychological conception of behaviour
(Shove, 2010), where policy interventions aim to affect either the
resources and economic incentives or the cognitive-motivational frame-
work of the individual clients involved, neglecting the socio-institutional
processes which contribute decisively to both. The individualistic models

of autonomous adaptation/mitigation actions are not necessarily incor-
rect, but clearly insufficient—considering the complex development
pathways described in the previous section. They suffer from what
Ostrom and Cox (2010) have called the panacea-problem, that is, uncrit-
ical use of reductionist models to apprehend complex interactive pro-
cesses which are ultimately inadequate to understand what really matters.2
An individualistic approach will not be able to sufficiently contribute to
the required profound processes of social ecological transformation
(Shove, 2010, p. 1278).
2.2 C
limate Change in Coffee and Cattle
in Nicaragua: Evolving Knowledge
Frames and Emerging Actions
We present two cases regarding microfinance and climate change in rural
Nicaragua. They are based on experiences of the Fondo de Desarrollo Local
(FDL), one of the largest microfinance institutions in Nicaragua. Our
analysis is based on field research in the coffee region of La Dalia (see
Bastiaensen, Huybrechs, Forcella, & Van Hecken, 2015; Huybrechs,
Bastiaensen, Forcella, & Van Hecken, 2016; Romero et al., 2016) and in
the cattle region of Matiguás-Rio Blanco (see Bastiaensen, Merlet, &
Flores, 2015; Polvorosa, 2013).
FDL has a strong commitment to agriculture, currently destining 35%
of its portfolio to primary agricultural production. Given its proximity to
peasants, FDL has constantly faced the risks of agriculture: variability in
yields and prices and a high vulnerability with regards to plagues, dis-
eases, weather conditions and conditions for access to markets (Peck &
Pearce, 2005). Rather than avoid agriculture, FDL has tried to reduce the
risk of its agricultural portfolio, in collaboration with the research and
development institute Nitlapan-UCA, specialised in providing technical
assistance and doing research in rural territories.
From the combined perspective of risk management and sustainable
development, FDL and Nitlapan-UCA have developed a Microfinance
Plus approach promoting silvopastoral and agroforestal production

systems. This model consists in delivering a combination of credit (by
FDL) and technical assistance (by Nitlapan-UCA). In some cases, this
model also included benefits such as reduced interest rates and/or eco-
nomic incentives conditional upon specific land-use changes (similar
to the idea of PES).
The increasing threat of climate change and its effect on rural families´
vulnerability motivates both institutions to keep expanding and improv-
ing their Microfinance Plus approach and to act towards climate change
mitigation and adaptation. In addition to their own efforts, both institu-
tions have also been involved in international projects which allowed
them to experiment and learn. These include the Global Environment
Facility’s Central American Markets for Biodiversity (CAMBio), a green
microfinance project which ran from 2009 until 2013. Nitlapan-UCA
also participated in the Regional Integrated Silvopastoral Management
Programme (RISEMP), which was a World Bank-Global Environment
Facility PES project running from 2002 to 2007.
2.2.1 Climate Change Adaptation in Coffee
2.2.1.1 Climate Change and Coffee Production
In the Nicaraguan coffee producing regions, climate change is mainly

considered from an adaptation perspective. Coffee production is less
often associated to deforestation and greenhouse gas (GHG) emissions,
especially in comparison to cattle raising. When associated to mitigation
efforts, coffee is rather perceived as a potential positive contribution to
carbon capture as compared to the rivalling food staple, vegetables or
cattle production.3 With regards to adaptation, the key perceived prob-
lem is the (present and expected) gradual rise in temperature and an
increase in rainfall variability. These evolutions make the future of coffee
production in Nicaragua increasingly uncertain. The International Center
for Tropical Agriculture (CIAT) estimated that with the forecasted annual
rise of 2.1 degree Celsius by 2050 it will no longer be possible to produce
coffee in Nicaragua under 700 m.a.s.l. (Läderach et al., 2012); a threshold
below which a large portion of coffee production is currently located.

This expected impact of climate change also threatens hundreds of liveli-
hoods in the coffee producing region of La Dalia, the municipality in
Northern-Central Nicaragua where our research took place. Subsequently,
it poses a risk to the predominant coffee portfolio of the MFIs in that
region. In 2016, around 90% of all FDL loans in La Dalia were destined
to coffee production.
Local farmers and MFIs no longer perceive climate change as a merely
hypothetical, medium-term threat. Indeed, present-day negative effects
on coffee yields are attributed to rising temperatures and rainfall variabil-
ity (Ramírez, Ordaz, Mora, Acosta, & Serna, 2010). Changes in tempera-
ture and rainfall patterns are expected to alter the production cycle
(flowering, berry growth and ripening), which in turn would result in
fewer, smaller and lower quality grains and consequently reduce coffee
income (Bedmar Villanueva, López Noriega, Bucardo, & van Zonneveld,
2016). Furthermore, altering weather conditions also induce a higher inci-
dence of diseases, leading to yield reductions and the loss of plantations.
The most severe crisis in the Central American coffee sector in years—the
roya or coffee leaf rust outbreak—is often being attributed to climate
change (Avelino et al., 2015; Mendoza, 2013). After the 2011–2012 har-
vest, the disease left 20% of national coffee in need of renovation (Avelino
& Rivas, 2013). Clearly, these real and perceived negative effects of cli-
mate change further complicate the inclusion of poor families in the coffee
sector. Indeed, coffee farmers were already facing vulnerability to price
fluctuations, challenging market access conditions and poor governance
and lack of transparency in cooperatives (Mendoza, 2017).
The dominant reaction to the coffee rust was the conversion of ‘old
plantations’ into intensified monoculture coffee plantations. It promoted
an aggressive change to higher yielding and more input-dependent coffee
varieties (in particular Catimor), also held to be resistant to roya. Given
higher plant densities, these varieties are also considered adequate to
address reduced productivity. An alliance of local (chemical) input sup-
pliers, mainstream processing-exporting companies and (larger) entrepre-
neurial coffee plantation owners promoted this conversion. This alliance
represents the actor coalition which engendered the ‘monoculture
athway’, which focuses on producing high volumes of undifferentiated

p
coffee.4 This is the dominant pathway in most of the coffee territories of
northern Nicaragua.
There are several factors that led to this monoculture production sys-
tem becoming the dominant development pathway. First, this produc-
tion system contributes to export-oriented production, where the
concentration of power in the value chain lies in few processing and trade
companies. Furthermore, the monoculture coffee production system
aligns with expectations of cheap food products and short-term produc-
tivity thinking (IPES-FOOD, 2016). However, field observation has
indicated that the promotion of this development pathway also implicitly
generates exclusion from significant outside support of families of smaller
peasant producers. They cannot or do not want to follow the monocul-
ture pathway. This can be due to having insufficient means to keep up
with the required technical-managerial practices and/or having other
complementary motivations and values, including food security and risk-
reducing strategies. One can also observe that the search for greater short-
term production does not consider the negative effects this generates in
the short and long term, in terms of land concentration, soil fertility, loss
of biodiversity, water pollution and so on, which are associated with the
high use of agrochemicals and the relative reduction of shadow trees
which the high yield, intensive monoculture requires.
2.2.1.2 Th
e Microfinance Plus Approach to Support Coffee
Farmers’ Adaptation to Climate Change
For its importance as one of the main export products in Nicaragua, the
coffee sector has a privileged position in credit markets. It is one of the
rural sectors with most access to credit, including microcredit. The coffee
rust crisis, exacerbated by production loss due to excessive rainfall during
coffee ripening in 2015, generated concern in the microfinance sector. As
a reaction to these crises, FDL expanded its Microfinance Plus package
(credit and technical assistance) to all its coffee credits. Already before the
coffee rust crisis, FDL and Nitlapan-UCA had been implementing this
model in the context of the CAMBio project. This project aimed at

enhancing biodiversity in Central America through environmentally
conditioned microfinance. In this project, FDL’s strategy consisted in
complementing credit and technical assistance with environmentally
conditioned grants (bio-premiums, as some sort of PES amounting to
14% of the loan principal) for the establishment of agroforest systems in
coffee plantations.
In 2013, because of the coffee rust crisis and the end of the CAMBio
project, the declared aim was adjusted. Whereas the earlier Microfinance
Plus package focused on promoting biodiversity and good environmental
practices, the strategy now mainly aimed at helping families cope with
climate change-related hazards. More concretely, FDL and Nitlapan-
UCA wanted to help reduce the families’ income vulnerability—related
to yield reductions or damages in the plantations—through the renova-
tion of coffee plantations and the improvement of their productivity. This
also included a change to the allegedly more rust-resistant Catimor vari-
ety. The yield-oriented Nitlapan-UCA technicians had already promoted
this variety before the crisis for its capacity to adapt to higher tempera-
tures and its higher yield when the right techniques are adopted. This
‘technification’ includes an increased exposure to sun and the utilisation
of more chemical inputs to guarantee at least two fertilisations and two
leaf protection cycles.
To make this technical assistance model financially sustainable for the
technical assistance provider, there are in principle four to five farm visits
per client of which the producers pay between 50% and 100% of the
costs, depending on the credit amount. The credit official’s performance
is measured in terms of credit allocation and repayment, and the techni-
cian’s performance is measured in terms of yields increased. Therefore, to
guarantee the expected targets, the topics of the visits are defined before-
hand by the technical staff, while the last visit consists in a yield estima-
tion. The model strives for a synergy between technical assistance and
client monitoring in view of risk reduction, both for the farmer and for
the MFI. Despite methodological difficulties to determine what changes
can be attributed to the financial and technical services, Marín (2016)
suggests that the FDL and Nitlapan-UCA model had a positive impact
on coffee yields during 2013–2016. Indeed, the model appeared to have a
clear positive impact for FDL in terms of repayment rates as well as a

reduction of its transaction costs. The visits of the extension officer—
mostly paid by the producers—reduce the number of visits by credit offi-
cials, moderate the incidence of coffee rust in existing plantations
(Mendoza, 2013) and make the renewal of loans faster (Marín, 2016;
Romero, Flores, & Bastiaensen, 2016).
The analysis of this experience in La Dalia showed that these policies
were very much in line with the reaction to the coffee rust crisis through
denser and input-intensive Catimor plantations, which is the kind of solu-
tion proposed by the powerful coffee stakeholders who are wed to the dom-
inant coffee monoculture pathway we discussed earlier. This has both
social and ecological implications. Socially, this approach tends to exclude
smaller farmers who face many constraints to adopt and benefit from coffee
monoculture (Bastiaensen, Huybrechs, Forcella, & Van Hecken, 2015;
Bastiaensen et al., 2016) and who often have more variegated motivations.
Therefore, they historically tend to prefer more diversified, agroforestry
production systems, crystallised in what could be identified as a subaltern
diversification pathway, which features cocoa, plantains, fruits, staples
and/or cattle besides coffee (sometimes ‘specialty’ coffee) as an integrated
production system (Bastiaensen, Huybrechs, Forcella, & Van Hecken, 2015).
FDL’s original Microfinance Plus package does not match well with
these peasant strategies, but rather illustrates how the dominant pathway
approach to the coffee rust crisis becomes ‘naturalised’ and garnered deci-
sive support from FDL/Nitlapan-UCA. This becomes clearer when we
look at the distribution of credits and evaluate the lenses with which
FDL/Nitlapan-UCA look at their credit and technical assistance strate-
gies. The proportion of clients classified as ‘subsistence farmers’5 in the
portfolio is higher in areas where also subsistence farmers are more prone
to focusing on coffee rather than the more diversified pathway (Romero
et al., 2016). This strong inclination towards supporting farmers who
have a stronger focus on coffee production rather than the ones with a
more diversified strategy indicates the problems of the narrow, stan-
dardised response model. It promotes only one exclusive strategy for
adaptation to the climate change-related coffee crisis and is incapable to
match the variegated strategies, motivations and perspectives of the dif-
ferent types of families present. Hence, poorer producers can be excluded
due to the observed mismatch between the promoted model and the
endogenous local dynamics. Additionally, they face barriers in being
recognised as creditworthy, as the cash crop coffee is the main measure for
estimating the repayment capacity.
From an ecological point of view, the adopted one-size-fits-all response
to the coffee rust outbreak and the expected changes in climatic condi-
tions pose serious questions and have generated concerns and doubts
about the adequacy of the initial standardised response, both in general
and in the FDL specifically. In light of the (uncertainty of ) climate change
impacts and the earlier-described preferences and limitations of small-
scale coffee/diversified farmers, increasing farmers’ dependency on coffee
and fertilisation could even be considered as a maladaptation strategy
(Fenton et al., 2017) or a negative adaptation strategy (Ellis, 2000). It
reduces households’ flexibility to adapt and it makes them more vulner-
able to future crises (Bacon, Sundstrom, Stewart, & Beezer, 2017). For
instance, there have been recent reports of coffee leaf rust affecting sup-
posedly more resistant Catimor plantations (Martínez, 2017). The inci-
dence of rust in this variety would be devastating given the increasing
importance and high presence that this variety currently has gained in the
region. Additionally, the focus on coffee rust has tended to have a blind-
ing effect on other diseases and plagues (such as ojo de gallo) which can
also be devastating to the roya-resistant varietals. Regardless of suchlike
sanitary crises, the future of coffee production in certain (lower) areas is
questionable altogether, hence the need to consider a more integrated
and radical approach (Bacon et al., 2017).
Another indication of a tendency to stick to a given intervention or pre-
scription is at the level of the technical assistance. If producers are not able
or willing to implement all the technician’s recommendations, and therefore
do not obtain the expected result, this is generally not interpreted as a failure
of the model to actually help them to adapt to climate change. Rather, it is
often attributed as a failure caused by alternative perceptions or motivations,
understood as a cultural resistance of the families to change and progress.
However, with the technical assistance topics being defined beforehand and
applied in a standardised fashion in different regions, the conditions for the
intended two-way dialogue between producers and technicians are absent.
In practice, this often ends up becoming a sort of monologue where the
technicians impose standardised one-size-fits-all prescriptions to all types of

producers. Receiving a relatively expensive ‘prescription’ (in terms of agro-
chemical inputs), with a high uncertainty of its pertinence to the farm’s
agro-ecological characteristics and the family’s needs, may restrict the farm-
ers to adopt the recommendations. In fact, an increasing number of clients
ask for the inclusion of soil studies as part of the technical assistance, in order
to have more appropriate and tailored recommendations. Moreover, fami-
lies take several factors into account, such as food security, availability of
family labour force or less stationary income. These variegated factors are at
the core of the diversification pathway and show that the adoption of a
given technology may not depend exclusively on the certainty of being func-
tional and/or profitable. The understanding of these factors requires more
dialogue than what has so far been observed in the technical assistance and
the policy prescription of credits for higher yielding coffee. It requires a bet-
ter match to the context-specific socio-institutional dynamics, while avoid-
ing the exclusive adoption of standard panacea.
This initial response might thus represent a classical example of the
“‘clos[ing] down’ too rapidly to a small set of decision alternatives” (Wise,
et al 2014: 327). It also makes an excellent case for ‘opening up’ “policy
processes to wider participation, thus increasing the diversity of values
and ideas in decision-making”. Clearly, the neglected ‘diversification
pathway’—which is preferred by the poorer peasant farmers and some-
times stigmatised as a remnant of an undesirable, backward past—regains
relevance. With increasing concerns over low altitude coffee, FDL and
Nitlapan-UCA are increasingly conscious of the risks that an exclusive
dependence on coffee engenders for their clients, their social mission and
their own financial sustainability. Currently, FDL is therefore opening up
space in its portfolio for diversification strategies and is trying to discern
changes needed in terms of financial products and practices. In La Dalia,
cocoa production now receives increasing attention as an alternative
adaptation strategy. However, there is a clear risk of falling into the trap
of thinking about this strategy with the same monoculture logic, pre-
tending that cocoa will become the coffee substitute in areas where it will
not be possible to continue producing coffee. It could be wiser to ‘open
up’ perspectives for different views on ‘pathways to sustainability’.
FDL and Nitlapan-UCA face strong short-term barriers to change the

tendency to define a small set of straightforward policies. One of them is
the financial sustainability of the provision of technical assistance. The
inclusion of other kind of studies, such as soil studies and spending more
time in the territories to be embedded of its dynamics, increases the
transaction and operational costs in the short term and further increases
the heavy work load of the technicians. Nevertheless, changes in the con-
ceptual framework of the model also imply changes in the operational
strategy, where less expensive and more collective strategies could be
found. As this would involve the inclusion of other stakeholders with
similar objectives this could contribute to a cost reduction.
2.2.2 M
itigation in Cattle Raising in Nicaraguan
Agricultural Frontier
2.2.2.1 The Environmental Pressure of Cattle Production
The problem of cattle raising in Nicaraguan agricultural frontiers is more

often related to mitigation than adaptation. This does not mean that
adaptation measures are not necessary; in the expanding dry zones, live-
stock is affected by water scarcity, climate irregularities and temperature
changes. However, one of the country’s biggest environmental problems
is the accelerated rate of forest destruction by the advancement of the
agricultural frontier, due to exploitation of timber and extensive livestock
production. This threatens some of the largest biosphere reserves in
Central America (the Bosawas Reserve in the North and the Indio-Maiz
Reserve in the South). The conversion of dense rainforest into crop land
and natural pastures contributes significantly to GHG emissions, while
methane emissions from poorly fed cattle add to the negative global
warming balance.
Historically, the availability of abundant land (Maldidier, 1993), the
consolidation of extractive export-oriented meat markets (Myers, 1981)
and patronage governance that involve dynamic interactions of actors
with unequal resources, knowledge and power (Maldidier & Martínez,
2015; Merlet, Collado Solís, Lemoine, & Polvorosa Narváez, 2015;
Polvorosa, 2013) are part of the institutional bricolage that engendered a

dominant ‘extensive livestock pathway’ in the agricultural frontier.
In this pathway, the guiding idea for a ‘successful producer’ is that of a
(very) large landholder with over hundred heads of cattle moving around
in huge farms in different climate zones, practising transhumance to deal
with rainy and dry seasons without any need to invest in on-farm fodder
production. A lack of (wage) labourers constrains on-farm intensifica-
tion, in particular in potentially more rewarding dairy production requir-
ing significant labour for milking, better grazing management and/or
fodder crop production and provisioning (Polvorosa, 2013).
The route to achieve this ‘success’ is the permanent accumulation of
land and cattle. As land remains relatively cheap (something that is slowly
changing), the strategy to accumulate land outcompetes any strategy to
produce more value added per unit of land (Maldidier, 1993). Economic
incentives and cognitive-motivational systems induce farmers to buy
more land, usually from more capital-constrained smaller farmers that
face soil fertility issues (due to insufficient rotation) and/or need more
land to guarantee their children’s heritage (Merlet et al., 2015). Given
capital constraints, these poorer farmers are dependent upon clientelistic
patronage relationships through which they can obtain cattle in a share-
cropping arrangement from the larger cattle ranchers allowing them to
make use of the underutilised pastures on eroded soils. These arrange-
ments consolidate the advantages of the larger cattle ranchers, appropri-
ating the larger part of the value added, thus often converting clientelistic
patronage into a prelude of future land purchase when the smaller farmer
decides to move further east (Merlet et al., 2015).
The influential capital-rich cattle ranchers that buy more land also
have decisive capacity to condition outsiders’ actions, thus generating
a socio-institutional dynamic that consolidates the ‘extensive live-
stock pathway’. This pathway generates opportunities for a small
number of large cattle ranchers, while imposing barriers to alternative
pathways for smaller producers. Many of them participate in an emer-
gent cocoa value chain, which offers a new high value-added alterna-
tive that allows consolidation of viable livelihoods with less land
(Maldidier & Martínez, 2015).
2.2.2.2 Microfinance Plus to Transform Extensive Cattle
More land-intensive, silvopastoral cattle systems offer an alternative to

reconvert livestock production into a more environment-friendly sys-
tem, while helping poorer farmers consolidate their farms and slow
down the advancement of the agricultural frontier (Pagiola et al.,
2007). The silvopastoral strategy consists in increasing trees in pastures
(fences and shadow trees), improving pasture management with more
divisions and rotational grazing as well as the cultivation of fodder
crops (Pagiola et al., 2007). The proposed change aims to increase on-
farm carbon capture and to protect water sources and soil fertility.
These technical changes produce better environmental outcomes (in
particular an expected decrease in farm expansion and deforestation)
and enhance livestock productivity, increasing the extremely low cur-
rent carrying capacity (less than one animal per hectare). Given these
perceived advantages, environmentally concerned stakeholders (non-
governmental organisations—NGOs; associations; cooperatives) pro-
mote the ‘silvopastoral package’ as both sound sustainable business
and good stewardship.
In line with these perspectives, FDL and Nitlapan-UCA have been
providing credit and technical assistance to cattle producers. Aware of the
environmental and social pitfalls of the ‘extensive livestock pathway’,
they are concerned about a risk to indirectly finance the advancement of
the agricultural frontier and have therefore adopted a stringent
‘Microfinance Plus’ approach in which cattle credit is conditioned to sil-
vopastoral intensification and related technical assistance. Policies pro-
hibiting loans for the purchase of land and a ‘no credit’ buffer zone of ten
miles around the Bosawas nature reserve were also adopted. There is also
increasing support for agroforestry diversification (including cacao). As it
has been argued that the improvement of farm profitability, protection of
water sources, carbon capture and soil fertility through introduction of
the silvopastoral systems may not be enough to incentivise producers
(Pagiola et al., 2007), the Microfinance Plus approach has been comple-
mented with PES when possible (in particular in the RISEMP and
CAMBio pilot projects).
However, the results of these policies are not satisfactory. Despite the
widely supported technical model, most producers seem to doubt its use-
fulness. It has indeed been difficult to find sufficient producers-clients,
really wanting to buy into the conditioned credit-technical assistance
package, with many of them often paying for, but not complying with
the silvopastoral technical recommendations.
The RISEMP project produced substantial changes towards silvopas-
toral practices, allowing it to be used as an exemplary success story pro-
moting market-based PES (Pagiola et al., 2007; Van Hecken, Bastiaensen,
& Windey, 2015). Yet, the changes cannot be strictly attributed to the
economic incentives or to the technical assistance provided by the project
(Van Hecken & Bastiaensen, 2010). The project indeed benefited from a
broader dynamic of dairy production caused by the expansion of collec-
tion centres and good prices for milk. The conclusion that behavioural
changes were solely motivated by individual PES-related economic incen-
tives and/or technical assistance was also mistaken. Indeed, the presence
of the project generated a collective local cognitive process, where infor-
mation and motivations were widely shared, thereby shaping all produc-
ers’ behaviour, as proven by the fact that producers who received neither
PES-incentives nor technical assistance introduced equally important
changes (Van Hecken & Bastiaensen, 2010). The significant temporal
presence of the PES project strongly influenced the overall acceptability
and circulation of silvopastoral ideas about the cattle-dairy production of
the future.
However, despite the substantial positive changes, the longer term
contribution of the project was less evident than suggested in its ‘epis-
temic circulation’ as a successful PES pilot project. Five years after the
RISEMP project ended, it was found that many silvopastoral investments
(pasture improvements and fodder crops) did not receive maintenance
and disappeared (Huybrechs, Bastiaensen, Forcella, & Van Hecken,
2015). After the end of the project, dominant ideas and narratives about
cattle raising slipped back in line with the extensive model (e.g. arguing
that water scarcity did not allow to sustain improved pastures and fodder
crops). Furthermore, the 2009–2010 negative shock in meat prices and
the related debt crisis (Bastiaensen, Marchetti, Mendoza, & Pérez, 2013)
brought farmers to re-exploit their ‘conservation areas’ and to sell trees

(Huybrechs, Bastiaensen, Forcella, & Van Hecken, 2015). Most significantly,
only five years after the termination of RISEMP, 29 of the 123 participants—
with one exception all small (18) and medium-sized (10) producers—had
sold their farms to larger producers and moved further east. So even while
RISEMP focused silvopastoral intensification upon smaller and medium-
sized producers, it was not successful in avoiding the logic of land concentra-
tion and the related out-migration of smaller farmers to the agrarian frontier
(Huybrechs, Bastiaensen, Forcella, & Van Hecken, 2015).
Following this analysis, we argue that change cannot be promoted
solely by a Microfinance Plus intervention at the individual client
level. Behavioural change will only be possible and sustained if broader
social, political and cultural transformations in the emergent territo-
rial pathways can be achieved, such that cognitive-motivational frame-
works as well as economic incentives are changed and no longer push
the expansion of the extensive system. To be more successful, the FDL/
Nitlapan-UCA Microfinance Plus approach should therefore articu-
late with other local and external stakeholders, helping to reshape ter-
ritorial governance. Producer strategies are not shaped from individual
interest and motivations only; local and national institutions and the
territorial governance also influence what is desirable and feasible for
them to do or not. Although credit, technical assistance and incentives
are definitely relevant and necessary to help families make the required
investments to consolidate alternative pathways, those services cannot
be simply designed from the outside and expect to succeed in promot-
ing change within the local communities (Van Hecken, Bastiaensen,
& Huybrechs, 2015). Policies looking to mitigate the environmental
impact of livestock production need to become embedded into the
existing socio-institutional arrangements which condition local gover-
nance and producer’s behaviour (Bermúdez et al., 2015; Van Hecken,
Bastiaensen, & Windey, 2015). Here it is needed to understand the
different dynamics and identify different pathways to sustainability,
according to families’ motivations, interest and values and the context
in which they develop their strategies.
2.3 Conclusions
Both the coffee and the livestock case-study have similar lessons about
the role of microfinance in climate change mitigation and/or adaptation.
First, it is clear that a merely individualistic technical-economic client-
oriented approach falls short of capturing the crucial links between the
changes in individual behaviour that are promoted through access to
financial and non-financial services and its interconnection with the
broader enabling and constraining territorial pathways in which they are
inevitably embedded. Substantial and sustainable changes in livelihood
strategies as well as impact on a required territorial scale depend upon the
transformation of these pathways.
Clearly, it is beyond the scope of microfinance alone, and even beyond
that of ‘Microfinance Plus’ alone, to engender such a more encompassing
transformation of the dominant territorial pathways. Yet, since microfi-
nance’s environmental impact depends upon the interaction of its ser-
vices with the pathway dynamics, it cannot ignore these. Also,
microfinance inevitably interacts with, and thus also affects, the relevant
power-laden processes of institutional bricolage. This implies that micro-
finance can (attempt to) play a role in helping to change prevailing local
perceptions and to redress prevailing power imbalances, allying with less
powerful stakeholders. Similarly, it could also fall into the trap of (un)
consciously siding with the more powerful stakeholders promoting the
dominant, naturalised pathways.
Another important dimension of the role of microfinance (plus) in the
context of climate change is that many issues concerning the best way to
adapt to the consequences and to mitigate the causes of climate change
are not clear at all. We face a challenge with intrinsic uncertainty as well
as variegated perceptions and interests with respect to what has to be
done. In the face of the perceived threat of rising temperature, should we
adapt coffee production systems or diversify and even abandon coffee
altogether? Or similarly, in the face of deforestation and cattle-related
GHG emissions, should we change the nature of monoculture cattle rais-
ing, promote diversified agricultural-cattle systems or even abandon cattle
altogether? Some of the emerging literature about the role of microfi-

nance in climate change adaptation or mitigation often implicitly (but
wrongly) assumes that the stakeholders know and agree what to do in
reaction to the emergent problems. In practice, this is clearly not the case
as we are in a realm of adaptive governance with gradual learning, evolv-
ing perceptions and shifting consensus, mediated by differentiated—local
and possibly global—(perceived) interests and motivations.
Here, one key lesson is to avoid “closing down too quickly and nar-
rowly” on specific strategies, in particular if they are promoted as ‘the’
solution to the emerging climate change problems or opportunities. This
risks a too exclusive support for particular (almost by definition more
powerful) stakeholders, while not necessarily contributing to increase the
capacity of local societies in the face of the climate change challenges. We
know from complexity science that resilience of a system depends upon
its diversity and its flexibility, not upon its being wed to one particular
rigid ‘solution’. A more diverse approach is thus advised and often will
have the additional advantage of being able to build on and ally with a
more diverse set of stakeholders, including the less visible and less power-
ful ones who tend to be neglected in the dominant territorial pathways.
This diverse approach can not only involve individual producers, techni-
cians and microfinance institutions; it requires the involvement of a more
diverse set of stakeholders, value chain actors, local and national govern-
ments, NGOs, cooperatives, among others, which shape overall local ter-
ritorial dynamics.
Notes
1. Concept taken from Bierschenk and Olivier de Sardan (1998, p. 240),
defined as “a place of concrete confrontation between social actors inter-
acting on common issues”. The concept underlines that bargaining pro-
cesses do not only take place within ‘political’ bodies, like parliaments or
village councils, but in every ‘real’ meeting place of actors around resources
or opportunities.
2. This resonates with the well-known finding that ‘microcredit alone’ is not
a panacea for poverty reduction and that attributing impact is difficult
due to complex interactions with mediating processes (Garikipati,

Johnson, & Guérin, 2017).
3. Although specialised studies (Van Rikxoort & Schroth, 2014) indicate
differences between different types of coffee production systems in terms
of their contribution to carbon capture—with a clear advantage for more
diversified, dense peasant systems over monocropping coffee planta-
tions—this is not very present in the current debate on the ground.
4. The alternative strategy, often promoted in the Fair Trade cooperative
movement, is to build upon differentiated ‘specialty’ coffee with social,
ecological or other attributes. This strategy is less focused on increasing
yield and aims to increase value added per unit of coffee produced.
5. FDL categorises its clients on the base of their estimated total capital:
campesino de subsistencia (subsistence peasant), campesino finquero (peas-
ant farmer), finquero (capitalised farmer) (Romero et al., 2016).
References
Agrawala, S., & Carraro, M. (2010). Assessing the Role of Microfinance in Fostering
Adaptation to Climate Change. OECD Environmental Working Paper No.
15, OECD Publishing.
Ambrosio-Albalá, M., & Bastiaensen, J. (2010). The New Territorial Paradigm of
Rural Development: Theoretical Foundations from Systems and Institutional
Theories. Antwerp: IOB Discussion Paper 2010.02.
Avelino, J., Cristancho, M., Georgiou, S., Imbach, P., Aguilar, L., Bornemann,
G., et al. (2015). The Coffee Rust Crises in Colombia and Central America
(2008–2013): Impacts, Plausible Causes and Proposed Solutions. Food
Security, 7(2), 303–321.
Avelino, J., & Rivas, G. (2013). La roya anaranjada del cafeto. Retrieved from
http://hal.archives-ouvertes.fr/hal-01071036
Bacon, C., Sundstrom, W., Stewart, I., & Beezer, D. (2017). Vulnerability to
Cumulative Hazards: Coping with the Coffee Leaf Rust Outbreak, Drought,
and Food Insecurity in Nicaragua. World Development, 93, 136–152.
Bastiaensen, J., Huybrechs, F., Forcella, D., & Van Hecken, G. (2015).
Microfinance Plus for Ecosystem Services: A Territorial Perspective on
Proyecto CAMBio in Nicaragua. Enterprise Development and Microfinance,
26(3), 292–306.
Bastiaensen, J., Marchetti, P., Mendoza, R., & Pérez, F. (2013). After the
Nicaraguan Non-Payment Crisis: Alternatives to Microfinance Narcissism.
Development and Change, 44(4), 861–885.
Bastiaensen, J., Merlet, P., & Flores, S. (2015). Rutas de Desarrollo en Territorios
Humanos: Las Dinámicas de la Vía Láctea en Nicaragua. Managua: UCA
Publicaciones.
Bastiaensen, J., Huybrechs, F., Bermúdez, M., Flores, M., Romero, M., &
Ramírez, J. (2016). Planting the Seeds of Inclusion: An Alternative Pathway
for Coffee Productions. Passerelles, 3, 19–25.
Bedmar Villanueva, A., López Noriega, I., Bucardo, E., & van Zonneveld, M.
(2016). Caficultura y cambio climático en Nicaragua: Reflexiones sobre el papel
de las políticas públicas y las cooperativas. Rome: Bioversity International.
Berdegué, J., & Schejtman, A. (2008). Towards a Territorial Approach for Rural
Development. Manchester: IDPM, School of Environment & Development.
Bermúdez, M., Flores, S., Romero, M., Bastiensen, J., Merlet, P., Huybrechs, F.,
et al. (2015). ¿Es posible financiar la gandería en la fronter agrícola de Nicaragua
de manera sostenible? Managua: Policy Brief, Nitlapan.
Bierschenk, T., & Olivier de Sardan, J. (1998). Les Pouvoirs au Village: Le Bénin
Rural entre Démocratization et Décentralisation. Paris: Karthala.
Bromley, D. W. (2012). Environmental Governance as Stochastic Belief
Updating: Crafting Rules to Live. Ecology and Society, 17(3), 14.
Cleaver, F. (2001). Institutional Bricolage, Conflict and Cooperation in Usangu,
Tanzania. IDS Bulleting, 32(4), 26.
Cleaver, F. (2012). Development Through Bricolage: Rethinking Institutions for
Natural Resource Management. Abingdon: Routledge.
Cranford, M., & Mourato, S. (2014). Credit-Based Payments for Ecosystem
Services: Evidence from a Choice Experiment in Ecuador. World Development,
64, 503–520.
De Haan, L., & Zoomers, A. (2005). Exploring the Frontier of Livelihoods
Research. Oxford: Institute of Development Studies.
Ellis, F. (2000). A Framework for Livelihood Analysis. In F. Ellis (Ed.), Rural
Livelihoods and Diversity in Developing Countries (pp. 28–51). Oxford:
Oxford University Press.
Fenton, A., Paavola, J., & Tallontire, A. (2015). Microfinance and Climate
Change Adaptation: An Overview of the Current Literature. Enterprise
Development and Microfinance, 26(3), 262–273.
Fenton, A., Paavola, J., & Tallontire, A. (2017). The Role of Microfinance in
Household Livelihood Adaptation in Satkhira District, Southwest
Bangladesh. World Development, 92, 192–202.
Garikipati, S., Johnson, S., Guérin, I., & Szafarz, A. (2017). Microfinance and
Gender: Issue, Challenges and the Road Ahead. The Journal of Development
Studies, 53(5), 641–648.
Hammill, A., Matthew, E., & McCarter, E. (2008, September). Microfinance
and Climate Change Adaptation. IDS Bulletin, 39(4).
Huybrechs, F., Bastiaensen, J., & Forcella, D. (2015). Guest Editorial: An
Introduction to the Special Issue on Green Microfinance. Enterprise
Development and Microfinance, 26(3), 211–214.
Huybrechs, F., Bastiaensen, J., Forcella, D., & Van Hecken, G. (2015).
Enfrentando la vía ganadera extensiva: potencialidades y limitaciones de los
pagos por servicios ambientales y de las microfinanzas verdes. In J. Bastiaensen,
P. Merlet, & S. Flores (Eds.), Ruta de desarrollo en territorios Humanos: Las
dinámicas de la vía láctea en Nicaragua (pp. 373–402). Managua: UCA
Publicaciones.
Huybrechs, F., Bastiaensen, J., Forcella, D., & Van Hecken, G. (2016). La
microfinance pour les services environnementaux: Enseignements en matiere
de politiques du Proyecto CAMBio au Nicaragua. Revue Tiers Monde, 125–154.
IPES-FOOD. (2016). From Uniformity to Diversity: A Paradigm Shift from
Industrial Agriculture to Diversified Agroecological Systems. International Panel
of Experts on Sustainable Food Systems.
Läderach, P., Zelaya, C., Ovalle, O., Garcia, S., Eitzinger, A., & Baca, M.
(2012). Escenarios del impacto del clima futuro en áreas de cultivo de café en
Nicaragua. Cali, Colombia: CIAT.
Leach, M., Scoones, I., & Stirling, A. (2010). Dynamic Sustainabilities:
Technology, Environment, Social Justice. London: Earthscan.
Maldidier, C. (1993). Tedencias actuales de la frontera agrícola en Nicaragua.
Managua: Nitlapan.
Maldidier, C., & Martínez, S. (2015). Apoyar una ruta campesina de desarrollo
en una antigua frontera agrícola: dinámicas territoriales y prácticas de desar-
rollo en Río Blanco. In J. Bastiaensen, P. Merlet, & S. Flores (Eds.), Rutas de
Desarrollo en territorios humanos: Las dinámicas de la vía Láctea en Nicaragua
(pp. 313–351). Managua: UCA-Publicaciones.
Marín, Y. (2016). Evaluando la sinergia entre los servicios financieros y no financie-
ros en la producción cafetalera. Managua: Nitlapan – UCA.
Martínez, L. E. (2017, Abril 23). Alerta por amenazas a plantíos de café en
Jinotega y Matagalpa. La Prensa. Retrieved from http://www.laprensa.com.
ni/2017/04/23/departamentales/2218524-alerta-amenazas-plantios-cafe-
jinotega-matagalpa
Mendoza, R. (2013). El Café en tiempos de la roya. Envío Ed. 372.

Mendoza, R. (2017). Toward the Reinvention of ‘Fair Trade’. Retrieved from
http://peacewinds.org/wp-content/uploads/2017/02/articulo-comercio-
justo-version1-ene-2017-1-eng.pdf
Merlet, P., Collado Solís, C., Lemoine, L., & Polvorosa Narváez, J. (2015).
Acceso a tierra y rutas de desarollo en el municipio de Río Blanco. In
J. Bastiaensen, P. Merlet, & S. Flores (Eds.), Rutas de desarrollo en territorios
humanos: Las dinámicas de la vía láctea en Nicaragua (pp. 191–228). Managua:
UCA Publicaciones.
Myers, N. (1981). The Hamburger Connection: How Central America’s Forest
Become North America’s Hamburgers. Ambio, 10(1), 3–8.
Ostrom, E., & Cox, M. (2010). Moving Beyond Panaceas: A Multi-Tiered
Diagnostic Approach for Social-Ecological Analysis. Environmental
Conservation, 37(4), 451–463.
Pagiola, S., Ramírez, E., Gobbi, J., de Haan, C., Ibrahim, M., Murgueitio, E.,
et al. (2007). Paying for Environmental Services of Silvopastoral Practices in
Nicaragua. Ecological Economics, 64, 374–385.
Peck, C., & Pearce, D. (2005). Gestión de riesgos y diseño de productos para las
microfinanzas agrícolas: Características de un modelo emergente. Estudio
Especiales CGAP, (11), 1–60.
Polvorosa, J. (2013). Opportunities and Constraints for Small and Medium-Sized
Farmers in the Context of Booming Dairy Value Chains in Nicaragua: Case-
Study of Matiguás. Antwerp: University of Antwerp, IOB.
Ramírez, D., Ordaz, J., Mora, J., Acosta, A., & Serna, B. (2010). Nicaragua:
Efectos del cambio climático sobre la agricultura. México: CEPAL.
Romero, M., Flores, S., & Bastiaensen, J. (2016). Entendiendo las complejidades
del territorio, las oportunidades de desarrollo y los desafíos para las microfinanzas
Plus. Managua: Cuaderno de Investigación No. 53. UCA Publicaciones.
Schuite, G., & Pater, A. (2008, December). The Triple Bottom Line for
Microfinance. Microfinance Getaway.
Shove, E. (2010). Beyond the ABC: Climate Change Policy and Theories of
Social Change. Environment and Planning, 42, 1273–1285.
Stirling, A. (2006). Analysis, Participation and Power: Justification and Closure
in Participatory Multi-Criteria Analysis. Land Use Policy, 23(1), 95–107.
Van Hecken, G., & Bastiaensen, J. (2010). Payments for Ecosystem Services in
Nicaragua: Do Market-Based Approaches Work? Development and Change,
13, 421–444.
Van Hecken, G., Bastiaensen, J., & Huybrechs, F. (2015). What’s in a Name?
Epistemic Perspective and Payments for Ecosystem Services Policies in
Nicaragua. Geforum, 63, 55–66.
Van Hecken, G., Bastiaensen, J., & Windey, C. (2015). Towards a Power-
Sensitive and Socially-Informed Analysis of Payments for Ecosystem Services
(PES): Addressing the Gaps in the Current Debate. Ecological Economics,
120, 117–125.
Van Rikxoort, H., & Schroth, G. (2014). Carbon Footprints and Carbon Stocks
Reveal Climate Friendly Coffee Production. Agronomy for Sustainable
Development, 34(4), 887–897.
Wise, R., Fazey, I., Stafford Smith, M., Park, S., Eakin, H., Archer Van Garderen,
E., et al. (2014). Reconceptualising Adaptation to Climate Change as Part of
Pathways of Change and Response. Global Environmental Change,
28, 325–336.
3
Remittances: Loan Funds for a Rural
Economy? Evidence from the Kayes
District (Western Mali)
Michel Namé and Philippe Lebailly
3.1 Introduction
Remittances from migrants represent a considerable financial windfall
according to the African Development Bank (AfDB, 2007, p. 61) and the
World Bank (2015, pp. 4–5; 2017, pp. 2–3). Most estimates of remitted
amounts are based only on remittances through formal channels (money
transfer providers, banks and post offices).1 The assessment of flows through
informal channels is usually done through banknote swaps (AfDB, 2007,
note3, p. 69). However, according to Freund and Spatafora (2005, p. 2),
migrant remittances represent between 50% and 250% of the estimated
amounts. Today, the concerns of international organizations, states and
financial institutions only focus upon the minimization of transfer costs in
order to reduce the use of informal channels and the risks associated with
them (Bourenane, Bourjij, & Lheriau, 2011, pp. 110–112).
M. Namé (*) • P. Lebailly

Faculty of Gembloux Agro-Bio Tech, University of Liège, Liège, Belgium
e-mail: jmname@doct.uliege.be; Philippe.Lebailly@uliege.be

40 M. Namé and P. Lebailly
Indeed, the reduction of these costs allows better quantification of the

flows and an increase in the positive impact of remittances on the stan-
dard of living of households according to the AfDB (2007, p. 6). Despite
their large volume, remittances may not be a substitute for a sustainable
and liberating development effort (Gupta, Patillo, & Wagh, 2007, p. 43),
as it is accepted that a considerable part of the windfall is consumed
instantly by beneficiaries including family members and local communi-
ties in the form of food, school and emergency health expenditures
(AfDB, 2007, p. 23). While long-term economic development, reducing
poverty and inequality in rural areas2 also require productive investment
(World Bank, 2006, pp. 149–150), Lucas (1985, p. 372) explains that
remittances have a negative impact on local agricultural production in
the short term. That said, when invested in the long run, they allow
increased cereal production and facilitate increased livestock numbers in
sub-Saharan African countries.
Notwithstanding this fact, Martin, Martin, and Weil (2002) cited by
Rocher and Pelletier (2008, p. 31) indicate that less than 10% of remit-
tances are allocated to local investment in the Comoros and Mali. The
incentive for productive investment by households is therefore through
financial inclusion policies (Rocher & Pelletier, 2008, p. 33) via microfi-
nance institutions (MFIs).3 Gupta et al. (2007, p. 42) have shown a posi-
tive and significant impact of remittances on financial development in
beneficiary environments. Remittances thus serve as a guarantee of access
to microfinance for the improvement of agricultural production equip-
ment, housing, health and education. But the prerequisites for providing
this type of service are a stable financial situation and a rigorous organiza-
tional structure, indeed what most MFIs in sub-Saharan Africa do not
have, according to Sander (2004, pp. 12–13). The usefulness of microfi-
nance lies in its ability to improve family budgets and stabilize small local
production companies, but does not claim to fight effectively against pov-
erty (Fouillet, Guérin, Morvant-Roux, Roesch, & Servet, 2007, p. 333).
As a result of major climactic shocks, productive investment suffers and
microfinance does not, until now, have any effective means of absorbing
this type of shock according to the French Agricultural Research Centre for
International Development (CIRAD) (2002, p. 4). For Nabeth and Levy
(2007, p. 186), guaranteeing a level of economic development would
3 Remittances: Loan Funds for a Rural Economy?… 41
inevitably lead to a decrease in vulnerabilities in the rural world. The solu-

tion may come from crop insurance.4 A portion of the remittances could be
insurance contributions paid to financial institutions, banks and insurance
companies, who are responsible for making the transfers.5
Nevertheless, the complexity of crop insurance requires the active par-
ticipation of governments and the full involvement of private insurers
and financial markets (Nabeth & Levy, 2007, p. 187). However, subsi-
dies for a crop insurance policy would be difficult for governments to
support because of the very high proportion of farmers in the population,
from 30% to 70% depending on the beneficiary country. This is quite
apart from private insurers who fear fraud, moral risk and problems of
adverse selection. In the same way, rural populations may often be reluc-
tant to pay an insurance premium for an unpredictable future event
(ibid.: 190–191).
3.1.1 A Customized Financial Product: Savings
For Gubert (2008, p. 53) and Dendir (2017, p. 14), if it is not disputed
that remittances from migrants have a resilience function against exoge-
nous shocks, in this case climactic, it should therefore be readily accepted
that these transfers differ from other capital sources since they are essen-
tially intended for unsustainable and expenditures that are not profit-
oriented. As a result, vulnerable households need savings and insurance
services more than traditional financial products (Fouillet et al., 2007,
p. 343). Certain subsidies would be more effective in the fight against
poverty if they supported the implementation of savings networks using
migrant remittances.6 In relation to this, Gupta et al. (op. cit.) showed
that of households benefiting from remittances, the propensity to save is
around 40%. But the long-term difficulty is to lead these savings to the
financing of productive investments as a large part of remittance flows do
not pass through formal channels in sub-Saharan African countries.
However, the opposite may be possible in Mali, where the savings of
the (banked) migrant invested at a rate of 3.5%7 could be used to finance
a locally selected development project in the form of a loan at a borrow-
ing rate of 14% (Gubert, 2008, p. 54). But the risks incurred would not
be bearable for the borrower in the event of a shock. It is, therefore, the
importance of the share of potential savings resulting from remittances,
both formal and informal8 after their use that determines the possible
economic development of the rural environment (Dendir, 2017, p. 3;
Gupta et al., 2007, p. 40). Accordingly, is there really a saving capacity of
households regularly receiving remittances from migrants regardless of
the methods of use? If so, how is it to be formalized and used for local
development? To try to answer these questions, we will focus on the
Kayes region, Mali’s largest emigration basin, where remittances are usu-
ally received informally (AfDB, 2007, p. 23).
There is little research on the possibility of savings from remittances
received by recipient households, especially in West Africa. This chapter
also attempts to fill this gap. To do this, we carry out an analysis of the
composition of the Kayes village savings and self-managed credit banks
(CVECAs9).
In this way, a review of the history of the microfinance sector as the
main instrument of financial inclusion of the poorest populations in the
sub-Saharan context seems to be the prerequisite before the investigation
of the presence of any kind of savings in CVECAs, considered as the basic
microfinance institution model, and finally, its impact on agricultural
productive investment.
3.2 Microfinance
In the countries of the South, a significant proportion of the population
does not have access to the financial services of commercial banks.
According to Gentil and Servet (2002, p. 730), this share is estimated to
represent 90% or even 100% in rural Africa. Also, in Latin America
(Brazil), 85% of the adult population is reported not to have a bank
account despite the importance of the geographical coverage of financial
services.10 One would immediately want to think that banks are not very
fond of this kind of clientele (rural or urban working class). However,
this lack of interest on the part of banks in this category of the population
is not a priori due to subjective aversion, but clearly responds to a quasi-
inadequacy of their activities in relation to the banking habits of these
populations. The constraints of immediate profitability and management

rigidity, such as the use of prudential ratios, are characteristic of the way
banks operate. These are believed to systematically prevent any deposit or
withdrawal activity and the granting of small amounts of loans without
substantial guarantees and with precise geographical location. It is the
absence of these last two socio-economic aspects added to cultural tradi-
tions and solidarity which have shaped the financial practices within
these populations who have been traditionally disadvantaged by formal
finance. This led to the name informal finance taken up by Lelart (2006).
Servet (1990, p. 92) even described it as insubordination from the for-
mal system.
According to Lelart (2006, p. 5), informal finance is presented as a set
of savings and credit practices that are almost flexible and amorphous
since they do not necessarily replicate a preconceived “modus operandi”
styled approach. These practices are fundamentally based on relationships
of trust and specific to the stakeholders—borrowers and lenders—who
are likely to know each other personally and who are likely to decide for
themselves how these financial exchanges will work. There are two dif-
ferent types.
On the one hand, there are individual practices exercised by the money
keeper or custodian as well as by the tontinier, also known as a travelling
banker. The money keeper is a person of proven integrity recognized by
the community (the village) to whom the savings of the inhabitants are
entrusted. This could be a religious leader, a former civil servant or even
a successful businesswoman who undertakes to return the amounts paid
by savers on request, at any time and without any increase in savings, or
any kind of remuneration for the service rendered. The custodian could
also make loans in Niger, according to Tinguiri (1990, p. 190). Unlike
the custodian, the tontinier in charge of savings moves to potential savers,
traders in this case, at the markets to regularly recover identical amounts.11
The refund of the sums paid is made in full and for the period agreed in
advance. In addition, the tontinier receives a payment for the security
service he offers, withholding the amount of one instalment from the
total amount returned to the investor.
On the other hand, there are collective practices: tontines12 or credit
mutuals such as rotational savings and credit associations (ARECs)
according to Bouman (1977) cited by Lelart (2006, note 11). They bring
together people from the same social background,13 from the same class,
in the same geographical environment. Its least complex form—since
there are some very sophisticated ones (see Lelart, 2006, pp. 10–11)—
consists of bringing together a group of people—12 to make it simple—
each month, so 12 times in a row that would each pay the same amount
each time. At these meetings, the amount is collected by one of the group
members as they go along until the end of the cycle. It is as if the first
more advantageous beneficiary receives a loan from the other members
and then undertakes to repay it, whereas the last least advantageous to
receive the funds raised only recovered their deposit without any interest
related to the waiting time. The warm and friendly atmosphere around
this practice gives it a social rather than a financial character. The resump-
tion of the cycle, as well as the enlargement of the circle by welcoming the
new members, should be unanimously agreed by the group. In addition
to this description, loan sharks14 could be added as an individual practice
together with self-help associations as a collective practice in the informal
sector (Camara, 1991, p. 3).
That said, the loans granted or the savings mobilized are not expected
to have a real impact on economic development and the fight against
poverty. For Haudeville (1990, p. 80), informal savings are thought to
finance only informal sector activities. Citing Servet (op. cit.), “the infor-
mal is self-financed by the informal” well summarizes the fact. The eco-
nomic development that is necessary for the poor population requires an
accumulation of sufficient and appropriate financial resources to support
sustainable productive investment (Soedjede, 1990, p. 129). For Lelart
(2006, p. 20), informal finance cannot play this role of accumulation
apart from banks. In the informal sector, companies are created, lasting a
few years for the most efficient and disappearing to give way to others of
the same kind. “The future of the informal sector is to remain informal”
Haudeville (1990, p. 81).
Thus, microfinance has the potential to break this vicious circle created
and fuelled by the informal sector. It can, therefore, be the intermediary
sector between informal finance and the traditional banking system, the
link or bridge between the working class and commercial financial
institutions.
Labie (1999) provides the following definition:
Microfinance is the provision of financial services (usually credit and/or sav-

ings) to people engaged in productive activity, most often handicrafts or trade,
who do not have access to commercial financial institutions because of their
socio-economic profile. (As taken from Koloma, 2007, pp. 3–4)
This definition seems to state the role, delimit the scope of interven-
tion and define the overall objective of microfinance. However, Poursat
(2001) proposes a much more detailed definition:
Microfinance is the provision of sustainable financial services to poor clients

(including small self-employed workers or micro-entrepreneurs) who do not
have access to the formal banking system. These financial services are most often
credit and savings, but they may also be other specialized services (insurance,
credit, leasing, etc.). (Also taken from Koloma, 2007, p. 3)
The idea should, in fact, be to be able to direct financial resources from

the informal sector to the market economy so that they can also contrib-
ute to economic development and poverty reduction. As informal finance
focuses on savings, microfinance is believed to place greater emphasis on
the provision of microcredits, its main activity according to Guérin,
2000, p. 405). Microcredit refers to an amount borrowed from a micro-
finance institution (MFI) that generally does not exceed 50% of the gross
national product (GNP) per capita. According to Lelart (2006, p. 25),
the World Bank recommends a 30% cap.
As a result, to achieve the objectives of economic development and
poverty alleviation, an MFI should be able to deploy rapidly in space
and time, to reach poor and highly disadvantaged populations and thus
to improve their living conditions more or less significantly (Copestake,
2007, pp. 1723–24; Jégourel, 2008, p. 200). However, for Montalieu
(2002, p. 28), microfinance alone is not sufficient to meet all these per-
formance requirements, let alone to achieve the noble objective of pov-
erty alleviation. Already, traditional microcredit appears to be aimed
only at vulnerable populations on a recurrent basis, not at the poorest of
the poor. It could allow the moderately poor to diversify their sources of
income by mobilizing capital, but not to directly mitigate the emergence

of risks15 (Guérin, 2004, p. 25). Hofmann and Marius-Gnanou (2005)
cited by Koloma (2007, p. 2) believe that microcredit is not likely to
improve the living conditions of the poorest people since it does not
reach them. It may even increase the depth of poverty in the event of
over-indebtedness. Furthermore, the less-populated rural environment
with the highest number of poor people has actually little or no access to
the financial services offered by MFIs (Labie, 2004, p. 21). They prefer
richer areas, accentuating inequalities between already well-defined ter-
ritories and leading to risks of saturation of the supply of services and
over-indebtedness of clients (Guérin et al., 2007, p. 106). However,
according to the same authors, in the case of effective targeting of the
very poor, microfinance services are not well adapted to demand, which
happens to be very heterogeneous and complex. Gentil and Servet
(Gentil & Servet, 2002, pp. 730–731) and Granger (2006, pp. 89–90)
show that, in the end, there is an insidious junction between the fields of
MFIs and commercial banks. Indeed, while some MFIs are working in
their development to keep in mind their original credo of providing
financial services to those excluded from the banking system, others ulti-
mately now prefer to open up to the formal sector by offering services to
the upper class of the urban informal sector. However, this could be an
alternative attributed to an operational restructuring due to poor man-
agement and out of a concern for greater profitability. In addition, there
has also been a descent of some commercial banks into the MFI field,
hoping to capture the resources of the informal sector as well (Granger,
2006, p. 90).
Notwithstanding these negative aspects of MFIs, however, microfi-
nance is still better adapted to the needs of the poor than the various
development projects initiated in rural areas. It allows the value of work
to be recognized, especially by women, and the reduction of inequalities
within populations (Nowak (2005) cited by Koloma (2007, p. 2)).
Microfinance will continue to be the ideal and appropriate instrument
for the ambitious formalization of the informal sector. As a result, it
represents the only option for the banking of the rural and urban work-
ing class that has long been ignored by commercial banks (ibid.: 6, 17).
3.3 A Readiness to Save

3.3.1 The Kayes District
The Kayes District is located northwest of Bamako, in western Mali. It

includes 29 municipalities, with its capital of the same name. Located in
the Sahel zone on both sides of the Senegal River, its low agricultural
production due to various climatic uncertainties is not able to meet the
cereal needs of the population of about two million,16 most of whom are
Soninke (Gorse, 2008, p. 28; Loveluck, 2008, p. 9). According to
Loveluck (ibid.), this is probably the same as that used in Delville (1991),
being one of the factors explaining the high rate of emigration character-
istic of this region of Mali, which was even more marked from the early
1970s according to Gorse (ibid.). In this way, remittances from its nation-
als constitute the only economic asset of the region, its development
being mainly supported by the nearly 400 village migrant associations
(VMAs) created in France (Besson, 2008, p. 19) and gathered within the
“Coordination des associations pour le développement de la région de
Kayes en France” (CaderKaF). Indeed, in terms of its immigration catch-
ment area, Kayes is said to be in the lead amongst other regions of Mali,
with 80% of its emigrants living on French territory. As a consequence,
in this mythical corridor (France-Mali), Soninke represent a little more
than 90% of Malian migrants in the Paris-Montreuil region according to
Gubert (2000b) as cited by Loveluck (2008, p. 12). Each year it accounts
for two-thirds of the remittances by volume to Mali (AfDB, 2007,
pp. 11–12).
It was in March 199817 that the CVECAs network of the Kayes-
Yélimané districts was created under the impetus of three inter-village
migrant associations from Diama Djigui, Ordik and Benkadi, all from
the region and established in France. According to Loveluck (2008,
p. 21), migrant groups in village associations share a twofold concern
being to facilitate the integration of new migrants in terms of mutual
aid18 and solidarity on French territory and to participate in development
actions in their villages of origin (see also Gubert on co-development
(2008)). Previously, these activities, which were carried out differently
according to each association, were limited to the construction of reli-

gious buildings (mosques). They then evolved into the establishment of
health centres, education support programmes, water and electricity ser-
vices and agricultural development with an increase in livestock numbers,
digging of wells, development of cultivable land,19 to name just a few. In
2016, more than €32 million was injected into the local economy from
abroad using only formal channels. One of the particularities of the net-
work is the granting of short-term loans20 with an obligation to provide
substantial material guarantees due to the high level of migration in the
region (CERISE, 2006, p. 2). It is estimated to have about 50 branches
in the Kayes district and about 20 in the Yélimané district (Ponsot &
Obegi, op. cit.). According to the authors’ estimates, the village mutuals
have approximately 15,000 members,21 including 1100 migrants regis-
tered as such and 8000 others living in France.
Gorse (2008, p. 28) estimates the network’s financial flows at around
€6.1 million per village each year. At the end of 2009, the Kayes network
reportedly recorded a net surplus of nearly 1.9 billion CFA francs or around
3 million euros in savings, so it had to find other gainful savings-related
uses (Ponsot & Obegi, op. cit.). For Loveluck (2008, p. 116), the migration
context is necessarily the main factor in the network’s self-sufficiency.
Relying on Gorse (op. cit.) and the World Bank (2015, p. 12), migrants
from France are said to be reluctant to deposit their savings directly into
village banks, which they believe are less reliable in terms of providing secu-
rity guarantees. It is also known that there is a growing consensus on the
allocation of remittances: smoothing of food consumption, health, educa-
tion/training expenditures, family and individual real estate, productive
investment and social projects (AfDB, 2007, pp. 43–44; Azam & Gubert,
2005, p. 1332; Generoso, 2012, p. 13; Gubert, 2000a: 27). However, in
Kayes there is another item for allocating amounts transferred according to
Loveluck (op. cit.) and Ponsot and Obegi (2010, p. 20), being the “savings”
of beneficiary families released from the remittances—in this way, “migrant
savings”22 supply the bulk of the banks in the Kayes network.
We then hypothesize that there is an ability in Kayes to save money
received from abroad and even that this can be easily assessed, as a marginal
propensity to save, by strengthening the organization23 in fundraising and
recipient data collection based on Brown’s (1994) studies in Western Samoa
and Tonga (South Pacific); Richard (2002) in rural Pakistan and Dendir
(2017) in Ethiopia and Kenya. However, this is not exactly the same as
consent to Keynesian-like saving but a completely different kind of saving,
as Lelart (1990, p. 57) describes in these words:
In Africa, the act of saving is not in time but in space. The peasant or worker
does not seek to secure his future alone by secretly investing his money today to
dispose of it tomorrow. He privileges his social relationships, he takes root in a
group and he finds the security he needs from others. Saving is not an attitude
towards the passing of time, it is an attitude towards those who are close.
Servet (1990, p. 94) adds to this that “the fact of accumulating from an
individual point of view goods, assets, in savings accounts not only, very
largely does not make sense, but becomes a negative practice in the societies
under consideration.”
Thus, this ability to save is more like a social act than a quasi-normal
economic reflex.
Most people in Kayes are reported to live on the margins of poverty
according to Ponsot (2007, p. 55). That said, the low level of agricultural
production and low incomes from operations are likely to be offset by
higher monetary value and high purchasing power in the region (Loveluck,
2008, p. 20). As a result, some public servants are reportedly asking to be
reassigned to another part of the country because of the higher prices of
everyday consumer goods and services. However, while it is widely accepted
that the ability to generate family savings is the first step in any develop-
ment process (Lelart, 1990, p. 45; Dendir, op. cit.), it should improve the
level of production of cultivated food, if only for migrant families.
3.3.2 M
igration and Productive Agricultural
Investment
3.3.2.1 My Family, My Community, My Homecoming!24
For De La Brière, De Janvry, Lambert, and Sadoulet (1997, p. 3), migra-

tion might be summarized as an implied insurance contract between the
migrant and his or her family of origin. Having emigrated, the migrant
should be responsible for providing for the financial needs of his relatives
in terms of food, health, education and so on. In this way, the migrant
has no greater concern than taking care of his family and building com-
munity structures through his tacit participation in a village association.
Any other objective is actually only secondary (Ndione & Lombard,
2004, p. 12). Also, according to Bloy and Dupuy (1990, p. 67):
“Membership in associations does not come from a voluntary process, it
is imposed on individuals. They are not expected to freely adhere to the
group’s objectives, but social pressure forces them to do so.” Does he
have a choice?
It is only when the migrant is able to ensure the basic needs of his fam-
ily and participate financially in the community life of the village that an
income-generating strategy might then result, aimed at achieving an
individual project. In this case, a priority tends to be real estate invest-
ment. The fulfilment of a housing project is likely to become the migrant’s
second objective according to Ponsot and Obegi (op. cit.). For Servet (op.
cit.), this makes sense, since it allows the wealth of the family unit to be
shown to others. Welcoming a large household, according to his words,
is seen to be a symbol of wealth and prosperity. Also, according to
Loveluck (2008, p. 16) and (Fouchard, 2017), the best finished and most
impressive buildings in Kayes are said to house exclusively migrant fami-
lies. The myth about migration and the quality of life it provides must
therefore be preserved, no matter what the cost, at the risk of falling into
shame and collective disrepute.
That which is far away is always better received in Africa, and agriculture, the
main element of the rural African existence, is no exception to the rule, espe-
cially in Keyes.
Investment in agricultural development projects is also reported to

remain a near-individual initiative. It seems more difficult for migrant
associations to design and implement large-scale projects with a signifi-
cant agricultural impact than to provide community facilities. This
requires appropriate technical skills and, therefore, partnerships that are
often difficult to mobilize. In addition, the means of production, the size
of the plots of land and the food requirements are highly specific to each
family25 in Kayes (Loveluck, 2008, p. 22). Most agricultural projects are

likely to be limited to the establishment of cooperative stores aimed at
reducing the prices of consumer goods so as to allow access for generally
imported foodstuffs for the population.
That said, since the majority of migrants in Kayes are rural, agriculture
is believed to be the second most important sector for individual produc-
tive investment and income-generating activities, far ahead of transport26
and on the heels of trade (Ndione & Lombard, op. cit.). In addition,
households that regularly receive remittances generate savings, which is
often accumulated in the form of livestock.27 In this way, livestock farm-
ing alone accounts for 49% of productive investment.
Basic farm investment, consisting of the purchase of farm equipment—
such as drill-seeders, ploughs, accessories, carts, oxen and donkeys—is
highly dependent on rainfall variations (Ndione & Lombard, op. cit.), the
absence of minimum farm infrastructure and the short-term nature of
loans obtained from local branches (Loveluck, 2008, p. 100). As a result,
households are more likely to invest their savings in off-season28 activities
that are not directly related to agricultural production. In this way, the
mobilization of family savings could ultimately serve to allow the emigra-
tion of another family member. This may often be perceived as the best
investment of all, as other migrants could find their expenses considerably
reduced, while increasing the savings capacity of the beneficiary house-
hold in the future (Gubert, 2002, p. 217). Such specific use of savings
generated can also be observed in rural China (Zhu, Wu, Peng, & Sheng,
2014, p. 1321).
3.4 Conclusion
Although widely used by migrants in Kayes to send remittances to their
families of origin, informal channels are said to carry some risk of non-
receipt in the channelling of these funds. This could be due to misappro-
priation by a migrant from the same region to whom money is handed
over for his relatives, especially when returning on holidays; liquidity
problems very often encountered, according to operators in the sector,
are typically experienced downstream; no traceability of the funds trans-
ferred, not to mention the psychological risk and stress suffered by the
sender (Ponsot & Obegi, 2010, p. 41). However, for most households,
the money is well received when viewed in terms of the accounts opened
by them in the CVECAs, the speed of network development and the
annual surpluses sometimes achieved. This last point is probably due to
remittances, since Kayes is one of the poorest regions in Mali. In this way,
there is an ability for recipients to save surplus money. But, it is unlikely
to invest much or very little in family farming activities. Citing Loveluck
(op. cit.), it may be that in reality productive agricultural investment is
not a priority for the villagers of Kayes or that the short-term loans
granted do not meet the sustainable financing needs of the population’s
agriculture. Nevertheless, the importance of financing a new emigration
project is still noticeable within households, especially among the young-
est, most influenced by the myth surrounding migration.
It is recognized that the PARMEC (Project to support the regulation of
mutual savings and loans) law does not authorize remittance transactions to
microfinance institutions, as these institutions are sometimes subagents of
commercial banks, which may themselves be exclusive partners of money
transfer providers. Accordingly, the interest shown by international organi-
zations, states and financial institutions in migrant remittances and the con-
tinuous efforts to formalize them may be in response to a concern to combat
money laundering and terrorist financing, but it may also be in response to
a real quest for profitability that may differ between the parties. Apart from
commercial banks and other financial institutions engaged in a constant
search for markets, we believe that donors, governments and NGOs have
more areas to explore so that the impact of remittances on poverty reduction
is no longer a theoretical objective to be achieved in the future.
Notes
1. Money Transfer Companies: Western Union (world leader) barely
matched by Money Gram.
2. “Rural poverty still explains 90% of national poverty,” Lachaud (2005,
pp. 2–3).
3. According to Rocher and Pelletier (2008, p. 35), the geographical cover-
age of MFIs and their proximity to poor-rural families is an additional
advantage over other operators in the remittance process.
4. Cover options for crop failure, price volatility and livestock mortality.
5. Inspired by the idea of insurance products designed by the AfDB (2007,
pp. 53–54) for the formalization of remittance channels.
6. “Safe saving is often a more important service than credit,” Fouillet et al.
(2007, p. 346).
7. Subsidies from the French Development Agency (AFD), shareholder of
the National Bank for Agricultural Development.
8. Remittances, whether formal or informal, are in all cases well received by
AfDB households (2007, pp. 35–37).
9. Data provided by the literature due to lack of up-to-date information.
10. Of the 5658 municipalities in the country, only 4% do not offer formal
banking services (Gentil & Servet, op. cit.).
11. It leaves the investor with a kind of “tracking card” with personal infor-
mation and checkboxes for each payment. The number and period of
payments are known in advance and correspond to the number of boxes
checked (Lelart, 2006, p. 6).
12. From the Neapolitan banker Lorenzo TONTI, he was at one time Louis
XV’s financial advisor and is said to have put forward the idea of the first
public tontines (Lelart, 2006, p. 9).
13. In this respect, tontines were not only the preserve of populations
excluded from the traditional banking system but also a cultural and
social practice. Indeed, according to Lelart (2006, p. 16), even among
African officials at the World Bank, tontines are practised.
14. Also known as street bankers, often solicited for short-term loans at
excessive interest rates (between 100% and 125%).
15. For example, risks of low harvests due to climate change.
16. Antony Fouchard (23 May 2017) for the newspaper “Le Monde Afrique”.
Available at: http://www.lemonde.fr/afrique/article/2017/05/23/dans-la-
region-de-kayes-au-mali-les-habitants-comptent-sur-la-diaspora-pas-sur-
l-etat_5132637_3212.html (Accessed June 2017).
17. The main financing body is the Agence Française de Développement
(AFD) and the technical support structure in charge of training network
agents, called the Centre d’Appui à la Microfinance et au Développement
(CAMIDE).
18. Living conditions during the first year are often very difficult for the new
newcomer.
19. Onions, tomatoes, bananas, Arabic gum (the purest in the Sahel) for
export (Fouchard, 2017).
20. Credits mainly granted for trade (67%), in fact without major climate
risk; agriculture covering only 4% of loans (agricultural equipment,
seeds, inputs, plant protection products) in 2006.
21. 10,069 savers in 2011. Available at: www.mixmarket.org (Accessed
February 2015).
22. Dedicated to family and community (Ndione & Lombard, 2004, p. 11).
23. More information provided by the applicant client than before.
24. Based on a migration slogan inspired by Ponsot and Obegi (2010, p. 18).
25. Number of dependent individuals per family because in Kayes, subsis-
tence agriculture (food crops) is reported to predominate (Loveluck,
2008, p. 18).
26. Lack of road infrastructure leading to considerable maintenance costs.
27. Very common among the Haalpularan (Peule and Toucouleur
population).
28. Three to four months of the rainy season and eight to nine months of the
dry season. The growing season is the rainy season (Ndione & Lombard,
2004, p. 17).
References
Azam, J.-P., & Gubert, F. (2005). Those in Kayes. The Impact of Remittances on
Their Recipients in Africa. Revue économique, Presses de Sciences-Po, 56(6),
1331–1358.
AfDB. (2007). Les transferts des fonds des migrants, un enjeu de développement.
Tunis, 85p.
Besson, E. (2008). Les migrants, acteurs du développement solidaire: Soutenir les
initiatives des migrants en faveur du développement de leur pays d’origine.
Premier ministre, Paris, 36p. Retrieved February 2015, from http://www.
ladocumentationfrancaise.fr/var/storage/rapports-publics/084000608.pdf
Bloy, E., & Dupuy, C. (1990). Adaptation des règles de gestion aux contraintes
du financement informel: Réflexion dans le cadre du contexte africain. In
G. Henault & R. M’Rabet (dir), L’entrepreneuriat en Afrique francophone:
Culture, Financement et Développement, Aupelf-Uref (éd.), (n°5, pp. 65–75).
Paris: John Libbey Eurotext.
Bouman, F. J. A. (1977). Indigenous Saving and Credit Societies in the Third
World: A Message. Savings and Development, 1(4), 181–219.
Bourenane, N., Bourjij, S., & Lheriau, L. (2011). Réduire les coûts des transferts
d’argent des migrants et optimiser leur impact sur le développement: outils et
produits financiers pour le Maghreb et la zone franc. Résumé exécutif du rap-
port, 8p. Retrieved February 2015, from https://www.afdb.org/fileadmin/
uploads/afdb/Documents/Publications/ESF_BAD_DGT_AFD_2011_
Transferts_argent_migrants_rapport_fr.pdf
Brown, R. P. C. (1994). Migrant’s Remittances, Savings and Investment in the
South Pacific. International Labour Review, 133(3), 347–367.
Camara, I. F. (1991). Services financiers et allègement de la pauvreté: structures
mutualistes d’épargne et de crédit dans l’Union Monétaire Ouest-Africaine
(UMOA). Document de travail n°8, Département du développement des
entreprises et coopératives, Bureau International du Travail (BIT),
Genève, 35p.
CERISE. (2006).CVECA de Kayes – Mali. Evaluation SPI – CERISE, 2p. [En
ligne]. Retrieved March 2015, from http://prospera-microfinance.org/IMG/
pdf/syntheseSPI-Mali-Kayes2006-2.pdf
CIRAD. (2002). Le financement de l’agriculture familiale dans le contexte de
libéralisation: quelle contribution de la microfinance? Résumé exécutif des
résultats du programme de recherche et actes du séminaire, Dakar, 35p.
Copestake, J. (2007). Mainstreaming Microfinance: Social Performance
Management or Mission Drift? World Development, 35(10), 1721–1738.
De La Brière, B., De Janvry, A., Lambert, S., & Sadoulet, E. (1997). Why Do
Migrants Remit? An Analysis for Dominican Sierra. FCND Discussion Paper,
n°37, IFPRI, 56p.
Delville, L. P. (1991). La rizière et la valise: Irrigation, migration et stratégies pay-
sannes dans la vallée du fleuve Sénégal. Séries: Collections ateliers du dével-
oppement, Paris, Syros-Alternatives, 231p.
Dendir, S. (2017). Saving Out of Remittances: Evidence from Ethiopia and
Kenya. International Migration, 55(4), 118–140.
Fouchard, A. (2017). Dans la région de kayes au Mali les habitants comptent sur
la diaspora pas sur l'Etat. Le Monde du 23 mai 2017. Retrieved June 2017,
from http://www.lemonde.fr/afrique/article/2017/05/23/dans-la-region-de-
kayes-au-mali-les-habitants-comptent-sur-la-diaspora-pas-sur-l-
etat_5132637_3212.html
Fouillet, C., Guérin, I., Morvant-Roux, S., Roesch, M., & Servet, J.-M. (2007).
Le microcrédit au péril du néolibéralisme et de marchands d’illusions.
Manifeste pour une inclusion financière socialement responsable. Revue du
MAUSS, 1(29), 329–350.
Freund, C., & Spatafora, N. (2005). Remittances: Transaction Costs, Determinants

and Informal Flows. Policy Research, Working Paper Series, Vol. 1, n° 3704,
World Bank, Washington, DC, 42p.
Generoso, R. (2012). Transferts de fonds et résilience des pays d’Afrique de l’Ouest
face à la variabilité des précipitations: une perspective macroéconomique. Cahier
du CEMOTEV n°2012-03, 60p.
Gentil, D., & Servet, J.-M. (2002). Microfinance: petites sommes, grands effets?
Introduction. Revue Tiers-Monde, Tome 43, n°172, pp. 729–735.
Gorse, F. (2008). Institutions de microfinance au Mali, évaluation rétrospective.
Département de la recherche, Division Evaluation et Capitalisation, AFD,
Série Evaluation et Capitalisation ex-post n°14, Paris, 55p. Retrieved February
2015, from https://issuu.com/objectif-developpement/docs/14-evaluation-
capitalisation
Granger, B. (2006). La microfinance risque de renier son inspiration humaniste.
Finance et bien commun, 25(2), 88–93.
Gubert, F. (2000a). Migration, Remittances and Moral Hazard. Evidence from the
Kayes Area (Western Mali). Working Paper E2000.17, CERDI, 32p.
Gubert, F. (2000b). Migration et gestion collective des risques. L’exemple de la
Région de Kayes. Thèse de doctorat, Université d’Auvergne et IRD, Clermont-
Ferrand, 288p.
Gubert, F. (2002). Le comportement de transfert des migrants est-il guidé par
un motif d’assurance ? L’exemple des Maliens de Kayes. Cahier du GEMDEV,
n° 28, Paris, pp. 197–220.
Gubert, F. (2008). (In)cohérence des politiques migratoires et de codéveloppe-
ment françaises. Illustrations maliennes. Politique africaine, 1(109), 42–55.
Guérin, I. (2000). Pratiques monétaires et financières des femmes en situation de
précarité: Entre autonomie et dépendance. Thèse de doctorat, Université
Lumière-Lyon-II, 588p.
Guérin, I. (2004). Taking Stock of Microfinance Progress. Report for the ILO
(Mid-Term Evaluation of Family Indebtedness with Microfinance Schemes
& Related Services (Ras/99/Mo1/Net), ILO, Geneva, 96p.
Guérin, I., Fouillet, C., Hillenkamp, I., Martinez, O., Morvant-Roux, S., &
Roesch, M. (2007). Microfinance: effets mitigés sur la lutte contre la pau-
vreté. In Annuaire suisse de politique de développement: Financer le développe-
ment par la mobilisation des ressources locales, IUED (Vol. 6, n°2,
pp. 103–119). Genève.
Gupta, S., Patillo, C., & Wagh, S. (2007). L’impact bénéfique des envois de
fonds sur l’Afrique. Finances & Développement, pp. 40–43.
Haudeville, B. (1990). Epargne et financement de l’entreprise productive. In

Hofmann, E., & Marius-Gnanou, K. (2005). Empowerment des femmes et
microfinance en Inde: Entre relativisme culturel et instrumentalisation. In:
I. Guérin, K. Marius-Gnanou, J. M. Servet, & T. Pairault (Dir.), Microfinance
en Asie: entre traditions et innovations. Karthala, IRD, Institut Français de
Pondichéry.
Jégourel, Y. (2008). La microfinance: entre performance sociale et performance
financière. Regards croisés sur l’économie, n°1, pp. 197–205.
Koloma, Y. (2007). Microfinance et réduction de la pauvreté en Afrique
Subsaharienne: quels résultats au Mali? Document de travail n°138, Lare-Efi,
GED, Université Montesquieu, Bordeaux IV, 38p.
Labie, M. (1999). La microfinance en questions. Limites et choix organisationnels
(116p). Bruxelles: Luc Pire.
Labie, M. (2004). Microfinance: un état des lieux. Mondes en développement,
2(126), 9–23.
Lachaud, J.-P. (2005). Transferts privés de Côte d’Ivoire, et pauvreté durable et
transitoire au Burkina Faso. Document de travail 93/2004, Centre d’Economie
du Développement, Université Montesquieu-Bordeaux IV, 14p.
Lelart, M. (1990). Les circuits parallèles de financement: Etat de la question. In
Lelart, M. (2006). De la finance informelle à la microfinance. AUF et Editions des
Archives Contemporaines, 62p. Retrieved February 2015, from https://
halshs.archives-ouvertes.fr/halshs-00009833/document
Loveluck, W. (2008). Potentialités de financement de l’économie agricole et rurale
dans une zone de migration: Les caisses autogérés de Kayes au Mali. Mémoire,
Montpellier SupAgro, 174p.
Lucas, R. E. B. (1985). Migration Amongst Botswana. Economic Journal,
95(378), 358–382.
Martin, P., Martin, S., & Weil, P. (2002). Best Practice Options: Mali.
International Migration, 40(3), 87–102.
Montalieu, T. (2002). Les institutions de microcrédit: entre promesses et doutes:
Quelles pratiques bancaires pour quels effets? Mondes en Développement,
Tome, 30(119), 21–32.
Nabeth, M., & Levy, U. (2007). Les dérivés climatiques, un paroxysme de la

mxicro-assurance? Autrepart, 4(44), 185–198.
Ndione, B., & Lombard, J. (2004). Diagnostic des projets de réinsertion
économique des migrants de retour: Etude de cas au Mali (Bamako, Kayes).
Revue Européenne des Migrations Internationales, 20(1), 169–195.
Nowak, M. (2005). On ne prête (pas) qu’aux riches, la révolution du microcrédit.
Paris: JC Lattès, 270p.
Ponsot, F. (2007). Le rôle des institutions de microfinance dans l’offre de produits
spécifiques aux migrants dans leur pays d’origine. Réseau français de la
Microfinance sous-commission n°1, 63p. Retrieved February 2015, from
https://www.microfinancegateway.org/sites/default/files/mfg-fr-publica-
tions-diverses-imf-et-produits-specifiques-aux-migrants-12-2007.pdf
Ponsot, F., & Obegi, B. (2010). Etude de capitalisation des initiatives et mécanismes
en matière de transferts de fonds au Mali, 108p. Retrieved February 2015, from
http://www.arabic.microfinancegateway.org/sites/default/files/mfg-fr-
etudes-de-cas-initiatives-transferts-de-fonds-au-mali-03-2010.pdf
Poursat, C. (2001). Les concepts clés de la microfinance. Publication Agridoc-GRET.
Richard, H. A., Jr. (2002). Precautionary Saving from Different Sources of Income:
Evidence from Rural Pakistan. Policy Research Working Paper 2761, 36p.
Rocher, E., & Pelletier, A. (2008). Les transferts de revenus des migrants: quel
impact sur le développement économique et financier des pays de l’Afrique
Sub-saharienne ? Bulletin de la Banque de France, (173), 27–38.
Sander, C. (2004). Capturing a Market Share? Migrant Remittances and Money
Remittancesas a Micro-Finance Service in Sub-Saharan Africa. Small
Enterprise Development, 15(1), 20p.
Servet, J.-M. (1990). Représentations de la monnaie et des supports d’épargne
et limites de la mobilisation de l’épargne informelle: Hypothèses de travail
pour une enquête au Sénégal. In G. Henault & R. M’Rabet (dir),
L’entrepreneuriat en Afrique francophone: Culture, Financement et
Développement, Aupelf-Uref (éd.), (n°7, pp. 87–96). Paris: John
Libbey Eurotext.
Soedjede, D. (1990). Politique de financement de l’entrepreneuriat au Togo. In
Culture, Financement et Développement, Aupelf-Uref (éd.), (n°11,
pp. 129–144). Paris: John Libbey Eurotext.
Tinguiri, K. L. (1990). Epargne et crédit informels en milieu rural au Niger:
L’activité des tontines et des gardes-monnaies villageois. In M. Lelart (dir), La
tontine, AUPELF-UREF (éd.), (n°8, pp. 177–208). Paris: John
Libbey Eurotext.
World Bank. (2006). Global Economic Prospects 2006: Economic Implications of

Remittances and Migrations. n°34320, 182p.
World Bank. (2015). Migration and Remittances, Recent Developments and
Outlook, Special Topic: Financing for Development. Migration and Development
Brief 24, 27p.
World Bank. (2017). Migration and Remittances, Recent Developments and
Outlook, Special Topic: Global Compact on Migration. Migration and
Development Brief 27, 44p.
Zhu, Y., Wu, Z., Peng, L., & Sheng, L. (2014). Where Did All the Remittances
Go? Understanding the Impact of Remittances on Consumption Patterns in
Rural China. Applied Economics, 46(12), 1312–1322.
4
Islamic Microfinance: Exploring
the Experience of Akhuwat
in Promoting Qard Hasan in Pakistan
Joana Silva Afonso and Ajaz Ahmed Khan
4.1 Introduction1
Islamic microfinance has received increasing recognition in recent years as
the number of institutions providing Shari’ah compliant microfinance has
grown. Some Islamic microfinance institutions (IMFIs) have gained
greater visibility and attracted the interest of practitioners, investors and
academics. It was not surprising, therefore, that the programme of the
fifth European Research Conference on Microfinance included a plenary
session on the topic.2 It gathered highly experienced and prominent fig-
ures within the sector, including Professor Malcolm Harper from Cranfield
Management School, Professor Habib Ahmed from the University of
Durham and Dr Mohammed Kroessin from Islamic Relief Worldwide.
J. Silva Afonso (*)
Portsmouth Business School, University of Portsmouth, Portsmouth, UK
e-mail: joana.afonso@port.ac.uk
A. A. Khan
CARE International, London, UK
e-mail: Khan@careinternational.org

62 J. Silva Afonso and A. A. Khan
Also present was Dr Amjad Saqib, Founder and Executive Director of

Akhuwat Islamic Microfinance (AIM), an IMFI from Pakistan that has
grown remarkably over the last decade and is now one of the largest IMFIs
in the world, with more than 940,000 active clients. AIM focuses exclu-
sively on providing qard hasan or ‘benevolent’ loans to low-income bor-
rowers. In 2017, it became the largest microfinance provider in Pakistan
in terms of number of active clients. The invitation to the conference
derived not only from this recent performance but also from the recogni-
tion of the institution’s distinctive features, and its ambitious plans to
expand the model to other non-Muslim-majority countries.
This chapter aims to explore the advantages and challenges associated
with the interest-free microcredit model promoted by AIM, as well as to
provide a better understanding of Islamic microfinance. Qard hasan is
defined by Obaidullah (2011) as an indigenous Islamic microfinance model.
This type of microcredit methodology attracts attention based on the expec-
tation that it can increase and improve access to formal financial services for
Muslim populations who can seem to refrain from participating in tradi-
tional interest-based microfinance programmes (Harper, 2017, p. 189).
The most recent data from the Global Findex Survey on Financial
Inclusion shows that many Muslim-majority countries have relatively
lower rates of financial inclusion, challenging the accepted positive asso-
ciation between financial and economic development (Beck, Demirgüç-
Kunt, & Levine, 2007). This is the case in Pakistan: the share of the adult
population who owned a formal bank account in 2017 (18%) or had
borrowed from a formal financial institution in the previous year (2.6%),
was far below the same indicators for the average of the lower middle-
income countries, which were 56.1% and 9.8%, respectively (World
Bank, 2018).3
During the conference session, the speakers emphasised the ethical
essence associated with a qard hasan microcredit model. Dr Amjad Saqib
explained that AIM has an explicit social mission and considers the provi-
sion of benevolent loans as a means to an end—alleviating poverty
through promoting financial inclusion. The panellists raised a number of
key questions to the development of Islamic microfinance. These related
to the demand, the sustainability, the replicability and the impact of the
qard hasan model. This chapter will explore these questions further, using
4 Islamic Microfinance: Exploring the Experience of Akhuwat… 63
AIM as an example. The information shared by Dr Amjad Saqib and

other speakers at the conference is complemented with literature on AIM
and Islamic microfinance more generally, and the results of an on-going
evaluation led by Lendwithcare, a microfinance crowdfunding platform
established by CARE International UK, which has been funding AIM
since 2013.4
The chapter starts with a brief introduction to Islamic microfinance.
Section 4.3 describes the development of AIM and the qard hasan model.
In Sect. 4.4, key questions are contextualised using the experience of
AIM. The chapter ends with the main conclusions and their implications
for the sector.
4.2 An Introduction to Islamic Microfinance

This section provides a brief introduction to the principles that govern
Islamic microfinance and some of the main financing methodologies
employed by IMFIs. Islamic microfinance differs from interest-based
microfinance because of the need to conform to certain religious princi-
ples. These are contained in the Qur’an and the Sunnah, or word and
living tradition of the Prophet Muhammad. While Islamic teachings do
sometimes provide quite specific guidance—for example, the longest
verse in the Qur’an deals with financial transactions and contracts—the
religious principles derive from the objective of promoting honesty,
transparency and, above all, fairness in economic activities and behaviour
between all parties, regardless of their relative power and status (Khan,
Kustin, & Khan, 2017).
There are four main Islamic finance principles that are of particular
importance. Firstly, riba, most commonly translated as interest or usury, is
forbidden. Lenders cannot expect to receive a predetermined, fixed sum,
regardless of the outcome of the enterprise in which the funds are invested.
A return on capital is allowed providing the lender participates in the pro-
ductive process and is exposed to risk. Secondly, gharar, which describes
any transaction that involves excessive uncertainty and risk, deceit or
fraud, is prohibited. Gharar refers to any transaction of items whose exis-
tence or description is uncertain due to lack of information or knowledge
of the ultimate outcome. For example, it is not permitted to sell some-

thing that one does not own, in the hope that it can be bought cheaper at
a later date—‘short-selling’ is therefore forbidden. Thirdly, Islam considers
certain activities as morally or socially harmful and prohibits investment
in these areas. The prohibited or haram areas include the production and
sale of alcohol, gambling and illegal drugs. Finally, financial transactions
should directly or indirectly be linked to a real, tangible, economic activity
or asset, as opposed to financial speculation or debt.
In addition to these general principles, Islamic teachings emphasise the
importance of honouring contracts and agreements. These should be
clear, by mutual agreement, and the responsibilities and benefits of all
concerned parties should be clearly detailed and for a known period and
price. Furthermore, there should be at least two witnesses present when a
contract is signed. To ensure ‘authenticity’ it is essential to seek approval
from qualified Islamic scholars as to whether the manner in which opera-
tions are structured and implemented are Shari’ah compliant and also to
conduct regular Shari’ah audits. In practice, this involves IMFIs request-
ing local religious authorities to provide confirmation that their opera-
tions conform to Shari’ah.
Although the Qur’an prohibits the use of interest and encourages legit-
imate commerce, trade and wealth creation, it does not promote any
specific type of contract. A number of Islamic financing techniques have
developed in accordance with Shari’ah and these have been adopted by
IMFIs. Broadly speaking, it is possible to distinguish between techniques
that promote partnerships, such as mudaraba and musharaka, and
arrangements that are essentially sales contracts, such as murabaha, ijarah
and bai salam. Islamic finance only permits one type of loan, namely
qard hasan.
Under mudaraba, an IMFI provides the capital required to fund a proj-
ect, while a micro-entrepreneur uses his or her skill to manage the project.
Profits are shared according to a predetermined ratio, usually determined
as a percentage of the profit. In the case of a loss, providing it was not due
to mismanagement or misconduct, the IMFI loses its capital while the
micro-entrepreneur loses the time and effort that he or she has expended
in the activity. Musharaka involves two or more parties contributing
towards financing a venture and sometimes managing the project as well,
and profits and losses are shared in accordance with each partners’ equity
participation. Partnership finance, which relies heavily on trust and confi-
dence, is largely absent from Islamic microfinance because most IMFIs
lack the time and capacity for robust monitoring and evaluation of clients’
activities, and in any case the small scale of such activities is likely to make
transactions uneconomic. At the same time, most micro-entrepreneurs
are, for a variety of reasons, unable to keep accurate accounts. As a result,
it is difficult to calculate the exact level of profits or losses.
Murabaha is the most widely used instrument in Islamic microfinance,
largely because it is relatively straightforward to structure, understand
and implement for all parties. In a murabaha contract, an IMFI will pur-
chase and deliver an asset or other items requested by the micro-
entrepreneur. It adds a ‘mark-up’ or profit margin to the sale price and the
micro-entrepreneur pays instalments to the IMFI over an agreed period
of time for this service. There is some scepticism towards this particular
method of financing, as the fixed mark-up appears similar to interest.
However, the mark-up is not compensation for time, rather it relates to
the efforts and expense of the IMFI in seeking out, negotiating, purchas-
ing and delivering the asset requested, at the best possible price.
Ijarah is similar to leasing and involves an IMFI purchasing an asset
that it then rents out at a price that enables it to recover its investment plus
a profit. More often than not, the micro-entrepreneur makes regular pay-
ments and becomes the owner of the asset once all instalments have been
made. Bai salam is a contract whereby the full price for an asset or com-
modity is agreed and paid up front in cash by the IMFI at the time the
agreement is made. The item is delivered at a specific time in the future.
The essential purpose is to ease cash liquidity shortages, most commonly
for smallholder farmers, by enabling them to receive advance payment.
Qard hasan is a cash loan that is repaid without interest, mark-up or
share in the venture for which the loan is used. Qard hasan has a particular
resonance for Muslims, as such loans are encouraged by Islamic teachings
as an effective way of helping poor people. Indeed, they are preferred to
the provision of outright charity. Qard hasan is considered a ‘benevolent
loan’ and this is generally interpreted to mean that a borrower cannot be
forced to make a repayment—in the event that a borrower is unable to
repay, the lender must accept the transaction as a charitable act. Sometimes
a small administrative charge is permitted for qard hasan loans. Importantly

though, this charge cannot be made proportional to the amount or term
of the loan. This distinction is what differentiates it from riba. Qard hasan
is generally more appropriate and appealing in most cases, as many poor
people prefer to receive cash to use at their own discretion, rather than
being tied to a particular commodity, as is the case with murabaha.
However, IMFIs that rely exclusively on this methodology will not be able
to cover their operational costs without other sources of income, typically
from charitable sources such as zakat, sadaqah and waqf.5
The prohibition on riba and gharar also impacts on savings and insur-
ance products. Since riba forbids offering a predetermined rate of return,
the relatively small proportion of IMFIs that do offer savings accounts
reward savers a share of the institution’s profits. Some IMFIs also offer
takaful which is a Shari’ah compliant alternative to conventional insur-
ance. In takaful, which translates as ‘guaranteeing each other’, all partici-
pants contribute to a mutual fund and this pool of contributions creates
a takaful account. The IMFI will manage this account and charge an
agreed fee to cover the operating costs. Any claims are paid out from the
account and any unused funds are returned to the participants at the end
of the agreed period.
Although partnership finance and qard hasan represent the traditional
forms of Shari’ah compliant finance, most IMFIs concentrate on sales
contracts, offering mainly murabaha and to a lesser extent ijarah. In a
recent book (Harper & Khan, 2017) that examined 15 different Islamic
microfinance programmes in 11 countries in Asia, Africa and the Middle
East, AIM was the only institution which exclusively employed qard
hasan methodology.
4.3 AIM and the Qard Hasan Model

4.3.1 The Institution
Akhuwat was founded in 2001 by a group of philanthropists with the

mission of alleviating poverty. Its operating philosophy is guided by five
principles. Firstly, and in line with Islamic teachings as outlined in the
previous section, it provides interest-free loans to the ‘economically active
poor’, mostly to help establish and develop their businesses. The organ-
isation views interest as a barrier to ‘widespread proliferation of capital,
and a violation of all moral and ethical codes’. It believes that ‘burdening
the poor with exorbitant interest rates’ undermines efforts towards pov-
erty alleviation. Secondly, it holds its meeting and loan disbursements
in local religious places—mostly mosques, but sometimes temples and
churches. The organisation believes that using the existing indigenous
infrastructure for operations allows it to minimise costs and also promote
greater transparency and accountability, as well as create a sense of good-
will amongst the local community. Thirdly, it encourages clients to
become donors through the ‘Member Donor Program’. It decided to
instigate this initiative when clients asked how they could contribute to
the organisation’s development. The donations are entirely voluntary
and, it is hoped, helps instil the value of helping others in need. Fourthly,
it promotes a spirit of volunteerism. It considers that there should be a
social contract between the privileged and the underprivileged, a sense of
duty between those who have resources towards those who do not. AIM
trains and employs volunteers on a regular basis. Finally, it is a
non-discriminatory organisation and works with all, regardless of ‘their
caste, colour, creed, gender, politics or faith’.
In its first decade of operations, the institution gradually increased its
scale and outreach, as many other IMFIs have done (Khan, Ishaq, Afonso,
& Akram, 2017). However, AIM has expanded operations markedly dur-
ing the past five years (Table 4.1). Impressive recent growth is partially a
result of increasing financial support from regional governments in Pakistan
that have sought to promote financial inclusion and entrepreneurship
Table 4.1 AIM main activity indicators

Gross loan portfolio Active borrowers PAR30
(USD) Annual variation (%) (No.) Annual variation (%) (%)
2014 24,849,495 60.3 235,517 44.2 0.48
2015 46,887,198 88.7 405,939 72.4 0.29
2016 76,632,330 63.4 567,761 39.9 0.33
2017 123,903,151 61.7 855,232 50.6 0.26
2018 129,371,850 4.41 941,782 10.1 0.24
Source: Adapted from Khan, Ishaq, Afonso, and Akram (2017), Haider (2017) and
personal communication with Project Manager, Akhuwat, on 1 July 2018. Data as of 30
June each year
among low-income populations. The partnership with provincial authori-

ties has provided AIM with significant funding for both loans and opera-
tional costs, complementing other national and international donations
gathered by the institution (Khan, Ishaq, Afonso, & Akram, 2017).
The growth of the credit portfolio has been accompanied by an equally
remarkable geographical expansion. As of the end of June 2018, AIM oper-
ated through 794 local branches throughout the country, even including
the remote regions of Gilgit-Baltistan in the north of the country and the
Federally Administered Tribal Areas (FATA) that border Afghanistan. These
are areas where many microfinance institutions are reluctant to operate
because of poor infrastructure and insecurity. AIM branches are small and
simple in order to minimise operational costs (Khan, Ishaq, Afonso, &
Akram, 2017), but the institution’s human resources are significant: as of
mid-2018 the microfinance institution employed around 4200 staff.6
In the last two years, AIM has become the largest Pakistani microfi-
nance institution by far, in terms of number of active borrowers. However,
the MFI ranks below Pakistani interest-based microfinance banks such as
the Khushhali Microfinance Bank, Tameer Microfinance Bank and the
National Rural Support Programme Bank, in terms of the market share
as measured by the gross loan portfolio (Haider, 2017). This reflects a
much lower average loan size than other microfinance providers and
strongly signals that AIM is targeting relatively poorer and more vulner-
able segments of the population.
Until 2017, the institution worked as a non-governmental organisa-
tion (NGO) which, beyond the provision of interest-free loans to low-
income entrepreneurs, offered other types of credit, including housing,
education, emergencies and debt payment. It also offered non-financial
services linked to education, health, and a clothes recycling programme
that exclusively hired employees from the transgender community, one of
the most marginalised groups in the country.
Regulatory changes in the microfinance sector in 2016 led to a separa-
tion of the institution’s microcredit operation, which is now implemented
by Akhuwat Islamic Microfinance, a non-bank financial institution regu-
lated by the Securities and Exchange Commission of Pakistan. The non-
financial activities remain under the sphere of Akhuwat, which is a
registered non-governmental institution.
4.3.2 The Qard Hasan Microcredit Model
During the conference session, the AIM manager put forward three main
reasons to explain adopting benevolent interest-free loans as their micro-
credit model. Dr Amjad Saqib stated that the two main motivations
behind the creation of Akhuwat were the implementation of a Shari’ah-
compliant microcredit methodology; and the rectification of a feature of
most conventional microfinance programmes that was deemed unfair—
namely charging higher interest rates to poorer clients compared to rela-
tively better-off borrowers. Professor Malcolm Harper added a third
reason, arguing that other Islamic financial instruments, including the
partnership and sales contracts described in Sect. 4.2, are not suited to
the financial needs of relatively poorer people who often require cash for
working capital or for making a series of relatively small purchases.
It is important to stress here the ambiguity associated with the con-
cepts of poverty and poor, which are often interpreted differently by
microfinance actors. Robinson (2001) talks about the economically
active poor, and this was an expression employed by Dr Amjad Saqib in
his intervention. Qard hasan loans are offered to low-income entrepre-
neurs, but engagement in a self-employed activity is a pre-requisite to
access the loan. Thus, it should be expected that, like most other produc-
tive credit programmes, AIM’s microcredit programme does not reach
some of the poorest segments of the population.
AIM’s microcredit model is based on the traditional group lending
methodology. Borrowers form solidarity groups of three to six members
living in the same neighbourhood and undertaking their own independent
businesses. The applications of all the members of the group are assessed
simultaneously, although the applications regard individual loans (Khan,
Ishaq, Afonso, & Akram, 2017). The microcredit product is called a Family
Enterprise Loan, which discloses one of its distinctive features—the process
expects the involvement of the household by supporting both the loan
application and the business. In practice, this translates into the inclusion
of another member of the household (often the spouse) as co-signatory of
the loan contract (Khan, Ishaq, Afonso, & Akram, 2017).
AIM demands that loan candidates should not have any other active
loans at the point of application. Once they finish repaying their first
microcredit loan, they may apply for a second larger loan. The initial
application process takes, on average, three to four weeks and includes

visits by the loan officer to the homes or businesses of each of the group
members. The final step in the application process is a compulsory group
meeting, usually held in a mosque or branch office, in which the branch
manager verifies that all conditions have been met and all members are
made aware of their responsibilities regarding the loan.
After approval, the disbursement of loans is scheduled for the following
public session, which again usually takes place at the local mosque—on a
small number of occasions when the majority of the borrowers are Christians
or Hindus then churches or temples are also used. At this event, borrowers
and their guarantors must be present to sign the loan contract (Khan,
Ishaq, Afonso, & Akram, 2017). During this public ceremony, which sev-
eral hundred people typically attend, AIM’s branch managers or other
senior staff emphasise the fact that qard hasan loans are promoted in accor-
dance with Islamic teachings and they discourage all present from engaging
in interest-based transactions and encourage prompt repayment so that
others might be able to benefit from the same funds in the future. In this
way, the moral obligation to repay the loan is reinforced both by the public
nature of the event (the presence of witnesses being one of the requisites of
Shari’ah), and by the attachment of ‘religious sanctity to the oath of return-
ing it on time’ (Obaidullah, 2011, p. 420).
From this, it can be concluded that the uniqueness of the AIM model
is not limited to not charging interest on loans or penalties for late repay-
ments. It is also rooted in its family approach and in the use of religious
sites and language to strengthen the moral commitment of borrowers to
honour the loan contract and, at the same time, reduce the operational
costs of the institution.
4.4 A
Working Hypothesis and Four Essential
Questions on Qard Hasan and Islamic
Microfinance
The main argument raised throughout the conference session referred to
the ethical nature of qard hasan loans, and their potential role in fulfilling
the social mission of microfinance institutions. This vision of the model
has several implications on the identified critical questions relating to

demand, sustainability, replicability and impact for the development and
growth of Islamic microfinance. It particularly influences the notions of
sustainability and replicability of microfinance programmes, leading dif-
ferent actors in the sector to think ‘outside the box’ of conventional
microfinance. This chapter will now explore the critical questions in each
of these four areas in turn.
4.4.1 Demand
Data on the microfinance sector published by the Pakistan Microfinance

Network (PMN) highlights the impressive performance of AIM in terms
of outreach. As mentioned previously, AIM’s growth over recent years has
been outstanding both in terms of number of active clients and geo-
graphical coverage. The overall penetration rate of the microfinance sec-
tor in the country has increased significantly from 11.8% in September
2014 to 25.4% in September 2017, even though Islamic microfinance
only comprises a relatively small proportion (16% by September 2017)
of the sector as a whole and is due in large part to the operations of AIM.7
Despite this positive evolution, estimates of the financially unserved
population in Pakistan, especially in rural areas, are relatively high. This
seems to be in line with the conclusions of the 2014 MIMOSA report,
which analyses microcredit markets worldwide in terms of market satura-
tion. Pakistan was classified as an under-served market, ranking 1 on a
scale of 1 to 5 (Javoy & Rozas, 2015). This is not surprising, considering
Pakistan’s financial inclusion figures mentioned earlier, and it can be per-
ceived as a mismatch between the current offer and the needs and con-
straints (including religious) of low-income Pakistanis.
In this analysis, an important and often neglected issue is the role of
informal finance. Although the Global Findex figures on borrowing from
informal sources seem to show a more modest use of informal loans in
Pakistan compared to its neighbours India and Bangladesh (World Bank,
2018), this might not reflect the whole story. Pakistanis are usually suspi-
cious and careful in revealing personal information, especially on such a
sensitive topic (Lieven, 2011). Thus, it should be expected that the
declared figures for informal finance are underestimated. Given that

informal finance can include both exploitative moneylenders and infor-
mal savings and credit groups (locally referred to as ‘committees’), it
would be useful to have a better understanding of the existing mecha-
nisms and their comparative relevance.
The unmet demand for qard hasan loans, and Islamic microfinance in
general, does not seem to be confined to Pakistan, as illustrated by the
case studies from 11 different countries included in Harper and Khan
(2017). In the conference session, Professor Malcolm Harper shared his
perception that the situation is geographically heterogeneous, giving the
example of the relatively more conservative Somalia where the preference
for Shari’ah compliant loans seems comparatively more significant than
in Bangladesh. The preference for Islamic financial products when similar
conventional products are available was also found in a randomised con-
trol trial implemented in Jordan. The Shari’ah-compliant product intro-
duced by the MFI was a murabaha loan, which was compared to a
conventional consumption loan destined to buy household assets
(El-Zoghbi, Karlan, Osman, & Shammout, 2016, p. 1).
4.4.2 Sustainability
Adopting a qard hasan microcredit model implies looking at the sustain-

ability of the institutions from a different perspective and challenging
traditional concepts that stipulate that costs should be covered by loan
charges and profits. Different speakers at the conference asserted that if
AIM still exists and continues to grow after continuously operating and
implementing the same basic model for the past 17 years, it ought to be
considered sustainable, regardless of its funding sources. Although this
perception does not fit with the traditional accounting definition of sus-
tainability, which is commonly used in the evaluation of the performance
of microfinance institutions, the question raised by the panel was, to
what extent does it have to fit? Considering the mixed results on social
impact obtained by conventional microcredit programmes, the pursuit of
different and innovative models should be encouraged, and its mis-
matches with conventional programmes accommodated to properly eval-
uate these models, both its advantages and weaknesses.
Khan, Ishaq, Afonso, & Akram (2017) offer insight into the funding
of AIM’s microcredit programme. In the fiscal year ending in June 2016,
the main funding sources of the institution were national and interna-
tional donations (including sadaqah or zakat), institutional funding from
regional governments (in particular the Government of Punjab) and
application fees (AIM charges a fixed and non-refundable fee of 200
Pakistani Rupees (just under US$2) for all loans). Some two years later,
the composition of the funding sources had not changed significantly.8
Although three additional international funders now support AIM
(British Asian Trust, United Nations Development Program and Louis
Berger), the amounts granted are relatively small and less than the main
international funder, Lendwithcare, which provided approximately
US$750,000 in 2017.
In addition to these funds, during 2015–2016 Akhuwat generated more
income from voluntary donations from its own active clients than from the
application fees charged during the same period (US$1.45 m compared to
US$1.13 m). In this year, as in the previous one, the income raised, even
not including the borrowers’ contributions, exceeded the total operational
expenses (cost coverage ratios of 109% and 131%). This was achieved due
to increased income and a strategy of low costs. Besides the maintenance of
a simple and low-cost logistical infrastructure (in terms of physical space
and equipment), average salaries for comparable positions are lower than
those in other microfinance institutions. They are, however, compensated
by career development opportunities, since AIM has a policy of promoting
from within; and the ‘faith’ factor, with many staff being committed
Muslims who value working for the organisation as it acts on Islamic teach-
ings of helping the poor while simultaneously not engaging in interest-
based transactions. Both factors seem to be key to achieving low rates of
staff turnover. The authors, thus, conclude that the provision of qard hasan
loans can be compatible with achieving sustainability.
Voluntary donations from clients are a distinctive feature of AIM’s
microcredit model, which Dr Amjad Saqib highlights. These donations
are encouraged by AIM from the beginning of the relationship with the
client. AIM believes that if an individual is able to take a loan without
interest, which it is hoped will help to improve his/her economic situa-
tion, when this person manages to get out of poverty at some future time,
he/she will have the moral obligation to assist others to do the same,
although there is no formal obligation to do so. Dr Amjad Saqib referred
to this as ‘Akhuwat’s social contract with its clients’ and more widely a
‘pact between the privileged and the underprivileged, a duty of those who
have resources towards those who do not’. From this perspective, AIM
should be seen not only as a microcredit programme, but also as a part-
nership with the community, in which the principles of solidarity and
sharing are central. AIM managers believe that if clients understand these
principles, they are inspired to contribute, even if their monetary contri-
tion is relatively small. Interestingly, in order to convey the message that
such donations are not a charge for its service but to promote solidarity
with other ‘deserving’ people, AIM has decided not to use the voluntary
donations from clients to cover operational costs, but rather as extra
loan capital.
4.4.3 Replicability
It might be assumed that the success of a qard hasan model is strongly

dependent on its religious component and can therefore only work when
implemented within Muslim-majority countries or communities. Dr
Amjad Saqib disagrees with this idea, remarking that the practice of interest-
free loans has long been present in many societies, most often among fam-
ily and friends.9 The results of the Global Findex Survey seem to corroborate
this argument showing that, in 2017, borrowing from family and friends
was the main source of credit in low- and middle-income countries (World
Bank, 2018). Dr Amjad Saqib argued that interest-free loans should be
perceived essentially as an ethical practice. Despite the fact that AIM con-
siders itself a faith-based institution, it does not consider itself as a prosely-
tising organisation. It works equally with Hindus and Christians, even
using their temples and churches to gather borrowers and disburse loans on
occasions (Khan, Ishaq, Afonso, & Akram, 2017).
Clearly, the ability of AIM to attract zakat and sadaqah donations from
economically better-off Muslims as well sadaqah donations from its own cli-
ents has been key to generating the necessary funds for its microcredit pro-
gramme, particularly prior to attracting institutional support. Furthermore,
the concept of qard hasan loans has a particular resonance for Muslims,10
whether they are private individuals or representing institutional donors,
such as the various provincial governments in Pakistan. Considering interest-
free loans as a purely ethical, rather than a faith-based, practice opens the
door to a more positive and open view on the possibilities of replicating the
model in other, particularly non-Muslim, contexts. However, there is doubt
as to whether the interest-free model would be able to attract the same level
of financial or moral support in a non-Muslim environment.
Khan, Ishaq, Afonso, and Akram (2017) briefly mention the imple-
mentation in Pakistan of 16 independent local replications of the AIM
model. However, these institutions have struggled to attract donations
and increase their scale and outreach. In fact, there are few successful
examples of IMFIs that rely exclusively on qard hasan. Analysing the
experience of START, a microfinance institution created in 2002 in
Kosovo by Islamic Relief, Khan and Zeqiri (2017) observe how, unable to
attract donations in an environment where Islamic religious practice is
not strong, the organisation gradually moved away from solely using qard
hasan to mostly employing murabaha (with relatively higher charges for
clients), reserving qard hasan only for the poorest clients. It is clear that
Akhuwat’s success is only partly due to the qard hasan model striking a
chord with donors and clients. Equally, if not of more importance, there
has been the personal charisma and leadership qualities of Dr Amjad
Saqib, and the ability to recruit very able staff and instigate strong opera-
tional policies and procedures. During the session, Dr Amjad Saqib
referred to AIM’s plans to implement replications of the model in other
(non-Muslim majority) countries during 2018 and beyond.
4.4.4 Impact
The considerations and arguments put forward while developing the

three previous questions build up to the final question relating to the
impact of Islamic microfinance. What evidence exists regarding the
impact of qard hasan models and Islamic microfinance overall? How does
this research compare with impact studies of conventional interest-based
microfinance? Is there enough evidence on AIM to justify the ethical
perspective associated with the model and its implications?
The common perception of the conference speakers is that there is

certainly much to do in this regard. In fact, there is relatively little research
on the impact of Islamic microfinance programmes, and the studies
undertaken so far suffer from the same deficiencies that conventional
microfinance has long been accused of, namely that the studies have used
weak methodologies and data. There is no conclusive impact study using
randomised control trials, and rigorous quasi-experimental studies
are scarce.
Regarding the impact of AIM microfinance programme, which is the
most well-known example of the qard hasan model, until recently, the
most rigorous of the studies was the social performance investigation car-
ried out for the European Union: ‘Pakistan Financial Services Sector
Reform Programme in 2005–2006’. The research aimed to ‘quantify and
demonstrate some of the outcomes of microfinance institutions’ and
included six Pakistani MFIs, with different missions and programmes.
The quasi-experimental methodology implemented was based on a
difference-in-difference approach, comparing the differences in the
incomes of clients and non-clients of the institutions in the locations
where the programme was implemented with control locations (Zaidi,
Jamal, Javeed, & Zaka, 2007, pp. ii–iii).
The researchers surveyed a sample of 220 active Akhuwat borrowers
and 120 non-borrowers. In the quantitative analysis, they did not find,
for most of the economic and social outcomes included, significant dif-
ferences between the clients and non-clients. They associated these non-
results mainly with the short period of participation in the programme.
Interestingly, they also found that only 15% of the sample clients were
considered to be poor according to the official poverty line, which was a
common result with the other participating MFIs in the study. To com-
plement the information collected through the quantitative survey, the
study also included two focus group discussions with a smaller sample of
clients in different loan cycles, which overall provided a very positive
perception of the borrowers regarding the impact of the microcredit loan
received from Akhuwat (Zaidi et al., 2007).
It should be noted that this study was implemented in 2005 when
Akhuwat had 13 branches located in the Punjab province and only 7150
active borrowers, a very different reality compared with the present situa-
tion. Moreover, the research was based on cross-sectional data, therefore

being more sensitive to selection bias and less suitable to measure changes
in the lives of the clients over time.
The first longitudinal study conducted on the impact of AIM micro-
credit programme was implemented within a broader evaluation project
led by Lendwithcare. The project, in which the University of Portsmouth
participates as academic consultant, started towards the end of 2014 and
included the implementation of two waves of a household survey in 2015
and 2017. Although also based on a quasi-experimental methodology
using a difference-in-difference model, data was available for two periods
in time, which allowed for a comparison between the changes in the
selected outcomes for the sample of clients and non-clients before and
after the participation in the programme. The main findings of the study,
which are included in the final project report, are presented next
(Afonso, 2018).11
The household survey was applied to a sample of 500 new AIM clients
from four branches in the city of Lahore supported by Lendwithcare and
100 non-clients from the same neighbourhoods. The survey included a
purpose-designed questionnaire and the application of the Poverty
Probability Index (PPI).12
A first result from the application of the baseline survey in 2015 was
the (expected) relevance of the ‘faith’ factor in the reasons for applying for
a business loan at AIM. Of the new clients interviewed, 62% stated ‘com-
pliance with Islamic teachings’ as their main motivation to choose AIM,
with only 29% choosing the cheaper cost of the loans in comparison with
other financial providers. In the second wave of the survey, there is also
evidence to corroborate the identification of AIM clients with the prin-
ciples of the institution, specifically the solidarity and sense of responsi-
bility towards others in worse conditions. Of the clients interviewed in
both surveys, 93% had made either regular or occasional voluntary dona-
tions to the institution during the previous year when they were repaying
their first microcredit loan.
The study results were, overall, positive at both business and household
levels, with quantitative and qualitative indicators showing that, on aver-
age, business sales, profits, net income and poverty scores, among other
indicators, had improved for AIM new clients in comparison with the
baseline survey. The observed variations of these indicators were stronger

than those verified for the group of non-clients. These results reflected in
the majority (68%) of the clients considering their quality of life to be
better in 2017, and in the significant decrease of clients identifying access
to capital as the main constraint for their businesses (only 20% in 2017
compared with 76.5% at baseline).
It is, however, important to note that the research also highlighted the
heterogeneity among the clients. There were significant differences,
namely in the indicators related to income, consumption and poverty,
between female and male clients, with female borrowers earning lower
personal incomes and registering lower PPI scores, on average, relative to
male clients in the two waves of the survey. These differences are not sur-
prising considering other studies conducted in the sector in Pakistan
(Asim, 2009; Safavian & Haq, 2013; Zulfiqar, 2017). It should be noted
that the gender differences in income and expenditure were not signifi-
cant at the household level.
Statistically significant differences were also identified between the cli-
ents from the four branches involved in the study: the clients living in the
two central urban areas were more likely to earn higher incomes and be
less poor. The differences were also verified in the composition of the
branch samples in regard to gender and literacy levels, with a stronger
presence of female and illiterate clients in the poorer areas. Interestingly,
the comparison between the different client segments showed that the
variation of the PPI scores was stronger for these two sub-groups (female
compared with male clients and illiterate compared with clients with for-
mal education), which indicates a potential reduction of the respective
poverty gaps, although this needs to be corroborated with further analysis.
In comparison to non-clients, the research found that the access to loans
had a positive impact for those respondents who experienced stronger
improvements in their poverty scores, but had no significant impact for
respondents who experienced a deterioration or no variation of the indi-
vidual PPI scores.13 Not forgetting that the quasi-experimental methodol-
ogy employed presents some limitations, it should be noted that these
limitations are not sufficient to overlook a consistent finding of the study,
the heterogeneity among AIM clients and its consequences on the effects of
the qard hasan loans provided by the institution. This acknowledgement is
important in the assessment of the AIM model and should also be taken in
consideration when analysing the existing, and at this stage, not very suc-
cessful replications of the model within Pakistan as well as the potential
future replications of the qard hasan model in other contexts.
4.5 Conclusion
The preceding discussion points towards an under-served demand for
Shari’ah-compliant finance in Pakistan. In the context of relatively high
rates of financial exclusion, we can assume that this demand will likely be
shared by other countries with significant Muslim populations. AIM’s
experience also demonstrates that there is a strong desire among Muslim
donors to support qard hasan programmes as a preferred means of help-
ing low-income borrowers. This willingness is based, in part, on Islamic
teachings and extends to borrowers themselves who support the organisa-
tion’s philosophy of solidarity with ‘others in need’ and make voluntary
donations to support the programme. Although it does not adhere to the
same evaluation criteria as employed in conventional interest-based
microfinance, the Akhuwat qard hasan model has been ‘financially sus-
tainable’ so far. How likely is it that this model can be replicated elsewhere?
The discussion during the conference and in this chapter shows that
AIM’s experience appears to be unique, particularly regarding the capac-
ity of the institution to attract donations. Microcredit models based on
qard hasan have either struggled to increase scale and outreach, as has
been the case with the many local replications in Pakistan, or have shifted
their emphasis from providing benevolent loans to a greater focus on
other Islamic microfinance products that generate greater income for the
institution, as was the case with START in Kosovo.
While acknowledging AIM’s innovative approach to cover its operat-
ing costs, it is also important to take into consideration the various strate-
gies adopted by the institution to maintain low operational costs and
high loan repayment rates, and promote a strong sense of commitment
and loyalty from its staff. Some of these strategies, such as the use of
mosques and references to Islamic teachings, incorporate a faith element.
However, others, such as careful client selection and rigorous loan
appraisal, are simply examples of good microfinance practice. Furthermore,

AIM has benefited from the leadership qualities of Dr Amjad Saqib, his
personal charisma and the professional connections he has developed
throughout his career, particularly with the various regional governments
in Pakistan.
Exploring the relative importance of faith and other factors in the suc-
cess of AIM and other Islamic microfinance methodologies, requires fur-
ther research including the application of qualitative methodologies.
Indeed, compared to interest-based microfinance, Islamic microfinance is
a vastly under-researched area. The longitudinal investigation led by
Lendwithcare represents a step in the right direction, especially consider-
ing the intention of the project partners to continue the research and
implement a third wave of the household survey in 2019. This will pro-
vide an opportunity to further investigate the differences encountered in
the outcomes of the microcredit programme for distinct segments of
AIM clients. Analysing client heterogeneity can contribute to better
understanding the change mechanisms associated with benevolent loans,
and consequently, identify critical factors to the successful replication of
the model. More generally, promoting further research and programme
evaluations that focus on the qard hasan model will enhance transparency
and accountability of IMFIs, and contribute to overcoming the prevailing
scepticism towards the financial sustainability of this type of model and
Islamic microfinance. At the same time, it will contribute to the develop-
ment of evaluative thinking and related skills within the institutions.
Notes
1. The authors acknowledge the speakers at the conference plenary for their
contributions to the discussion, Muhammad Shakeel Ishaq and Shahzad
Akram from Akhuwat Islamic Microfinance, for providing information
on the organisation’s operations, and Elise Aston for her comments
and suggestions.
2. The fifth European Research Conference on Microfinance took place in
Portsmouth, UK, from 12 to 14 June 2017. The conference was hosted
by the University of Portsmouth, who co-organised the event with the
European Microfinance Platform (Luxembourg) and the Research

Centre for Microfinance (Belgium), known by its acronym CERMi.
3. It should be noted that despite this comparative worse performance, the
figures show a clear improvement regarding the data from the previous
waves of the survey in 2014 and 2010. Data from the Global Findex
Survey on Financial Inclusion is available at https://globalfindex.world-
bank.org/
4. The University of Portsmouth is one of the partners participating in the
Lendwithcare evaluation project. For more information about
Lendwithcare, please see https://www.lendwithcare.org
5. Zakat is an obligatory almsgiving for all Muslims who meet the neces-
sary criteria of wealth. While zakat funds can be used to cover the opera-
tional costs of organisations that assist the poor, they cannot be used to
fund loans. Sadaqah is a voluntary charitable donation and can be used
for both loan capital and to cover operational costs. Waqf is an endow-
ment (usually a building or plot of land) or a trust set up for charitable
purposes. It involves tying up a property or fund in perpetuity so that it
cannot be sold, donated or inherited by anyone. An IMFI can, for exam-
ple, use the rent from letting a property to finance its operations.
6. According to personal communication with Project Manager, Akhuwat,
22 June 2018. Altogether, Akhuwat employs approximately 7500 staff.
7. The information is available online in the quarterly editions of
MicroWatch (No. 33 and 45), a publication of the Pakistan Microfinance
Network (http://www.pmn.org.pk/publications/category/MicroWatch).
8. Personal communication with Project Manager, Akhuwat, 24 May 2018.
9. Khan, Kustin, and Khan (2017) also observe that interest-free loans are
not exclusively Islamic either. For example, gemach or ‘acts of kindness’
is a Jewish interest-free loan fund with easy repayment terms and similar
to qard hasan.
10. There are several verses in the Qur’an that encourage Muslims to provide
qard hasan loans to ‘those who need them’. For example, ‘Establish regu-
lar prayer and give regular charity and give Allah qard al hasan’ (Surah
73, Verse 20). Indeed, qard hasan loans are considered as if they were
made to Allah, rather than the borrowers, to encourage lenders to part
with their wealth.
11. The final report of the project is available at http://www.careevaluations.
org/wp-content/uploads/Impact-Assessment-Report-Pakistan-Joana-
Afonso.pdf
12. The Poverty Probability Index (PPI) is a poverty assessment tool initially
developed by Mark Schreiner for the Grameen Foundation, and cur-
rently managed by the Innovation for Poverty Action (IPA). PPI is based
on a country-specific questionnaire with 10 multiple-choice questions
on household characteristics and assets ownership. More information
available at https://www.povertyindex.org/
13. The analysis was based on the application of quantile regression methods
to assess the impact of different factors, including the access to the
microcredit loan, in the variation of the PPI scores. The access to the
microcredit loan was statistically significant (at 1% significance level) for
the group of respondents in quartile 3, meaning those who have experi-
enced stronger improvements in the PPI scores.
References
Afonso, J. S. (2018). Lendwithcare Assessment Project: Akhuwat Islamic
Microfinance Report. Project Report. Portsmouth: University of Portsmouth.
Asim, S. (2009). Evaluating the Impact of Microcredit on Women’s Empowerment
in Pakistan (CREB Working Paper 03-09). Lahore: CREB, Lahore School of
Economics.
Beck, T., Demirgüç-Kunt, A., & Levine, R. (2007). Finance, Inequality and the
Poor. Journal of Economic Growth, 12, 27–49.
El-Zoghbi, M., Karlan, D., Osman, A., & Shammout, N. (2016). Understanding
Demand for Sharia-Compliant Loans: Results from a Randomised Experiment in
Jordan. Washington, DC: CGAP.
Haider, M. (2017). MicroWatch (Issue 45). Islamabad: Pakistan Microfinance
Network.
Harper, M. (2017). What Do the Cases Tell Us? In M. Harper & A. A. Khan
(Eds.), Islamic Microfinance: Shari’ah Compliant and Sustainable?
(pp. 185–201). Rugby, UK: Practical Action Publishing.
Harper, M., & Khan, A. A. (2017). Islamic Microfinance: Shari’ah Compliant
and Sustainable? Rugby, UK: Practical Action Publishing.
Javoy, E., & Rozas, D. (2015). MIMOSA 2.0: Mapping the (Micro)Credit Cycle.
Retrieved from MIMOSA Project: http://mimosaindex.org/wp-content/
uploads/2015/11/MIMOSA-White-Paper.pdf
Khan, A. A., & Zeqiri, V. (2017). Pioneering Islamic Microfinance in Kosovo:

The Experience of START. In M. Harper & A. A. Khan (Eds.), Islamic
Microfinance: Shari’ah Compliant and Sustainable? Rugby, UK: Practical
Action Publishing.
Khan, A. A., Kustin, B., & Khan, K. (2017). Islamic Financing Principles and
Their Application to Microfinance. In M. Harper & A. A. Khan (Eds.),
Islamic Microfinance: Shari’ah Compliant and Sustainable? (pp. 1–16). Rugby,
UK: Practical Action Publishing.
Khan, A. A., Ishaq, M. S., Afonso, J. S., & Akram, S. (2017). Is It Possible to
Provide Qard Hasan and Achieve Financial Self-Sustainability? The Experience
of Akhuwat in Pakistan. In M. Harper & A. A. Khan (Eds.), Islamic
Microfinance: Shari’ah Compliant and Sustainable? (pp. 17–31). Rugby, UK:
Practical Action Publishing.
Lieven, A. (2011). Pakistan – A Hard Country. London: Penguin Books.
Obaidullah, M. (2011). Islamic Microfinance: The Way Forward. In K. Hassan
& M. Mahlknecht (Eds.), Islamic Capital Markets: Products and Strategies
(pp. 415–427). New Jersey: Wiley.
Robinson, M. (2001). The Microfinance Revolution. Washington, DC: The
World Bank.
Safavian, M., & Haq, A. (2013). Are Pakistan’s Women Entrepreneurs Being Served
by the Microfinance Sector. Washington, DC: The World Bank.
World Bank. (2018). The Little Data Book on Financial Inclusion 2018.
Washington, DC: The World Bank.
Zaidi, S. A., Jamal, H., Javeed, S. & Zaka, S. (2007, April). Social Impact Assessment
of Microfinance Programmes. Draft Report. Retrieved from https://www.research-
gate.net/profile/Sarah_Javeed/publication/266874152_Social_Impact_
Assessment_of_Microfinance_Programmes/links/56275aea08aed3d3f13a3a91/
Social-Impact-Assessment-of-Microfinance-Programmes.pdf
Zulfiqar, G. (2017). Does Microfinance Enhance Gender Equity in Access to
Finance? Evidence from Pakistan. Feminist Economics, 23(1), 160–185.
5
Relationship Lending in Microfinance:
Do Women Benefit As Much As Men?
Mathilde Bauwin
5.1 Introduction1
The term “financial inclusion” has been gaining importance since the
early 2000s, especially following a speech given on 29 December 2003 by
the former General Secretary of the United Nations, Kofi Annan, who
said: “the stark reality is that most poor people in the world still lack
access to sustainable financial services, whether it is savings, credit or
insurance. The great challenge before us is to address the constraints that
exclude people from full participation in the financial sector”. Since then,
media have been highlighting the number of people excluded of financial
services, and financial inclusion has gradually become one of the primary
objectives of international institutions such as the World Bank or the
various agencies of the United Nations.
M. Bauwin (*)
Université Paris-Dauphine, PSL Research University, IRD, LEDa, [UMR
225], DIAL, Paris, France
Institut National d’Etudes Démographiques, Paris, France

86 M. Bauwin
As a consequence, attention has been focused for about a decade on

the number of people holding a bank account, or more recently a mobile
account thanks to technological progress, with microfinance becoming
only a tool to help increase this number. However, these striking num-
bers, showing the progress made towards financial inclusion, not only
demonstrate the increasing reach of microfinance worldwide, with even
more new clients or “banked” people every year, they also exhibit another
concomitant phenomenon, which is the retention of older clients. Once
people get access to microfinance, they remain “financially included”,
meaning that they keep returning to these products and services.
So far, client retention in the microfinance sector has not appeared as
a major issue of interest. Instead, being inherent to microfinance’s modus
operandi, it is included in impact assessment studies as a way to control
for the duration of inclusion in a microfinance programme. However,
client retention is at the core of some recent scandals about the “mission
drift” of microfinance. Indeed, over-indebtedness, in particular, is more
likely to occur after a client has received several loans than after they have
received the first loan—all the more so as loan amounts usually increase
over credit cycles, which is called “progressive lending”. Thus, this chap-
ter aims at more deeply analysing the conditions of loan renewals.
Additionally, it has been shown by Agier and Szafarz (2013) in relation
to Brazil as well as in one of our previous work on Tunisia that women are
not necessarily favoured in the microcredit allocation process, particu-
larly in terms of amounts granted. This chapter therefore focuses on the
conditions of loan renewals from a gender perspective. In particular, the
objective is to analyse the policy of progressive lending by the main
Tunisian microfinance institution (MFI) Enda inter-arabe to check if the
initial gap observed between amounts granted to new male and female
clients that we observed in our previous work is persistent or not over
credit cycles.
Indeed, this initial gap seems to be accounted for by great information
asymmetry between new applicants and credit officers, which is usual in
any lender–borrower relationship; because of such an asymmetry, officers
tend to refer to stereotypes about women and their projects, most prob-
ably leading to statistical discrimination. In the same way, the main
hypothesis in this chapter is that the same kind of stereotypes may have
5 Relationship Lending in Microfinance: Do Women Benefit… 87
an effect on the application of a progressive lending policy. Since gender

division of labour within the household remains significant in Tunisia
(MAFFEPA, 2005),2 female clients may still be considered to have less
time to dedicate to their projects whatever the credit cycle. As a conse-
quence, we assume that loan officers are likely to conclude that if women
invest less time on average in their activity, their project is likely to evolve
less quickly, and women should also need less money at any credit cycle.
Such general consideration would lead to statistical discrimination taking
the form of a slower progressive lending policy for women that cannot be
justified by project characteristics, or by different risky behaviours
between men and women as clients.
The alternative hypothesis is that credit officers gain information on
clients as clients renew their loans, which should reduce the moral hazard
issue. To that extent, the relationship between credit officers and microfi-
nance clients would be similar to any lending relationship between finan-
cial intermediaries and firms, and several theoretical papers show that
such lending relationships enable creditors to produce information about
borrowers and to use it in their next credit decisions (Campbell & Kracaw,
1980; Diamond, 1991; Leland & Pyle, 1977). As a consequence, MFIs’
credit officers could also learn from their relationships with their clients
that female clients are eventually as able as men to manage their projects
and even less risky in terms of repayment. In such a case, the progressive
lending policy should be at least fairly applied between men and women.
To check our hypotheses, we use longitudinal client data from Enda
and analyse Enda’s progressive lending policy by examining amounts
granted over credit cycles on the one hand and the growth rate of these
amounts granted on the other hand, while correcting the selection bias
and taking former repayment behaviour into account.
The main result is that although female clients are less risky in terms of
repayment behaviour, loan amounts granted to women grow more slowly
from one credit cycle to another than those granted to men, all things
being equal.
Section 5.2 reports how progressive lending is applied in practice and
considered in the literature, Sect. 5.3 describes the data, Sect. 5.4 details
the empirical method, states the results and discusses them, and Sect. 5.5
concludes.
88 M. Bauwin
5.2 P
rogressive Lending in Practice
and in the Literature
5.2.1 Progressive Lending in Microfinance
Few questions are generally raised, other than about the issue of impact,
concerning what happens after clients have received their first micro-
credit. Yet, what happens is very specific to the microfinance sector. As
Armendáriz and Morduch (2010) explain, microfinance institutions aim
at serving vulnerable people, and hence offer very low loan amounts,
leading them to face higher transaction costs than traditional banks. As a
strategy to reduce these costs, MFIs implement “progressive lending”
(Armendáriz & Morduch, 2010, p. 143): they progressively increase the
loan amounts over credit cycles, provided that the client has demon-
strated good repayment behaviour. This enables MFIs to remain profit-
able as their transaction costs progressively decrease relative to loan
amounts. In a broader perspective, one of the strategies implemented by
MFIs is to encourage client retention by creating “good dynamic incen-
tives […] through attractive long-term relationships with clients”
(Armendáriz & Morduch, 2010, p. 161). Progressive lending is one of
these good dynamic incentives designed to encourage clients to keep
resorting to the MFI. Finally, progressive lending is also what enables
MFIs to avoid potentially large losses as, in practice, loan officers can test
borrowers’ repayment behaviour with small loans at first before allowing
them to climb up the loan scale.
As a result, client retention and progressive lending are part and parcel
of the microfinance system. Client retention has recently been the subject
of higher attention: it is today considered as an indicator of social perfor-
mance in the Universal Standards for Social Performance Management,
which has led to its integration in practical tools of social performance
assessment such as SPI4. However, no additional recommendation is
provided in these standards and tools about how progressive lending is
supposed to be implemented. More generally speaking, the conditions of
what would be a responsible policy of client retention and progressive
lending are not detailed.
5.2.2 Progressive Lending in the Literature
With regard to academic literature, progressive lending is usually not ana-

lysed in itself; instead, the time component of client retention is some-
times considered, and appears mostly in impact studies as a way of
distinguishing between treatment and control groups to assess the effects
of benefiting from microfinance services. For instance, Banerjee, Duflo,
Glennerster, and Kinnan (2009) analyse the impact of microfinance on
several economic and social indicators with a randomized experiment in
India, in which the average loan amount enables to control for how long
clients have been benefiting from a microfinance programme, as loan
amounts are supposed to increase gradually with credit cycles. In a more
recent version of their study (Banerjee, Duflo, Glennerster, & Kinnan,
2015), the authors expect to assess the impact of microfinance by com-
paring the treatment group consisting of people benefiting from microfi-
nance services for a longer time (three years), and the control group
consisting of new clients. Client retention is thus used to estimate the
potential impact of microfinance after a certain time. It is considered as a
phenomenon logically resulting from the first step, which is accessing
microfinance services. Once again, the issue of access remains the main
focus, and no question is raised about what happened during the three-
year period in terms of number of loans, amount increase, or variation in
credit cost (interest rates may differ according to loan amounts, low
amounts usually being more expensive than higher amounts).
In the same way, Weber and Ahmad (2014) compare women in higher
loan cycles of a Pakistani microfinance institution with those in the first
loan cycle to assess the possible impact of microfinance on women’s
empowerment. The treatment group consists only of women having
taken part in the microfinance programme for five years, and if the num-
ber of loans they got is provided in descriptive statistics as additional
information, it is not included in the analysis itself. Once again, the treat-
ment and control groups are distinguished only by time, and the impact
of microfinance is supposed to appear only over time, whatever the num-
ber of loans or the growth rate of loan amount.
In the study by OECD (Organisation for Economic Co-operation and
Development) on the effects of microfinance on poor rural households and
the status of women (OECD, 2007), several “treatment variables” are used
90 M. Bauwin
to estimate the impact of microfinance services, these variables being

“availed program loan” (yes or no), the number of months the loan pro-
gramme has been available, the cumulative total amount of loans (which
blurs the possible evolution in loan amounts over cycles), and finally the
number of loans. However, the last two treatment variables were eventually
not used in final discussions as they did not appear statistically significant.
This non-significance could have raised some new questions for further
research, but classical treatment variables were used for the impact
study instead.
The meta-analysis achieved by Chliova, Brinckmann, and Rosenbusch
(2015) is very meaningful in this respect. They gathered the maximum
possible number of quantitative studies about the impact of microfinance
since 1980, ending up with 91 studies. In most of these studies, the inde-
pendent variable of interest is dichotomous and represents the participa-
tion, or not, in a microfinance programme, that is receiving at least one
loan. Chliova et al. (2015) also used some other studies (representing a
minority) in which participation is captured by a continuous variable and
measured by time since the reception of the first loan. Nothing other
than time is used to consider client retention in impact analyses, and the
other features of progressive lending are not examined.
Some recent studies focus on client retention from the MFI’s point of
view, such as Epstein and Yuthas (2013), who show that client retention
is higher in rural parts in Malawi than in urban regions, or Pearlman
(2014), who focuses on the determinants of dropouts, but again the other
aspects of progressive lending such as growth rate of loan amounts or
decrease in credit cost are ignored.
Thus, to the best of our knowledge the conditions of loan renewal have
not been analysed yet.
5.3 Data
5.3.1 Data Preparation and Management
Enda provided a complete panel dataset containing information about all

new clients from June 2012 to December 2013 and about all the loans
they received from June 2012 to March 2016. It was decided to limit the
dataset to new clients up to the end of December 2013 as the situation in

the country changed in 2014, with the entry of new actors in the micro-
finance sector leading to the possibility that new clients in 2014 may have
been selected in a different way.
The whole dataset consists of 69,301 clients (63.5% of whom are
women) who received a total of 183,109 loans. One client can hold two
loans concurrently but not two project loans. The loans dedicated to per-
sonal projects or market opportunity (a tiny minority though, as they
represent only 2636 loans, that is 1.42% of all the loans granted over the
period) were removed from the dataset. However, as holding two credits
at the same time could influence the evolution of projects and/or clients’
financial situation, a dummy variable was created to take this informa-
tion into account.
5.3.2 Descriptive Statistics
5.3.2.1 Evolution of Clients’ Projects
As already mentioned, all the loans considered in our dataset were dis-
bursed to finance a project, which may be a low-scale income-generating
activity (IGA) or a micro or very small enterprise. In order to analyse the
progressive lending policy applied by an MFI, it is necessary to consider
how clients’ projects evolve, as the way loan amounts grow is likely to
depend at least partly on the evolution of clients’ projects.
The average age of projects when clients receive their first loan is
5.8 years, without statistical difference between projects led by men or
women. The main activity sector is agriculture among both male and
female clients, followed by trade. However, women are more likely to
lead projects in the production sector (i.e. mainly textile production,
food production, or handicrafts) whereas men are more likely to work in
services (especially transport or mechanics).
As often observed in microfinance, female clients tend to lead smaller
projects than men. Here, the classification concerns the type of financial
products which are intended to be tailored to each type of project. When
clients receive their first loan, women are relatively more likely to receive
a product designed for income-generating activities, or “micro projects”,
92 M. Bauwin
whereas men are relatively more likely to receive credit for very small
enterprises, especially in the non-agricultural sector. In addition, a spe-
cific financial product is designed for young people only (under 35 years
of age) to enable them to start an activity, and men are more represented
in this category than women. The financial products differ, in particular,
in terms of maximum amounts and interest rates. Although they are sup-
posed to be tailored to the size and type of clients’ projects, the choice of
financial product is at the discretion of loan officers. For instance, if a
loan officer estimates that a high amount (above 3000 TND) should be
granted considering the project characteristics, he or she has no other
choice than granting a financial product for very small enterprises. We
therefore cannot conclude with certainty that a client’s project exactly
corresponds to the category the product is supposed to be designed for—
this classification only reflects the assessment of loans officers.
A striking gender difference concerns the evolution of financial prod-
ucts over credit cycles. If we estimate that the financial product granted
actually corresponds to the project’s size and type, a micro project may
turn into a very small enterprise whether in the agricultural sector or not,
or a project may regress and a small enterprise may decline into a micro
project. In the same way, the creation of an activity by a young client may
then turn into a micro project or a very small enterprise. In any case, the
evolution of financial products from one credit cycle to another reflect at
least the way officers see the evolutions of clients’ projects, if not actual
evolutions. Men who receive a first credit for a micro project are more
likely to receive subsequent credits for small enterprises than women,
who are more likely to keep receiving credits for micro projects. By con-
trast, women receiving first credits for small enterprises seem more likely
to regress in terms of financial product as compared to men, that is to say
to receive next credits for micro-projects.
This could reflect the fact that women’s projects develop less quickly
than men’s, possibly because of the gender division of labour in the house-
hold, differences in priorities and preferences, inequalities in access to
resources and mobility, or starting inequalities in education, training and
skills, and so on. The second possibility is that this evolution reflects the
evolution of loan officers’ assessments, especially of their clients’ financial
needs, as financial products are distinguished not only by activity sector
but also by their maximum amount. This is why we turn to other more
objective indicators to take the size and type of projects into account in the
econometric analyses, such as fixed assets, current assets, or monthly profit.
The dataset has three more indicators for non-agricultural loans only,
which are being part of the formal sector or not (which means the activity
is officially registered), the location of the project (at home or in indepen-
dent premises), and monthly profit. It has two other indicators for agri-
cultural loans, which are the useful area for the activity and three categories
of project size assessed by the value of fixed assets.
About non-agricultural loans, when clients receive their first loans,
only 19.4% of their projects are officially registered on average, but this
is even less the case for women, as 11.9% of them work in the formal
sector against 33.7% of men. Additionally, gender differences exist in
transitions as well: men working in the informal sector are respectively
more likely to evolve towards the formal sector than women, as 18.6% of
them make the transition against only 6.3% of women, and less likely to
be back into the informal sector when being officially registered (25.1%
lost the registration against 41.7% of women).
Concerning agricultural loans, the useful area for men’s projects is
3.83 ha on average against 2.99 ha for women’s projects. The MFI also
classifies the projects according to the value of fixed assets and considers
that the project is an income-generating activity (IGA) if fixed assets
worth less than 8000 TND, a micro enterprise if they worth between
8000 and 100,000 TND, and a very small enterprise above 100,000
TND. Again, men are more represented in very small enterprises (2.58%
against 0.56% of women’s projects), whereas women are more likely to
run income generating activities (82.1% against 59.3% of men’s proj-
ects), for both culture and breeding, even though breeding tend to be
smaller projects for both men and women.
In the same way, 89.7% of women’s projects did not have any employee
against 83.9% of men’s projects, since 10.5% of men’s projects resorted
to regular employees, 2.4% to seasonal workers, and 3.3% to both,
against respectively 6.4%, 1.3%, and 2.7% of women’s projects.
These indicators tend to show that women’s projects would start from
further behind and then develop less quickly. However, as project devel-
opment also depends on financial investment, the role of the progressive
lending policy in these evolutions remains unknown at this point.
94 M. Bauwin
5.3.2.2 Evolution of Clients’ Financial Situations
When receiving a loan, clients should provide a guarantee, which can

take several forms as is usual in microfinance, since MFIs are supposed to
be more flexible with people excluded from the traditional banking sys-
tem. Once again, the types of collateral offered by clients vary according
to gender. First, the share of unique guarantors (a loyal client acting as a
guarantor, parental engagement or the client’s own credit background) is
similar among men and women. However, women tend to resort more to
reciprocal guarantee: this includes joint surety, which involves several
current clients, and mutual guarantee, which involves only one other cli-
ent. Conversely, men have more recourse to financial or physical guaran-
tees (salary or pledging of equipment). This could reflect the existing
gender inequalities in terms of access and control over resources. In par-
ticular, most female clients are married and have another member of their
household who is active, these proportions being higher among women
than men. We can therefore expect women to be at least as likely as men
to offer salary as collateral, but salary is the most common collateral
offered by men and not by women, which would imply that women can-
not use their household’s resources as collateral or prefer not to.
With regard to specific financial indicators, if the household’s financial
situation does not differ much between men and women, the project’s
financial indicators are higher for men than women when all credit cycles
are taken together. Households’ median expenses and revenues are c om-
parable, whereas median fixed assets, current assets, and monthly profits
(applicable to non-agricultural projects only) are higher for men’s proj-
ects than for women’s.
To take a first look at the evolution of these indicators over credit
cycles, we consider only clients who got four credit cycles (15,572 clients
from our dataset) to avoid selection bias and compare comparable clients
(Fig. 5.1). The evolutions in terms of value and of ratio (with the baseline
being the value of the indicator when the client took his or her first loan)
show that all financial indicators evolve positively for both male and
female clients, but some gender differences may be observed: fixed assets
of men’s projects increase more quickly than those of women, which wid-
ens the initial gap. Concerning current assets, the gap in value between
Fig. 5.1 Evolution of financial indicators, in value and ratio (medians)

96 M. Bauwin
Fig. 5.1 (continued)
men’s and women’s projects persists, but current assets of women’s proj-
ects still increase more quickly. In the same way, the existing gap between
men’s and women’s monthly profits (for clients getting non-agricultural
loans) slightly widens, but women’s profits grow more quickly. This ques-
tions the assumption that women’s projects grow more slowly: the initial
gaps are sizeable between men’s and women’s projects, but then men and
women seem to manage their projects differently and to make different
choices in terms of investments: men seem in particular more likely to
invest in fixed assets. With regard to households’ financial indicators,
both revenues and expenses increase slightly more quickly for women
than for men, which could also indicate different choices in terms of
allocation of resources. Given these contrasted figures, both values and
evolutions in ratios of financial indicators will be taken into account in
the analysis of loan renewals.
5.3.2.3 Loan Amounts over Credit Cycles
As it is the case for most microfinance institutions, Enda applies a policy of

progressive lending: amounts granted go from an average of 678 TND for
the first loan up to 2364 TND for the fifth loan (Table 5.1). Not surpris-
ingly, the amounts are higher for male clients, which could be explained by
the differences between men’s and women’s projects in terms of size, type,
or financial indicators. The econometric analysis will attempt to check if
these differences totally explain the gaps observed in amounts granted or
not. It also seems that amounts granted increase more quickly over credit
cycles for men than for women, as the gaps between amounts granted to
men and women become increasingly higher over cycles (Fig. 5.2).
To consider the evolution of these loan amounts in greater detail, we
again consider ratios; this time, as loan amounts are limited (the ceiling
being 5000 TND), we do not expect extreme values and use average
ratios. The evolution of loan amounts over credit cycles is defined as the
growth rate of loan amounts from one credit to the next one.
First, it should be noticed that if loan amounts increase from a credit cycle
to another in 95% of the cases, they still may decrease: thus, the minimum
ratio from a cycle to the next one is 0.10 while the maximum is 12.5. The
evolution of growth rate is represented in Fig. 5.3. Unsurprisingly, if the
growth rate is substantial from the first credit to the second, it tends to be
lower afterwards. Indeed, the leeway for increasing the amount is high after
the first loan and then decreases. The evolution of the growth rate is similar
for men and women. Nonetheless, knowing that the amounts at the first
credit cycle are much lower for women, such similar growth rates can result
in increasing gaps in terms of loan amounts, as seen in Fig. 5.2. Moreover,
as financial indicators evolve differently for men and women, we cannot
know at this stage if these similar growth rates represent a fair progressive
lending policy which takes the evolution of projects into account.
Table 5.1 Average loan amount by credit cycle and by gender

Cycle 1 2 3 4 5
Men 882 1401 1899 2329 3058
Women 560 924 1274 1577 1912
Total 678 1093 1494 1838 2364
98 M. Bauwin
Fig. 5.2 Average amounts granted over credit cycles by gender
Fig. 5.3 Evolution of loan amounts from one credit cycle to another by gender
(in ratios over the previous amount)
5.4 Methods and Results

The aim of the chapter is to analyse the conditions of loan renewals and,
in particular, to check if the loans are renewed in a fair manner between
men and women, given the evolution of their respective projects and situ-
ations. However, the first emerging issue is the fact that not all clients
renew their loans. There is a significant amount of natural attrition in our

dataset,3 which corresponds to the clients who left the MFI. As dropouts
may have different characteristics, whether they left the MFI after default-
ing or not, we suspect that the selection (whether it is self-selection by
clients themselves or exclusion by the MFI) is not random. Therefore, the
analysis includes a correction for selection bias on panel data.
The variable of interest is first the amount granted when a loan is
renewed, and second the growth rate of loan amounts.
Whatever the variable of interest considered, the selection bias should be
corrected. In order to do that, we follow Wooldridge’s procedure (Semykina
& Wooldridge, 2010; Wooldridge, 1995) to correct selection bias in panel
data models. Indeed, unobservable variable(s) could have an effect on both
the probability of renewing a loan and the loan amount granted, in level or
growth rate. These characteristics could be tenacity or perseverance which
could push the client to renew their loan to keep their activity running, as
well as insisting that the loan officer should increase loan amounts more
significantly; they could also be better entrepreneurial skills in general. These
unobservable variable(s) could also be correlated to the client’s or project’s
observable characteristics, particularly the project’s financial indicators such
as fixed assets, current assets, or profits for non-agricultural projects. This
possible correlation is an allowed hypothesis in Wooldridge’s procedure.
The procedure is composed of three steps. The first equation may be
written in latent form as follows:
S* = Z δ +ε
it +1 it +1 t +1 it +1 (5.1)
s =1 if Zit δ + ε >0
 it +1 +1t t +1 it +1
And 
s = 0 otherwise
 it +1

The selection equation being: sit = Zitδi + εit > 0
With
• sit = 1 if the client renews a loan at the end of the credit cycle t,
with 1 ≤ t ≤5
• Zit representing the client’s characteristics (some being time-varying, i.e.
changing from one credit cycle to another, and others being time-invariant)
as well as the characteristics of the project and the loan (being time-varying)
100 M. Bauwin
• δ a vector of parameters
• εit following a normal distribution
• Corr(εit, εit + 1) = ρtt + 1
After estimating Eq. (5.1) with T standard probit models, the second
step consists in computing T inverse Mills ratios for sit = 1, and these
ratios are included in the subsequent equation to correct the selection
bias. The exclusion variable is the fact that the last repayment of the pre-
vious loan was made during the last week of the month, since, in this
case, the probability of a loan being renewed is much lower. Indeed, this
week is supposed to be dedicated to collecting late repayments, and offi-
cers are even not supposed to proceed to any renewal during the period.
If a closing date falls during this period, officers are expected to anticipate
and launch the renewal procedure before. If they do not, there is higher
risk that there is no renewal at all. Consequently, the probability of a loan
being renewed is negatively correlated with the previous loan’s closing
date falling during this period. By contrast, if a loan is renewed, the fact
that the last instalment was paid during the last week of the month is not
expected to have any effect on the next amount granted.
5.4.1 T
he Amounts Granted After the First
Credit Cycle
We first focus on the amounts granted in level, in order to check if the gap
observed between amounts granted to men and women at first cycle persists
or not for the next cycles. There are two main differences between the grant-
ing procedure of the first amount and of the next ones: first, officers better
know their clients from the second cycle, and do not resort to a moral
enquiry any longer, since clients’ repayment behaviour during the first cycle
is a sufficient indicator to anticipate moral hazard; second, there is no official
amount requested by clients after the first cycle. Indeed, at the time of our
study, the renewal procedure consisted in officers discussing with clients
before the end of the previous cycle, and determining a loan amount together
given clients’ financial conditions and needs. Consequently, there is no offi-
cial amount requested by clients which can be taken into account in the
estimation of amounts granted from cycle 2.
Following Wooldridge’s procedure, loan amounts from cycle 2 are esti-

mated using Eq. (5.2) and a pooled ordinary least squares estimator with
bootstrapped standard errors:
yit = ψ xi + β X it + γ Request + ϕYear + δ Nbdelay + ρλit + εit

i1 1 it −1
(5.2)
With
• yit the loan amount granted for each individual i at a time t, t being a
credit cycle >1
• x̄i the vector of average values of Xit by individual
• Xit the matrix of time-varying independent variables
• Requesti1 the amount requested at cycle 1
• Year1 a dummy variable indicating the year (2012 or 2013) when the
first credit cycle was granted, as amounts tend to grow from a year
to another
• Nbdelayit-1 the number of days of delay of the previous credit
cycle (in log)
• λit the inverse Mills ratios
• εit the idiosyncratic error term
The matrix of independent variables consists of constant socio-

demographic characteristics (age, marital status, household size, educa-
tion), and the loan’s time-varying characteristics (credit cycle, time interval
between two loans, collateral), the project’s time-varying characteristics
(activity sector, age, employees), financial time-varying characteristics
(household expenses, fixed assets, current assets), and organizational time-
varying features (new officer compared to the previous loan or not, offi-
cer’s gender, branch fixed effect, branch’s age and branch’s rate of rural areas).
The model is run on all loans first (Table 5.2), and then on non-
agricultural and agricultural loans separately (Table 5.3), in order to
introduce the additional information available for each type of financial
product. The coefficients of average values of Xit by individual have been
dropped from the table for more clarity.
Concerning all loans first, a first model is run including a linear effect
for credit cycles (Model 1), a second one introduces dummy variables for
102
Table 5.2 Estimation of amounts granted from cycle 2 (all loans)

Model (1) Model (2) Model (3)
Female −113.9*** (4.353) −113.7*** (4.289) −12.18** (5.214)
Credit cycle (linear effect) 365.4*** (2.586)
M. Bauwin
Credit cycle (vs. 2)

3rd cycle 387.7*** (3.890) 481.4*** (8.067)
4th cycle 712.0*** (7.159) 882.1*** (12.69)
5th cycle 1116*** (18.38) 1367*** (30.86)
Female # credit cycle
Female # 3rd cycle −143.2*** (8.534)
Female # 4th cycle −260.0*** (13.31)
Female # 5th cycle −400.5*** (34.37)
Loan
Previous number of days overdue (log) −144.6*** (3.099) −144.0*** (3.039) −146.0*** (2.907)
Number of days between two cycles −0.224*** (0.0400) −0.236*** (0.0332) −0.223*** (0.0370)
Requested amount at cycle 1 (100 TND) 26.6*** (0.282) 26.6*** (0.299) 26.5*** (0.264)
First amount received in 2013 (vs. 2012) 35.11*** (4.808) 34.47*** (4.210) 33.35*** (4.442)
Parallel personal or opportunity loan 124.7*** (35.39) 120.9*** (36.53) 108.9*** (35.05)
Collateral (vs. unique guarantor)
Reciprocal guarantee −5.913 (8.905) −7.539 (8.317) −6.479 (7.983)
Physical guarantee −30.00*** (10.83) −30.48*** (10.88) −19.88** (9.934)
Credit use (vs. working capital)
Investment 354.4** (166.6) 353.9** (168.9) 347.3** (159.9)
Creation −175.6*** (64.02) −176.0*** (67.07) −177.1** (70.55)
Other −114.2 (152.6) −114.5 (165.7) −130.7 (154.2)
Project
Activity sector (vs. agriculture)
Trade 68.35*** (13.70) 68.01*** (13.19) 63.31*** (14.09)
Production 30.91** (13.23) 30.98** (13.74) 22.48 (15.04)
Services 122.1*** (20.10) 122.5*** (21.01) 118.7*** (22.97)
Not documented −17.14 (18.13) −16.30 (19.93) −14.87 (22.29)
Age of project −0.834*** (0.312) −0.847*** (0.263) −0.885*** (0.285)
Employees (vs. none)
Seasonals only 68.62*** (24.66) 68.87*** (23.46) 60.96*** (23.14)
Regular workers only 108.3*** (13.47) 108.8*** (13.15) 102.3*** (13.31)
Both 109.0*** (23.06) 109.0*** (21.81) 105.3*** (23.12)
Fixed assets (100 TND) 0.268*** (0.0256) 0.268*** (0.0262) 0.257*** (0.0279)
Current assets (100 TND) 2.19*** (0.118) 2.19*** (0.120) 2.19*** (0.123)
Socio-demographic profile
Young (<35) −43.66*** (16.24) −42.91*** (16.42) −39.29** (15.99)
Education (vs. illiterate)
Primary 17.86*** (5.981) 17.86*** (5.231) 17.56*** (5.388)
Secondary 58.52*** (6.619) 58.57*** (6.358) 58.76*** (6.382)
Higher 139.6*** (10.39) 139.9*** (10.82) 139.7*** (10.46)
Housing (vs. tenant)
Free lodging 22.28*** (7.585) 22.33*** (7.133) 23.04*** (7.191)
Owner 29.40*** (6.128) 29.20*** (5.976) 29.69*** (5.637)
Other active member in household −43.62*** (5.713) −43.92*** (5.505) −44.51*** (5.527)
Household size −2.061* (1.134) −2.017* (1.118) −1.989* (1.087)
Single 4.096 (5.356) 4.572 (5.063) 5.151 (5.267)
Household monthly expenses (100 TND) 2.50* (1.46) 2.52* (1.45) 0.0293* (0.0160)
5 Relationship Lending in Microfinance: Do Women Benefit…
(continued)
103
104
Table 5.2 (continued)
Model (1) Model (2) Model (3)
M. Bauwin
Officer and branch

Other officer than previous loan 48.00*** (6.132) 47.82*** (6.080) 48.87*** (5.739)
Officer gender 9.901 (8.868) 10.18 (9.103) 11.30 (7.945)
Officer experience (years) 8.210*** (0.879) 8.239*** (0.808) 8.342*** (0.840)
Branch rate of rural areas (vs. less than 0.07%)
0.07–0.35% 201.0*** (16.22) 200.7*** (15.00) 202.4*** (15.87)
0.35–0.55% 394.2*** (16.96) 392.1*** (15.54) 390.3*** (17.12)
>0.55% 601.4*** (21.55) 596.7*** (21.59) 595.2*** (23.13)
New branch −66.64*** (11.03) −64.02*** (10.83) −66.19*** (10.12)
Inverse Mills ratio −485.5*** (10.52) −495.7*** (10.92) −507.7*** (11.76)
Constant −26.62 (159.1) 697.4*** (165.4) 650.2*** (157.7)
All average Xi included Yes Yes Yes
Branch fixed effect Yes Yes Yes
Observations 110,608 110,608 110,608
Adjusted R-squared 0.647 0.647 0.650
Bootstrapped standard errors in parentheses
***p < 0.01, **p < 0.05, *p < 0.1
Table 5.3 Estimation of amounts granted from cycle 2 (non-agri. and agri. loans
separately)
Non-agricultural Agricultural
loans (4) loans (5)
Female 71.04*** (7.199) 3.600 (7.348)
3rd cycle 481.9*** (9.873) 461.0*** (11.61)
4th cycle 864.9*** (14.34) 884.3*** (19.86)
5th cycle 1260*** (37.78) 1422*** (40.82)
Female # 3rd cycle −132.6*** (10.91) −141.5*** (13.49)
Female # 4th cycle −234.3*** (15.19) −250.5*** (21.98)
Female # 5th cycle −312.5*** (41.15) −370.8*** (55.63)
Loan
Previous # of days overdue (log) −134.7*** (3.247) −146.5*** (4.226)
Number of days between two cycles −0.216*** (0.0423) −0.230*** (0.0669)
Requested amount at cycle 1 (100 TND) 22.2*** (0.309) 26.1*** (0.462)
First amount received in 2013 (vs. 2012) 3.246 (4.618) 88.96*** (7.708)
Parallel personal or opportunity loan 96.06*** (34.53) 128.1** (62.33)
Reciprocal guarantee −8.177 (10.00) 7.923 (14.97)
Physical guarantee −13.15 (11.98) −13.05 (18.72)
Investment 182.2 (169.4) 573.3* (326.7)
Creation −194.9*** (70.65) −18.41 (150.7)
Other 107.9 (204.9) −189.7 (268.2)
Project
Trade 25.54 (17.39) NA NA
Production −16.19 (17.86) NA NA
Services 67.30*** (24.51) NA NA
Not documented −17.33 (23.01) NA NA
Age of project −1.640*** (0.314) −0.751 (0.460)
Seasonals only 80.23*** (27.08) 1.957 (43.71)
Regular workers only 75.76*** (13.71) 18.03 (24.14)
Both 87.42*** (22.03) 43.76 (37.56)
Fixed assets (100 TND) 0.234*** (0.0563) 0.260*** (0.0309)
Current assets (100 TND) 2.07*** (0.148) 1.67*** (0.158)
Monthly profit (100 TND) 9.03*** (0.855) NA NA
Formal sector 81.70*** (10.82) NA NA
Independent premises 53.27*** (8.798) NA NA
Culture (vs. breeding) NA NA 64.10*** (20.75)
(continued)
106 M. Bauwin
Non-agricultural Agricultural
loans (4) loans (5)
Young (<35) −64.69*** (18.13) −1.342 (27.55)
Primary 19.89*** (6.909) 19.71*** (7.358)
Secondary 59.52*** (7.748) 51.23*** (8.719)
Higher 135.3*** (12.26) 138.5*** (20.60)
Free lodging 35.65*** (6.970) −19.42 (23.62)
Owner 39.97*** (5.791) −16.41 (21.45)
Other active member in household −26.09*** (7.090) −45.41*** (7.534)
Household size −3.860*** (1.319) 3.911** (1.699)
Single 10.58* (5.840) 7.520 (8.798)
Household monthly expenses (100 1.88 (1.94) −3.19 (2.29)
TND)
Officer and branch
Other officer than previous loan 45.38*** (6.301) 55.89*** (10.29)
Female officer 23.80*** (8.897) −5.699 (15.48)
Officer experience (10 years) 8.032*** (0.942) 7.116*** (1.201)
Branch rate of rural areas (vs. less than 0.07%)
0.07–0.35% 165.0*** (15.09) 206.6*** (72.98)
0.35–0.55% 229.2*** (18.84) 556.5*** (73.44)
>0.55% 535.5*** (67.59) 622.6*** (74.57)
New branch −43.94*** (10.34) −78.27*** (16.46)
Inverse Mills ratio −469.4*** (12.76) −424.1*** (20.54)
Constant 578.6*** (210.6) 512.2* (281.2)
All average Xi included Yes Yes Yes Yes
Branch fixed effect Yes Yes Yes Yes
Observations 75,226 41,285
Adjusted R-squared 0.673 0.671
***p < 0.01, **p < 0.05, *p < 0.1
cycles 2 to 5 (Model 2), and a third one keeps dummy variables and
includes an interaction term between women and each cycle (Model 3).
Indeed, introducing dummy variables enables the effects of credit cycles
to be heterogeneous, which seems relevant in our case given the results.
Looking at Model 1, the first striking result is that Enda indeed applies a
progressive lending policy: loan amounts increase by 365 TND on aver-
age from a cycle to another, all other things being equal, which means
that this increase is not due to project evolution but only to the fact that
clients start an additional credit cycle. Model 2, with dummy variables
for credit cycles from 2 to 5, confirms this result, and shows that the
increase from a cycle to another seems regular. This implies that the rela-
tionship building between clients and the MFI all along clients’ credit
history enables these clients to benefit from better loan conditions, and
especially from higher loan amounts than if it had been their first credit.
Relationship lending takes the form of progressive lending in the case of
this MFI, since other credit conditions such as interest rates cannot
change for a specific financial product, whatever the credit cycle.
These two models also show that the effect of being a woman is to get
113 TND less than a man on average, all other things being equal, and
especially for a same repayment behaviour, represented by the number of
days overdue during the previous credit cycle (which has a strong nega-
tive effect on amounts granted, as expected).
Model 3 gives more details on this gender effect, and shows that even
though women benefit from the progressive lending policy as well, the
negative effect associated with being a woman is increasingly stronger
over credit cycles. Looking at Models 4 and 5 enables us to get a more
accurate view of the gender effect by introducing the additional informa-
tion available for non-agricultural and agricultural loans. Interestingly,
women do not seem to get lower amounts at cycle 2, since those receiving
non-agricultural loans even appear advantaged in terms of amount; how-
ever, from cycle 3 the negative effect associated with being a woman
appears as increasingly stronger, as illustrated by Fig. 5.4.
With regard to other determining factors of loan amounts, having
requested higher amounts at cycle 1 has a positive effect on future loan
amounts, so has a first credit received in 2013 compared to 2012 for
agricultural loans. Indeed, over the last years, Enda has implemented a
policy consisting in encouraging officers to grant higher amounts than
before, all other things being equal; as a consequence, loans granted in
2013 were likely to be greater than those granted in 2012, even though
they were first loans. The introduction of this dummy variable enables us
to isolate this effect.
As already mentioned, having showed a bad repayment behaviour with a
high number of days overdue during the previous credit cycle has a negative
108 M. Bauwin
Fig. 5.4 Predictions of amounts granted from cycle 2

effect on amounts granted. Interestingly, collateral does not seem as deter-

mining as for the first loan, as there is no significant effect of any type of
guarantee. By contrast, higher financial indicators such as fixed assets, cur-
rent assets, or monthly profits still have a positive effect on amounts granted,
as well as being officially registered or having independent premises: this
indicates that the progressive lending policy is not just an automatic increase
in loan amounts, but also takes project characteristics into account. Being
young still has a negative effect as well but for non-agricultural loans only,
while it is not determining for agricultural loans.
Finally, concerning officers, being received by a new officer compared
to the previous loan has a positive effect on amounts granted, which may
reflect officers’ willingness to get back and keep the clients previously fol-
lowed by their colleagues. More experienced officers also tend to grant
higher amounts, which might correspond to higher confidence and less
risk aversion for these officers. Moreover, just for first loans, female offi-
cers do not tend to grant lower amounts, as there is no effect of officer
gender for agricultural loans, and even a slightly positive effect for
non-agricultural ones.
5.4.2 T
he Evolution of Amounts from a Credit Cycle
to the Next One
We now focus on the evolution of the loan amount from one credit cycle
to the next, implying that the dependent variable is the ratio between the
on-going credit amount and the previous amount received. Indeed, look-
ing at the progressive lending policy in terms of growth rate may provide
additional information: if loan amounts grow over credit cycles, the
growth rate may remain constant all along clients’ credit history, or
increase itself, which would represent an even stronger advantage for cli-
ents resulting from their relationship with the MFI.
Technically, the magnitude of the ratio should depend on the previous
amount: we expect the ratio to be higher if the previous amount was low,
as the MFI would have more leeway to increase the amount given the
credit ceiling fixed at 5000 TND. Therefore, we would expect these ratios
to decrease over credit cycles if previous amounts were not included,
since amounts increase over credit cycles. This is why we include previous
110 M. Bauwin
amounts in explicative variables to isolate this mechanical effect, and to

properly observe the effect of the growing relationship between clients
and the MFI through credit cycles.
As a consequence, growth rates are estimated using Eq. (5.3):
y = ψ χ + β X + γ Request +ϕYear +τ Amount
it i it i1 1 it −1
+α Install +δ Nbdelay + ρλ + ε
it −1 it −1 it it (5.3)
With
• yit the ratio between the on-going loan amount and previous loan
amount observed for each individual i at a time t, t being a credit cycle
• x̄i the vector of average values of Xit by individual
• Xit the matrix of time-varying independent variables
• Requesti1 the amount requested at cycle 1
• Year1 a dummy variable indicating the year (2012 or 2013) when the
first credit cycle was granted
• Amountit-1 the amount received at the previous credit cycle
• Installit-1 the duration in months of the previous credit cycle
• Nbdelayit-1 the number of days of delay of the previous credit cycle
• λit the inverse Mills ratios
• εit the idiosyncratic error term
The matrix Xit includes the same independent variables as before, add-
ing the growth rates of financial indicators in addition of their levels.
The model defined by Eq. (5.3) is estimated by adding an interaction
term between women and credit cycle in dummies, as we are especially
interested in the effect of gender on the growth ratio over credit cycles. The
model is estimated by separating non-agricultural loans (Model 6), and
agricultural ones (Model 7). Indeed, we have additional information (in
particular, monthly benefit, location of the activity, and official registration
or not for non-agricultural projects, and activity for agricultural projects)
about the projects for these specific types of loans and consider that such
information is relevant as it could have an impact on the growth rate of the
amount of loan and should be included. Furthermore, separating loans
enables us to check whether the observed effects are similar for all types of
loans or not. The results of the two models are presented in Table 5.4.
Table 5.4 Growth rate of loan amount from one credit cycle to another (in ratio)
Non-agricultural
loans (6) Agricultural loans (7)
Female 0.0583*** (0.00555) 0.00690 (0.00862)
3rd cycle 0.128*** (0.00889) −0.0529*** (0.00981)
4th cycle 0.302*** (0.0146) −0.0178 (0.0139)
5th cycle 0.475*** (0.0239) 0.100*** (0.0234)
6th cycle 0.754*** (0.0968) 0.183*** (0.0337)
Female # 3rd cycle −0.0658*** (0.00710) −0.0285*** (0.00940)
Female # 4th cycle −0.112*** (0.00935) −0.0388*** (0.0120)
Female # 5th cycle −0.131*** (0.0191) −0.108*** (0.0250)
Female # 6th cycle 0.0300 (0.152) 0.0107 (0.0748)
Loan
Previous loan amount −0.321*** (0.00562) −0.343*** (0.00680)
(TND 1000)
Previous loan term (months) −0.110*** (0.00234) −0.0501*** (0.00302)
Previous number of days −0.117*** (0.00391) −0.121*** (0.00552)
overdue (log)
Number of days between two −0.00110** (4.74e-04) −0.000298 (6.99e-04)
cycles (10 days)
First amount received in 2013 0.0322*** (0.00358) 0.0567*** (0.00581)
(vs. 2012)
Requested amount at cycle 1 0.0505*** (0.00274) 0.0980*** (0.00366)
(1000 TND)
Parallel personal or 0.0494** (0.0244) 0.116** (0.0528)
opportunity loan
Reciprocal guarantee −0.0400*** (0.00799) −0.0538*** (0.00913)
Physical guarantee 0.0285*** (0.00887) 0.0382*** (0.0139)
Investment −0.0697 (0.170) 0.290 (0.366)
Creation 0.245*** (0.0629) 0.274* (0.150)
Other −0.0892 (0.249) −0.0140 (0.327)
Project
Trade 0.0309** (0.0145) NA NA
Production −0.00332 (0.0151) NA NA
Services 0.0713*** (0.0186) NA NA
Not documented −0.0330* (0.0191) NA NA
Age of project −0.000231 (0.000269) −0.000357 (0.000413)
(continued)
112 M. Bauwin
Non-agricultural
Seasonals only 0.0468*** (0.0181) 0.0582** (0.0248)
Regular workers only 0.0353*** (0.00924) 0.00397 (0.0213)
Both 0.0200 (0.0147) −0.00310 (0.0316)
Fixed assets (1000 TND) 0.00140*** (2.68e-04) 0.00210*** (2.19e-04)
Current assets (1000 TND) 0.0163*** (0.00103) 0.0141*** (0.00128)
Evolution of fixed assets (ratio) −1.43e-08 (1.11e-06) 5.90e-06*** (1.53e-06)
Evolution of current assets −7.71e-06 (5.13e-06) 1.28e-05* (7.01e-06)
(ratio)
Monthly profit (1000 TND) 0.0727*** (0.00501) NA NA
Evolution of profits (ratio) 1.56e-05 (2.32e-05) NA NA
Formal sector 0.0570*** (0.00851) NA NA
Independent premises 0.0612*** (0.00613) NA NA
Culture (vs. breeding) NA NA 0.0907*** (0.0175)
Young (<35) −0.0149 (0.0135) −6.30e-05 (0.0194)
Primary 0.00773 (0.00700) 0.00424 (0.00639)
Secondary 0.0104 (0.00740) 4.36e-05 (0.00828)
Higher 0.0193* (0.0103) 0.00777 (0.0154)
Free lodging −0.00223 (0.00606) 0.00201 (0.0161)
Owner 0.00265 (0.00447) −0.0106 (0.0140)
Other active member −0.000339 (0.00467) −0.0117** (0.00576)
in household
Household size 0.000569 (0.00108) 0.00420*** (0.00136)
Single 0.00684 (0.00470) 0.0183*** (0.00682)
Household monthly expenses 0.0340*** (0.0109) 0.0324* (0.0168)
(1000 TND)
Evolution of expenses (ratio) −0.00101 (0.00130) 0.000351 (0.00205)
Officer and branch
Other officer than previous 0.0173*** (0.00473) 0.0258*** (0.00790)
loan
Officer gender −0.00187 (0.00752) 0.00455 (0.0145)
Officer experience (years) 0.000843 (0.000671) 0.00214* (0.00114)
Branch rate of rural areas
(vs. less than 0.07%)
0.07–0.35% 0.0886*** (0.0126) 0.103 (0.0715)
0.35–0.55% 0.0660*** (0.0155) 0.220*** (0.0698)
(continued)
Non-agricultural
>0.55% 0.0352 (0.0352) 0.168** (0.0691)
New branch 0.0581*** (0.0107) 0.0228* (0.0119)
Inverse Mills ratio −0.0858*** (0.0205) −0.123*** (0.0250)
Constant 1.230*** (0.254) 1.285*** (0.343)
All average Xi included Yes Yes
Branch fixed effect included Yes Yes
Observations 70,707 40,918
Adjusted R-squared 0.386 0.315
Bootstrapped standard errors in parentheses
***p < 0.01, **p < 0.05, *p < 0.1
First and foremost, in accordance with what was expected, we observe

that the effect of the amount previously received is statistically significant
and negative in the two models: the higher the previous amount, the
lower the ratio between the on-going amount and the previous one.
Having taken account of the effect of the previous amount, we observe
significant and clear positive effect of credit cycles for non-agricultural
loans (Model 6). Looking at the interaction term with female, women
seem to benefit from increasing growth rates just as men, even though the
effect of credit cycles is significantly less positive for women. This means
that at equal amount previously received, the growth rate is higher
between two later credit cycles than between two early ones, for both
men and women. Descriptive statistics could not have suggested such an
effect, as ratios directly depend on the previous amount. In some way,
this positive effect of credit cycle could reflect the increasing trust of the
MFI in its clients over time, especially as the ratio is estimated all other
things being equal, including financial characteristics. This result is dis-
cussed in more detail in the next section. Overall, this means that the
progressive lending policy is indeed applied to both men and women, not
only on average but also on all other things being equal, and that it is
even increasingly progressive, since not only amounts but also growth
rates increase. However, considering only the coefficient of the i nteraction
term between female and credit cycle, it reveals that if ratios increase over
credit cycles for both men and women, this increase is less substantial for
women, as the coefficient of the interaction term is statistically significant
114 M. Bauwin
and negative. As a consequence, this result implies that the progressive

lending policy would be applied differently for men and women, and
more precisely that it would be more favourable to men (Fig. 5.5).
Looking at the results of Model 7, if growth rates also eventually
increase for agricultural loans, it is not the case for second and third
cycles. This does not mean that clients receiving agricultural loans do not
benefit from the progressive lending policy, since amounts in levels
increase over credit cycles as showed previously; however, the progressive
lending policy is not increasingly progressive from a cycle to another as
for non-agricultural loans: the amounts grow more slowly between sec-
ond and third cycles than between first and second ones, but they still
increase. This difference between agricultural and non-agricultural loans
may be due to the specificity of agricultural activity, and to the fact that
financial needs and repayment capacity may be less regular than for cli-
ents running non-agricultural projects. Looking at the interaction effect
with women, it is also less positive (or even more negative for third cycles)
than for men.
With regard to the other characteristics of the loan, the effect of the
duration of the previous loan is also controlled for and is significantly
negative, meaning that clients who receive longer previous loan terms
also receive lower new amounts in terms of ratio. This is logical, as the
duration is taken into account at equal previous amount: this means that
a client with a previous longer loan but with the same amount benefited
from a longer duration in order to enable him or her to repay. As a con-
sequence, his or her capacity for repayment, which is not directly
observed, was most probably lower. Therefore, such a client is more likely
to still show a lower capacity for repayment at the end of the previous
credit period and then to get a lower new amount in terms of ratio.
Regarding clients’ repayment behaviour, in accordance with descriptive
statistics and the analysis of loan amounts in levels, clients who accumu-
late more days of payment being overdue during the previous cycle see
their credit amount growing less rapidly than others, as the number of
days overdue has a significant and negative impact on the ratio in the
two models.
Fig. 5.5 Predictions of growth rates of loan amounts from a cycle to another (in
ratio)
116 M. Bauwin
Concerning financial characteristics, both current and fixed assets in lev-

els, have a significant and positive effect on the ratio: the higher the current
and fixed assets, the higher the growth rate of loan amounts. This is under-
standable as fixed and current assets are indicators of the project size and of
financial needs. We expect that clients with greater projects and/or higher
needs at a given time get greater amounts in terms of ratios. Interestingly,
the evolution of these indicators is not determining: officers rather refer to
current financial information than to their evolution to grant specific
amounts. This is in keeping with procedures, which do not detail how loan
officers would be supposed to consider projects’ history in their application
of the progressive lending policy. Officers just know that they are supposed
to increase loan amounts over credit cycles, and hence certainly consider
immediately available information only: indeed, they use the same sheet to
carry out the financial analysis of projects for first credits cycles as for the
next ones; there is no specific sheet for cycles following the first ones, which
would request officers to consider project evolution.
For non-agricultural loans, higher monthly benefits, a registered activ-
ity, and independent premises logically tend to increase the growth rate,
as these variables also reflect greater projects. For all types of loans, having
some employees also increases the growth rate.
With regard to projects’ other characteristics, the growth rate is signifi-
cantly higher for activities in trade or services compared to agriculture,
while for agricultural loans only, it is higher for culture compared to
breeding. In terms of collateral, offering physical guarantee has a positive
effect on the growth rate compared to unique guarantor, whereas recipro-
cal guarantee still seems considered as the least secure kind of guarantee.
About officer characteristics, being served by a different officer from
the previous loan tends to increase the growth rate of the loan amount in
all cases. This could be explained by the fact that a new officer will do his/
her best to keep the client. Indeed, as seen in the previous chapter, being
served by a new officer decreases the probability of renewing the loan as
clients probably feel a less strong relationship with the MFI when the
officer changes. As a consequence, a new officer is more likely to apply a
more generous progressive lending policy, even in terms of growth rates,
in order to make sure the client will renew the loan at the end of the term.
5.4.3 Discussion of Initial Hypotheses
The main result of this analysis is that the progressive lending policy is
applied not only on average but also on all other things being equal: this
means that at equal characteristics, amounts as well as their growth rates
are more generous between two late cycles than between two early cycles.
In other words, the more loans a client takes out, the higher the loan
amounts, and the faster the increase, therefore the more generous the
progressive lending policy. Referring to contract theory in a situation of
imperfect information, this is in keeping with the idea that the moral
hazard issue between the principal and the agent decreases in the case of
repeated games or contracts: the necessity to acquire or maintain a repu-
tation for agents comes into play (Kreps & Wilson, 1982; Milgrom &
Roberts, 1982). Applied to the credit market, this means that as the lend-
ing relationship between the creditor and the borrower lasts and strength-
ens, creditors get to know their clients better and obtain proves of their
clients’ ability to manage projects and repay.
In the microfinance sector, interest rates are fixed by financial by-
product, and cannot vary from a client to another according to their
credit history only. Consequently, our result may be interpreted this way:
the gain in terms of information for the MFI induced by such a long-
term lending relationship is rather translated into an increasingly gener-
ous progressive lending policy. Another way of seeing it is that the value
of assets and/or collateral required to get a specific amount decreases as
clients take out more loans. Simple descriptive statistics confirm this idea
with regard to collateral, since a client’s own credit history or background
may serve as guarantee from the second credit cycle, the share of such a
guarantee quickly increasing with credit cycles.
Finally, it should be noticed that in the case of Enda, renewing loans is
the only way to keep clients: indeed, Enda is not allowed to offer saving
products since the Tunisian legislation only allows banks to do it; as a
consequence, contrary to other MFIs which may retain their clients with
several kinds of financial products, Enda can only grant other credits to
do so. A generous progressive lending policy is therefore a way to foster
client retention.
118 M. Bauwin
The second main result is that despite the progressive gain in informa-
tion on clients as clients take out more loans, the gap between amounts
granted to men and women persists over time and credit cycles: whereas
the gap concerning first loans could have been accounted for by great
information asymmetry, such a supposition does not hold after several
credit cycles, but women still receive lower amounts all other things being
equal. Three explanations for this result remain plausible and are not
exclusive: first, the stereotypes on women, their projects, and their ability
to fulfil their roles of project manager and housekeeper at the same time
persist over time as well, leading officers to apply slower progressive lend-
ing policy to female applicants, all other things being equal, because they
think that women need or are able to manage smaller loans.
A second explanation is that, since the common belief within Enda is
that women are more loyal than men, it might be possible that officers
make more effort to attempt to keep male clients than female clients: they
would offer higher amounts to men to encourage them to keep resorting
to their services.
The third possible explanation applies to all credit cycles, including
the first one, and may add to the other two. Despite the observable
characteristics of their projects, their better repayment behaviour and
their lower propensity to default, female clients may still convey a bad
signal about their projects: indeed, some officers reported that women
were less likely than men to pound on the table if they were offered less
than 100% of their requests for first loans, or to show their dissatisfac-
tion if they do not get what they expect at the time of renewal.
Consequently, if women also tend to accept lower amounts for a similar
project in such a case of information asymmetry, this could be consid-
ered as a bad signal as theorized by Akerlof (1970) or Leland and Pyle
(1977): their general tendency to accept lower amounts would signal
that they want to invest less in their projects, which hence would imply
that their projects are of lower quality or less profitable. Such a signal,
even if it does not reflect the real quality of their projects, could rein-
force existing stereotypes.
5.5 Conclusion
In the microfinance sector, progressive lending is a very commonly
applied policy as it enables the burden of transaction costs to be reduced
and favours client retention as it is a good dynamic incentive. It also helps
MFIs to avoid large losses by testing clients’ repayment behaviour before
lending high amounts. However, the way such a progressive lending pol-
icy is applied has rarely been analysed and this chapter has attempted to
fill this gap. In particular, as some recent studies have revealed that women
may face less favourable loan conditions than men while being favoured
in terms of access to microcredit, the analysis has focused on the condi-
tions of loan renewals from a gender perspective within the main Tunisian
microfinance institution Enda inter-arabe.
Having taken account of the selection bias since not all clients renew
their loans, and of repayment behaviour, the conditions of loan renewal
were analysed in terms of amounts granted and growth rates of these
amounts, defined as ratios between on-going amount and previous one.
All things being equal, including previous or first loan amount, amounts
and ratios tend to increase over credit cycles, reflecting an increasingly
favourable progressive lending policy over credit cycles. This could be
explained by an increasingly trusting relationship between the MFI and
the clients over cycles as MFIs tend to know their clients better from one
contract to another. Logically, amounts and ratios tend to be lower for
clients who repaid their previous loans with more days overdue, confirm-
ing the importance of trust in the application of the progressive lending
policy. However, the most striking result concerns our item of interest, as
the progressive lending policy appears to be less favourable to women, all
other things being equal. Indeed, if amounts and ratios increase over
cycles, women still receive lower amounts all other things being equal,
and these amounts increase less quickly. As a result, the initial gap between
the first loan amounts granted to men and women found in a previous
work persists over time and even increases: women benefit from a pro-
gressive lending policy just as men do, but it is less favourable.
Consequently, existing inequalities between men and women can only be
reproduced: indeed, women already start from a lower position as they
120 M. Bauwin
tend to run smaller projects in terms of assets and profits, and then tend
to request lower first amounts. If Enda inter-arabe, as other MFIs, actu-
ally favours women in terms of access to credit in attempting to counter-
balance these starting inequalities, they fail to do so entirely, as we still
observe a gap between amounts granted, which increases over time as the
progressive lending policy is less favourable to women, all other things
being equal. As a consequence, if women keep receiving lower amounts,
their projects will evolve less quickly too, and catching up with men in
terms of economic power will become almost impossible.
Notes
1. This research work was supported by UNU-Wider (United Nations
University) and conducted in cooperation with the Tunisian microfinance
institution Enda inter-arabe, which enabled the author to access and use
its microdata and provided all necessary information. Any opinions,
findings, conclusions, or recommendations are those of the authors
and do not necessarily reflect those of the mentioned organizations.
2. The time-use survey, carried out by the Tunisian Ministry of Women
Affairs in 2005 (MAFFEPA, 2005) on men and women in Tunisia and
regularly referred to in reports on gender issues, indicates that women
dedicate more than five hours a day to domestic work against an average
of 39 minutes for men.
3. Here attrition does not correspond to data collection issues but to an
actual phenomenon.
References
Agier, I., & Szafarz, A. (2013). Microfinance and Gender: Is There a Glass
Ceiling on Loan Size? World Development, 42, 165–181. https://doi.
org/10.1016/j.worlddev.2012.06.016
Akerlof, G. A. (1970). The Market for “Lemons”: Quality Uncertainty and the
Market Mechanism. The Quarterly Journal of Economics, 84(3), 488–500.
https://doi.org/10.2307/1879431
Armendáriz, B., & Morduch, J. (2010). The Economics of Microfinance (2nd
ed.). Cambridge, MA: MIT Press.
Banerjee, A., Duflo, E., Glennerster, R., & Kinnan, C. (2009). The Miracle of
Microfinance? Evidence from a Randomized Evaluation. Working Paper. IFMR
Research, Centre for Micro Finance. Consulté à l’adresse https://pdfs.seman-
ticscholar.org/689e/4c2e75a1ae6c2e8f2f8683e6e991a6d31901.pdf
Banerjee, A., Duflo, E., Glennerster, R., & Kinnan, C. (2015). The Miracle of
Microfinance? Evidence from a Randomized Evaluation †. American Economic
Journal: Applied Economics, 7(1), 22–53.
Campbell, T. S., & Kracaw, W. A. (1980). Information Production, Market
Signalling, and the Theory of Financial Intermediation. The Journal of
Finance, 35(4), 863–882. https://doi.org/10.2307/2327206
Chliova, M., Brinckmann, J., & Rosenbusch, N. (2015). Is Microcredit a
Blessing for the Poor? A Meta-Analysis Examining Development Outcomes
and Contextual Considerations. Journal of Business Venturing, 30(3),
467–487. https://doi.org/10.1016/j.jbusvent.2014.10.003
Diamond, D. W. (1991). Monitoring and Reputation: The Choice Between
Bank Loans and Directly Placed Debt. Journal of Political Economy, 99(4),
689–721. https://doi.org/10.1086/261775
Epstein, M. J., & Yuthas, K. (2013). Rural Microfinance and Client Retention:
Evidence from Malawi. Journal of Developmental Entrepreneurship, 18(01),
1350006. https://doi.org/10.1142/S1084946713500064
Kreps, D. M., & Wilson, R. (1982). Reputation and Imperfect Information.
Journal of Economic Theory, 27(2), 253–279. https://doi.
org/10.1016/0022-0531(82)90030-8
Leland, H. E., & Pyle, D. H. (1977). Informational Asymmetries, Financial
Structure, and Financial Intermediation. The Journal of Finance, 32(2),
371–387. https://doi.org/10.2307/2326770
Milgrom, P., & Roberts, J. (1982). Predation, Reputation, and Entry
Deterrence. Journal of Economic Theory, 27(2), 280–312. https://doi.
org/10.1016/0022-0531(82)90031-X
Ministère des Affaires de la Femme, de la Famille, des Enfants et des Personnes
âgées (MAFFEPA). (2005). Time-Use Survey 2005. In Profil Genre Tunisie
2014 (Agence Française de Développement).
OECD. (2007). Effect of Microfinance Operations on Poor Rural Households and
the Status of Women. Special Evaluation Study No. SST: REG 2007-19.
Pearlman, S. (2014). Dropouts, Defaulters, and Continuing Borrowers: Client
Exit from Microfinance: Dropouts, Defaulters, and Continuing Borrowers.
The Developing Economies, 52(4), 301–321. https://doi.org/10.1111/
deve.12055
122 M. Bauwin
Semykina, A., & Wooldridge, J. M. (2010). Estimating Panel Data Models in
the Presence of Endogeneity and Selection. Journal of Econometrics, 157(2),
375–380. https://doi.org/10.1016/j.jeconom.2010.03.039
Weber, O., & Ahmad, A. (2014). Empowerment Through Microfinance: The
Relation Between Loan Cycle and Level of Empowerment. World Development,
62, 75–87. https://doi.org/10.1016/j.worlddev.2014.05.012
Wooldridge, J. M. (1995). Selection Corrections for Panel Data Models Under
Conditional Mean Independence Assumptions. Journal of Econometrics,
68(1), 115–132. https://doi.org/10.1016/0304-4076(94)01645-G
6
What Happens When Microfinance
Programmes Are Withdrawn?
A Conceptual Framework for Analysing
Causal Effects
Nargiza Alimukhamedova
6.1 Introduction and Motivation1

Microfinance is defined as the provision of a broad spectrum of financial
services including microloans, savings, and other banking products to
low-income households in support of their entrepreneurship activities.
First introduced in the 1970s, microfinance has a noble intention of
helping poor people to escape poverty and creating employment oppor-
tunities. For the past five decades, microfinance has been hailed as an
important policy instrument to promote access to finance and reduce
poverty. With global recognition, the microfinance model has become
N. Alimukhamedova (*)
CERGE-EI, Charles University and the Czech Academy of Sciences,
Prague, Czech Republic
University of Economics in Prague, Prague, Czech Republic
Westminster International University, Tashkent, Uzbekistan
e-mail: nargiza@cerge-ei.cz

124 N. Alimukhamedova
popular and many countries have adopted it in various forms, including,

classical Grameen-type group liability, small and medium-sized finance,
individual, and so on. In some areas, microfinance has demonstrated
rapid expansion and sustainable growth, for example in Bangladesh,
India, Latin America, and other countries in South Asia. Unfortunately,
however, in some developing countries the microfinance model did not
succeed, as demonstrated by the failure of certain microfinance institu-
tions (MFIs), for example the liquidity shock in the microfinance sector
in Nigeria in 2005, the severe downturn of the microfinance sector dur-
ing the global financial crisis in Bosnia and Herzegovina in 2009, the
Karnataka (2009) and Andhra Pradesh (2010) crises in India, and public
protests against microcredit loan enforcement in Nicaragua in 2011
(Marconi & Mosley, 2005; Microfinance Focus, 2011; Mujkovic, 2010).
As microfinance is being promoted rapidly across the globe and extends
deeply within nations, the probability of other crises arising is potentially
high. Therefore, measuring the consequences and impact of closing the
microfinance sector in different settings represents an important policy
agenda at the international level.
With the global expansion of microfinance, most academic literature
has been attempting to measure its impact on households and business
development indicators. Most of the impact studies have focused on esti-
mating the effects of introducing new or expanded branches of microfi-
nance institutions and various microfinance lending features. Randomised
Control Trial (RCT) has become the most reliable experimental method
of estimating this impact. Most RCTs find a quite modest effect of micro-
finance loans on business and households (see e.g. Bauchet, Marshall,
Starita, Thomas, & Yalouris, 2011; Duvendack et al., 2011; Fouillet,
Hudon, Harriss-White, & Copestake, 2013; Guerin, Labie, & Servet,
2015 for a summary of the main findings).
In comparison, little or almost nothing is known about the reverse
effect, that is, what happens when a microfinance institution is closed or
the entire microfinance sector is withdrawn. Undoubtedly, measuring the
consequences of microfinance sector closure is an important policy agenda
because the potential harm from closing MFIs could be far more devastat-
ing than the positive effects of introducing new ones. Banerjee, Breza,
Duflo, and Kinnan (2017) and Breza and Kinnan (2018) p rovide the first
6 What Happens When Microfinance Programmes… 125
empirical evidence on these effects from India, where microfinance insti-

tutions ended their operations. The authors find a heterogeneous effect on
existing entrepreneurs and new start-ups. They also find changes in house-
hold spending activities.
What is the effect of terminating MFIs in other countries? How large
is the microfinance sector and does it create more harm than intended
benefits? How does it affect local communities and infrastructures? Does
it create a long-lasting harmful effect and, if so, on whom? To answer
these challenging questions, we propose a comprehensive framework and
approach for analysing the consequences at different levels in order to
obtain conclusive evidence. The proposed analytical framework is thus
suitable for both academic scholars aimed at establishing a causal effect
using robust econometric methodologies, as well as policy makers respon-
sible for tailoring the microfinance pathway and decision-making. To
highlight the importance of integrating these two sides, failure to find
statistically significant causal estimates of MFI termination could lead to
misleading decision-making. On the other hand, working with only
qualitative and descriptive inferences on measuring the impact of with-
drawals could also be a misguided approach. The proposed conceptual
framework therefore calls for a wider perspective and comprehensive
approach for measuring the impact. To that end, we attempt to reconcile
both sides and suggest a multi-dimensional approach to impact
assessment.
Specifically, the suggested framework is in the form of a “research
onion” in which the impact is assessed at several layers and in consecutive
order. The first layer is macro, that is, the general economic environment
for microfinance operations. At this layer, the particular microfinance
niche in the macroeconomy is analysed and assessed. The perceived mis-
sion of microfinance and its evolution in the particular country is deter-
mined. At the second, the so-called meso layer, that is, the banking and
finance sector and communities, the role and participation of MFIs are
assessed in terms of creating competition, its market share and niche, and
how deeply it is integrated with domestic economies. Finally, at the third
micro layer, the effect of MFI termination is assessed on borrowers and
nearby households. For each of these layers, we also define specific research
objectives and the corresponding methodology for impact assessment.
Another advantage of the developed framework is that it enables us to

measure not only the direct effect on beneficiaries, but also on a general
segment of the population and long-term contributions to institutional
development. Institutional economics and the role of institutions nowa-
days are new and important branches in economics that help to explain
variations in economic growth.
An ideal and unbiased way of impact assessment could be achieved by
designing and implementing RCTs, but it could be very time-consuming
and costly. The second-best solution could be based on benefiting from
natural-type events in which MFIs were closed or terminated their activi-
ties. Therefore, we build the methodological side based on primary sur-
vey and primary data analysis. It should be emphasised that the proposed
conceptual framework for measuring microfinance withdrawal effects
should be not taken as a standardised boilerplate approach, but mainly as
a general guide. Impact assessment is highly specific to the particular
country setting, the size of the microfinance sector, MFI operations, and
other macro-institutional dimensions. At the end of this chapter, we pres-
ent the country case of Uzbekistan as a practical example for the applica-
tion of the developed framework and analysis of the sector
withdrawal effects.
The chapter is further structured as follows: Sect. 6.2 briefly defines the
contribution of this project to the mainstream literature. Section 6.3
defines the multi-dimensional framework and detailed vision of each
impact level. Section 6.4 provides a practical example. The last section
concludes.
6.2 R
elevant Literature and Specific
Contributions
This research contributes to the academic literature in several ways.
First, in the general microfinance literature, this research falls under
“impact assessment” studies. An overview of RCTs and other impact
studies can be found in Bauchet et al. (2011), Duvendack et al. (2011),
Fouillet et al. (2013), Guerin et al. (2015), and Van Rooyen et al. (2012)
for the Sub-Saharan region. In contrast to most of these studies, we aim

to measure the reverse effect of the closure or termination of activities of
microfinance institutions. To the best of our knowledge, there are only
two relevant studies by Banerjee et al. (2017) and Breza and Kinnan
(2018), both of which benefit from a natural termination of MFIs in
India. A further group of researchers have developed the microcredit cri-
sis prevention dashboard to analyse the reasons behind crises in microfi-
nance (Microcredit Crisis Prevention Dashboard, 2016). In our research,
we represent an analytical framework showing the importance of analys-
ing and reflecting on not only the direct beneficiaries of microfinance
programmes, but also a general segment of the population while measur-
ing the closure effects.2 This is important to know as closure could poten-
tially bring more harm compared to opening MFI(s). Harmful effects
could also spread to local communities and the set of institutions, such as
trust and values in societies, which are difficult to restore and rebuild
once damaged. Thus, we go beyond analysing the reasons only and
attempt to measure the consequences.
Second, our research can be placed under the branch of modern “insti-
tutional economics”. Relevant studies include the contribution of finan-
cial sector development (FSD) to long-run economic growth (King &
Levine, 1993; Madsen & Ang, 2016), access to finance for small entre-
preneurship, and the role of high-quality institutions for their successful
operations (Beck, Demirgüç-Kunt, & Maksimovic, 2008). Social trust
also captures informal institutions and relevant studies on the influence
of trust on economic growth (Berggren, Bergh, & Bjørnskov, 2012;
Berggren & Bjørnskov, 2017; Berggren, Daunfeldt, & Hellström, 2016;
Berggren, Elinder, & Jordahl, 2008; Berggren, Jordahl, & Poutvaara,
2017; Bergh & Bjørnskov, 2011; Bjørnskov & Méon, 2015; Camussi,
Mancini, & Tommasino, 2018; Guiso, Sapienza, & Zingales, 2004; Zak
& Knack, 2001).
In particular, most microfinance studies to date have focused on mea-
suring direct or so-called immediate effects, that is, what would happen
to households’ welfare and businesses if they took out a loan from MFIs.
Almost no studies have analysed the long-term impact of microfinance
institutions and what happens when microfinance integrates deeply into
societies. This is argued by the fact that in order to properly implement
microfinance programmes, it is necessary to make significant deep trans-

formations in the country’s financial sector, legal base, and relevant credit
infrastructure. Similar changes could also take place in nearby areas where
MFIs are located. Therefore, when there is a (negative) shock from the
microfinance side, such as a closure, crisis, or other complications, these
could create harmful effects on a whole set of institutions in a chain reac-
tion. The literature, and more importantly, empirical studies establishing
the causal link and quantitatively measuring these effects, is largely miss-
ing. In this regard, our analytical framework may serve as a helpful toolkit
for analysing the long-term effects of microfinance programmes.
Many countries have adopted microfinance with a high hope and
anticipation of achieving positive promised effects. However, little is
known about what happens to microfinance programmes when they are
implemented over long periods and when microfinance is deeply embed-
ded in national economic systems or dominates the financial sector.
According to the market failure hypothesis, microfinance fills the niche
where formal financial institutions are weak (Vanroose & D’Espallier,
2013). Once countries grow and develop, microfinance programmes
should shrink. Therefore, an observed “closure” could be the long-term
result of a country’s growth and “graduation” of microfinance clients. It
could also indicate the consolidation and opening of the market to com-
mercial banks that are more adept and have a wider outreach for serving
clients compared to individual MFIs. Thus, the long-term effects differ
based on the strength of the external socio-economic environment, the
quality and maturity of formal institutions, and how deeply microfinance
programmes are incorporated. For this reason, in our analytical frame-
work we call for a comprehensive evaluation not only of the impact of
closure itself, but also of the long-term impacts.
Third, our research is relevant to studies that analyse the effect of
macro and external factors on microfinance operations. In particular,
Ahlin, Lin, and Maio (2011), Imai, Gaiha, Thapa, and Annim (2012),
Vanroose (2007, 2008), Vanroose and D’Espallier (2013), and Ahlin
and Jiang (2008) explore the impact of various socio-economic factors
on MFI operations, including economic growth, foreign direct invest-
ment, population density, poverty, and financial sector development
indicators. Our specific contribution to this strand of the literature
should provide a better understanding of the mechanisms through

which microfinance interacts with the broader economy. Although this
is a popular issue in microfinance policy debates, the literature on the
aggregate effect of microfinance programmes is rather thin.
Therefore, this research project provides a timely and important contri-
bution to measuring the impact of microfinance and its long-term conse-
quences. Being mainly in the form of an analytical framework, the project
will be followed up with a set of specific studies providing empirical evi-
dence on defined research objectives along a multi-dimensional framework.
6.3 M
ethodology: Analytical Framework
for Measuring Withdrawal Effects
The general vision of the multi-dimensional framework is presented in
Fig. 6.1 in which all layers are in the form of a “research onion” (Saunders,
Lewis, & Thornhill, 2009). The main intuition behind this structure is
that the findings of each layer are interdependent and aimed at providing
a conclusive evaluation of the impact of microfinance programme termi-
nation from different angles.
The specific details of each level’s methodology are as follows.
Fig. 6.1 Summary of the multi-dimensional framework

6.3.1 Level 1. Macro-Level Impact
6.3.1.1 Research Objectives
• To define and measure the size of the microfinance sector as a share of

the general financial sector; characterise its evolution and history.
• To describe the microfinance sector and its niche: types of MFIs, port-
folio, outreach, lending programme of MFIs, customer base, regula-
tory body, supervision, legal base; identification of key microfinance
stakeholders and beneficiaries; comparison with regional peers and
international perspective.
• To describe what happened to the microfinance segment after MFI(s)
termination: clients, personnel of MFIs, other beneficiaries.
• To identify the main reasons for MFI terminations: sector-specific,
external factors.
• To identify whether MFI terminations or sector closures were expected
or unexpected. This is a critical departure point as further impact anal-
ysis could have two diverse paths as defined in Table 6.1.
6.3.1.2 Methodology
For this level, primarily a mixed approach could be suitable. First, a

detailed descriptive analysis based on historical data on microfinance evo-
lution could be used—an ethnographic type description of the MFI
Table 6.1 Understanding the reasons for MFI(s) termination

Rationale Expected closure Unexpected closure
1 Whether commercial banks Harm could be less or Harm could be
could lobby and reveal non-existentas banks greater
that they want to enter and other institutions
the sector, therefore MFI could be more
sector should be closed prepared for this
2 Credit portfolio of Increased before Increased after
commercial banks and closure → respective closure →
deposit implications for respective
assessing harm on implications for
households (Level 3 assessing harm on
impact) households (Level 3
impact)
activities and closure procedures. Second, to understand the true reasons

that lead to MFI terminations, a qualitative approach could be used and,
where applicable, regression analysis for disentangling effects. A sug-
gested list of respondent groups for in-depth interviews may include the
groups as summarised in Table 6.2. This list should be adjusted for the
particular country setting and specific nature of microfinance operations.
Table 6.2 Respondents and objectives for in-depth interviews

No. Group Respondents Objective
1 MFI side Managers, owners, To understand reasons for
shareholders loan closure from the MFI side. To
officers, credit assess the potential
personnel, field officers, perspective of re-opening
other admin staff from the MFI side
2 Commercial Managers, loan officers, To understand whether clients of
banks employees working non-bank MFIs moved to
with clients commercial banks. To assess how
much of the non-bank MF sector
and lending was conceptually
different from commercial
lending. To assess any potential
competition between these
types of institutions
3 Regulators Representatives of the To understand reasons for
central bank, ministry sector closure from the
of finance, credit regulator side. To assess
bureau and other whether the sector could be
regulatory agencies re-opened
4 Potential Households and To assess the potential need for
clients entrepreneurs residing non-bank MFIs among various
close to non-bank MFIs, segments of the population,
the so-called eligible including entrepreneurs,
non-participants ordinary households, pensioners
(for deposits), and so on
5 Other Donors, international To assess whether the non-bank
beneficiaries experts, development MF sector had some other
specialist, mahallas, positive benefits for the
community society and local communities
representatives, and so and other relevant parties
on. Also including
various licence and
collateral-assessing
offices that issue
required documents for
taking out microloans
To analyse in-depth interviews, NVivo software could be used. NVivo

is an advanced qualitative research software that analyses various types of
data, including documents, images, PDFs, audio, video, spreadsheets,
web pages, and social media marketing (SMM) data. NVivo generates
word trees and word clouds, hierarchical visualisations, revisualisations,
cluster analysis, project, concept, and mind maps. Third, it is important
to collect national statistics on regional development data, including pop-
ulation, banking, employment, and others and then reconcile qualitative
and quantitative results to generate conclusive evidence on real reasons
for sector closure. In addition to qualitative research, regression analysis
could be used to determine the dominant factors that led to sector clo-
sure. This should enable us to better understand whether it was mainly
due to reasons specific to the microfinance sector, issues on the regulation
side and on the demand side, or other factors.
6.3.1.3 Expected Results
Expected findings at this macro level should provide a “global picture” of

the microfinance role in the national economy. Thus, the consequences
of the withdrawal of MFIs or the entire sector could be judged accord-
ingly. The findings at this level should provide clear reasons for sector
closure and a description of the closing procedures. This is a missing gap
in the literature given the lack of “closure” events internationally. In addi-
tion, the findings should enable us to generate lessons learnt from the
international perspective. The findings at this level could be further
beneficial for a comprehensive evaluation of the impact of terminating
MFI activities and inform Level 2 and Level 3 impacts.
6.3.2 Level 2. Meso-Level Impact
After determining the reasons for sector closure at the macro level, the
next step is to analyse the consequences on financial sector development
(FSD) and household financial portfolios.
• To determine the consequences of microfinance sector closure for gen-

eral financial sector development and access to finance.
• To define any changes in the market structure and composition,
including the role of other financial sector participants.
• To determine the role of microlending in the household portfolio of
borrowing and savings.
6.3.2.2 Methodology
For this level, mainly a quantitative approach is more suitable using

econometric modelling. First, to determine the consequences of sector
closure for FSD, a theoretical framework on the potential effect of non-
bank MFI termination and hypotheses should be developed. This could
be nested within existing partial and general equilibrium theories devel-
oped by Buera, Kaboski, and Shin (2017). Second, the determinants of
the placement of MFIs in the country should be established. This could
be undertaken using probit and Tobit models for predicting the probabil-
ity of appearance and the number of MFIs in districts. This is mainly
applicable when MFIs have random appearance in regions and uneven
distribution within a country. In probability models, the dependent vari-
able should be the MFI dummy equal one if there is an MFI in the dis-
trict and zero if none. The list of independent variables could consist of
relevant groups of district-level determinants such as socio-demographics,
infrastructure, economic structure of districts and regions, and others
depending on the country context.
To understand the household “financial portfolio”, which is comprised
of borrowings and savings (Collins, Morduch, Rutherford, & Ruthven,
2009), a household-level dataset or customised surveys could be used.
For example, the European Bank for Reconstruction and Development
(EBRD) Life in Transition Survey (LITS) waves and the World Bank
Living Standards Measurement Study (LSMS) surveys may be relied
upon, thereby benefitting from a reliable standardised dataset and coun-
try representative samples. Depending on budget availability, customised
surveys could be initiated. The main objective of this analysis is to classify

households into four broad categories: (1) those who borrow mainly, (2)
those who save mainly, (3) those who borrow and save, and (4) who do
neither. This could be undertaken using the maximum likelihood function
to predict the probability of a household to belong to either of (1), (2), or
(3) categories of borrowing and saving. After estimating the maximum
likelihood function, in the second stage the estimation of the actual
amount of borrowing and savings of households is appropriate. This can
be conducted by estimating the simple regression conditional that a
household is in either of (1), (2), or (3) previously identified categories.
The suggested set of controls includes age, household head, education,
marital status, number of children, education attainment, bargaining
power within the household, main and additional occupation, migrant
family members and remittances, rural location, and other determinants.
6.3.2.3 Expected Results
The findings at this level should help us to understand whether MFI or

sector closure is part of general financial sector decision-making or an
unexpected event. In addition, the findings can help us determine what
consequence closures have for other players, for example whether they
“absorbed” and accommodated the clients of MFIs. These are highly
important questions for both academic research and policy makers. The
results should enable us to determine whether microfinance programmes
are likely to be terminated due to political reasons, MFI operation fail-
ures, issues on the household side, or the development stage of the
financial sector. This remains a highly debated issue in microfinance lit-
erature to which we aim to contribute. Meso-level results should also
provide insights on the consequences of sector closure on regional com-
munities and local infrastructure. We hypothesise that even if MFIs
stopped their activities, households and other beneficiaries have long-
term “learning” and spillover effects as shown by Alimukhamedova, Filer,
and Hanousek (2017). Finally, the identification of a microfinance niche
in the household portfolio should enable us to assess the potential loss
that could be caused after the suspension of access to microlending.
6.3.3 Level 3. Micro-Level Impact
At this final stage, the main objective is to identify those who were hurt
by sector closure and to what extent. This is essentially the core part of
the entire impact evaluation in which the primary unit of analysis is the
household. Impact evaluation results at the previous macro and meso
levels should be helpful in making a final judgment regarding the impact
on borrowers and households.
• To identify the socio-economic profile of households most affected by

sector closure.
• To estimate the extent of the harm.
• To identify the heterogeneous effect of sector closure based on experi-
ence in borrowing, savings, and entrepreneurship activity. Therefore, it
is important to understand and establish empirical results at earlier
levels, that is, linking the nature of sector closure (expected vs. unex-
pected, Level 1) and the household portfolio of borrowing and savings
(Level 2).
6.3.3.2 Methodology
Given the ambitious objective of estimating the causal effects using robust
methodologies, at this level the choice of econometric approach is highly
dependent on data availability and the nature of microfinance programme
terminations. The ideal is when comprehensive country representative
household-level datasets are available before and after microfinance pro-
gramme termination. As a second-best option, impact evaluation could
be benefited by RCTs and then a survey ex-post MFI termination. This is
how Banerjee et al. (2017) and Breza and Kinnan (2018) built their
impact evaluation strategy.
In this research, we present the methodology assuming before- and
after-closure datasets. The impact evaluation is then based on a “quasi”
difference-in-differences (DiD) approach, which is suitable for studying
the differential effect of treatment on a “treatment group” versus “control

group” as conducted in experimental research designs. The term “quasi”
implies the absence of pure panel data tracking households before and
after sector closure.
We consider the following two-stage identification. In the first stage,
we predict the probability of borrowing from all sources, that is, formal
(banks, MFIs, other financial institutions) and informal (i.e. family,
friends, neighbours). Once we define how the probability of borrowing
has changed since sector closure, we attempt to quantify the following
outcomes: (1) What happens with those who did not borrow from any
sources? (2) What are the characteristics of those who did not borrow? (3)
How did the characteristics of those who did not borrow change after the
sector was closed? (4) We then create a profile of those who were harmed
by sector closure; and (5) Identify the proportion of households that were
squeezed out of the market as a result of sector closure.
The main identification strategy is built on the following strategy: the
coefficients from the before-closure survey (defined as t = 1) are used to
predict what group each household would have been in the after-closure
survey (defined as t = 2) if access to MFIs had still been allowed. In other
words, we construct an “artificial” control group as Eq. (6.1). In the first
stage, we estimate the probability of borrowing from MFIs in t = 1 survey
as defined in Eq. (6.1):
P [ Borrowing t =1 = 1xi ] = Φ ( β ′xi ) (6.1)

yi ,t =2 = Borrowing t = 1 + b X i + ui (6.2)
where Xi are individual- and household-level controls such as age, edu-

cation, occupation, business ownership, financial literacy, risk aversion,
and so on as defined before; Borrowingt=1 is a fitted value from Eq. (6.1);
and ui is an error term with zero mean. To acknowledge that the economy
of the country has also changed between two surveys, the dependent vari-
able should be a difference between two variables: Δ Business Indicators
and Δ Consumption Indicators. Table 6.3 outlines DiD estimations.
y = β + β T + β2 S + β3T ⋅ S + ε (6.3)
0 1
Table 6.3 Difference-in-differences estimator of the impact of microfinance sec-

tor closure
Treatment Control group
ystyst group (s = 2) (s = 1) Difference
t = 2 (after closure) y22 y12 y12 − y22
t = 1 (before closure) y21 y11 y11 − y21
Change: y21 − y22 y11 − y12 (y11 − y21) − (y12 − y22)
where T is a dummy variable for t = 2 (after closure) and S is a dummy

variable for s = 2 (treatment group). T·S term is a composite dummy vari-
able capturing the situation in which S = T = 1.
A well-known application of DiD by Card and Krueger (1994) esti-
mated minimum wage comparing employment in the fast food sector in
New Jersey and Pennsylvania. Since borrowing from MFIs is not observed
in the after-sector closure survey, we use fitted values from the before-
closure equation:
y=β  + βX +u y = β  + βX +u β0 = ( y | T = 0, S = 0 )
0 i i 0 i i
β1 = ( y | T = 1, S = 0 ) − ( y | T = 0, S = 0 ) β2 = ( y | T = 0, S = 1) − ( y | T = 0, S = 0 )
β3 = ( y | T = 1, S = 1) − ( y | T = 0, S = 1)  − ( y | T = 1, S = 0 ) − ( y | T = 0, S = 0 )  β3 = ( y11 − y21 ) − ( y12 − y22 )
Treatment effect → Impact of closure (6.4)
Additional robustness checks could be based on dummy regressions,

regression discontinuity design (RDD). Ideally, all different approaches
should provide consistent estimates of causal impacts. For the particular
set of countries, EBRD Life in Transition Survey (LITS) survey rounds
(Life in Transition III (2016), Life in Transition II (2010), and Life in
Transition I (2006)) could be used for robustness checks.
6.4 P
ractical Example: The Case of Non-Bank
MFIs in Uzbekistan
In this section, we provide a practical example for the application of the
suggested analytical framework in measuring the consequences of micro-
finance programme withdrawal. The example is based on a case study of
the closure of nearly the entire non-bank MFIs in Uzbekistan, which

occurred on one day in October 2011. It should be emphasised that
Uzbekistan is not the only country where microfinance encountered dif-
ficulties. The other six major microfinance crises include the Nicaragua
microfinance crisis of 2008, Kolar microfinance crisis of 2009 in India,
Andhra Pradesh microfinance crisis of 2010 in India, Pakistan floods in
2010, liquidity crisis in Nigeria in 2005, and the Bosnia and Herzegovina
microfinance crisis of 2009 summarised in Microfinance Focus (2011).
In contrast to other crises, what is valuable about the Uzbekistan example
is that it enables us to analyse the consequences of entire sector closure as
opposed to a single MFI, which is the case in other countries. Therefore,
we provide the following example from this wider perspective.
6.4.1 Microfinance Sector Evolution in Uzbekistan
Uzbekistan is a lower middle-income country3 located in the heart of the

Central Asia region. With a population of around 33 million, Uzbekistan
is the most populous country in the region. Uzbekistan obtained its inde-
pendence from the former Soviet Union in 1991. Microfinance pro-
grammes were first introduced in the country in 1998 by two pilot
projects with the support of UNDP (United Nations Development
Programme) to smooth the transition process and ensure poverty
eradication. These pilot microfinance programmes were based on a
Grameen-type joint liability group lending model. After the success of
these probing models, microfinance developed rapidly. International
donor-funded microfinance programmes such as Nederlandse Organisatie
voor Internationale Bijstand (NOVIB; Netherlands Organization for
International Development Cooperation), FINCA International, UNDP,
and others spread rapidly. Microfinance gained official recognition, which
was heralded by the adoption of a legal base for microfinance operations,
including the “Law on Credit Unions” in 2002, and the “Law on
Microfinancing” and “Law on Microcredit Organisations” in 2006. As a
result, the number of non-bank microfinance institutions (MFIs)
increased dramatically. During the short period between 2006 and 2011,
the microfinance sector expanded rapidly in both depth and outreach,
with the number of credit unions (CUs) reaching 121 and microcredit
organisations reaching 34 by 2011. The non-bank microfinance sector

served more than 200,000 clients. See Alimukhamedova (2014) for a
more detailed description of the non-bank microfinance sector during its
entire history over 1998–2011.
Three factors contributed to the rapid growth of the sector during the
2007–2011 period before closure: (1) new commercial models of micro-
credit organizations (MCOs) and credit unions (CUs) which boosted the
demand for microcredit among the population, (2) the adoption of the
“State Program of Microfinance Development for 2007–2010” in which
the sector was acknowledged as an important segment of the country’s
financial system and household welfare improvement, and (3) the estab-
lishment of a specialised “Mikrokreditbank” as a leading bank-MFI with
extensive countrywide branches offering individual and group micro-
credit at subsidised interest rates below the market average. Over a rela-
tively short period, the number of CUs increased dramatically, reaching
121 by 2011. The number of MCOs also grew, albeit moderately due to
constraints on deposit attraction and restrictions on external donor sup-
port. Rapid expansion of the non-bank microfinance sector in terms of
depth and outreach spurred its integration into centralised credit bureaus.
Figure 6.2 summarises the entire evolution of the microfinance sector in
the country, illustrating three important periods in market evolution.
Fig. 6.2 Cumulative growth of non-bank MFIs in Uzbekistan, 1998–2011

As of January 2011, before closure, the profile of the microfinance sec-

tor in Uzbekistan was represented by the following financial institutions:
downscaling banks, specialised “Mikrokreditbank”, and non-bank MFIs
such as credit unions and microcredit organisations (Table 6.4). As shown
in Table 6.4, non-bank MFIs had their “own niche” and already served a
significant share of the population.
6.4.2 Sector Closure
In October 2011, the non-bank sector was officially terminated when

credit unions and microcredit organisations were at their peak of opera-
tions and public support. The “closure” effect is defined by the decision
of the Central Bank of Uzbekistan to revoke the licences of all credit
unions (their number constituted 121 before closure) and some of the 34
microcredit organisations in October 2011. Before the official closure in
2011, the non-bank microfinance sector represented by CUs and MCOs
had been playing an important role in the country’s financial sector devel-
opment (Alimukhamedova, 2014). Households and business representa-
tives had been actively taking out microcredit loans that were fostering an
increase in business size and household consumption patterns
(Alimukhamedova et al., 2017).
An intriguing question arises as to why the sector was closed, particu-
larly after almost two decades of tedious efforts to establish it. Preliminary
research and in-depth interviews with the main stakeholders indicate
that there were various reasons that led to the unexpected sector closure.
Some reasons are relevant to microfinance sector specifics and the way it
was adopted from the “international model”, as well as extremely high
interest rate charges by CUs that contradicted their original pro-poor
mission. Other reasons include flaws in the supervision side and the
capacity of the regulator. Another reason could be due to the rudimental
legal base for microfinance. In the extreme case, fraudulent behaviour in
some MFIs may have shattered trust in the entire non-bank microfi-
nance sector.
Table 6.4 Microcredit and microdeposit services in Uzbekistan
Loan Average Monthly
Microfinance Profit Legal No. of No. of portfolio, Average loan No. of deposit, interest rate
providers status status inst. borrowers ‘000 USD balance, USD depositors USD on loans
Specialised Profit Bank 1 51,074 165,001 3231 56,540 1511 1.2%
“Mikrokreditbank”
Downscaling Profit Bank 2 7478 37,409 5003 n/a n/a n/a
commercial banks
Credit unions Profit Non- 121 52,965 121,792 2300 153,063 654 3.7%
bank
Microcredit Profit Non- 34 9574 3853 402 0 0 4.8%
organisations bank
Total 138 121,091 328,055 10,936 209,603 2165
Source: Microfinance Information eXchange (MIX), National Association of Microcredit Organizations and Credit Unions
(NAMOCU), UNDP (2011)
n/a indicates that data is not available
6 What Happens When Microfinance Programmes…
141
6.4.3 Impact Assessment4
Preliminary results from the macro impact (Level 1) based on probit and
truncated Poisson models suggest that non-bank MFIs in Uzbekistan fol-
lowed general economic principles before their activities were terminated.
The sector played an important role in maintaining a particular niche
between traditional commercial banks and informal loan sharks.
Historical changes in the legal framework and other exogenous changes
did not affect the functioning of MFIs based on free market principles.
This is particularly critical in environments with heavy regulation and
state dominance. More than microcredits and deposits, MFIs also created
“learning hubs” and credit-related infrastructure. Therefore, the closure
of the sector was harmful for rural communities. In addition, there was a
potential loss of “microfinance knowledge networks” as geographical
access to MFIs in remote areas was shown to have had a causal effect on
business and consumption outcomes (Alimukhamedova et al., 2017).
More rigorous econometric analysis is in progress, comparing regions and
districts with and without non-bank MFIs.
Preliminary results from the meso impact (Level 2) are based on a cus-
tomised survey of households in six geographic-economic zones of
Uzbekistan. Empirical evidence suggests that both regular households and
entrepreneurs maintain a sophisticated “portfolio” of lending, borrowings,
and also participate in informal savings clubs as a reliable cushion against
various risks. Households are found to have more lending to others com-
pared to borrowing outside the family. Borrowing and lending patterns
differ among entrepreneurs and regular households. Entrepreneurs are
found to re-invest in business and cut down on consumption, have less
savings, and invest more in human capital. Households are found to spend
more on non-durables and participate more in savings clubs. After sector
closure in 2011, most households switched to borrowing from banks
(36%), relatives (34%), friends, and neighbours (26%). Entrepreneurs
mainly switched to bank lending even though it is more costly and requires
more paperwork due to high loan sizes and the terms. Therefore, we con-
clude that when having access to non-bank micro lending, households
were actively re-investing borrowed funds for business and entrepreneur-
ship activities and spending less on non-durables and consumption. The
cut in the supply of microcredit associated with sector closure had a detri-
mental effect on household borrowing and lending patterns. Microcredits
were found to be an important source for start-up and entrepreneurship
activities. They were also important for the learning behaviour of house-
holds and the wise re-allocation of family funds towards more productive
activities, that is, reducing unnecessary spending and re-investing into
durable and financially profitable activities.
Preliminary results from the micro impact (Level 3) are in the process of
estimation due to the complexity of econometric analysis and various
robustness checks.
In the meantime, however, based on preliminary results we can infer
that the consequences of closing the microfinance sector could cause het-
erogeneous effects on business performance and household consump-
tion. More than the direct effect on beneficiaries, microfinance has been
shown to produce an important effect on educating households on effec-
tively running a business enterprise, re-allocation of household funds,
and other important development dimensions. In the long-term, when
microfinance programmes become deeply embedded in regional econo-
mies and societies, the termination of their activities could have detri-
mental effects not only on direct beneficiaries, but via spillover, also on
non-borrowers. Therefore, careful consideration is needed when actively
promoting the expansion of MFIs.
6.5 Conclusion
Measuring the withdrawal effects of microfinance programmes has
become a prominent research agenda, as some countries and MFIs face
various challenges and pressures from increased competition in the mar-
ket. Other challenges include issues on the regulation side, external
shocks, and other reasons. Proper measuring of microfinance termination
effects is important and should be undertaken beyond focusing on house-
holds and end beneficiaries only, as has been the agenda of most canoni-
cal microfinance impact studies. Addressing the gap in the literature, we
develop a multi-dimensional framework on measuring these effects.
It is also important to measure the long-term effects of microfinance

rather than direct effects on borrowers and direct beneficiaries. Statistical
and econometric analysis could be limited in finding negative or detri-
mental effects; there may be other explicit effects on neighbourhoods and
infrastructure that could remain, producing long-lasting effects even if
MFIs terminated their activities. Therefore, it may not be entirely accu-
rate to claim that the withdrawal of MFI causes only harm. Indeed, in
some respects, withdrawal could in fact be beneficial from financial sector
perspectives.
Notes
1. The author is grateful to Ariane Szafarz, Randall K. Filer, Jan Hanousek,
the participants of the fifth European Research Conference
on Microfinance, Portsmouth, United Kingdom, June 12–14, 2017
and the 12th Russian Summer School on Institutional Economics,
Moscow, Russia, June 30–July 6, 2018 for valuable comments and sug-
gestions. The views expressed in this chapter are entirely those of the author
and do not necessarily represent the views of affiliated institutions. All
the remaining issues are the responsibility of the author.
Financial support from the Czech Science Foundation (GA ČR, grant
19-19158S) is greatly acknowledged.
2. In this chapter, we do not present a detailed and extended econometric
analysis of causal effects as this merits separate, stand-alone academic
papers that are in progress.
3. Based on UN and Work Bank statistics as of 2018.
4. In this version, we present mainly generalised results. Detailed estimation
procedures for each level are forthcoming in a separate paper.
References
Ahlin, C., & Jiang, N. (2008). Can Micro-Credit Bring Development. Journal
of Development Economics, 86, 1–21.
Ahlin, C., Lin, J., & Maio, M. (2011). Where Does Microfinance Flourish?
Microfinance Institution Performance in Macroeconomic Context. Journal
of Development Economics, 95, 1–16.
Alimukhamedova, N. (2014). Microfinance Environment in Uzbekistan: Analysis

of Supply and Demand, Institute for East and Southeast European Studies. IOS
Working Project No. 344, Regensburg, Germany: Institute for East and
Southeast European Studies.
Alimukhamedova, N., Filer, R., & Hanousek, J. (2017). The Importance of
Geographic Access for the Impact of Microfinance. Development Policy
Review, 35, 1–13.
Banerjee, A. V., Breza, E., Duflo, E., & Kinnan, C. (2017). Do Credit Constraints
Limit Entrepreneurship? Heterogeneity in the Returns to Microfinance.
Heterogeneity in the Returns to Microfinance (September 1, 2017), Buffett
Institute Global Poverty Research Lab Working Paper,
(17–104).
Bauchet, J., Marshall, C., Starita, L., Thomas, J., & Yalouris, A. (2011,
December). Latest Findings from Randomized Evaluations of Microfinance.
In Access to Finance Forum (Vol. 2). Washington, DC: CGAP.
Beck, T., Demirgüç-Kunt, A., & Maksimovic, V. (2008). Financing Patterns
Around the World: Are Small Firms Different? Journal of Financial Economics,
89(3), 467–487.
Berggren, N., Bergh, A., & Bjørnskov, C. (2012). The Growth Effects of
Institutional Instability. Journal of Institutional Economics, 8(2), 187–224.
https://doi.org/10.1017/S1744137411000488
Berggren, N., & Bjørnskov, C. (2017). The Market-Promoting and Market-
Preserving Role of Social Trust in Reforms of Policies and Institutions.
Southern Economic Journal, 84(1), 3–25.
Berggren, N., Daunfeldt, S., & Hellström, J. (2016). Does Social Trust Speed
Up Reforms? The Case of Central-Bank Independence. Journal of Institutional
Economics, 12(2), 395–415.
Berggren, N., Elinder, M., & Jordahl, H. (2008). Trust and Growth: A Shaky
Relationship. Empirical Economics, 35(2), 251–274.
Berggren, N., Jordahl, H., & Poutvaara, P. (2017). The Right Look: Conservative
Politicians Look Better and Voters Reward It. Journal of Public Economics,
146(February), 79–86.
Bergh, A., & Bjørnskov, C. (2011). Historical Trust Levels Predict the Current
Size of the Welfare State. Kyklos, 64(1), 1–19.
Bjørnskov, C., & Méon, P. (2015). The Productivity of Trust. World Development,
70, 317–331.
Breza, E., & Kinnan, C. (2018). Measuring the Equilibrium Impacts of Credit:
Evidence from the Indian Microfinance Crisis. (No. w24329). National Bureau
of Economic Research.
Buera, F., Kaboski, J., & Shin, Y. (2017). The Macroeconomics of Microfinance.
NBER Working Paper Series, No. 17905.
Camussi, S., Mancini, A., & Tommasino, P. (2018). Does Trust Influence Social
Expenditures? Evidence from Local Governments. Kyklos, 71(1), 59–85.
Card, D., & Krueger, A. (1994). Minimum Wages and Employment: A Case
Study of the Fast-Food Industry in New Jersey and Pennsylvania. American
Economic Review, 84(4), 772–793.
Collins, D., Morduch, J., Rutherford, S., & Ruthven, O. (2009). Portfolios of the
Poor: How the World’s Poor Live on $2 a Day. Princeton University Press.
Duvendack, M., Palmer-Jones, R., Copestake, J., Hooper, L., Loke, Y., & Rao,
N. (2011). What Is the Evidence of the Impact of Microfinance on the Well-
Being of Poor People? London: EPPI-Centre, Social Science Research Unit,
Institute of Education, University of London.
Fouillet, C., Hudon, M., Harriss-White, B., & Copestake, J. (2013). Microfinance
Studies: Introduction and Overview. Oxford Development Studies, 41(sup1),
S1–S16. https://doi.org/10.1080/13600818.2013.790360
Guerin, I., Labie, M., & Servet, J.-M. (Eds.). (2015). The Crisis of Microcredit.
Zed Books.
Guiso, L., Sapienza, P., & Zingales, L. (2004). The Role of Social Capital in
Financial Development. The American Economic Review, 94(3), 526–556.
Imai, K., Gaiha, R., Thapa, G., & Annim, S. (2012). Microfinance and Poverty –
A Macro Perspective. World Development, 40, 1675–1689.
King, R., & Levine, R. (1993). Finance and Growth: Schumpeter Might Be
Right. The Quarterly Journal of Economics, 108(3), 717–737.
Madsen, J., & Ang, J. (2016). Finance-Led Growth in the OECD Since the
Nineteenth Century: How Does Financial Development Transmit to
Growth? Review of Economics and Statistics, 98(3), 552–572.
Marconi, R., & Mosley, P. (2005). Bolivia During the Global Crisis 1998–2004:
Towards a Macroeconomics of Microfinance. Sheffield Economic Research
Paper Series, No. 2057, 1–45.
Microcredit Crisis Prevention Dashboard. (2016). Microfinance in Crisis.
Research and Policy Briefs, No. 9.
Microfinance Information eXchange, National Association of Microcredit
Organizations and Credit Unions, United Nations Development Program.
(2011). Анализ и сравнительные показатели микрофинансирования:
Узбекистан 2010 [Analysis and the Benchmarking of Microfinance:
Uzbekistan 2010].
Microfinance Focus. (2011). 6 Microfinance Crises That the Sector Does Not
Want to Remember. Microfinance Focus, April 22. Retrieved from http://
www.microfinancefocus.com/6-microfinance-crises-sector-does-not-
want-remember
Mujkovic, A. (2010). Microcrediting in Bosnia and Herzegovina: What Went
Wrong? Retrieved from Microfinance Gateway Publications, CGAP: http://
www.microfinancegateway.org/p/site/m//template.rc/1.9.42370
Saunders, M., Lewis, P., & Thornhill, A. (2009). Research Methods for Business
Students. New York: Pearson.
Vanroose, A. (2007). Microfinance and Its Macro-Environment. MicroBanking
Bulletin, No. 14, 22–24.
Vanroose, A. (2008). What Macro Factors Make Microfinance Institutions Reach
Out? Solvay Business School Working Paper No. 08/036, 1–28.
Vanroose, A., & D’Espallier, B. (2013). Do Microfinance Institutions
Accomplish Their Mission? Evidence from the Relationship Between
Traditional Financial Sector Development and Microfinance Institutions’
Outreach and Performance. Applied Economics, 45(15), 1965–1982.
Van Rooyen, C., Stewart, R., & De Wet, T. (2012). The Impact of Microfinance
in Sub-Saharan Africa: A Systematic Review of the Evidence. World
Development, 40(11), 2249–2262.
Zak, P., & Knack, S. (2001). Trust and Growth. Economic Journal, 111(470),
295–321.
7
Good Customer Service Versus
Bad Regulation
Mark Hannam
7.1 Introduction
The failings of microfinance institutions (MFIs) have been widely docu-
mented in the academic literature: there are failings in the treatment of
customers, failings in the development of sustainable business models
and failings to make lasting, measurable impacts on the lives of customers
and their communities. One common response to these failings is to call
for greater regulation of MFIs. For example, Lesley Sherratt (2016)
writes that:
The efforts towards self-regulation of the microfinance industry repre-

sented by the Smart Campaign go nowhere far enough to prevent deeply
unethical practices in the industry or the recurrence of repayment crises.
The ethical issues microfinance raises are too engrained in its model … for
M. Hannam (*)
Institute of Philosophy at the School of Advanced Study,
University of London, London, UK
e-mail: mark.hannam@sas.ac.uk

150 M. Hannam
tinkering around the edges to succeed. The protection to clients offered by

the Smart Campaign may emerge from the best of intentions. There is no
need to question that. What is needed, however, is independent regula-
tion. (p. 181)
And, Philip Mader (2015) writes that:
One perfectly logical option for policy makers would be to discontinue

their direct and indirect support … for microfinance systems, until the day
their beneficence can be proven. Another logical possibility would be to
regulate the sector in such a way as to ensure that the interests of investors
and MFIs cannot supersede those of their clients. Yet the sector remains
deeply averse to such regulation with teeth—such as interest rate caps or
loan usage restrictions—and will likely strongly resist it by employing
arguments that emphasize poor people’s right to choose between different
credit sources and modalities which would be curtailed by regula-
tion. (p. 204)
We should be careful what we wish for: more regulation does not nec-
essarily mean better regulation, nor does it guarantee better outcomes for
customers. Regulation is not a free good. It comes with a price-tag and
ultimately this price is paid by the customer.
Economists have described some of the generic effects of regulation on
economic activity (Stigler, 1971). These include the raising of barriers to
market entry, which embeds the power of existing market participants at
the expense of potential new entrants to the market; this, in turn, reduces
the pressure for innovation in product development and customer service
within regulated market-places. Further, compliance with regulation
increases producer costs, which are frequently passed on to customers,
either in the form of higher prices or in the form of reduced service
quality. In addition, as James Broughel (2017) observes, “… regulations
are unique in that—unlike taxes and spending—the vast majority of
their effects are not captured in government budgets. In this sense, the
effects of regulation are invisible to the public” (p. 2).
The regulation of MFIs, and other sub-prime credit providers, has
tended to focus on product features and product price. Many regulatory
regimes for MFIs and sub-prime credit impose price caps on the amount
7 Good Customer Service Versus Bad Regulation 151
that can be charged for credit. Since compliance with regulation tends to
increase costs for the provider of services, where these additional costs
cannot be passed on to customers through increased prices, they can only
be mitigated by reductions in the resources available for service delivery.
Credit providers are incentivised by the regulatory system to lower the
costs, and therefore the quality, of their customer service.
The pressure to reduce the costs of delivering services to customers has
been a major driving force in the move away from relationship banking
to transactional finance. Just as supermarkets and airlines have sought to
improve their margins by the provision of “no frills” services, so too
financial institutions have redesigned their products and services to
become simpler, standardised and automated. This trend towards low-
cost, low-touch, low-margin products has been supported by the increased
use of information technology to deliver services. The growing use of data
analysis to predict customer buying behaviour has helped providers deter-
mine which products to promote to which customers without the need
to consult the customer directly.
The automation of risk assessments and product delivery works well
for customers who want standard products and who need only generic
advice. However, low-cost, transactional finance does not work well for
customers who need non-standard products delivered with a high-touch
customer service. To the extent that regulation pushes MFIs and sub-
prime credit providers to become more like mainstream banks, with a
greater reliance on automation and standardised transactional relation-
ships, it also pushes them to reduce the richness of their customer rela-
tionships. This, in turn, will tend to worsen both repayment rates and
customers’ credit histories. Bad regulation tends to embed financial
exclusion more deeply in the financial system.
7.2 Part I
There is a well-developed body of research that examines the role of
relationships in the mainstream banking sector. The starting insight is
that borrowers face a trade-off when they seek to raise funds for their
business. Aside from the fact that borrowing directly from the capital
152 M. Hannam
markets—through the issuance of bonds or commercial paper—tends

to be possible primarily for larger firms, there is also an information
asymmetry problem. New companies, without a track record of gener-
ating income and managing themselves well, are less attractive to “arms-
length” lenders. Debt markets work best for larger firms and for firms
with a track record of competence. Newer and smaller firms are there-
fore more dependent on direct lending, often from banks.
Arnoud Boot (2000) defines relationship banking as the provision of
financial services by a financial intermediary, which invests in obtaining
customer-specific information, often proprietary in nature, and which
evaluates the profitability of these investments through multiple interac-
tions with the same customer over time and/or across products. According
to Boot, the two principal benefits to the credit provider of building
relationships in this way are, first, the acquisition of re-usable informa-
tion about the customer and, second, the ability to introduce special con-
tractual features, specific to individual borrowers. Over time, these
benefits lead to safer lending for the credit provider and, correspondingly,
lower costs for the borrower.
Banks are willing to lend to new borrowers if they believe that they will
be able to build up a store of private information about the business that
will allow them to exert some influence or control over the borrower,
making it less likely that they will default on the loan and more likely that
they will pay for additional banking services in the future (Rajan, 1992).
Greater competition between banks, or a shift from a relationship to a
transactional banking model, threatens to reduce the amount and quality
of private information that is acquired by the bank, and to reduce the
value of the relationship to the bank. This, in turn, makes it less likely
that credit will be available to smaller and newly established businesses.
Research on the US small business market supports these insights.
Relationships between banks and businesses are valuable to both, and
they tend to increase the total amount of credit available. The private
information that is shared in a close relationship with a borrower creates
greater assurance for the lender that resources are being well managed
and appropriately allocated. Banks use this private information to adjust
contractual terms for loan agreements, which would not be possible in a
purely transactional relationship. Lenders pass on some of the lower costs
to the borrower in the form of lower prices, or reduced demands for col-
lateral (Berger & Udell, 1995). In addition, the value of private informa-
tion is such that more credit is made available both at the early stage and
at subsequent stages of the relationship (Petersen & Rajan, 1994).
There is some evidence that the greater the market power of a lender
(i.e., the less constrained it is by competitive pressures) the more likely it
is to lend to firms with lower credit quality and at lower prices. From the
borrowers’ point of view, therefore, location in a highly concentrated
credit market is an advantage in terms of access to credit. This is particu-
larly true for new firms, which tend to borrow more when there are fewer
lenders active in the market (Petersen & Rajan, 1995). The explanation
for these findings appears to be that lenders smooth their interest rates
over time, charging less than the risk profile of the borrower might imply
at the start of the relationship, but charging more than the risk profile
implies at later stages of the relationship. While the absolute cost of credit
falls as firms become better established, the risk-adjusted margin secured
by the borrower rises, to compensate for the higher risks taken at the
outset of the relationship.
Some of these features of the small business lending markets have also
been found in the personal credit markets. In a study that examined con-
sumer credit card lending in the US, Agarwal and co-authors (Agarwal,
Chomsisengphet, Liu, & Souleles, 2009) found that there were “… sub-
stantial potential benefits from relationship lending, through lower
default risk, lower attrition and increased utilisation” (p. 5). The authors
observed that information gathered through the wider banking relation-
ship (e.g., the use of savings products) helped the lender better to predict
a borrower’s changing demand for credit.
In mainstream finance, it is evident that the quality of the relationship
between borrower and lender is crucial to the efficiency of the flow of credit
to new businesses and to individuals. Where the quality of these relation-
ships deteriorates, the ability of the lender and borrower to share the sur-
pluses generated by business activity declines; consequently, the supply of
credit reduces and the cost rises. Do these same factors apply to MFIs?
There have been two recent attempts to measure the importance of
relationship banking within microfinance, one drawing on research
from Ecuador and the other from Bangladesh. Using data from
154 M. Hannam
ProCredit Ecuador, Jan Schrader (2009) argues that competition

between lenders leads to higher risk-taking, that is, banks providing
loans to customers with a higher probability of repayment problems.
Banks that view lending as a purely transactional business tend to have
lower interest rates, because their costs are lower, but correspondingly
they have weaker relationships with their borrowers and less private
information about the businesses to which they lend. As a result, where
there is strong competition between transactional lenders, a higher
number of risky loans are made and this tends to contribute to over-
indebtedness among borrowers.
Schrader also noted that when borrowers are struggling to pay back
multiple loans, they will give priority to loans from relationship lenders,
at the expense of loan repayments to transactional lenders. In some cases,
borrowers took out additional short-term loans from transactional lend-
ers to ensure that they could repay their relationship loans. Borrowers
appeared to place a higher value on relationship lending, despite it being
a more expensive form of borrowing. The borrowers saw value in the
relationship over and above the cost of credit.
In their study of lending in Bangladesh, Chakravarty and Shahriar
(2010) found evidence that potential borrowers who had developed a
relationship with an MFI, and who already used non-credit services such
as savings accounts, were more likely to apply for, and were more likely to
be approved for, a loan. The lenders were more willing to extend credit to
borrowers with whom they had established a credit history and with
whom they had a wider (i.e., multi-product) financial relationship. Their
study also suggested that, whereas in developed markets, such as the US,
it is the larger lenders that are most likely to place reliance on formal
financial records and objective customer data, in Bangladesh the larger
MFIs were more likely to build relationships with their customers and to
rely on relationship information in their lending policies. Small MFIs in
Bangladesh, unlike small banks in the US, did not have the resources to
build relationships with customers, nor to develop bespoke lend-
ing products.
In summary, there is some evidence that the valuable role of private
information in building relationships between lenders and borrowers
operates in the MFI sector just as it does in mainstream banking. There
is also evidence that intense competition among lenders can lead to looser
lending policies, greater risk-taking and, in due course, greater over-
indebtedness. The value of relationship banking, both for the business
seeking credit and for the lender providing credit, seems to be consistent
across the mainstream and the MFI sectors.
7.3 Part II
Thus far I have considered the value of relationship banking from the
point of view of the lender, who gathers private information to reduce the
risk of loan delinquencies and to cover the initial cost of loans to new
businesses; and from the point of view of the borrower, who benefits
from greater access to credit and, initially, lower rates than might be
expected from their relationship lender. There is an additional, very
important aspect to relationship banking, which likewise appears consis-
tent across the mainstream and the MFI sector. This is the benefit to the
borrower of regular contact with and support from their lender. The per-
sonal aspect of relationship banking turns out to be highly beneficial to
borrowers, especially for small businesses and for individuals who experi-
ence financial exclusion.
Antoinette Schoar (2012) conducted a study of the small business cli-
ents of the Industrial Credit and Investment Corporation of India
(ICICI), the largest commercial bank in India, between July 2007 and
April 2009. The evidence she gathered suggested that personal interac-
tions between the borrower and the lender reduced the propensity of
borrowers to default on their loans, which is of benefit to both borrowers
and lenders.
Schoar divided her sample of around 1300 clients into four groups. All
clients were provided with a 12-month overdraft facility with an average
size of just over $20,000. The control group received a standard SMS text
message from the bank when a payment was due; this was ICICI’s nor-
mal level of interaction with its customers for this overdraft product.
Another quarter of the customers received a telephone call, from a ran-
domly chosen bank employee, when a payment was due, but with no
additional support or services; a third quarter received a telephone call
156 M. Hannam
twice a month, from one of three or four different employees, which

addressed any service issues or customer questions, as well as reminding
customers about forthcoming payments; the final quarter of customers
received the high-touch fortnightly calls, but in this case always from the
same employee.
There was a clear difference in the payment histories between the cus-
tomers who received the high-touch, fortnightly calls and those who did
not; and there was a modest, but not significant, difference between those
customers who received calls always from the same bank employee and
those who received calls from one of a small group of bank employees.
Schoar remarks that “these results suggest that the interactions between
the relationship manager and client not only prevented late payments
overall, but in particular reduced the likelihood of repeated late payment
spells” (p. 14). Investment by the bank in building relationships with
borrowers was rewarded by greater loyalty and reliability, which improved
the risk prolife of the bank’s loan book. While these improvements were
clearly of benefit to the bank, they were also of significant benefit to the
borrowers, who become eligible for improved access to credit and other
banking products.
In a very different study, Lisa Servon (2017) went to work for a cheque
casher—RiteCheck—in South Bronx, New York, and then for a payday
lender—Check Centre—in Oakland, California. Working as a teller and
a loan collector for sub-prime financial companies in the US, Servon
observed first-hand the customers who make extensive use of products
that are outside the mainstream and which are often regarded as exploit-
ative by academics and policy makers. Her study is an ethnography rather
than a randomised experiment; but for that reason, it contains a wealth
of information and insight into the motivations and behaviours of the
customers of sub-prime credit providers.
Of the many lessons that Servon draws from her work experiences, two
stand out. First, many of the customers of sub-prime lenders chose them
in preference to mainstream financial service providers because they offer
a more flexible and transparent service. It makes sense for some people to
cash a cheque at only 98.5% of the face value and get the cash immedi-
ately, rather than wait three days for the cheque to clear through the
banking system. Likewise, it makes sense for some people to pay a $2 flat
charge to take $10 in cash from an electronic benefits card, if they have
$15 in their account and the local ATM (automatic teller machine) only
dispenses $20 notes. Although these transaction costs appear to be high
(by comparison with mainstream products), they allow easy and immedi-
ate access to cash which is what the financially excluded need most; and
these costs are transparent, which allows customers better control over
their limited financial resources.
Second, many sub-prime service providers offer a high-touch per-
sonal service. They build relationships with their customers and take
time to understand their financial needs and habits. Most mainstream
financial companies provided financial services this way fifty years ago,
but today they prefer to automate as many of their services as possible,
from automatic teller machines (ATMs), to online banking, to algo-
rithmic reviews of credit profiles. The only “advice” that mainstream
banks provide is when they try to cross-sell or up-sell a new product;
or, if the customer is wealthy enough to be served by a private bank,
when they provide helpful advice on tax management. Sub-prime pro-
viders, by contrast, tend to hire staff who look and sound like their
customers, and who understand the challenges faced by people who
are financially excluded. When RiteCheck hires its staff, the key quali-
ties it looks for are friendliness, patience and a customer orienta-
tion (p. 22).
These research studies by Schoar and Servon help to demonstrate
that while relationship banking might be under threat in the main-
stream sector, as automation drives a shift to transactional banking ser-
vices, in many parts of the sub-prime sector relationships still matter a
great deal. The private information gathered through relationships
improves the quality of the services provided by the lender and improves
the outcomes for the borrower. Relationship banking increases the
chances that lending will be responsible, and that access to credit will
be beneficial. For those who are concerned about the potential damage
of irresponsible lending and the over-supply of credit, relationship
banking appears to be a valuable mechanism with which to mitigate
these risks.
158 M. Hannam
7.4 Part III

Let us now return to our starting point, the demand for greater regula-
tion of MFIs. One of the general consequences of regulation is that bar-
riers are raised to new market entrants, which tends to allow established
market participants to build and maintain a dominant position in the
market-place. While this might not be a spur to innovation and competi-
tion, it does help preserve some of the benefits of relationship banking.
There is evidence that where competition pressures on lenders were not
great, it was more likely that borrowers were able to access credit and that
the costs of credit were lower. In other words, a well-regulated market is
likely to provide a more conducive environment for relationship banking
to flourish and this is likely to increase the amount of credit available to
new borrowers, especially in the small business sector. This outcome is
also likely in the MFI sector, since a well-regulated environment will sup-
port relationship lending that reduces the likelihood of the mispricing of
lending risks. In this respect, then, regulation can be beneficial to both
lenders and borrowers.
There are, however, good reasons to be alert to the potential of greater
regulation to lead to maleficent outcomes for poorer customers. James
Broughel (2017) describes the benefits of product innovation: when an
innovation occurs, consumers pay the same price for a better product,
and the innovative producer captures market share from competitors,
increasing profits. However, the second effect is time-limited, because
other producers will eventually copy the innovative product to win back
market share. As Broughel points out, this process has the consequence
that the benefits of innovation are permanent for consumers but only
temporary for producers (p. 85). It is important to recognise that the
converse is also true. When ill-conceived regulation inhibits the ability of
producers to innovate, customers continue to pay the existing price for a
product that is no longer optimal, and producers are unable to gather
additional profits through increased market share, due to innovation. In
this case, the opportunity costs of regulation for businesses are tempo-
rary, but the higher prices and lower quality of products available to con-
sumers are permanent regulatory costs.
When the specific form of the regulation involves price caps on lend-
ing products, which is common in the MFI and sub-prime sectors, then
the impact of regulation is even less favourable for customers. By capping
the prices of products, the regulator reduces the ability of the service pro-
vider to manage their income. Unable to increase income—to cover,
inter alia, rising costs of regulation—by raising prices, the service pro-
vider is forced instead to reduce the costs of delivering products to the
customer, resulting in more standardisation, more automation and fewer
resources invested in customer service functions.
It is often the poorest customers—those who suffer most from finan-
cial exclusion—who are most affected by increased regulatory costs
(Zywicki, Manne, & Morris, 2014). They include smaller businesses with
poor credit histories, who struggle to find banks willing to fund them,
because the loan sizes they need are often considered sub-scale for the
bank’s profitability. Antoinette Schoar’s (2012) evidence from India sug-
gested that these customers would be more likely to repay on time and to
remain loyal customers of the bank, if they received regular calls from a
customer service employee. However, this is not the standard service
offered by the ICICI (or other banks, for that matter) because it is expen-
sive. Automated SMS messages are the standard service, despite Schoar’s
evidence that they are ineffective for both lender and borrowers.
Likewise, Lisa Servon’s (2017) account emphasises the importance of
the relationship between the teller, the loan collectors and the customers.
She notes that while the headline costs of cheque cashing and payday
loans might look high, they are costs that customers are willing to accept
for the speed, transparency and quality of the service they receive. If ser-
vice providers are not able to offer such a high-touch customer service, it
is the customers who suffer. They are already, in many cases, excluded
from the mainstream: price caps further reduce their options by making
it harder for sub-prime lenders to offer their customers much valued,
highly personal customer service.
Good customer service that seeks to build a strong and transparent
relationship between borrower and lender is likely to lead to more credit
being available to borrowers, and is likely to lead to better repayment rates
by borrowers. Good regulation should therefore seek to improve customer
service, not to curtail it. When academics—and others—demand more
160 M. Hannam
regulation, they should therefore be very careful to specify what sort of

regulation they want. If the aim is to encourage lenders to provide better
services to customers, and to ensure that the needs of financially excluded
people are met, then a regulatory regime that requires evidence of good
service should be specified. That would almost certainly mean abandoning
regulation by price, which would be a very good outcome for those who
suffer financial exclusion.
References
Agarwal, S., Chomsisengphet, S., Liu, C., & Souleles, N. (2009). Benefits of
Relationship Banking: Evidence from Consumer Credit Markets.
Retrieved October 6, 2017, from https://papers.ssrn.com/sol3/papers.
cfm?abstract_id=1440334
Berger, A., & Udell, G. (1995). Relationship Lending and Lines of Credit in
Small Firm Finance. Journal of Business, 68(3), 351–381.
Boot, A. (2000). Relationship Banking: What Do We Know? Journal of Financial
Intermediation, 9, 7–25.
Broughel, J. (2017). Regulation and Economic Growth. Arlington, VA: Mercatus
Centre, George Mason University.
Chakravarty, S., & Shahriar, A. Z. (2010). Relationship Lending in Microcredit:
Evidence from Bangladesh. Retrieved April 28, 2017, from https://www.frbat-
lanta.org/-/media/documents/news/conferences/2010/international-devel-
opment/paper/Shariar.pdf
Mader, P. (2015). The Political Economy of Microfinance. Basingstoke: Palgrave
Macmillan.
Petersen, M., & Rajan, R. (1994). The Benefits of Lending Relationships.
Journal of Finance, 49(1), 3–37.
Petersen, M., & Rajan, R. (1995). The Effect of Credit Market Competition on
Lending Relationships. Quarterly Journal of Economics, 110(2), 407–443.
Rajan, R. (1992). Insiders and Outsiders. Journal of Finance, 47(4), 1367–1400.
Schoar, A. (2012). The Personal Side of Relationship Banking. Retrieved April 28,
2017, from https://www.povertyactionlab.org/sites/default/files/publica-
tions/804%20ASchoarPaper_WB.pdf
Schrader, J. (2009). The Competition Between Relationship-Based Microfinance
and Transaction Lending. Retrieved April 28, 2017, from https://papers.ssrn.
com/sol3/papers.cfm?abstract_id=1522422
Servon, L. (2017). The Unbanking of America. New York: Houghton

Mifflin Harcourt.
Sherratt, L. (2016). Can Microfinance Work? Oxford: Oxford University Press.
Stigler, G. (1971). The Theory of Economic Regulation. The Bell Journal of
Economics and Management Science, 2(1), 3–21.
Zywicki, T., Manne, G., & Morris, J. (2014). Price Controls on Payment Card
Interchange Fees: The US Experience. Retrieved May 3, 2017, from https://
papers.ssrn.com/sol3/papers.cfm?abstract_id=2446080
8
Reframing Microfinance and Financial
Inclusion Research: Case Studies
and Synthesis
James Copestake
8.1 Introduction
Growing reference to ‘financial service providers’ and ‘financial inclusion’
has usefully broadened debate away from a narrow focus on the impact of
specialist microfinance institutions (MFIs) on poverty. To illustrate the
difference, the 2017 Global Findex Database suggests that 3.8 billion peo-
ple (or 69% of adults) have access to a bank account, whereas 1.7 billion
do not (World Bank, 2018). In contrast, Balkenhol (2018, p. 78) suggests
that specialist MFIs reach approximately 200 million clients, with the
highest level of country “penetration” (defined as the ratio of MFI bor-
rower accounts to the active poor population) being Bangladesh with
25%. Of course, these figures merely beg further questions about how
access and use of different kinds of financial services affect the reproduc-
tion of poverty and inequality among users and non-users. The “ fin-tech
J. Copestake (*)
Department of Social and Policy Sciences, University of Bath, Bath, UK
e-mail: j.g.copestake@bath.ac.uk

164 J. Copestake
revolution” in potential access may be radically changing the scope of this

central microfinance problem, but it certainly has not eliminated it.
In reviewing the scope of microfinance research it is important to
acknowledge the fluidity of the ideas competing to influence policy and
practice, which can be characterised as a struggle between overlapping
conglomerations of ideas and people that Abbott (2000) refers to as
‘knowledge lineages’. Which ideas survive and which thrive depends on
their internal coherence, historical timeliness, congruence with more or
less powerful interests, and resilience in the face of public scrutiny, criti-
cism and debate. The mimetic success of the term financial inclusion, for
example, can be attributed to its congruence with both a neo-liberal
belief in the benign effects of properly regulated market integration, and
with its technical tractability as a Specific, Measurable, Assignable,
Realistic and Timebound (SMART) objective for the international devel-
opment industry. In contrast, the history of microfinance (narrowly
defined) teaches us that lack of a strong body of independent evidence of
positive impact is unlikely to constrain growth of a new area of aid activ-
ity, whereas strong case studies of negative or contested impact (such as
the Andhra Pradesh crisis of 2014) can do so.
One aspect of the evolution of new knowledge lineages is a tendency
for them to grow outwards from one or more relatively narrow starting
points. In the case of microfinance and financial inclusion, it is useful to
distinguish between the product or service categories, presented by col-
umns in Table 8.1. Much of the early interest centred on highly specific
and relatively simple retail services, particularly group-based microcredit
(incorporating both the Grameen solidarity model and the Finca village
banking model). This focus can be attributed to the importance attached
(a) to lack of capital as a feature of poverty, and (b) to achievement of
economies of scale as a precondition for achieving financial self-sufficiency
on the supply side. In tension with this was more limited recognition of
the potential complementarities to be gained from bundling together
complementary financial services (credit plus) and non-financial services
(credit plus plus), represented by columns to the right of Table 8.1. For
example, a relatively distinct body of literature emerged that was con-
cerned with the link between financial services—particularly micro-
insurance—and health. The scope of microfinance research only later
expanded to cover payment and transfer services (incorporating digital
8 Reframing Microfinance and Financial Inclusion Research… 165
Table 8.1 The scope of microfinance and financial inclusion research

Retail: Individual, household and/or business
Wholesale services
Ancillary
Relatively Insurance goods
Corporate, Payment simple and more and
public and and savings complex services
development transfer and credit financial (e.g. BDS,
finance services services services health)
Technical
Organisational
and economic Inner core
Social and
political
Cultural and
historical
information and communication technology), as well as wholesale

mechanisms for financing financial service providers, including impact
investing, as indicated by the columns to the left of Table 8.1.
The rows in Table 8.1 expand the framework by distinguishing
between different theoretical and disciplinary framing of research. At
its inception, the core of the field was concerned primarily with eco-
nomic and social aspects of organising saving and credit contracts.
Research then expanded to address ever more complex back-office
organisation and technology underpinning financial service provision
(top two rows), often with a strong economics and management empha-
sis on how technical innovation could contribute to lowering transac-
tions costs. At the same time, microfinance and financial inclusion
attracted researchers who brought with them a broader range of social
science perspectives, covering the full gamut from social anthropology
to international relations via management and development studies. As
a counterpoint to the dominant discourse upholding the benefits of
financial inclusion, a growing body of these academic researchers came
to regard microfinance more critically, as a form of financialisation or
integration of clients into a market-dominated global financial system
on adverse terms (e.g. Bateman, 2010).
Copestake et al. (2016) suggest that disciplinary framing does help in
part to explain the tendency towards polarisation or even schism in
166 J. Copestake
microfinance research after the turn of the Millennium, although it is

important to note that early advocates of a neo-liberal approach to micro-
finance also did a good line in political economy as well as technical
analysis (e.g. Adams & von Pischke, 1992). More importantly, they warn
against the way generalised argument for and against microfinance can
crowd out an understanding of microfinance not as a universal phenom-
enon but as various practices and experiences in specific geographical and
historical contexts. This is illustrated by Table 8.2.
Table 8.2 Mainstream, alternative and plural disciplinary perspectives on

microfinance
Mainstream Critical Plural
Historical Widespread poverty Reproduction of Context specific
dimension traps, excluding most poverty through and path
(how the poor people from adverse dependent
world is) regulated services incorporation into patterns of
delivered through markets, financial
competitive markets particularly through eco-system
over-indebtedness evolution
Normative Universal access to A stronger voice and Services that
dimension financial services better terms for combine
(how it offered through financial service innovation with
should competitive and users, partly through adaptation to
be) regulated markets as solidarity economy diverse local
a means to promote alternatives to contexts and
self-employment, commercial suppliers social norms
liquidity smoothing
and risk management
Gap closing Accelerated innovation Civil society and state Largely
or action and financial mobilisation to endogenous
dimension integration through promote fairer and adaptation of
(how it selective state more democratic services to meet
could be) activism (smart services locally specific
subsidies) and needs and
enabling regulation opportunities
Leading CGAP; mainstream UNRISD; civil society Locally rooted
advocates financial institutions; activists; heterodox practitioners;
neoclassical social scientists social
economists anthropologists
Source: Adapted from Copestake et al. (2016). The mainstream model might also be
referred to by its critics as a neo-liberal model. It also subsumes within it ‘World
Bank Consensus’ and ‘NGO Minimalist’ views on credit subsidy policy set out in
Copestake (1996). In contrast, Gravesteijn (2014) distinguishes between ‘business
first’, ‘development first’ and ‘social enterprise’ models of microfinance
Another important lesson from this review of the framing of microfi-

nance research is to avoid hasty generalisation about whether financial
inclusion is likely to be positive or negative in its overall welfare effects in
different contexts. This is reflected in the long-standing academic litera-
ture on the concept of social inclusion, which depicted it not only as the
opposite of social exclusion, but also in tension with the concept of
adverse incorporation (Copestake, 2007; Hickey & du Toit, 2007; Wood,
2003). Likewise, it is useful to locate the evolution of competing knowl-
edge lineages about financial services as expanding from microfinance
(narrowly defined) onto a larger canvas of financial integration that
entails clients being included, excluded and adversely incorporated.
Having set the scene for thinking about microfinance current research,
this chapter next reflects, first, on how the specific contributions to this
book are situated within it, and then how they collectively contribute to
the field.
8.2 L ocating This Book in the Context

of Wider Research into Microfinance
8.2.1 The Chapter Case Studies
Bastiaensen, Romero and Huybrechs’ case study of subsidised credit for

cattle and coffee farmers in Nicaragua reviews two ‘green microfinance
plus’ programmes, combining subsidised credit with technical assistance,
from a broad interdisciplinary perspective (see Chap. 2). It illustrates how
the wider potential of credit can often only be fully appraised through a
place-specific systems perspective linking financial access to real out-
comes, in this case encompassing both livelihood effects and their envi-
ronmental effects. The entry point for Namé and Lebailly’s case study
from western Mali are self-managed village saving and credit banks, also
viewed from a broad interdisciplinary perspective (see Chap. 3). They
also situate financial integration on a broader canvas, emphasising how
these institutions facilitate the wider circulation of money entering the
168 J. Copestake
region as remittances, thereby enabling labour migration to have positive

as well as adverse effects on rural livelihood resilience. Both these chapters
highlight the need for detailed and contextualised description of financial
services if their full role and effects are to be accurately understood. This
point is further emphasised by Afonso and Khan’s case study of interest-
free or qard hasan credit provided by Akhuwat Islamic Finance in Pakistan
(Chap. 4). They highlight how conventional analytical concepts for
assessing microcredit services—demand, sustainability, replicability and
impact—can only be fully understood when viewed from the more
explicitly ethical perspective of Islamic Finance. Only through an appre-
ciation of its moral and cultural dimensions is it possible to understand
its resilience and scale.
Bauwin’s empirical assessment of Enda’s microcredit programme in
Tunisia takes us back to the core of microfinance, focusing on progressive
lending, or the tendency for repeat loans to get larger (Chap. 5). This is a
frequently observed, but rarely analysed feature of minimalist micro-
credit: important both to borrower discipline (by establishing dynamic
incentives to repay) and to lenders’ financial self-sufficiency (by facilitat-
ing portfolio growth). But the relatively technical and quantitative analy-
sis nevertheless generates a powerful and culturally deeply embedded
challenge: controlling for other borrower characteristics, why does pro-
gressive lending dynamically discriminate against women? The authors
hint at a powerful feedback loop here, with slower growth in the value of
loans sanctioned to women reflecting their more limited business oppor-
tunities and debt capacity, perhaps, but also helping to reproduce their
disadvantage. This may be bad for the women, and also acts as a break on
the MFI’s portfolio growth potential.
Alimukhamedova (Chap. 6) reminds researchers of the benefits to
being opportunistic. In 2011, the non-bank sector of Uzbekistan, com-
prising credit unions and MFIs, was officially closed down. While hardly
a random event or natural experiment the shock nevertheless created an
opportunity to assess the role the sector played, with early evidence sug-
gesting clients were switch to borrowing from a combination of banks,
relatives, friends and neighbours. But the case is also a reminder of the
complexity and subtlety of repercussions of losing access to credit: on
social networks and opportunities for learning as well more immediate
effects on liquidity, cost of banking, consumption and investment.

Hence, in a roundabout way the chapter reinforces the potential value of
having a diverse range of impact evaluation methods to draw upon.
Hannam’s review of customer service and regulation (Chap. 7) is also a
reminder of the value of widening the research lens beyond individual
clients and providers, focusing not only on the meso or market level, but
also on the incentive effects, costs and benefits of heavier or lighter touch
regulation.
8.2.2 Viewing the Case Studies Together
It is instructive to juxtapose the chapters of this book (a non-representative

sample of nearly 90 presentations made at the Portsmouth conference)
with growing investment in systematic synthesis of evidence about the
effects of microfinance and financial inclusion. Such research has grown
rapidly in recent years in response to demand from governments and
donors for the ‘distillation’ of evidence to inform public investments in
different areas of development. This can be viewed as a logical and laud-
able attempt to realise the full benefit of discrete case studies by extracting
from them the most credible and useful generalisations possible. It can
also be viewed more sceptically, as an attempt to render the world in all
its glorious diversity and political complexity to a level of simplicity and
technical measurability that highly centralised bureaucracies can cope
with. Doubtless the truth lies somewhere between. Here we look at two
recent examples.
Starting with the demand side, Duvendack and Mader (2018) report
on a highly ambitious systematic ‘review of reviews’ covering 11 meta-
studies of impact assessments, covering credit (mostly), savings, insur-
ance and payment services. These were selected on the basis of
methodological quality out of 32 meta-studies they identified. Areas of
impact were divided into economic, gender, social, behavioural and
macro-structural. They conclude that financial services have diverse
impacts on the lives of poor clients, more positive than negative (particu-
larly for savings promotion) but generally small. The evidence available is
also generally short-term (12 months), limited in scope and weak in what
170 J. Copestake
it reveals about heterogeneity of outcomes as well as the effects of com-

bining different services.
Turning to the supply side, Hermes and Hudon (2018) systematically
selected and reviewed 169 mostly quantitative papers in peer-reviewed
English language journals concerned with understanding the financial
and/or social performance of microfinance institutions. Their goal was to
explore scope for generalising about how performance was affected by
measures of efficiency, the influence of different sources of funding, indi-
cators of organisational governance and market conditions. Cross-country
analysis generally suggests that the financial performance of larger and
older MFIs is stronger, and that social performance tends to be stronger
for non-profit compared to for-profit MFIs. Many other variables have
ambiguous and/or country-specific effects. They also conclude that evi-
dence of a trade-off between social and financial performance is mixed,
with most studies confirming it does exist, but some also finding a syner-
gistic relationship or no link. A major problem with the evidence is reli-
ance for measuring social performance on rather weak proxy indicators of
outreach, which in turn may reveal very little about client level impact.
Overall, both reviews highlight the difficulty of generalising about
how better access to financial services (even in the form of relatively
intensive microfinance interventions) affects the welfare of clients. As a
field of development practice, the financial sector is simply too complex
and diverse for the importance and magnitude—direction even—of
causal mechanisms explored in one country and/or sector context to be
useful on their own in planning activities in another (Cartwright &
Hardie, 2012). Direct evidence on wider economic and societal effects is
even scarcer.
This is bad news for policy makers seeking replicable models for reduc-
ing poverty. It also helps to explain why contradictory views about the
impact of microfinance and financial inclusion persist. However, if this
places a premium on local understanding as the basis for context-specific
policy responses then this is not necessarily such a bad thing. A more
realistic basis for policy may then be a body of theory informed by evi-
dence, but not mechanically or inductively derived from it. Such theory
may also be strengthened by building on unruly case studies such as
presented in this book (unruly in the sense of being less amenable to sys-
tematic review).
The idea that middle range theory can play an important role in policy
formulation is not new; it is integral to realist evaluation and synthesis,
for example (see Pawson, 2013). It does not need to be mutually consis-
tent or comprehensive to be useful: indeed, in confronting complexity
messiness may be a strength. Let us consider the following propositions:
• Credit can contribute towards nudging farmers to reduce carbon emis-

sions and adapt to climate change, but rarely on its own.
• Remittance inflows create opportunities for strengthening local finan-
cial markets.
• Progressive lending is important to sustainable provision of financial
services, but can reinforce social discrimination.
• Religiously informed and ethically based microcredit can be very effec-
tive, but hard to replicate.
• Most borrowers can access a range of financial services, and can adjust
to the radical curtailment of some of them, but not without adverse
long-run consequences.
• Trade-offs between consumer protection and the cost of borrowing
need to be assessed closely.
These are all lessons that can be drawn from this book. They do not
constitute a joined-up theory of best practice in microcredit, their useful-
ness will vary in different contexts and there will always be scope for add-
ing to them and refining them. But absorbing such insights and reflecting
selectively on their relevance to new contexts is likely to improve policy
and practice. That perhaps is the most we can hope for.
8.3 Conclusions
This concluding chapter has reflected on the scope of research into micro-
finance as a knowledge lineage that has evolved and expanded into a
wider field of research into financial inclusion, exclusion and adverse
incorporation. I have emphasised the complexity of the task, as reflected
172 J. Copestake
in the multiple disciplinary perspectives that can be brought to bear upon

it and the scope for divergence in people’s views over its potential and
actual impact. I then briefly reviewed the empirical case studies in this
book, contrasting their unruly heterogeneity with a couple of more recent
exercises in systematic evidence synthesis. But I conclude that messy
research that generates disparate and disconnected findings still has an
important contribution to make in generating insights and middle range
theories to inform a devolved and pluralistic approach to financial sector
development.
References
Abbott, A. (2000). Chaos of the Disciplines. Chicago: University of Chicago Press.
Adams, D., & von Pischke, J. (1992). Microenterprise Credit Programs: ‘deja
vu’. World Development, 20, 1463–1470.
Balkenhol, B. (2018). Microfinance: Research, Debates, Policy. Abingdon and
New York: Routledge Focus.
Bateman, M. (2010). Why Doesn’t Microfinance Work? The Destructive Rise of
Local Neoliberalism. London and New York: Zed Books.
Cartwright, N., & Hardie, J. (2012). Evidence-Based Policy: A Practical Guide to
Doing It Better. Oxford: Oxford University Press.
Copestake, J. (1996). NGO-Donor Collaboration and the New Policy Agenda –
The Case of Subsidised Credit. Public Administration and Development,
16(1), 21–30.
Copestake, J. (2007). Poverty and Exclusion, Resources and Relationships:
Theorising the Link Between Social and Economic Development. In
I. Gough & J. A. McGregor (Eds.), Well-Being in Developing Countries: From
Theory to Research. Cambridge: Cambridge University Press.
Copestake, J., Johnson, S., Garcia Cabello, M., Goodwin-Groen, R., Gravesteijn,
R., Humberstone, J., et al. (2016). Towards a Plural History of Microfinance.
Canadian Journal of Development Studies, 37(3), 279–297.
Duvendack, M., & Mader, P. (2018). Impact of Financial Inclusion in Low and
Middle Income Countries: A Systematic Review of Reviews. New Delhi; London;
Washington, DC: International Initiative for Impact Evaluation.
Gravesteijn, R. (2014). Models of Social Enterprise? Microfinance Organisations as

Promoters of Decent Work in Central Asia. Doctoral Thesis (Unpublished),
University of Bath, UK.
Hermes, N., & Hudon, M. (2018). Determinants of the Performance of
Microfinance Institutions: A Systematic Review. Journal of Economic Surveys.
https://doi.org/10.1111/joes.12290
Hickey, S., & du Toit, S. (2007). Adverse Incorporation, Social Exclusion and
Chronic Poverty. Chronic Poverty Research Consortium, Working Paper 81.
Retrieved October 31, 2018, from http://www.chronicpoverty.org/uploads/
publication_files/WP81_Hickey_duToit.pdf
Pawson, R. (2013). The Science of Evaluation: A Realist Manifesto. London: Sage.
Wood, G. (2003). Staying Secure, Staying Poor: The “Faustian Bargain”. World
Development, 31(3), 455–471.
World Bank. (2018). Global Findex Report 2018. Retrieved from https://glo-
balfindex.worldbank.org/
Index1
A Assets, 2, 47, 64, 65, 72, 82n12, 93,

Access, 5, 9, 16, 18, 20, 21, 31, 40, 94, 96, 99, 101, 109, 116,
42, 45, 46, 51, 62, 69, 78, 117, 120
82n13, 85, 86, 89, 92, 94, Attrition, 99, 120n3, 153
119, 120, 120n1, 123, 127,
133, 134, 136, 142, 153,
155–158, 163, 164, 167, B
168, 170, 171 Banana Skins, 5, 11n1
Accountability, 67, 80 Benevolent loans, 62, 65, 79, 80
Adaptive governance, 15, 32 Bottom line, 3, 11, 17
Agricultural frontier, 26
Agricultural loans, 93, 101, 107,
109, 114, 116 C
Agriculture, 18, 50–52, 91, 116 Cattle raising, 19, 26, 31
Akhuwat Islamic Microfinance Causal effects, 123–144
(AIM), 62, 68 Client protection, 150
Asset and Liability Management Client retention, 86, 88–90, 117, 119
(ALM), 2–3 Climate change, 9, 13–32, 171
Note: Page numbers followed by ‘n’ refer to notes.

1

in Microfinance and Financial Inclusion, https://doi.org/10.1007/978-3-030-05261-4
176 Index
Coffee production, 19–21, 23, 24, Evidence, 5, 9, 10, 39–52, 75, 77,
31, 33n3 125, 129, 132, 153–155,
Commercialization, 3, 10 158–160, 164, 168–170, 172
Competition, 125, 131, 143, 152, Extensive livestock pathway, 27, 28
154, 155, 158
Complexity, 8, 14, 15, 32, 41, 143,
168, 169, 171 F
Consultative Group to Assist the Financial development, 40
Poor (CGAP), 3 Financial exclusion, 79, 151, 155,
Credit, 5–7, 9, 14, 17, 19, 21–23, 159, 160
25, 28, 30, 42, 43, 45, 68, 69, Financial inclusion, 1, 3, 4, 7, 10,
72, 74, 85–92, 94, 97–120, 40, 42, 62, 67, 71, 85, 86,
128, 130, 131, 138–140, 163–172
150–159, 165, 167–169, 171 Financialisation, 165
Customer service, 149–160, 169 Financial self-sufficiency, 164, 168
Financial services providers, 157
Financial sustainability, 25, 26, 80
D Fondo de Desarrollo Local (FDL), 14,
Decision-making, 15, 25, 125, 134 18–26, 28, 33n5
Deforestation, 19, 28, 31 Formal financial services, 62
Demand, 6, 46, 62, 69, 71–72, 79, Funding sources, 72, 73
132, 139, 153, 158, 159,
168, 169
Disciplinary framing, 165 G
Discrimination, 9, 10, 168, 171 Gender, 10, 67, 78, 86, 87, 92–94, 98,
Diversification, 3–4, 23, 25, 28 101, 107, 109, 110, 119, 169
Diversification pathway, 23, 25 Gharar, 63, 66
Donors, 3, 4, 7, 11, 52, 67, 75, 79, Global Findex Survey, 62, 74, 81n3
131, 139, 169 Governance, 4–5, 7, 20, 26, 30, 170
Dropouts, 90, 99 Green microfinance, 3, 10, 17, 19
Group lending, 6, 69, 138
E
Ecological outcomes, 16 H
Econometric analysis, 97, 142–144, History, 4, 42, 107, 109, 116, 117,
144n2 130, 139, 154, 164
Environmental impact, 10, 17, 30, 31 Household survey, 77, 80
Ethics, 4 Human rights, 5
Index 177
I Loan renewals, 86, 90, 96, 98, 119

Ijarah, 64–66 Local communities, 30, 40, 67, 125,
Impact, 4, 10, 17, 20, 22–24, 30, 127, 131
31, 32n2, 40, 42, 44, 50, 52, Longitudinal studies, 77
62, 66, 71, 72, 75–79, 86, Low-income entrepreneurs, 68, 69
88–90, 110, 114, 124–130,
132–137, 142–143, 149, 159,
163, 164, 168–170, 172 M
Impact investing, 165 Macro, meso and micro levels, 135
Informal finance, 43–45, 71, 72 Mali, 39–52, 167
Information asymmetries, 86, Markets, 3, 7, 8, 10, 16, 18, 26, 41,
118, 152 43, 45, 52, 68, 71, 91, 117,
Interest-based microfinance, 62, 68, 125, 128, 133, 136, 139, 142,
75, 79, 80 143, 150, 152–154, 158, 164,
Interest-free microcredit, 62 169–171
Interest rate caps, 150 Market withdrawal, 125, 128, 133,
Interest rates, 5, 19, 67, 69, 89, 92, 136, 139, 142, 143
107, 117, 139, 140, 153, 154 Microcredit, 3, 5, 11, 21, 45, 46, 62,
Islamic microfinance institutions 68–70, 72–74, 76, 77, 79, 80,
(IMFIs), 61–67, 75, 80, 81n5 82n13, 86, 88, 119, 124, 127,
Islamic microfinance models, 62 138–143, 164, 168, 171
Islamic teachings, 63–66, 70, 73, Microfinance, 1–11, 13–32, 40,
77, 79 42–46, 61–80, 85–120,
123–144, 149, 150, 153,
163–172
K Microfinance crisis, 138
Knowledge, 14–30, 63, 90, 127, 142 Microfinance institutions (MFIs), 2,
Knowledge lineages, 164, 167, 171 13, 68, 86, 124, 163
Micro-insurance, 164
Mission drift, 3, 86
L Mobile banking, 4, 10
Lending markets, 153 Monoculture (coffee) production,
Liquidity smoothing, 166 20, 21
Livelihood strategies, 31 Mudaraba, 64
Loan amounts, 86–91, 97–101, 106, Multi-dimensional approach, 125
107, 109–117, 119 Murabaha, 64–66, 72, 75
Loan officers, 6, 10, 70, 87, 88, 92, Musharaka, 64
99, 116, 131 Muslim countries, 79
178 Index
N R
Neo-liberal, 164, 166 Randomised control trials (RCTs), 4,
Nicaragua, 13–32, 124, 138, 167 72, 76, 124
Non-bank microfinance sector, Regulation, 4, 8, 132, 142, 143,
139, 140 149–160, 169
Non-governmental organisation Regulation costs, 158, 159
(NGO), 6–7, 28, 32, 52, 68 Relationship banking/lending,
Normative, 166 85–120, 151–155,
157, 158
Remittances, 10, 39–52, 134,
O 168, 171
Obaidullah, M., 62, 70 Repayment behaviour, 87, 88, 100,
Ohio-State University, 7 107, 114, 118, 119
Operating expenses, 26, 73 Replicability, 62, 71, 74–75, 168
Outreach, 67, 71, 75, 79, 128, 130, Research, 8, 18, 20, 42, 75–78, 80,
138, 139, 170 90, 120n1, 125–130,
Over-indebtedness, 5, 46, 86, 132–136, 140, 143, 151–153,
154, 155 157, 163–172
Resilience, 17, 32, 41, 164, 168
Resources, 14, 16, 17, 26, 32n1,
P 44–46, 67, 68, 74, 92, 94, 96,
Pakistan, 49, 61–80, 138, 168 151, 152, 154, 157, 159
Pathway to change, 15 Riba, 63, 66
Payment services, 43, 164, 169 Risk, 2, 5, 9, 11, 11n1, 15, 17, 18,
Plural, 15, 166 20, 22, 25, 28, 32, 39, 41, 46,
Portfolio, 2, 6, 11, 18, 20, 23, 25, 50, 51, 53n15, 54n20, 63,
68, 130, 132–135, 142, 168 100, 109, 136, 142, 151, 153,
Portsmouth, 7, 80n2, 144n1, 169 155–158
Private information, 152–155, 157 Rural microfinance, 13, 17–18
Producers, 16, 21–25, 27–30, 32,
150, 158
Progressive lending, 6, 86–91, 93, S
97, 106, 107, 109, 113, 114, Sadaqah, 66, 73, 74, 81n5
116–120, 168, 171 Saqib, A., 62, 63, 69, 73–75, 80
Savings, 3, 5, 8, 41–45, 48, 49, 51,
66, 72, 85, 117, 123,
Q 133–135, 142, 153, 154, 165,
Qard hasan, 61–80, 81n10, 168 167, 169
Qualitative/quantitative approach, Savings and Credit Cooperative
76, 77, 80, 90, 131–133, 168 Societies (SACCOs), 6–7
Index 179
Sector closure, 124, 130–138, 140, Transparency, 20, 63, 67, 80, 156,
142, 143 157, 159
Selection bias, 77, 87, 94, 99, 100, 119 Tunisia, 86, 87, 120n2, 168
Self-regulation, 149
Service quality, 150
Shari’ah, 61, 64, 66, 70, 72 U
Smart Campaign, 149 Use (of financial services), 5, 163
Socially embedded, 16 Uzbekistan, 126, 137–143, 168
Social performance, 76, 88, 170
Stakeholders, 4, 14, 23, 26, 28,
30–32, 43, 130, 140 V
Statistical discrimination, 86, 87 Village banking, 9, 48, 164
Sub-prime providers, 157 Vulnerability, 17–20, 22, 41
Subsidies, 2, 7, 8, 10, 41, 139, 167
Supervision, 4, 8, 130, 140
Sustainability, 15, 30, 62, W
71–74, 168 Women’s empowerment, 10, 89
Systematic synthesis, 169 Wooldridge’s procedure, 99, 101
T Y
Technical assistance, 14, 17–19, Yunus, Muhammad, 5, 10–11
21–26, 28–30, 167
Technology, 4, 11, 16, 25, 151, 165
Time inconsistency, 8–9 Z
Transactional finance, 151 Zero default, 11

Emerging: Challenges

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Emerging: Challenges

Uploaded by

Copyright:

Available Formats

Emerging

ISBN 978-3-030-05260-7 ISBN 978-3-030-05261-4 (eBook)

Cover illustration: Witthaya Prasongsin / Getty Images

Microfinance is an unusual discipline, with one foot in the world of busi-

chapters into the context of broader research in the sector—thus setting

European Microfinance Platform Daniel Rozas

This publication results from a selection of outputs from researchers who

2 Addressing Climate Change with Microfinance Plus:

3 Remittances: Loan Funds for a Rural Economy? Evidence

4 Islamic Microfinance: Exploring the Experience of

5 Relationship Lending in Microfinance: Do Women Benefit

6 What Happens When Microfinance Programmes Are

7 Good Customer Service Versus Bad Regulation149

8 Reframing Microfinance and Financial Inclusion Research:

Nargiza Alimukhamedova received a PhD in economics in 2014 from

Mathilde Bauwin graduated with a PhD in development economics in

in order to analyse a particular Green Microfinance project which com-

Philippe Lebailly has a PhD in agricultural sciences. He is Professor of

environmental degradation. Her research is mainly based on the experi-

and Voluntary Sector Quarterly, Oxford Economic Papers, Quarterly Review

Fig. 5.1 Evolution of financial indicators, in value and ratio (medians) 95

Table 4.1 AIM main activity indicators 67

The fifth European Research Conference on Microfinance was organized

© The Author(s) 2019 1

Likewise, we are happy to celebrate the fifth edition of the European

so-called minimalist approach—or combining them with financial liter-

requirements, as well as the due diligence of debt providers. Corporate

Originally created in the nineteenth century in Germany and Britain,

pline financial impulses. While discipline can be a useful commitment

This chapter reflects on the role of rural microfinance in the context of

J. Bastiaensen (*) • M. Romero

© The Author(s) 2019 13

the importance of path-dependent local embeddedness. We illustrate the

uncertainty in planning for sustainability in the current era” (Wise et al.,

of dynamic interactions of global and local actors with different ideas

The relatively recent debate about the role of microfinance in climate

processes which contribute decisively to both. The individualistic models

Plus approach promoting silvopastoral and agroforestal production

2.2.1 Climate Change Adaptation in Coffee

2.2.1.1 Climate Change and Coffee Production

In the Nicaraguan coffee producing regions, climate change is mainly

below which a large portion of coffee production is currently located.

­ athway’, which focuses on producing high volumes of undifferentiated

model in the context of the CAMBio project. This project aimed at

clear positive impact for FDL in terms of repayment rates as well as a

technicians impose standardised one-size-fits-all prescriptions to all types of

FDL and Nitlapan-UCA face strong short-term barriers to change the

2.2.2.1 The Environmental Pressure of Cattle Production

The problem of cattle raising in Nicaraguan agricultural frontiers is more

Polvorosa, 2013) are part of the institutional bricolage that engendered a

2.2.2.2 Microfinance Plus to Transform Extensive Cattle

More land-intensive, silvopastoral cattle systems offer an alternative to

brought farmers to re-exploit their ‘conservation areas’ and to sell trees

altogether? Some of the emerging literature about the role of microfi-

due to complex interactions with mediating processes (Garikipati,

Mendoza, R. (2013). El Café en tiempos de la roya. Envío Ed. 372.

M. Namé (*) • P. Lebailly

© The Author(s) 2019 39

Indeed, the reduction of these costs allows better quantification of the

inevitably lead to a decrease in vulnerabilities in the rural world. The solu-

3.1.1 A Customized Financial Product: Savings

populations. The constraints of immediate profitability and management

Labie (1999) provides the following definition:

European Microfinance Platform Daniel Rozas

2 Addressing Climate Change with Microfinance Plus:

3 Remittances: Loan Funds for a Rural Economy? Evidence

4 Islamic Microfinance: Exploring the Experience of

5 Relationship Lending in Microfinance: Do Women Benefit

6 What Happens When Microfinance Programmes Are

7 Good Customer Service Versus Bad Regulation149

8 Reframing Microfinance and Financial Inclusion Research:

Fig. 5.1 Evolution of financial indicators, in value and ratio (medians) 95

Table 4.1 AIM main activity indicators 67

2.2.1 Climate Change Adaptation in Coffee

2.2.1.1 Climate Change and Coffee Production

athway’, which focuses on producing high volumes of undifferentiated

2.2.2.1 The Environmental Pressure of Cattle Production

2.2.2.2 Microfinance Plus to Transform Extensive Cattle

3.1.1 A Customized Financial Product: Savings

3.3 A Readiness to Save

3.3.2.1 My Family, My Community, My Homecoming!24

4.2 An Introduction to Islamic Microfinance

4.3 AIM and the Qard Hasan Model

4.3.2 The Qard Hasan Microcredit Model

5.2.2 Progressive Lending in the Literature

5.3.2 Descriptive Statistics

5.3.2.1 Evolution of Clients’ Projects

5.3.2.2 Evolution of Clients’ Financial Situations

5.3.2.3 Loan Amounts over Credit Cycles

5.4 Methods and Results

The matrix of independent variables consists of constant socio-