Scaling and institutionalization within agricultural innovation systems: the case of cocoa farmer field schools in Cameroon

Abstract

The farmer field school (FFS) concept has been widely adopted, and such schools have the reputation of strengthening farmers’ capacity to innovate. Although their impact has been studied widely, what is involved in their scaling and in their becoming an integral part of agricultural innovation systems has been studied much less. In the case of the Sustainable Tree Crops Programme in Cameroon, we investigate how a public–private partnership (PPP) did not lead to satisfactory widespread scaling in the cocoa innovation system. We build a detailed understanding of the key dimensions and dynamics involved and the wider lessons that might be learned regarding complex scaling processes in the context of agricultural innovation systems. Original interview data and document analysis inform the case study. A specific analytical approach was used to structure the broad-based exploration of the qualitative dataset. We conclude that scaling and institutionalization outcomes were impeded by: the lack of an adaptive approach to scaling the FFS curriculum, limited investments and genuine buy-in by extension actors, a failure to adapt the management approach between the pilot and the scaling phase, and the lack of strategic competencies to guide the process. Our findings support suggestions from recent literature that pilots need to be translated and adapted in light of specific contextual and institutional conditions, rather than approached as a linear rolling-out process. These findings are relevant for the further spread of similar approaches commonly involved in multi-stakeholder scaling processes such as innovation platforms.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tags20
International Journal of Agricultural Sustainability
ISSN: 1473-5903 (Print) 1747-762X (Online) Journal homepage: http://www.tandfonline.com/loi/tags20
Scaling and institutionalization within agricultural
innovation systems: the case of cocoa farmer field
schools in Cameroon
Sander Muilerman, Seerp Wigboldus & Cees Leeuwis
To cite this article: Sander Muilerman, Seerp Wigboldus & Cees Leeuwis (2018):
Scaling and institutionalization within agricultural innovation systems: the case of cocoa
farmer field schools in Cameroon, International Journal of Agricultural Sustainability, DOI:
10.1080/14735903.2018.1440469
To link to this article: https://doi.org/10.1080/14735903.2018.1440469
© 2018 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group
Published online: 23 Feb 2018.
Submit your article to this journal
Article views: 91
View related articles
View Crossmark data

Scaling and institutionalization within agricultural innovation systems:
the case of cocoa farmer field schools in Cameroon
Sander Muilerman
a
, Seerp Wigboldus
a,b
and Cees Leeuwis
a
a
Knowledge, Technology and Innovation group, Wageningen University, Wageningen, The Netherlands;
b
Wageningen Centre for
Development Innovation, Wageningen University and Research, Wageningen, The Netherlands
ABSTRACT
The farmer field school (FFS) concept has been widely adopted, and such schools have
the reputation of strengthening farmers’capacity to innovate. Although their impact
has been studied widely, what is involved in their scaling and in their becoming an
integral part of agricultural innovation systems has been studied much less. In the
case of the Sustainable Tree Crops Programme in Cameroon, we investigate how a
public–private partnership (PPP) did not lead to satisfactory widespread scaling in
the cocoa innovation system. We build a detailed understanding of the key
dimensions and dynamics involved and the wider lessons that might be learned
regarding complex scaling processes in the context of agricultural innovation
systems. Original interview data and document analysis inform the case study. A
specific analytical approach was used to structure the broad-based exploration of
the qualitative dataset. We conclude that scaling and institutionalization outcomes
were impeded by: the lack of an adaptive approach to scaling the FFS curriculum,
limited investments and genuine buy-in by extension actors, a failure to adapt the
management approach between the pilot and the scaling phase, and the lack of
strategic competencies to guide the process. Our findings support suggestions from
recent literature that pilots need to be translated and adapted in light of specific
contextual and institutional conditions, rather than approached as a linear rolling-
out process. These findings are relevant for the further spread of similar approaches
commonly involved in multi-stakeholder scaling processes such as innovation
platforms.
KEYWORDS
Africa; Cameroon; cocoa;
agricultural extension;
innovation
© 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/
licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not
altered, transformed, or built upon in any way.
CONTACT Sander Muilerman sander.muilerman@wur.nl
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY, 2018
https://doi.org/10.1080/14735903.2018.1440469

1. Introduction
The farmer field school (FFS) concept and such
schools’positive impact on agricultural develop-
ment and wider societal benefits (e.g. poverty alle-
viation) have been studied by many scholars (e.g.
Davis et al., 2012; Feder, Murgai, & Quizon, 2004;
Friis-Hansen & Duveskog, 2012; Larsen & Lilleør,
2014; Phillips, Waddington, & White, 2014;Tripp,
Wijeratne, & Piyadasa, 2005; van den Berg &
Jiggins, 2007; Yorobe, Rejesus, & Hammig, 2011).
Waddington et al. (2014) report FFS as one of the
most common approaches to rural adult education
and agricultural extension involving 10–20 million
people in more than 90 countries. This means that
use of the FFS approach has scaled up significantly
since its emergence in the 1980s, although the
nature and quality of its application may vary (Sher-
wood, Schut, & Leeuwis, 2012). Neither individual
studies nor comprehensive reviews such as that
conducted by Waddington et al. (2014)address
systematically what is involved in processes of
scaling-up the application of the FFS approach so
that it becomes an integral part of agricultural inno-
vation systems. The focus is generally on what the
FFS is about and what it effects are. Although in
numbers FFS has evidently gone to scale, questions
remain regarding what is involved in the success
and failure of purposefully attempting to scale it
up and institutionalize it in agricultural extension
systems.
We were particularly confronted with this question
when assessing the Sustainable Tree Crops Pro-
gramme (STCP). This public–private partnership (PPP)
initially focused on designing, testing, and validating
an innovative cocoa FFS curriculum –designed to
augment cocoa farmers’income by sustainably
increasing the yield and quality of their crops –and
in a next phase set out to take the approach to
scale. STCP was the first large PPP focusing on
scaling agricultural innovations for cocoa in sub-
Saharan Africa (David, 2007,2011; Gockowski,
Asamoah, David, Gyamfi, & Kumi, 2010). The pro-
gramme was implemented in Cameroon, Côte
d’Ivoire, Ghana, and Nigeria (and to a lesser extent in
Liberia) between 2001 and 2011. Although in each
country except Liberia FFS was introduced in virtually
the same manner and with similar staffing, the sca-
ling-up processes led to significantly different results
(Muilerman, Vellema, & Schut, 2018). In Cameroon, a
country with stable leadership and a stable
institutional landscape for cocoa, FFS went to scale
in terms of numbers of schools but in fact spread to
only a dozen emerging cocoa cooperatives. By 2011,
the programme had trained virtually all the
members of these relatively small cooperatives,
often including recruited non-members. During an
internal regional STCP management meeting in early
2011 (first author’s notes), the regional management
assessed that the focus on a limited number of coop-
eratives with limited membership would not enable
FFS scaling. Impact was not significant, there were
important quality concerns, and the sustained adop-
tion of cocoa FFS in the national innovation system
in Cameroon was deemed highly unlikely. Nor was
there clear proof of adoption of core FFS principles
by government or by national NGOs. This analysis
was corroborated by interviews with government offi-
cials in July 2010.
Côte d’Ivoire, a war-torn country in crisis, managed
to take FFS to scale through institutionalization in the
national professional extension services (Muilerman &
Vellema, 2017). A general retrospective, a comparative
analysis of what happened in the four STCP countries,
provided several clues regarding differences in the
extent to which, and the reason why, scaling did –
or did not –occur in Cameroon, Côte d’Ivoire,
Ghana, and Nigeria (Muilerman et al., 2018). This
general comparison showed that in Cameroon (con-
trary to particularly Côte d’Ivoire and Nigeria) STCP
faced a more challenging institutional context. The
programme experienced decidedly less favourable
interactions with the national agricultural institutions.
A lack of involvement with the dominant national
cocoa system meant that virtually no room was
made for the FFS innovation, and this contributed to
FFS never leaving its protected niche environment.
The general analysis indicated that the necessary pre-
conditions for the scaling of FFS were simply not
present, nor evolving in the right direction. However,
the case of scaling cocoa FFS in Cameroon remained
substantially more ambiguous and more difficult to
interpret in terms of the specific mechanisms and
factors that led to the limited level of scaling and
the failure to institutionalize FFS.
Here, we propose to analyse the Cameroon case in
more depth, using an analytical framework that has
the potential of helping to uncover a broad range of
potential factors and dynamics that may have played
a role in impeding the scaling of cocoa FFS and in
the failure to institutionalize it in the agricultural inno-
vation system. This may also provide relevant insights
2S. MUILERMAN ET AL.

for scaling similar participatory approaches and multi-
stakeholder processes, such as innovation platforms
and innovation labs (e.g. Kilelu, Klerkx, & Leeuwis,
2013). Going to scale is an important theme in the
FFS literature. Discussions in STCP focused on
approaches to scaling, changes to the methodology
in the course of expansion (Schut & Sherwood,
2007), and modalities for ensuring financial sustain-
ability (Feder, Murgai, & Quizon, 2008). Investing in
an intervention is a key element for reaching scale,
especially if a donor-funded extension-led FFS is to
be followed by sector-funded and/or community-led
initiatives (Settle, Whitten, Dilts, & OOI, 1998), although
this idealistic model of fiscal sustainability has been
strongly criticized (Feder et al., 2008). Worldwide,
the combined start-up and recurrent FFS costs are
highly variable, ranging between US$ 10 and 80 per
participant for FFS on food crops (van den Berg &
Jiggins, 2007;Duveskog,per.comm., 2011), with
cost depending also on the type and scope of the
implementing organization and the length of the
training.
This study builds on findings from earlier impact
studies that focused on the programme (e.g. David,
2011). The PRactice-Oriented Multi-level perspective
on Innovation and Scaling (PROMIS) methodological
approach (Wigboldus et al., 2016,2017) was selected
because of its suitability for understanding a range
of dimensions and dynamics involved in scaling pro-
cesses (see section 3). This choice is in line with find-
ings by Van de Fliert, Christiana, Hendayana, and
Murray-Prior (2010) who emphasized the need for a
systems perspective when introducing innovations.
PROMIS builds on the multilevel perspective (MLP)
on socio-technical innovation (Geels, 2002) and the
theory of modal aspects (e.g. Brandon & Lombardi,
2010), enabling the development of integrative,
analytical, and strategic perspectives on scaling initiat-
ives. Selecting key elements from this approach, we
developed a conceptual framework that matches the
needs of this study, which sets out to answer three
main research questions:
1. What factors and dynamics provide the best expla-
nation for the failure to achieve scaling and institu-
tionalization in the case of STCP and cocoa FFS in
Cameroon?
2. To what extent does the broad analysis as applied
in this study provide additional insights that lead to
a deeper understanding of factors and dynamics
involved in scaling and institutionalizing FFS?
3. What wider lessons can we learn from this case for
the design and implementation of future scaling
initiatives?
Section 2 briefly describes the relevant context.
Section 3 elaborates on the conceptual framework
and methodology used in this study, using illustrations
from section 2 to clarify this paper’s orientation. Section
4 provides a narrative account of the findings, and
section 5 provides an analytical account of the findings.
These two research angles are then further discussed in
section 6, which revisits the three research questions,
discusses possibilities for the wider application of the
findings, and presents conclusions from this study.
2. Context and background to the case
2.1. History and environment of the cocoa
sector
Cocoa was introduced into Cameroon as early as 1890
(Monga, 1996). Agriculture contributes to nearly a
quarter of Cameroon’s GDP (World Bank, 2015)andis
the main source of employment (UNdata, 2015).
Reportedly, 600,000 smallholder farmers produce
cocoa. FAO data (FAOSTAT, 2015) show that, at the
onset of STCP, cocoa productivity was low (∼375 kg/
ha). The STCP baseline on Cameroon (IITA/ODECO,
2003) attributed this mainly to an aging tree stock
(av. 32 years; 2 years over what is considered the
maximum optimal productive age), high farmer age
(av. 50 years), and widespread prevalence of pests
and diseases. Most cocoa smallholders grow their
own food, but cocoa sales constitute their primary
source of revenue. The STCP baseline also showed
that 24% of cocoa farmers were members of some
formal rural organization and that 35% had had
contact with governmental extension workers in the
previous three months (IITA/ODECO, 2003). Although
Cameroon was better serviced by governmental exten-
sion in 2001, as compared to major cocoa producers
Côte d’Ivoire (31%) and Ghana (23%) (IITA/KNUST,
2003; Kouadjo, Keho, Mosso, & Toutou, 2002), extension
focused primarily on larger farmers. International multi-
nationals became increasingly worried about this situ-
ation. A dozen years later, Cameroon’s production of
over a quarter of a million tonnes of cocoa continued
to be based on the use of vast areas of land, not on
intensification. Smallholder households continued to
obtain low yields (∼400 kg/ha) (FAOSTAT, 2015)on
hundreds of thousands of relatively small plots.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 3

2.2. Cocoa extension and innovation system in
Cameroon
In this paper, we focus in the agricultural extension
institutions for cocoa in Cameroon. Private and NGO
extension was practically non-existent. Previously,
cocoa farmers could call upon specialized cocoa
extension agents from the Cocoa Development Cor-
poration. However, even though the government of
Cameroon was sluggish in adopting reforms, in the
1980s and 1990s liberalization and various IMF and
World Bank programmes, including structural adjust-
ment programmes, progressively put an end to large
state and parastatal extension services (IITA/ODECO,
2003). For many years, MINADER (le Ministère de
l’Agriculture et du Développement Rural: Ministry of
Agriculture and Rural Development) in Cameroon
had been running World Bank-promoted Training &
Visit (T&V) programmes. Each of its extension agents
focused on multiple crops. From 2002 onwards,
MINADER adopted a single crop focus. All of this tran-
spired during the presidency of Paul Biya, which was
characterized by a socio-political status quo (Ngwafu,
2014). In 2010, President Biya, having already served
for 29 years, was re-elected in 2010 ‘against a back-
ground of general indifference’; and the IMF reported
that same year: ‘although [Cameroon] has plentiful
resources, its economic results do not match its econ-
omic potential because of the government’s reluc-
tance to adopt reforms’(Pigeaud, 2011, p. 1).
By 2010, MINADER’s PNVRA (Programme National
de Vulgarisation et de Recherche Agricoles: National Pro-
gramme for Agricultural Extension and Research) still
had the mandate for agricultural extension, including
for cocoa. Interview data suggest that the technical
modules on integrated crop and pest management
and good agricultural practices that featured in FFS
were in themselves considered to be very valuable
and could perhaps have been integrated into
PNVRA’s programmes if they had not been treated
as part of a fixed package. As a result, although
MINADER and IRAD (Institut de Recherche Agricole
pour le Developpement: Institute for Agricultural
Research for Development) contributed to the techni-
cal curriculum, MINADER/PNVRA did not feel that it
had specific ownership of the FFS approach. The
STCP/FFS programme included a number of principles
and components (sections 3 and 4), but Cameroonian
enthusiasm for FFS outside partner cooperatives
related not so much to FFS as process principles, but
rather to the technical curriculum and what quality
training could do for farmers (i.e. responsible use of
chemicals, social organization, enhanced profitability).
The inflexible FFS approach came with package deals
such as the reduction of pesticide use. A conflict of
interest therefore arose as MINADER was promoting
pesticide use. The lack of flexibility in applying the
FFS concept meant that STCP did not focus on the
purpose behind FFS –a purpose that might have
been achieved in other ways as well (e.g. innovation
platforms or other forms of participatory [technology]
development).
A new development in 2006 was the inception of
the National Cocoa and Coffee Sub Sector Develop-
ment Fund (FODECC), a national fund based on
export levies, which in a complex configuration was
to be managed collaboratively by three different min-
istries. FODECC struggled to become operational and
by 2010 was yet to develop into a more serious finan-
cier of cocoa extension. The STCP country manager
reported that STCP never made a functional connec-
tion with FODECC and therefore did not manage to
secure national funding for FFS.
2.3. The STCP cocoa FFS initiative
In the 1999 Paris Declaration, the chocolate industry,
donor agencies, trade organizations, producer
groups, and major research institutions made a com-
mitment to sustainable cocoa production. In 2000,
STCP –worth US$ 8 million (Velarde & Tomich, 2006)
–was set up as a broad PPP platform to address
farmer and business support services, research and
technology transfer, policy change and implemen-
tation, and market and information systems. It
adopted a focus to ‘identify, test, and validate inno-
vations in tree crop systems that could serve to
guide future development investments in tree crop
sectors’(STCP, 2006, p. ii). In a typical West-African
context of weak public and private agricultural exten-
sion systems and inadequate training approaches, a
pilot was conducted in Cameroon, Côte d’Ivoire,
Ghana, and Nigeria (David, 2007). Funded mainly by
the global chocolate industry, represented by the
World Cocoa Foundation (WCF), and by the United
States Agency for International Development, the pro-
gramme was hosted by the International Institute of
Tropical Agriculture (IITA). This is an agricultural
research institute with a mandate to conduct research
to support development initiatives, although not
necessarily to implement development activities on
a large scale. STCP’s organizational structure consisted
4S. MUILERMAN ET AL.

of a regional team of a programme manager and tech-
nical experts and country teams consisting of a
country manager and technical field staff (Velarde &
Tomich, 2006). Programme direction was largely
decided at country level. The regional staff focused
on developing methodologies, technical backstop-
ping, monitoring and evaluation, and scientific
production.
The cocoa FFS programme in Cameroon needs,
therefore, to be considered as part of a larger regional
STCP. STCP Cameroon’s connection to the wider
regional programme is visualized in Figure 1.It
shows that STCP Cameroon was an entity largely oper-
ating independently, having a less than optimal con-
nection with public and private partners, directly
intervening in cooperatives. However, as it operated
under a regional PPP programme, the four different
country management teams met at least twice a
year during the programme’s Executive Committee
meetings, where experiences were shared and dis-
cussed and cross-fertilization was expected to take
place. Participatory observation during several of
these bi-annual meetings, however, found that these
could not be characterized as (academic) critical
spaces, but rather as classic implementer–donor
reporting.
Figure 2 presents the type of organization(s)
driving the scaling phase in each country after the
initial research-led pilot phases (2001–2005) and
how Cameroon remained primarily research-led (per-
centages are used because the otherwise similar
country programmes differed in size). During the
STCP scaling phase, the same PPP under WCF leader-
ship started the Cocoa Livelihoods Programme, similar
to STCP and initially also implemented by IITA but later
by WCF itself.
2.4. STCP’s cocoa farmer field schools and their
attributes
FFSs were designed and first implemented by the FAO
in the late 1980s, as a participatory and experimental
learning approach focusing on integrated pest man-
agement and food crops in Asia (Simpson & Owens,
2002), rapidly expanding to multiple crops and geo-
graphies. After some FFS success with cocoa in Asia
(Mangan & Mangan, 2003), FFSs for the perennial
crop cocoa in West Africa were set up only in a small
pilot in Ghana (Asare, 2005) before STCP and national
experts adapted, tested, and validated cocoa FFS
between 2003 and 2005 for use on a wide range of
topics (STCP, 2003) (Asare & David, 2011; STCP,
2004). The approach constituted a considerable
break from the norm because of the altogether differ-
ent nature of extension provision and underlying prin-
ciples, as presented in Table 1.
Figure 1. Simplified overview of flow of decisions from regional to local for STCP Cameroon and its key partners with regard to service provision
to cocoa farmer and (the scaling of) FFS.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 5

In 2005, recommendations from an external review
(STCP, 2005a,2005b) of STCP activities identified FFS,
among other things, as a key innovation to be scaled
in collaboration with national partners. The STCP PPP
itself had been identified as an important innovation
‘from which all stakeholders derive value. This unique
partnership has never existed for the cocoa sector,
which is the most important tree-based commodity in
West Africa’(STCP, 2005a,p.1).Theprogramme’s
second phase focused on three desirable scaling out-
comes: (i) increased numbers of FFS organized in line
with the key principles, (ii) increased applications of
the principles underpinning FFS in (local) cocoa inno-
vation systems, and (iii) increased use of a cocoa FFS
curriculum (STCP, 2006,p.iv).
3. Conceptual framework and methodology
In this section, we explain the conceptual framework
and methodology used for retrospective analysis of
STCP’s FFS scaling initiative in Cameroon to ascertain
why it did not achieve the desired results.
3.1. Conceptual framework used in the analysis
Innovations are embedded in broader societal pro-
cesses that influence –and simultaneously are influ-
enced by –scaling processes. To better understand
how the scaling-up of FFS in Cameroon unfolded
and what factors and dynamics were at play, we
chose to apply and test an analytical approach
specifically designed to analyse and interpret
scaling processes (Wigboldus et al., 2016). Called
the PROMIS, it can be used to enhance learning
from, and planning for, scaling initiatives. PROMIS
involves two main methodological elements: the
MLP (Geels, 2002) and the theory of modal aspects
(e.g. Brandon & Lombardi, 2010). The first element
is particularly useful for making sense of dynamics
in innovation and scaling, and the second helps to
unpack the multifaceted nature of innovation and
scaling.
Figure 2. FFS leadership by country and by programme type during STCP scaling phase (2006–2011), based on STCP annual reports.
Table 1. Underlying principles of STCP’s cocoa farmer field schools
(based on David, 2004; Schut & Sherwood, 2007).
Principles of farmer field schools
.Adult education that acknowledges experience
.Interactive self-help group training approach
.Focus on field-based and concrete experimental learning
.Trained and competent farmer facilitators
.Practical curriculum based on natural (crop) cycle and emerging
issues
.Quality programme management and monitoring and evaluation
.Sustainable financing
6S. MUILERMAN ET AL.

The MLP is particularly useful for ‘reading’and
interpreting the unfolding FFS scaling process in
Cameroon. Geels (2011, p. 26) describes the MLP as
‘a middle-range theory that conceptualizes general
dynamic patterns in socio-technical transitions.’
Within the MLP, three analytical levels are used to
describe non-linear processes and interrelated devel-
opments. At the lowest level, we find (i) innovation
niches or protected spaces. These may eventually
challenge (ii) the stable socio-technical regime(s)
with their established rules and practices. A (iii)
socio-technological landscape (the wider context)
influences both these levels. Niche and landscape
influence, and are influenced by, the regime level,
which involves constant interactive alignment, adjust-
ment, and reconfiguration of processes. MLP has been
used to analyse the scaling of participatory extension
approaches before (e.g. Minh, Larsen, & Neef, 2010;
Schut et al., 2016). Figure 3 is a simplified way of
expressing, within an MLP, what the FFS scaling pro-
cesses were aiming to do in Cameroon.
It is particularly in the interactions between the
niche and the regime, during reconfiguration pro-
cesses, that we expect PROMIS to be able to help
further unpack and interpret context and process
dynamics at play in scaling processes. As already
stated, PROMIS uses the theory of modal aspects,
which comprises an ordered collection of 15 modal
aspects of experienced reality (Table 2). Innovation
and change processes are considered to involve a
reconfiguration of these different aspects (e.g.
Leeuwis, 2013). Scaling processes involve a multitude
of such reconfiguration processes as they take place in
a variety of specific contexts. This relates directly to,
and expands, MLP dimensions such as industry,
markets, and science, which characterize innovation
systems in socio-technological regimes (see Figure
3). A niche, a regime, and a landscape level can be
characterized along the lines of those same aspects
of experienced reality. ‘Experienced reality’is not
about this or that person’s way of experiencing
reality; rather, it a term that refers to a general
concept of reality as it can be experienced.
Together, the MLP and the theory of modal aspects
provide a rich framework for analysing both the pro-
cesses involved in scaling initiatives and the dynamics
of how these play out between the niche and the
regime level. The PROMIS approach further identifies
13 types of such dynamics related to what may
hinder and what may help an envisaged scaling
process. For the purpose of this paper, we have trans-
lated general categories from PROMIS into a set of four
simplified analytical categories (Table 3): (i) social
dynamics (interpersonal and group interactions), (ii)
system dynamics (interrelationships and interconnec-
tions between system dimensions within the cocoa
innovation system in Cameroon and related emerging
outcomes), (iii) scale dynamics (how social and system
Figure 3. Application of a simplified MLP to the case of cocoa FFS in Cameroon (based on Geels, 2002; Geels & Schot, 2007).
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 7

dynamics play out at different scale levels, including
the temporal scale), and (iv) management dynamics
(managerial arrangements, including roles and
responsibilities and related capacities and competen-
cies for task achievement).
3.2. Method of data collection and analysis
Field research was performed between 2010 and
2011, towards the end of STCP, by the first author
who was not previously connected with the pro-
gramme. The process tracing method was used for ret-
rospective analysis involving both narrative and
analytical categories. Qualitative research, using
semi-structured interview guides, was performed, tar-
geting key process actors in Cameroon. A total of 16
in-depth interviews were held in English, French, or
a local language (assisted by a translator) with 12
different persons or farmers’groups from STCP, pro-
gramme partners, and four partner cooperatives. All
interviews were recorded and transcribed. In addition,
STCP’s extensive electronic and paper archive
1
was
reviewed. STCP did not continue, leaving the data
unused. This study takes up the exercise again as an
ex-post study.
The following two sections present two different
and complementary accounts of our findings. In
section 4, we present a narrative account that allows
for developing a perspective on how things evolved
over time. In section 5, we present an analytical
account, applying the analytical categories as
described in Table 3. The discussion (section 6) syn-
thesizes findings from the two research angles, includ-
ing a discussion of findings along the MLP lines as
visualized in Figure 3.
4. Narrative account of findings and their
interpretation: evolution of the scaling
initiative
In this section, we examine the STCP chronology,
building on the research data and background on
Table 3. Analytical categories derived from the PROMIS approach
(abbreviated and adapted from Wigboldus et al., 2016).
Analytical categories Description
Social dynamics Stakeholder decision-making dynamics
Stakeholder diversity and social impact
System dynamics Complexity of scaling
Dominance/deviance dynamics
System/practice inertia
System/practice instability
Scale dynamics Path dependence and past imprinting
Anticipated futures
Cross-scale, cross-domain dynamics
Management
dynamics
The process of engaging stakeholders
The scope of analysis and evaluation, and
preparation/anticipation
The connection between strategy and
situational reality
The capacity to facilitate convergence
Table 2. Aspects of experienced reality that can in various ways be affected by, or affect, innovation and scaling processes (adapted and
abbreviated from Wigboldus et al., 2016).
Aspects of experienced reality What the aspects typically relate to
Natural and physical aspects
Quantitative, spatial, kinematic,
physical
Numbers, location, atmosphere, climate, water, soil, natural forces, chemistry, transportation, infrastructure,
buildings, equipment
Biotic, sensitive Plants, animals, birds, fish, organic processes, ecosystem, biodiversity, forest, desert, habitat, farm, crops,
livestock, animal behaviour
Human aspects
Biotic, sensitive Awareness, health, physical and mental abilities, emotion, personality, disposition, passion, observation,
population dynamics, safety
Analytical–logical Knowledge, theory, logic, conceptual framework, science, research, education
Formative Construction, creativity, skill, computer software, design, power (in relationship): technology, strategy,
methodology, innovation, adaptation
Social and financial aspects
Lingual, Social Symbols, signs, language, communication, information, media
Relationships, roles, social cohesion, competition, collaboration, organization, societies, alliances,
partnerships
Economic Resource management, conservation, stewardship, exchange of goods and services, transactions, efficiency,
sustainability, economy, land use, market, value chain, firm, employment
Cultural, political, and moral aspects
Juridical Rights, law, responsibility, appropriateness, policy, legal system, constitution, mandate, police, the state,
democracy, ownership
Aesthetic, ethical, certitudinal Appeal, beauty, enjoyment, leisure, sports, art
Attitude, care, sharing, goodwill, integrity, equity, being right, solidarity identity, belief, trust, faith, vision,
commitment, aspiration, worldview, ideology, paradigm
8S. MUILERMAN ET AL.

the landscape, dominant extension regime, and the
niche intervention, as presented in section 2.
4.1. Phase one: introduction of FFS and the
pilot process
Interview data and interviews show that STCP’s pilot at
first focused on a partnership with a vast loosely orga-
nized farmer-based organizational network, FORCE. By
1996, FORCE had emerged through a merger of 25
associations, combining 300 farmer groups or 20,000
farmers (FAO, 2002). Although it aligned with the
national policy direction to support micro projects,
the pilot did not involve MINADER/PNVRA or any
other national cocoa institution (except for resource
persons from IRAD). This pragmatic choice was
based on achieving optimal pilot quality, not on an
analysis of necessary capacity among prospective
partners for running or scaling FFS later on. The
choice to pay FFS farmer facilitators during the pilot
served the same purpose. A few MINADER extension
agents became facilitators, but only because they
had a pre-existing and constructive training relation-
ship with a specific cooperative. A general agreement
was signed in 2000 and 2002 (MINADER, 2009), but
this service agreement did not focus on building an
institutional relationship with the objective of later
scaling-up nationwide.
The pilot employed mainly externally hired expert
staff. According to the country manager, the initial
objective was to organize a high-quality pilot, not to
influence the existing socio-technical regime or to
get national acceptance of the principles underpin-
ning the FFS approach. From the interviews, it is
clear that MINADER demonstrated a keen interest in
the quality of technical training under FFS, but the
approach was not felt to connect with a pressing
need for innovation within existing extension practice.
Because STCP marketed FFS as the sole methodology
to accompany the curriculum, in interviews senior
MINADER officials called STCP dogmatic.
4.2. Phase two: the transition process from
pilot to scaling-up
The FFS pilot ran roughly from 2002 to 2005, after
which an external review of the regional STCP pro-
gramme (STCP, 2005a) advised the scaling-up of FFS.
When the pilot ended and conflict arose with the
farmer network, IITA and the expatriate programme
manager felt that the idea of a representative
network of farmer-based organizations as principal
partner had been implemented prematurely. IITA
took over full management. When funding was rela-
tively easily granted by the United States Department
of Agriculture, the focus was put on working directly
with a dozen (emerging) cooperatives, sustainably
strengthening their capacities and scaling FFS
among the membership.
This meant investing considerable resources in
slow, small-scale, local processes. National manage-
ment pointed out that the United States Department
of Agriculture’s funding was ‘so rigid that farmers
had no voice. But [cooperatives] were benefiting
from the STCP activities. [Therefore] we just contin-
ued.’The country manager confirmed: ‘Yes, people
are saying they are our baby,’although he disagreed.
This approach resulted almost automatically from
STCP’s close partnership with the Canadian NGO
Société de Coopération pour le Développement Inter-
national (SOCODEVI), which had the objective of build-
ing cooperatives and establishing independent
cooperative service provision. The SOCODEVI manage-
ment confirmed this restricted vision: ‘When I die I
hope to leave at least two or three cooperatives
behind that at least are not so clumsy.’This decision
may have closed scaling pathway opportunities later
on.
A lack of true partnership –a result of different
orientations, different expectations, and a lack of
shared effort –in effect meant that the STCP/FFS in
this programme was not in partnership with
MINADER/PNVRA. The data suggest that the extension
system (dominant regime) was not ready to embrace
FFS seriously, though possibly it might have accepted
the curriculum. FFS was introduced perhaps to supply
a rigid extension tool rather than to involve an insti-
tutional innovation in any meaningful way.
4.3. Phase three: management and guidance of
scaling process
During this scaling phase, STCP organized a high
quality FFS programme (Velarde & Tomich, 2006).
However, STCP’s scaling strategy (or lack thereof) cer-
tainly was criticized. The resolute belief in coopera-
tives did not result in the desired level of buy-in
from the dominant regime. Managing the scale-up
alone was a logistical challenge, especially when exter-
nal master trainers left, resulting, among other things,
in a failure to monitor and evaluate, despite heavy
research involvement.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 9

By 2010, five cooperatives had dropped out and
three new cooperatives had joined, and several coop-
erative members voiced their concerns about training
pressure. Targeted training numbers (scaling) were
finally obtained through a questionable tripling of
training numbers through farmer-to-farmer training
(one trainee transferring knowledge to two non-trai-
nees). The country manager was hesitant about
reporting this to the donors. Interviews show diver-
ging perspectives on what scaling FFS would involve.
MINADER agents started to co-facilitate, supervise,
and monitor FFS on an expensive consultancy basis,
and in competition with other initiatives. Under
private sector influence (key donors), this increasingly
also happened in ‘unstructured’communities, result-
ing in low participation. This case is therefore not
about success or failure of scaling-up FFS in general,
but rather about scaling-up the STCP-type of cocoa
FFS in Cameroon. It was clear that MINADER/PNVRA
could never support FFS. It could not even sustain
old-school T&V. The rigid donor-driven framework
for implementation in effect reduced farmers to ben-
eficiaries, undermining the whole FFS ideology. FFS
was increasingly reduced to a tool for scaling-up the
adoption of a technology, and not primarily for
addressing farmers’participation. The emerging
outcome contradicted the country manager’s original
objective. The participatory process of setting up sus-
tainable farmer-led FFS programmes within coopera-
tives was finally not reflected in the partnership and
scaling-up processes.
Table 4 outlines the direct cost of FFS, taking into
account that an FFS facilitator in STCP Cameroon
performed on average three training cycles. STCP
Cameroon incurred a direct cost of US$ 84 per
farmer trained. Low participation numbers, resulting
from working only with cooperative members (on
average 20), meant that the cost could have been
US$ 56 at an optimal participation of 30.
Muilerman and David (2011) outline several
options to reduce STCP’s implementation cost for
public and private sector operators interested in
taking cocoa FFS further to scale, including more train-
ing cycles per facilitator, bulk procurement of
materials, and farmer contributions. Arguments to
justify the relatively high cost of FFS include the
long-term sustainable impact of improved decision
making, the benefits of strengthening human and
social capital, and the ability to use FFS facilitators’
skills for other development initiatives. FFS costs
must therefore be rated against both immediate and
long-term development goals.
4.4. Phase four: institutionalisation and phase-
out
Research data do not support STCP’s claim in reports
that it pursued a two-pronged approach with coop-
eratives and MINADER/PNVRA. STCP shied away from
a structured relationship with the national extension
service and never managed to transfer non-training
tasks such as monitoring and evaluation to the 20–
30 MINADER agents with whom it had worked. In
2010, STCP had lost contact with nearly all of them.
Research data gave reason to doubt the manage-
ment’s belief in, and commitment to, the chosen
approach towards the end, but, with mounting
donor pressure on the programme, giving up on it
was not an option. This led to the over-pressuring of
cooperatives that had reached their training limits
and to a continued discourse that MINADER would
still come on board through a ripple effect.
MINADER, on the contrary, carefully considered the
consequences of adopting FFS. Rather than aiming for
the rapid scale-up of novelties, MINADER/PNVRA con-
sidered the lack of resources and the absence of a
clear political decision, and held on to its T&V-inspired
approach. STCP allegedly aimed for a long-term
approach with MINADER by integrating FFS in the cur-
riculum of agricultural colleges. Surprisingly, and only
after serious probing, it became clear that cocoa FFS
did not make it into the revised 2010/2011 school cur-
riculum. MINADER appears to have tolerated STCP’s
FFS, as funds were made available for something
Table 4. Farmer field school cost per farmer.
Item
Amount in
US$
Start-up: Training of trainers for one facilitator
a
:
Resource persons; training venue and equipment,
facilitators and trainers; food, accommodation, and
expenses; per diems and stipends; materials and
stationary; transportation
291
Recurring: Implementation per farmer field school
b
:
Materials, supervision, facilitator
1385
Cost per farmer in Cameroon under STCP
c
Training of trainer + implementation cost.
84
a
Assumes that the programme starts with new facilitators, trained and
supervised during the first training cycle, who run two more cycles
as experienced facilitators.
b
Assumes a 20% cost reduction for supervision costs during the
second and third years.
c
Based on average recorded participation of 20, although the target
for FFS is 30. Includes the costs borne by the cooperatives.
10 S. MUILERMAN ET AL.

beneficial that Cameroon could otherwise not afford.
Real space for institutionalization and scaling did not
exist and, from inception, there was no purposeful
engagement with key institutions. A MINADER direc-
tor stated: ‘The vision of coops is very tiny. If you
reach [a dozen] cooperatives in a county with
about 1.8 million agricultural households, you have
done nothing!’During an internal regional STCP
management meeting in early 2011, key regional
STCP experts opined that the cooperative scaling
model was flawed, training numbers were too ambi-
tious, and FFS impact in Cameroon (in terms of
farmers reached) was insignificant (first author’s
notes).
Outside STCP’s control, the envisaged future cocoa
fund mechanism FODECC did not become operational
until STCP was starting to wind down. Even then, it still
promised to be a complex procedure involving appli-
cation to three ministries at the same time, with no
clarity about whether a cooperative might be
fundable.
5. Analytical account of findings and their
interpretation: considering conditions for
scaling
Complementing the narrative account of the previous
section, this section presents an analytical account
along the lines of the dimensions and analytical cat-
egories as introduced in section 3 (Tables 2 and 3).
Section 5.1 focuses on the kinds of dimensions (as
aspects of experienced reality) that contributed to
the performance outcomes of the scaling initiative,
and section 5.2 complements this perspective with a
focus on the kinds of dynamics that did so.
5.1. Dimensions of conducive and constraining
factors
Findings from the analysis of conditions for scaling (as
summarized in Table 5) demonstrate the existence of
both conducive and constraining dimensions.
On the one hand, we found a situation character-
ized by farmers strongly motivated to make a long-
Table 5. Summary overview of findings in relation to key dimensions involved in the performance of the STCP/FFS scaling initiative.
Group Summary characterization of how this played a role
Natural and physical aspects
Quantitative, spatial, kinematic, physical;
Biotic, sensitive
Cameroon’s agro-ecology has a potential for high yields of quality cocoa, grown by tens of
thousands of smallholder famers. However, formidable challenges exist, such as high incidence of
pests and diseases, limited access to improved planting materials, lack of extension provision, and
poor road infrastructure. Unsustainable exploitation of fertile forest soils for cocoa causes
significant deforestation, causing the loss of biodiversity, and global climate change is
increasingly impacting agriculture and the natural landscape.
Human aspects
Biotic, sensitive; Analytical–logical;
Formative
Growing cocoa is motivated primarily by the need for security of income, land tenure, and general
livelihood support. Not all smallholder producers have an entrepreneurial focus. Because STCP
focused on entrepreneurship, general participation levels were low, resulting in a bias of the
programme towards already organized farmers. STCP’s management came from a development
NGO background and built on that specific skillset, showing confidence in government
institutions and favouring cooperatives. At the same time, cocoa being an underresearched crop,
the research-led STCP often let academic data needs, knowledge acquisition, or theory
development prevail over a focus on development outputs.
STCP’s service proposition was based on the proven FFS approach with strong underlying
principles and came with an expert-written, high quality technical cocoa curriculum. The quality
of the innovation was not contested, but the lack of flexibility in the curriculum and its role-out
left little room for creativity and alignment with prevailing conditions.
Social and financial aspects
Lingual, social; Economic
The FFS methodology is uniquely focused on applying a participatory, experimental learning
language, which facilitates communication and helps cement relationships and roles at farmer
level. STCP initially tried to work with hundreds of smaller farmer associations instead of
cooperatives. However, hardly any partnerships were developed with the cocoa institutions.
Although cost-effective in the long term, FFSs are relatively resource-intensive. STCP could not
manage more than 12 cooperatives, mainly owing to challenges regarding the necessary
resources, cooperative contributions, logistics, and timing. STCP did build its own expert
institutional capacity and memory, but it was unable to retain this in the long term.
Cultural, political, and moral aspects
Juridical; Aesthetic, ethical, certitudinal
Cocoa plays a central role in farming systems and associated livelihoods in large parts of STCP’s pilot
area, and effective cocoa cooperatives and cocoa officials are viewed as providing pivotal services
in communities. Cocoa farming is important for them and seen as investment over generations.
However, Cameroon’s socio-political situation was characterized by a persistent political status
quo, strong hierarchy, little incentive for innovation, and under the long-term rule of its president.
Resource appropriation by government officials through farmer organizations undermined
participation, and smallholder farmers did not expect much support from the government in
general.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 11

term living from cocoa farming in an agro-ecological
zone that is very suitable to cocoa and by a programme
introducing an in-principle very suitable approach for
addressing challenges faced by cocoa farmers. This pre-
sented a good point of departure for STCP. On the
other hand, the pathway to scale required dealing
effectively and appropriately with the prevalent socio-
political conditions and with related organizational
structures and institutions, which through their
history had made farmers wary about working with
government officials. STCP, despite having strong
(research) capacity and related values, proved
unready to navigate and adjust to socio-political and
socio-cultural conditions; this led to a mismatch
between the chosen strategic and operational
approach and what was actually needed. This demon-
strates the need to consider a wide range of dimen-
sions in assessing readiness to guide innovation
pathways to scale. An early wider-ranging institutional
analysis and feasibility study could have alerted the
programme to the need for more caution in the
design phase and for flexibility and adaptive manage-
ment in the scaling phase, even though a strong pilot
seemed to indicate that the time was appropriate to
take cocoa FFS to scale.
5.2. The interplay of conducive and
constraining dynamics
Having presented results in terms of the dimensions
that proved to be important to take into account
and how these played out interactively, in this
section we focus on the complexity of the interacting
processes and related dynamics that shaped the pro-
gramme’s outcome. The research findings and their
interpretation are summarized in four steps: (1)
social dynamics, (2) system dynamics, (3) scale
dynamics, and (4) management dynamics, reflecting
the analytical categories derived from the PROMIS
approach (see Table 3).
5.2.1. Social dynamics
The essential findings regarding social dynamics
(summarized in Table 6) demonstrate how an
initially strong partnership in relation to the intro-
duction of FFS turned into malfunctioning
decision-making processes in terms of appropriate
participation and discussion of alternative perspec-
tives. STCP’s choices with regard to scaling pro-
cesses soon became rigid, allowing hardly any
functional connection with the national extension
actors. The obligation to donors and the private
sector to achieve a fast return on investment in
the form of thousands of trained farmers was a
major cause of this.
5.2.2. System dynamics
The essential findings regarding system dynamics
(summarized in Table 7) demonstrate how the funda-
mental absence of institutional space for taking FFS to
scale continuously crippled the potential of FFS as
such, the potential of the programme, and the
Table 6. Summary of interpreted findings on social dynamics involved in the performance of the STCP/FFS scaling initiative.
Analytical category Summary of interpreted findings
Stakeholders’decision-making
dynamics
STCP represented key research, public, and private sector actors. After the FFS curriculum was adapted
for Cameroon with broad stakeholder participation through a large network of farmer-based
organizations, only a few emerging cooperatives became part of the strategy for the pilot and
possible scaling. MINADER/PNVRA was shunned, except for specific technical staff. Initially, private
sector involvement was limited. A large donor’s requirements drove STCP to gain technocratic
control over the process and to further decrease stakeholders’decision making. With external NGO
support, existing and newly created cooperatives were nurtured into incorporating FFS, limiting
ongoing feedback opportunities. Finally, quantitative scaling –training numbers –became the key
driver, under donor pressure, but increasingly also under private sector pressure due to growing
impatience.
Diversity of stakeholders’perspectives,
social impact
The research-led STCP did not facilitate discussion between conflicting perspectives on extension
delivery options and the scaling thereof. STCP and SOCODEVI as key implementers believed in the
approach of scaling through cooperatives, although IITA and donors increasingly expressed their
reservations about this. Everyone involved strongly appreciated the value of the technical curriculum,
but STCP would not consider a different methodology to accompany it in Cameroon. MINADER/
PNVRA in principle valued FFS but could not realistically support or responsibly scale it. PNVRA always
regarded STCP as a small closed pilot around a valuable but otherwise unadoptable innovation.
Fragile and under-supported cooperatives were happy with donors’interventions but behaved as
beneficiaries. Five dropped out, and those that remained expressed doubts about the strategy and
complained about pressure.
12 S. MUILERMAN ET AL.

potential for emerging partnerships. They also
demonstrate how the situation in a country or locality
cannot be taken at face value, as initially Cameroon
appeared to be more suitable than Cote d’Ivoire (see
reference to STCP in Cote d’Ivoire in the introduction),
but subsequently proved to have a more rigid
Table 7. Summary of interpreted findings on system dynamics involved in the performance of the STCP/FFS scaling initiative.
Analytical category Summary of interpreted findings
Complexity and
connectedness
Implementation of cocoa FFS is not particularly complex, nor does it involve many actors. The intervention
conflicted with the existing mode of extension provision, the existing regime. Faced with this stalemate, STCP
chose to persist, not reorient. STCP minimized external complexity by creating minimal linkages. Internally, the
complexity and connectedness was high because of the logic of working within the values, practices, and facilities
typical of cooperatives.
Dominance/deviance
dynamics
Extension institutions look to politicians for direction. STCP’s interactions with either institutions or politicians were
almost non-existent. Niche (FFS) and regime (extension system) did not respect each other’s discourse (without
contesting the innovation’s value). The regime was not prepared to integrate FFS because it not was considered
scalable within the socio-political context of Cameroon.
System/practice inertia The regime was rigidly stable, showing institutional and political lock-in, with disincentives for actors to change. In
the absence of political decisions or fund allocation, only the status quo could persist.
System/practice instability The Cameroonian socio-political landscape has been extremely stable, with practically no mobility. True, in the
recent past, the dominant regime had experienced deficiency shocks, particularly after the structural adjustment
programmes. Thus, with no resources and no alternative capacities to fall back on, the extension system had not
sufficiently recovered to adopt novel technologies like FFS.
Table 8. Summary of interpreted findings on scale dynamics involved in the performance of the STCP/FFS scaling initiative.
Analytical category Summary of interpreted findings
Path dependence, past
imprinting
Cameroon had previously adopted new approaches, particularly T&V. This logic persisted within its operations,
despite resource shortages inducing a shift to micro projects. Earlier pilots on food FFS were evaluated
positively.
Anticipated futures Management did not successfully anticipate ongoing processes or what a scaling-up scenario would entail. It
erroneously counted on an emerging cocoa/coffee fund, FFS integration into agricultural colleges, and
improving relations with MINADER/PNVRA after management rejuvenation. Soon, STCP’s scaling strategy
reached maximum capacity, when all cooperative members had been trained.
Cross-scale, cross-domain
dynamics
The powerful STCP, supported by the private sector and increasingly by WCF, integrated the latest technologies.
FFS was designed to address common challenges and respond to international (consumer) pressure on
sustainability and certification and particularly on issues of child labour and pesticide residues. Sector-wide
calls for increased smallholder support and training resonated in Cameroon, although some recommendations
(e.g. input use) conflicted with existing policy.
Table 9. Summary of interpreted findings on management dynamics involved in the performance of the STCP/FFS scaling initiative.
Analytical category Summary of interpreted findings
The process of engaging stakeholders Sustainably investing in broad stakeholder interaction was not a priority, particularly with the
public and private sector. Working with a broad network of farmer-based organizations was
initially unsuccessful. A subsequent focus on a rather restricted group of immature cooperative
partners and individual government agents during scaling caused a paternalistic style of
interaction. STCP’s unsubstantiated vision resonated well with the cooperatives’mandate and
strengthened their services but failed to attract large numbers of new members or create new
farmer-based organizations. Management seemed to lack the institutional and entrepreneurial
skills set to successfully navigate a route across the broader institutional landscape and take on
the existing socio-technological regime –even if that were possible.
The scope of analysis and evaluation, and
preparation/anticipation
At the level of STCP Cameroon, the scaling process was clearly not by design. No scaling process
tracing or monitoring was performed. This precluded any opportunity for evaluation, learning,
and/or reorientation.
The connection between strategy and
situational reality
This connection was practically non-existent, and the observation of high levels of ‘wishful
thinking’was consistently made by most stakeholders outside the STCP team. What remains
unclear, however, is why regional researchers and the increasingly powerful private sector did
not manage (or dare) to reorient a clearly struggling initiative, but rather let it run its course
and expire.
The capacity to facilitate convergence No evidence of any activity other than impromptu, unstructured high-level dialogue on the topic
has been recorded.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 13

institutional context for taking innovations such as FFS
to scale; and key stakeholders did not consider this an
attractive proposition.
5.2.3. Scale dynamics
The essential findings regarding scale dynamics (sum-
marized in Table 8) demonstrate a mismatch between
prevalent institutional conditions and emerging insti-
tutional realities on the one hand, and STCP’s strategic
approach and operations on the other. They demon-
strate the need to pay attention, and connect, to the
dynamic history in which programme are to be
located, and the need for strategic foresight, neither
of which were addressed appropriately by STCP in
Cameroon.
5.2.4. Management dynamics
The essential findings regarding management
dynamics (summarized in Table 9) demonstrate
serious deficiencies with regard to the process of
engaging with partners and stakeholders, in the
capacity for adaptive management, and in the
capacity to facilitate the convergence of competing
perspectives and orientations. These were not part
of STCP’s original design, but were also not prioritized
over the programme’s lifespan in Cameroon later on.
Even though during a 10-year programme there
was no clear outlook for improvement in the State’s
(financial) capacity to take FFS forward, this did not
lead at any point to a no-go decision; this further
demonstrates the lack of capacity for adaptive
management.
6. Discussion and conclusions
This section discusses the research findings in terms of
an integral consideration of the various research
angles. In section 6.1, we identify the key factors and
dynamics that were considered of particular impor-
tance for understanding what happened in the
STCP/FFS scaling process in Cameroon, and we
examine which of these provide the best explanation
for the failure to achieve the desired results. In
section 6.2, we discuss why the multifaceted nature
of scaling processes requires matching comprehen-
sive (analytical) perspectives. In section 6.3, we
discuss how learning from this study may benefit
other (research) initiatives, and in section 6.4 we
draw our general conclusion and make suggestions
for further research.
6.1. Main factors and dynamics causing the
programme not to achieve the desired results
In this section, we answer our first research question
regarding main factors and dynamics causing the
failure of the programme. This is a synthesis of insights
emerging from research findings as described in pre-
vious sections. We discuss related issues in light of
the MLP as visualized in Figure 3.
6.1.1. Context for scaling
Cameroon, with its political status quo, started the
new millennium with top-down agricultural extension
services that had been scaled down during the struc-
tural adjustment programme and liberalization, and
continued using outmoded extension approaches.
STCP, with its focus on FFS, a participatory learning
approach, in effect proposed a socio-technological
transition to another extension approach (or para-
digm). This proved to be unattainable by the govern-
ment services, which lacked the resources and
incentives to challenge the status quo. STCP proved
insufficiently equipped to navigate institutional
obstacles and complications. We may therefore con-
clude that FFS in the case of STCP Cameroon never
really scaled up from its niche environment because
of an absence of interaction and alignment processes
with the dominant socio-technical regime. The cocoa
sector, and agriculture in general, were stagnant
during the STCP era, and the dominant socio-techno-
logical regime was still recovering from system shocks
experienced in the 1980s and 1990s. The private
sector focused mostly on the main cocoa producers,
Côte d’Ivoire and Ghana, and powerful national insti-
tutions helped to maintain a socio-political status
quo in the absence of clear political intent to
provide more resources for extension services (innova-
tive or existing). This was also demonstrated by delays
in setting up mechanisms for internal funding. Thus,
there were few external pressures or internal incen-
tives for change.
The above characterization of the situation under-
scores the importance of contextualizing pro-
grammes, particularly in scaling processes (Klerkx,
Seuneke, de Wolf, & Rossing, 2017). FFS had a strong
track record in terms of its potential for engaging
farmers in participatory ways to address challenges
in agriculture (Braun, Jiggins, Roling, van den Berg, &
Snijders, 2006). STCP was meant to capitalize on this
potential, but management paid insufficient attention
to the fact that even a model as successful as FFS
14 S. MUILERMAN ET AL.

needs to become institutionally embedded through
contextualized processes (Chuluunbaatar & Yoo,
2015).
6.1.2. Readiness for scaling
The programme suffered from general management
issues that became an impediment to scaling. STCP
Cameroon underestimated what was needed man-
agerially, after the pilot, to appropriately embed an
innovation and manage an inclusive scaling
process that was not technology-driven. Appropriate
guidance is essential, and the idea of rolling out an
innovation after successful piloting was –at least in
this case –inappropriate. The regional coordi-
nation’s role can also be called into question. The
required capacities within STCP Cameroon and
among its national scaling partners for managing
the scaling process were not properly assessed
and consequently not catered to as needed (Schut
et al., 2016). We therefore conclude that a successful
pilot is not an automatic assurance of successful
scaling. The financial picture as presented in Table
5further undermined readiness for scaling because
of the high costs associated with FFS. MINADER/
PNVRA and cooperatives, who were meant to co-
invest, confirmed the cost concern that other
authors reported earlier (e.g. Braun & Duveskog,
2008;Davis,2010). The issue of readiness for
scaling can also be considered from a wider per-
spective on the intended role of FFS. FFS can be a
extension approach to convey particular knowledge
packages to farmers (which can be useful for the
private sector if they connect to particular products
and services), but it can also be an approach to
empower farmers, enabling farmers to make their
own decisions on the basis of increased access to
knowledge, a capacity to innovate, and a capacity
to cooperate (Van de Fliert, Dilts, & Pontius, 2002).
We argue that, in the case of cocoa FFS in Camer-
oon, the model focus was stronger than the means
focus. This has led STCP in Cameroon to adopt a
model-replicating mode and the associated rigidity
and lack of flexible navigation of institutional reali-
ties. This was further exacerbated by the external
pressures to perform along the lines of the model.
This is in line with learning from the case of
farmer livestock schools in Vietnam, which led
Minh et al. (2010) to emphasize the need to assess
the existing innovation system and then gradually
and systematically introduce matching institutional
innovations.
6.1.3. Scaling strategy
The scaling strategy was chosen on the basis of per-
sonal judgement rather than a systematic assessment
of all potential options. The choice to spend signifi-
cant resources on building up a small number of
emerging cooperatives appears to have been inap-
propriate. The data on the three other STCP countries
show that these involved national extension through
master trainers, right from the start. This confirms the
essential role played by skilled facilitators and trai-
ners in FFS and the associated difficulty of extending
benefits from FFS beyond the initial groups of
farmers and initial pilots, such as discussed by e.g.
Braun and Duveskog (2008) and Waddington et al.
(2014).
MINADER/PNVRA did not have the necessary politi-
cal and budgetary backing to adopt FFS, and no other
entity could realistically take FFS to scale. It would
have been conceivable to put scaling on hold and
invest in a lobbying exercise with the government of
Cameroon (a member of STCP’s regional Executive
Committee). STCP’s private sector partners were poss-
ibly in a stronger position to put pressure on govern-
ment, being part of the same PPP. More management
foresight might have led rather to a no-go decision on
scaling, paying more attention to lobbying and other
pressing STCP themes such as ‘planting materials’
and ‘pests and diseases’. This underscores the empha-
sis placed on understanding and engaging with pro-
cesses of FFS institutionalization (e.g. by
Chuluunbaatar & Yoo, 2015).
An important condition for sustainable scaling
success –sustainable national financing mechanisms
–did not materialize (at least not fast enough) in
Cameroon. STCP’s core partners all put pressure on
the programme in diverse ways: SOCODEVI was key
to the decision to focus exclusively on cooperatives;
IITA researchers insisted on intensive data collection;
the main donor, the United States Department of Agri-
culture, dictated ideal type FFS implementation
through a rigid log frame approach; and indirectly
the private sector put disproportionate pressure on
the cooperatives to train beyond their membership
base. These dynamics created strong path depen-
dence for STCP’s country management who see-
mingly attempted to meet all the above
expectations, conceivably against their better judge-
ment regarding scaling and institutionalization. This
again raises the issue of the way in which FFS is per-
ceived: as a model or as a means (Van de Fliert et al.,
2002).
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 15

6.1.4. Scaling approach
The MLP suggests that the objectives of management
and donors need to converge, become aligned
somehow, with the objectives of all prospective
scaling partners (Geels & Schot, 2007). The inter-
national (and particularly private) donors’objective
to achieve quantifiable impact –training numbers –
prevailed however. This should not have come at
the expense of paying attention to the complex
dynamics involved in processes related to institutiona-
lization, scalability, and sustainability in the long term
(this relates closely to the discussion in section 6.1.1 on
FFS as a model to be replicated). The process became
paternalistic rather than participatory for the coopera-
tives (which van de Fliert (2010) already identified as
key factor), and STCP did not purposefully engage
with the other stakeholders in the cocoa extension
arena. STCP’s claim that a partnership with
MINADER/PNVRA was emerging was not substan-
tiated by the data, even though the related wider insti-
tutional buy-in was important information for donors
and regional management.
Several underlying core principles of the
approach were not satisfied, nor accepted at the
regime level. The understanding of MINADER/
PNVRA about cocoa FFS was in fact close to business
as usual. STCP could have insisted in developing
broad regime support for more than the technical
curriculum and the label. Instead, in the eyes of
the regime actors, it adhered to the promotion of
its ‘brand’type of the FFS approach and the techni-
cal cocoa curriculum as a dogmatic package.
Working with a high-quality single crop curriculum
was in principle quite compatible with MINADER/
PNVRA, who expressed appreciation of the curricu-
lum while at the same time making it abundantly
clearthatitcouldnotadoptFFS.Failingtolinkup
and capitalize on this institutional reality may have
been a missed opportunity. This situation points to
the importance of (i) flexibility of approach, (ii) the
search for alignment, and (iii) the development of
joint national ownership for both the piloting and
a potential scaling process.
6.2. The need for multidimensional
understanding of scaling initiatives
Scaling and institutionalizing FFS involves a double
layer of complexity (Chuluunbaatar & Yoo, 2015). First,
there is the approach as such with its core principles
(as outline in Table 1) and the intricacies of what in
this approach connects to prevalent conditions in the
agricultural sector in general and to prevalent
approaches to extension specifically. Also, FFS institu-
tionalization involves both institutionalization of prin-
ciples underpinning FFS and institutionalization of FFS
as an approach (Braun & Duveskog, 2008). Second,
scaling processes introduce additional dynamics to be
taken into account as discussed in the previous
section. The findings from this research help to clarify
this double layer of multifaceted change dimensions
and dynamics (Wigboldus et al., 2016)andillustrate
how the scaling of a well-defined model involves any-
thing but a model roll-out process. This supports find-
ings by Van de Fliert et al. (2010) who emphasized
the importance of a participatory approach, appropri-
ate capacity of researchers and facilitators, and owner-
ship by relevant stakeholders. It also demonstrates the
important role of a comprehensive analysis, illustrating
how –in the midst of many positive conditions (as was
the case for cocoa FFS in Cameroon) –other complicat-
ing factors may still outplay their effects. A sufficiently
broad-ranging analysis is not just useful for doing a ret-
rospective assessment as we did in this paper, but, as
argued in section 5.1, would be even more important
as part of preparations for the design and management
of scaling initiatives.
6.3. Wider application of findings
The findings from this study may be relevant for
initiatives worldwide focused on the wider use of
FFS or similar participatory learning approaches, but
also more broadly in relation to taking agricultural
innovations to scale. For instance, the same conditions
for, and impediments to, change may apply to the
setting up and organization of innovation platforms
(e.g. Kabambe et al., 2012; Van Paassen, Klerkx, Adu-
Acheampong, Adjei-Nsiah, & Zannoue, 2014). More
widely, the findings underscore the need to take
scaling processes in agricultural innovation seriously;
this involves appropriate critical analysis, strategic
competencies, collaboration, and creative manage-
ment capacity (Westley, Antadze, Riddell, Robinson,
& Geobey, 2014; Wigboldus et al., 2016; Wigboldus &
Leeuwis, 2013). Many studies on scaling agricultural
innovations focus on the achievements of the
scaling process (e.g. in terms of adoption or dissemi-
nation) and the scaling mechanisms involved (e.g.
farmer-to-farmer extension). The approach used here
further broadens the scope of analytic dimensions
16 S. MUILERMAN ET AL.

and dynamics that affect, or are affected by, scaling
processes such as socio-cultural and socio-political
conditions. It can help decision makers to consider
what needs to be taken into account in the design,
management, and monitoring and evaluation of
scaling initiatives, all of which continue to be key chal-
lenges in agricultural research and innovation.
The findings support suggestions from recent lit-
erature pointing to the need to translate and adapt
pilots to specific context conditions and ‘best-fit’
options, rather than approaching scaling-up as a
mere rolling-out process (Garb & Friedlander, 2014;
Giller et al., 2011; Shiferaw, Okello, & Reddy, 2009).
Consideration must be given to the implications of
future scaling from a stage as early as innovation
design and piloting (Ghiron et al., 2014). The need
for specific competencies for guiding scaling initiat-
ives –competencies that go beyond those involved
in implementing pilots –has also been noted by
others (Hermans, Stuiver, Beers, & Kok, 2013; Spruijt
et al., 2014; Wigboldus & Leeuwis, 2013; Wittmayer &
Schäpke, 2014).
6.4. Conclusions
The case of STCP and cocoa FFSs in Cameroon demon-
strates that, when an innovation (in this case FFS) has
been selected because of its attractive attributes, the
scaling-up process involves more than the mere
implementation of an organizational roll-out plan. A
structured analysis of the institutional landscape,
including scenario planning, is needed to identify
opportunities and elucidate what a successful scaling
process might involve. This would need to include
an assessment of regime and context characteristics,
including past, present, and anticipated (future)
aspects and dynamics; how an initiative would need
to engage (adaptively) with identified constraints
and opportunities; and the capacities and competen-
cies that would need to be in place to support organ-
izational and relational processes. Guidance on FFS
preparation and implementation processes along
these lines would enhance the situational effective-
ness of FFS. In this, the primary goal would not be to
scale FFS as a model to be replicated or as a curriculum
to be rolled out, but rather to scale it for the benefit of
farmers as (one of the possible) means to empower
them.
This paper discussed some dynamics that could be
studied more generally, such as the extent to which
choosing a particular initial entry mechanism (e.g.
small-farmer organizations) to pilot an innovation
creates path dependence from which it is hard to
escape during the process of scaling and institutiona-
lization, and how funding mechanisms may create
perverse incentives that undermine adaptive manage-
ment capacity and partnership processes. This also
relates to the tension between FFS as a flexible
approach (e.g. FAO, 2016) and the high expectations
of its instrumentality for achieving impact at scale,
which funders of development efforts hold and com-
municate to programmes. The resulting drive for
speed and short-term results may jeopardize one or
more of the principles upon which the approach is
based, and similar tensions may occur in relation to
similar participatory approaches such as innovation
platforms.
Note
1. This included among other things: baseline report; pilot
report; transition report; summary external impact
review; the monitoring, financial, and training databases;
5 years of annual and semi-annual reports, work plans,
budgets and results frameworks, and country summary
reports for the scaling phase; 15 relevant studies or
working papers; 9 impact briefs/reports; 30 newsletters;
22 collaboration agreements, 7 training manuals, 31 mis-
cellaneous project documents, and 3 speeches by coop-
erative leaders.
Acknowledgements
This work was supported by the CGIAR Research Program on
Integrated Systems for the Humid Tropics (Humidtropics). We
thank two anonymous reviewers who provided important com-
ments that helped us improve an earlier version of this article.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This work was supported by the CGIAR Research Program on
Integrated Systems for the Humid Tropics (Humidtropics).
List of acronyms
FFS Farmer field school
FODECC National Cocoa and Coffee Sub Sector
Development Fund
GIZ Deutsche Gesellschaft für Internationale
Zusammenarbeit
IITA International Institute of Tropical Agriculture
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 17

IRAD Institut de Recherche Agricole pour le Dévelop-
pement: Institute for Agricultural Research
for Development
MINADER Ministère de l’Agriculture et du Développe-
ment Rural: Ministry of Agriculture and
Rural Development
MLP Multilevel perspective
PNVRA Programme National de Vulgarisation et de
Recherche Agricoles: National Programme
for Agricultural Extension and Research
PPP Public–private partnership
PROMIS PRactice-Oriented Multi-level perspective on
Innovation and Scaling
SOCODEVI Société de Coopération pour le Développe-
ment International: Cooperative Society for
International Development
STCP Sustainable Tree Crops Programme
T&V Training & Visit
WCF World Cocoa Foundation
ORCID
Sander Muilerman http://orcid.org/0000-0001-9103-3294
References
Asare, R. (2005). Cocoa agroforests in West Africa: A look at activi-
ties on preferred trees in the farming systems. Copenhagen:
Forest & Landscape Denmark (FLD).
Asare, R., & David, S. (2011). Good agricultural practices for sustain-
able cocoa production: A guide for farmer training. Manual no.
1: Planting, replanting and tree diversification in cocoa systems,
sustainable tree crops programme. Accra: International
Institute of Tropical Agriculture. July 2011 version. Retrieved
from http://biblio.iita.org/documents/U11ManAsarePlanting
NothomNodev.pdf-66a7d381ce34c6f69507cc1a51506a2f.pdf
Brandon, P. S., & Lombardi, P. (2010). Evaluating sustainable devel-
opment in the built environment. Chichester: Wiley-Blackwell.
Braun, A. R., & Duveskog, D. (2008). The FFS approach history,
global assessment and success stories. Background Paper for
the IFAD Rural Poverty Report 2011. Rome: IFAD.
Braun, A. R., Jiggins, J., Roling, N., van den Berg, H., & Snijders, P.
(2006). A global survey and review of farmer field schools experi-
ences. Nairobi: International Livestock Research Institute (ILRI).
Chuluunbaatar, D., & Yoo, J. (2015). A shift in global perspective.
Institutionalizing farmer field school. Rome: Food and
Agriculture Organization (FAO).
David, S. (2004). Managing the scaling up of STCP farmer field
schools: Five questions and hopefully some answers. Yaoundé:
International Institute of Tropical Agriculture (IITA).
David, S. (2007). Learning to think for ourselves: Knowledge
improvement and social benefits among farmer field school
participants in Cameroon. Journal of International
Agricultural and Extension Education,14(2), 35–49. doi:10.
5191/jiaee.2007.14203
David, S. (2011). An assessment of the effectiveness of a structured
system of farmer-to-farmer diffusion linked to cocoa ICPM
farmer field schools in Côte d’Ivoire, Ghana and Liberia.
Impact Brief. Accra: STCP.
Davis, K. (2010). Extension in Sub-Saharan Africa: Overview and
assessment of past and current models and future prospects.
Journal of International Agricultural Education and Extension,
15,15–28.
Davis, K., Nkonya, E., Kato, E., Mekonnen, D. A., Odendo, M., Miiro, R.,
&Nkuba,J.(2012). Impact of farmer field schools on agricultural
productivity and poverty in East Africa. World Development,
40(2), 402–413. doi:10.1016/j.worlddev.2011.05.019
FAO. (2002). Etude de cas d’aménagement forestier exemplaire en
Afrique centrale: les systèmes agroforestiers cacaoyers,
Cameroun [A study on an exemplary case of forestry manage-
ment in central Africa: Cocoa agroforestry systems,
Cameroon]. Document de travail FM/12F. Rome: FAO.
FAO. (2016). Farmer field school guidance document. Planning for
quality programmes. Rome: FAO.
FAOSTAT. (2015). Retrieved from http://faostat3.fao.org/home/E
Feder, G., Murgai, R., & Quizon, J. (2004). Sending farmers back to
school: The impact of farmer field schools in Indonesia. Review
of Agricultural Economics,26(1), 45–62. doi:10.1596/1813-
9450-3022
Feder, G., Murgai, R., & Quizon, J. (2008). “Investing in farmers—
The impacts of farmer field schools in relation to integrated
pest management”—A comment. World Development,36,
2103–2106. doi:10.1016/j.worlddev.2008.04.011
Friis-Hansen, E., & Duveskog, D. (2012). The empowerment route
to well-being: An analysis of farmer field schools in East Africa.
World Development,40(2), 414–427. doi:10.1016/j.worlddev.
2011.05.005
Garb, Y., & Friedlander, L. (2014). From transfer to translation:
Using systemic understandings of technology to understand
drip irrigation uptake. Agricultural Systems,128(0), 13–24.
doi:10.1016/j.agsy.2014.04.003
Geels, F. W. (2002). Technological transitions as evolutionary
reconfiguration processes: A multi-level perspective and a
case-study. Research Policy,31(8–9), 1257–1274. doi:10.1016/
S0048-7333(02)00062-8
Geels, F. W. (2011). The multi-level perspective on sustainability
transitions: Responses to seven criticisms. Environmental
Innovation and Societal Transitions,1(1), 24–40.
Geels, F. W., & Schot, J. (2007). Typology of sociotechnical tran-
sition pathways. Research Policy,36(3), 399–417.
Ghiron, L., Shillingi, L., Kabiswa, C., Ogonda, G., Omimo, A.,
Ntabona, A., …Fajans, P. (2014). Beginning with sustainable
scale up in mind: Initial results from a population, health and
environment project in East Africa. Reproductive Health
Matters,22(43), 84–92. doi:10.1016/S0968-8080(14)43761-3
Giller, K. E., Tittonell, P., Rufino, M. C., van Wijk, M. T., Zingore, S.,
Mapfumo, P., …Vanlauwe, B. (2011). Communicating com-
plexity: Integrated assessment of trade-offs concerning soil
fertility management within African farming systems to
support innovation and development. Agricultural Systems,
104, 191–203. doi:10.1016/j.agsy.2010.07.002
Gockowski, J., Asamoah, C., David, S., Gyamfi, I., & Kumi, M. A.
(2010). An evaluation of farmer field school induced
changes in Ghanaian cocoa production. Journal of
International Agricultural and Extension Education,17(3), 43–
56. doi:10.5191/jiaee.2010.17304
Hermans, F., Stuiver, M., Beers, P. J., & Kok, K. (2013). The distri-
bution of roles and functions for upscaling and outscaling
innovations in agricultural innovation systems. Agricultural
Systems,115, 117–128. doi:10.1016/j.agsy.2012.09.006
18 S. MUILERMAN ET AL.

IITA/KNUST. (2003). Overview of the cocoa sector in Ghana:
Findings from the STCP baseline study conducted in 2001.
Draft document. Yaoundé: IITA/KNUST.
IITA/ODECO. (2003). Overview of the cocoa sector in Cameroon:
Findings from the STCP baseline study conducted in 2001.
Draft document. Yaoundé: IITA/ODECO.
Kabambe, V. H., Chilimba, A. D. C., Ngwira, A., Mbawe, M.,
Kambauwa, G., & Mapfumo, P. (2012). Using innovation plat-
forms to scale out soil acidity-ameliorating technologies in
dedza district in central malawi. African Journal of
Biotechnology,11(3), 561–569. doi:10.5897/AJB10.2227
Kilelu, C. W., Klerkx, L., & Leeuwis, C. (2013). Unravelling the role of
innovation platforms in supporting co-evolution of inno-
vation: Contributions and tensions in a smallholder dairy
development programme. Agricultural Systems,118,65–77.
doi:10.1016/j.agsy.2013.03.003
Klerkx, L., Seuneke, P., de Wolf, P., & Rossing, W. A. H. (2017).
Replication and translation of co-innovation: The influence
of institutional context in large international participatory
research projects. Land Use Policy,61, 276–292.
Kouadjo, J. M., Keho, Y., Mosso, R. A., & Toutou, K. G. (2002).
Production et Offre du Cacao et du Café en Côte d’Ivoire –
Rapport d’Enquête [Production and supply of cocoa and
coffee in Côte d’Ivoire –Survey report]. Abidjan: Ministre de
l’Enseignement Supérieur et de la Recherche Scientifique,
Ecole Nationale Supérieure de Statistique et d’Economie
Appliqué (ENSEA) and the International Institute of Tropical
Agriculture (IITA).
Larsen, A. F., & Lilleør, H. B. (2014). Beyond the field: The impact of
farmer field schools on food security and poverty alleviation.
World Development,64, 843–859. doi:10.1016/j.worlddev.
2014.07.003
Leeuwis, C. (2013). Coupled performance and change in the
making. Inaugural lecture upon taking up the post of
Professor of Knowledge, Technology and Innovation at
Wageningen University, The Netherlands, on 6 June 2013.
Mangan, J., & Mangan, M. S. (2003). FFS for tree crops. LEISA
Magazine,19(1), 30–31.
MINADER. (2009). Agreement between the International Institute
of Tropical Agriculture and the Regional Delegation of
MINADER –South West (MINADER –SW). 24 April 2009.
Yaoundé, Cameroon: IITA/STCP.
Minh, T. T., Larsen, C. E. S., & Neef, A. (2010). Challenges to insti-
tutionalizing participatory extension: The case of farmer live-
stock schools in Vietnam. The Journal of Agricultural
Education and Extension,16, 179–194. doi:10.1080/
13892241003651449
Monga, Y. D. (1996). The emergence of Duala cocoa planters
under German rule in Cameroon: A case study of entrepre-
neurship. In W. G. Clarence-Smith (Ed.), Cocoa pioneer fronts
since 1800: The role of smallholders, planters and merchants
(pp. 119–136). Basingstoke: Macmillan.
Muilerman, S., & David, S. (2011). Costs associated with farmer field
schools and video viewing clubs on cocoa integrated crop and
pest management: The experience of STCP. STCP Impact Brief
8. Ibadan: International Institute of Tropical Agriculture.
Muilerman, S., & Vellema, S. (2017). Scaling service delivery in a
failed state: Cocoa smallholders, farmer field schools, persist-
ent bureaucrats and institutional work in Côte d’Ivoire.
International Journal of Agricultural Sustainability,15,83–98.
doi:10.1080/14735903.2016.1246274
Muilerman, S., Vellema, S., & Schut, M. (2018). Scaling co-learning
in global commodity chains: A comparative analysis of cocoa
farmer field school programs and socio-technical transition in
West-Africa. In Preparation. Wageningen: Wageningen
University and Research centre.
Ngwafu, P. A. (2014). The United States and democracy in Africa:
Discrepant orientations of Anglophone and Francophone
Africa toward democratic practices (good governance,
administrative reforms and human rights) (posted 19
November 2014). 2015 National Conference of Black
Political Scientists (NCOBPS) Annual Meeting. Retrieved
from http://ssrn.com/abstract=2528103 or http://dx.doi.org/
10.2139/ssrn.2528103
Phillips, D., Waddington, H., & White, H. (2014). Better targeting of
farmers as a channel for poverty reduction: A systematic
review of farmer field schools targeting. Development
Studies Research: An Open Access Journal,1(1), 113–136.
doi:10.1080/21665095.2014.924841
Pigeaud, F. (2011). Cameroon (Vol 8, 2011). In A. Mehler, H.
Melber, & K. van Walraven (Eds.), Africa yearbook online (pp.
184–191). Consulted online on 20 February 2018.
Schut, M., Klerkx, L., Sartas, M., Lamers, D., Mc Campbell, M.,
Ogbonna, I., …Leeuwis, C. (2016). Innovation platforms:
Experiences with their institutional embedding in agricultural
research for development. Experimental Agriculture,52, 537–
561. doi:10.1017/S001447971500023X
Schut, M., & Sherwood, S. (2007). FESs in translation: Scaling up in
name, but not in meaning. LEISA,24(4), 28–29.
Settle, W. H., Whitten, M., Dilts, R., & OOI, P. (1998, September).
Developments in community IPM for irrigated rice in Asia.
Proceedings sixth australasian applied entomological
research conference, Brisbane, Australia.
Sherwood, S. G., Schut, M., & Leeuwis, C., 2012. Learning in
the social wild: Encounters between farmer field schools
and agricultural science and development in Ecuador. In H.
R. Ojha, A. Hall, & V. R. Sulaiman (Eds.), Adaptive collaborative
approaches in natural resources governance: Rethinking partici-
pation, learning and innovation (pp. 102–137). Oxford:
Routledge.
Shiferaw, B. A., Okello, J., & Reddy, R. V. (2009). Adoption and
adaptation of natural resource management innovations in
smallholder agriculture: Reflections on key lessons and best
practices. Environment, Development and Sustainability,11(3),
601–619. doi:10.1007/s10668-007-9132-1
Simpson, B. M., & Owens, M. (2002). Farmer field schools and the
future of agricultural extension in Africa. Journal of
International Agricultural and Extension Education,9(2), 29–36.
Spruijt, P., Knol, A. B., Vasileiadou, E., Devilee, J., Lebret, E., &
Petersen, A. C. (2014). Roles of scientists as policy advisers
on complex issues: A literature review. Environmental Science
& Policy,40,16–25. doi:10.1016/j.envsci.2014.03.002
STCP. (2003). STCP Newsletter, 3 June. Ibadan, Nigeria: IITA.
STCP. (2004). Pilot phase: Project status and year three (2004/2005)
outlook. Summary note. Yaoundé: IITA.
STCP. (2005a). External review endorses STCP. STCP Newsletter,10
March–April 2005. Ibadan, Nigeria: IITA.
STCP. (2005b). The sustainable tree crops program (STCP) external
review synopsis. Ibadan: IITA/STCP.
STCP (2006). The sustainable tree crops program (STCP) phase II
program document –core program overview five-year new
phase program. Version 15 July 2006 revised. Ibadan: IITA.
INTERNATIONAL JOURNAL OF AGRICULTURAL SUSTAINABILITY 19

Tripp, R., Wijeratne, M., & Piyadasa, V. H. (2005). What should we
expect from farmer field schools? A Sri Lanka case study.
World Development,33(10), 1705–1720. doi:10.1016/j.
worlddev.2005.04.012
UNdata. (2015). Cameroon. Retrieved from http://data.un.org/
CountryProfile.aspx?crName=cameroon
Van de Fliert, E., Christiana, B., Hendayana, R., & Murray-Prior, R.
(2010). Pilot roll-out: Adaptive research in farmers’worlds.
Extension Farming Systems Journal,6(1), 63–71.
Van de Fliert, E., Dilts, R., & Pontius, J. (2002). Farmer researcher
teams, farmer field schools and community IPM: Different
platforms for different research and learning objectives. In
C. Leeuwis, & R. Pyburn (Eds.), Wheelbarrows full of frogs:
Social learning in rural resource management (pp. 121–133).
International Research and Reflections. Assen: van Gorcum.
van den Berg, H., & Jiggins, J. (2007). Investing in farmers—The
impacts of farmer field schools in relation to integrated pest
management. World Development,35(4), 663–686. doi:10.
1016/j.worlddev.2006.05.004
Van Paassen, A., Klerkx, L., Adu-Acheampong, R., Adjei-Nsiah, S., &
Zannoue, E. (2014). Agricultural innovation platforms in West
Africa: How does strategic institutional entrepreneurship
unfold in different value chain contexts? Outlook on
Agriculture,43(3), 193–200. doi:10.5367/oa.2014.0178
Velarde, S. J., & Tomich, T. P. (2006). Sustainable tree crops pro-
gramme in Africa. ASB Impact cases 1. Nairobi: ASB-
Partnership for the Tropical Forest Margins and World
Agroforestry Centre.
Waddington, H., Snilstveit, B., Hombrados, J., Vojtkova, M.,
Phillips, D., Davies, P., & White, H. (2014). Farmer field
schools for improving farming practices and farmer
outcomes: A systematic review. Campbell Systematic Reviews.
doi:10.4073/csr.2014.6
Westley, F. R., Antadze, N., Riddell, D. J., Robinson, K., & Geobey, S.
(2014). Five configurations for scaling up social innovation:
Case examples of non-profit organizations from Canada.
Journal of Applied Behavioral Science,50, 234–260. doi:10.
1177/0021886314532945
Wigboldus, S., Hammond, J., Xu, J., Yi, Z. F., He, J., Klerkx, L., &
Leeuwis, C. (2017). Scaling green rubber cultivation in south-
west China: An integrative analysis of stakeholder perspec-
tives. Science of the Total Environment,580, 1475–1482.
doi:10.1016/j.scitotenv.2016.12.126
Wigboldus, S., Klerkx, L., Leeuwis, C., Schut, M., Muilerman, S., &
Jochemsen, H. (2016). Systemic perspectives on scaling agri-
cultural innovations. A review. Agronomy for Sustainable
Development,36, 46. doi:10.1007/s13593-016-0380-z
Wigboldus, S., & Leeuwis, C. (2013). Toward responsible scaling up
and out in agricultural development. An exploration of concepts
and principles. Wageningen: Wageningen University and
Research centre.
Wittmayer, J. M., & Schäpke, N. (2014). Action, research and par-
ticipation: Roles of researchers in sustainability transitions.
Sustainability Science,9(4), 483–496. doi:10.1007/s11625-014-
0258-4
World Bank. (2015). World Bank databank. Retrieved from http://
databank.worldbank.org/data/views/reports/tableview.aspx
Yorobe, J. M.Jr., Rejesus, R. M., & Hammig, M. D. (2011). Insecticide
use impacts of integrated pest management (IPM) farmer field
schools: Evidence from onion farmers in the Philippines.
Agricultural Systems,104(7), 580–587. doi:10.1016/j.agsy.
2011.05.001
20 S. MUILERMAN ET AL.