Limits to the applicability of the innovation platform approach for agricultural development in West Africa: Socio-economic factors constrain stakeholder engagement and confidence by Ashley D. Sparrowa, (CSIRO Land and Water, Private Bag 5, Wembley, WA 6913, Australia) & Adama Traoré (Association pour la Promotion de l'Elevage au Sahel et en Savane (APESS), General Secretariat, 04, BP 590 Ouagadougou 04, Burkina Faso) https://doi.org/10.1016/j.agsy.2017.05.014

1. Contents lists available at ScienceDirect
Agricultural Systems
journal homepage: www.elsevier.com/locate/agsy
Limits to the applicability of the innovation platform approach for
agricultural development in West Africa: Socio-economic factors constrain
stakeholder engagement and confidence
Ashley D. Sparrowa,⁎
, Adama Traoréb
a
CSIRO Land and Water, Private Bag 5, Wembley, WA 6913, Australia
b
Association pour la Promotion de l'Elevage au Sahel et en Savane (APESS), General Secretariat, 04, BP 590 Ouagadougou 04, Burkina Faso
A R T I C L E I N F O
Keywords:
Agricultural innovation systems
IAR4D
Poverty traps
Sahel
A B S T R A C T
Integrated Agricultural Research for Development (IAR4D) is a mode of participatory action research for de-
velopment that aims to improve the well-being of smallholder farming households by facilitating networks
between farmers and marketplace actors through groups called ‘innovation platforms’. In 2012, the Association
for the Promotion of Livestock in the Sahel and Savannas (APESS, an NGO with rural membership) established
nine innovation platforms in the Sahelian zone in Senegal, Burkina Faso, Cameroon and Chad, with the aim of
engaging member livestock producers with meat and milk processors and traders. In late 2013 and early 2014,
each innovation platform reviewed its perception of “performance” or “functionality”, using a consensus-based
group assessment scored against nine criteria. The aim of this study is to test the extent to which perceived
functionality was related to the activities of the innovation platform, as well as current farm productivity and
profitability, household well-being, and regional context, using data gathered independently by APESS as part of
its broader mandate to improve the lives of members.
Across the nine innovation platforms, there were strong positive correlations between the responses to all self-
assessment criteria i.e. members perceived that their innovation platform was performing strongly or weakly in
all respects. The criterion scores were correlated positively with the number of meetings held, but not with
numbers or gender of attendees, training opportunities, nor actions agreed at meetings. Innovation platforms
self-assessed more positively in southern, higher rainfall regions than in the northern, lower rainfall regions.
Performance was positively correlated with many measures of farm productivity and household well-being,
including median area of farms, value of crop and hay production, sales of animals or animal products, ex-
penditure on agricultural inputs, current engagement with markets, and estimated number of months per year
that the family is able to feed itself.
We conclude that the IAR4D approach is perceived to work well in high rainfall zones, for smallholders with a
relatively high current level of well-being, who have larger and more productive farms, and who are already
engaged with markets. The approach does not appear to work well in more arid regions for poorer smallholders,
who are often not able to provide food for their families for 12 months of every year. The IAR4D focus on social
networking into markets appears to be relevant for producers who wish to strengthen or diversify existing
market relationships, but is not perceived to be sufficiently relevant by smallholders for whom achieving reliable
subsistence and averting starvation remains the priority.
1. Introduction
The “first” Green Revolution is generally regarded as having by-
passed or failed sub-Saharan Africa (McIntyre et al., 2009; Pingali,
2012), in sharp contrast with eastern and southern Asia and with South
America. During the period 1960–2007, and even more recently, pro-
duction of most types of agricultural commodities in sub-Saharan Africa
on a per unit area or per unit labour basis showed no substantial trend
or, in the worst cases, declined (Pretty et al., 2011). Food supply for the
growing human population has been met primarily by increasing the
area of land under cultivation each year and/or decreasing the duration
of fields' fallow periods in proportion to population, a strategy that both
fails to increase per capita production at the level of households, and in
the long term has an ecological absolute limit. In east and south Asia,
http://dx.doi.org/10.1016/j.agsy.2017.05.014
Received 29 February 2016; Received in revised form 13 April 2017; Accepted 28 May 2017
⁎
Corresponding author.
E-mail address: ashley.sparrow@csiro.au (A.D. Sparrow).

2. the area under cultivation has increased relatively little since 1960, but
yields per unit area have increased through an intensified agricultural
model based on the use of improved varieties, irrigation, fertilisers,
herbicides, and pesticides, which has ultimately flowed through to
decreased poverty (Pretty et al., 2011).
The food system sensu lato at national, regional or global scales is
highly complex, and many political, social, cultural and technical fac-
tors may provide opportunities and constraints for agricultural pro-
duction (for example, see Pinstrup-Andersen and Watson, 2011,
Poulton, 2014). However, one school of thought lays the blame for the
lack of growth in productivity in sub-Saharan Africa squarely at the feet
of researchers or, more specifically, the interface between research
technologies and producers. The so-called “linear model of technology
transfer”, in which the extension agencies of governments mediate in a
uni-directional manner between researchers and agricultural producers,
is widely perceived to have failed in sub-Saharan Africa and to be a
source of failure in technological uptake (Chambers et al., 1989), with
the factors that meant that linear tech transfer was not successful in
sub-Saharan Africa, while it seemed to succeed in Asia and South
America, identified as insufficient and inadequate services and sup-
porting institutions (Hounkonnou et al., 2012).
Reflection upon the causes of the apparent failure of linear tech
transfer in Africa led to a search for an alternative model, which con-
verged upon more participatory processes in which agricultural pro-
ducers are directly engaged in two-way conversation with researchers
(e.g. Byerlee, 1998), so that researchers better understand the context
and needs of producers, and the producers are able to participate di-
rectly in research activities and receive more targeted, nuanced and
relevant information from researchers and experiments alike. Many
authors have subsequently extended the notion of participatory agri-
cultural research to include entire agricultural “innovation systems”
consisting of researchers, producers, merchants engaged in all aspects
of agri-business, financiers, and consumers (Hall et al., 2001, 2004,
2006, Clark, 2002, Sumberg, 2005, Hall, 2011, Rajalahti et al., 2008,
Hawkins et al., 2009, Nederlof et al., 2011, Sanyang et al., 2014,
Sanyang et al., 2016). This field of development practice is now suffi-
ciently well advanced and embedded in international development ef-
forts for experts to be asking questions of the range of nuanced op-
portunities and constraints that are becoming apparent (e.g. Thiele
et al., 2011, Mulema and Mazur, 2016, Schut et al., 2015).
Management of scope-creep associated with uncritically expanding
to encompass the full diversity of participants within food systems has
demanded more exact specification of the foci and conceptual under-
pinnings for participatory agricultural research. Thus, with a specific
interest in enhancing market engagement by producers, FARA devel-
oped a particular brand of participatory research and innovation in sub-
Saharan Africa that focuses on integration along “market value chains”
and coined the term “integrated agricultural research for development”
(IAR4D) for its concept (Hawkins et al., 2009). In IAR4D, the partici-
patory groups focused on value chains are called “innovation plat-
forms” and are encouraged to include the combination of producers,
input suppliers, transporters, wholesalers, retailers, manufacturers,
credit suppliers and government policy and implementation staff most
relevant to the problem or “entry point” that catalyses engagement
(Adekunle and Fatunbi, 2012). This definition of innovation platforms
is much more specific than some recent usage in which the term has
been used to describe almost any participatory planning or problem-
solving initiative and its tools (e.g. Anttiroiko, 2016; Ojasalo and
Tahtinen, 2016).
In its operational plan 2008–2013, the West and Central African
Council for Agricultural Research and Development (CORAF/
WECARD), the agricultural research and development agency of the
Economic Community of West African States (ECOWAS), adopted
IAR4D as the best-bet approach to systemic innovation for agricultural
development in West and Central Africa and sought international aid
donors to support IAR4D-based research. During 2011–2014, with
funding from the Australian Department of Foreign Affairs and Trade,
and in partnership with Australia's Commonwealth Scientific and
Industrial Research Organisation (CSIRO), CORAF/WECARD estab-
lished a programme of six multi-country projects that took IAR4D and
innovation platforms as central methodological tools to address im-
provements in aspects of crop and livestock production. Given the ex-
plicit selection criteria for the funding programme, the design of these
projects all applied IAR4D principles, making the assumption that
IAR4D is a more impactful approach to research for development than,
say, linear technology transfer, even though success of IAR4D had not
been demonstrated rigorously at the time of project initiation in 2011.
Literature published during the period of project implementation has
been inconsistent about the relative benefits of IAR4D (e.g. Adekunle
et al., 2012, Nkonya et al., 2013, Sanyang et al., 2013, Siziba et al.,
2013, Pamuk et al., 2014, Sanyang et al., 2014, Pamuk and van Rijn,
2015, Adjei-Nsiah and Klerkx, 2016, Sanyang et al., 2016) and further
examination of the potential advantages and benefits remains critical
before future faith in the approach can be considered secure.
One CORAF/WECARD project was coordinated by the Association
for the Promotion of Livestock in the Sahel and Savannas (APESS).
Formed in 1989, after the extended Sahelian drought of the 1970s and
1980s, APESS is an international-funded, member-based association
that works towards improved environmentally and economically sus-
tainable animal husbandry practice by traditional herders and for
greater involvement of animal producers in the economic, political and
social development of West African countries. The project focused on
opportunities for enhanced meat and milk output and profitability by
animal producers in nine case studies in the Sahel. Due to time, space
and financial constraints of the project, formal testing of the hypothesis
that IAR4D is a more effective means of communicating information
and facilitating innovation in the agricultural sector sensu lato than
other approaches to agricultural innovation (such as the linear model of
technology transfer) was not feasible. It was not possible to compare
independent locations, groups in the agricultural sector and entire
market value chains under conditions of IAR4D and one or more other
approaches to innovation in production and market systems.
Furthermore, the two-year period after formation of the innovation
platforms was not sufficient to find durable innovation adoption
amongst producers and their markets, as concluded previously by
Triomphe et al. (2013).
Nevertheless, across the nine innovation platforms established
during the project, there was sufficient variation of experience to
permit testing of three secondary hypotheses about whether the man-
agement and activities of the innovation platform or the circumstances
of the participants beyond the innovation platform per se determine the
functionality of innovation platforms perceived by participants, where
“functionality” is defined as the merit, value, usefulness or success of
the experience for participants given their expectations and needs:
H1. Functionality of innovation platforms is affected by the number of
meetings, the number of training activities and/or the number of production
and value chain actions agreed and undertaken, including the number of
people and the gender representation under each of these activities.
H2. Functionality of innovation platforms is affected by the production
system, socio-economic circumstances and well-being of the families of the
producer members.
H3. Functionality of innovation platforms is affected by regional
characteristics external to the situation of individual members such as
agro-ecological zone (here indexed as mean annual rainfall) and local
population density.
In this paper, we test these specific secondary hypotheses as de facto
tests of the primary hypothesis about the efficacy of IAR4D, cognisant
of the requirement for the additional assumption that participant-per-
ceived functionality of innovation platforms is an indicator of likely
effectiveness in terms of potential future adoption of innovations in
A.D. Sparrow, A. Traoré

3. agricultural value chains.
2. Methods
In 2012, the APESS-led project entitled “Sustainable Intensification
of Integrated Agriculture-Livestock Systems to Increase Agropastoral
Productivity and Food Security in West and Central Africa” (ISIAE)
established and facilitated nine innovation platforms according to the
principles of IAR4D described by Hawkins et al. (2009) and Adekunle
and Fatunbi (2012). Those authors conceived as IAR4D as a set of good
practices that synergistically add value to existing research and devel-
opment processes, rather than a particular research and development
approach, framework or process. Its defining principles are:
1. IAR4D integrates the perspectives, knowledge and actions of dif-
ferent stakeholders around a common theme or “entry point” that
represents a research and development challenge previously iden-
tified by one or more stakeholders.
2. IAR4D integrates the learning that stakeholders achieve through
working together within a “social learning” process.
3. IAR4D integrates analysis, action and change across the full spec-
trum of environmental, social and economic dimensions of devel-
opment.
4. IAR4D integrates analysis, action and change at different levels of
spatial, economic and social organisation.
5. IAR4D requires a set of individual, organizational and institutional
capacities that enables these activities in practice.
There were three innovation platforms in Burkina Faso, three in
Cameroon, one in Chad and two in Senegal (Table 1). APESS employed
a trained local as a facilitator of each innovation platform. The kernel of
each innovation platform was a local chapter of APESS livestock pro-
ducers that had previously identified a meat or milk production and
market challenge. Initial membership of the innovation platforms was
comprised of active members of the local APESS chapters, and was
entirely self-selected/volunteered. As preliminary discussion of the
scope of the local value chain challenge progressed, innovation plat-
forms realised that additional input and expertise were essential, and
thus invited or co-opted additional members who represented more
distant markets, input suppliers, veterinarians, transport companies,
financial institutions, government agencies (with regard to questions of
policy, regulation and implementation) and researchers.
With oversight and support from APESS, all innovation platform
members were engaged in a series of social learning and training ac-
tivities, and facilitated informal group discussions with the objective of
improving meat and/or milk production and availability in markets,
and the economic and broader well-being of all participants and their
households. Thus whilst the implied catalytic objective of each in-
novation platform was initially improvement of the production and
well-being of local APESS-member households, the self-organised ob-
jective of the evolved innovation platform related as much to commu-
nity-wide benefit as to benefit for APESS members. The project financed
experts to speak to groups and covered the costs of participants to at-
tend meetings, training activities and implementation activities, but did
not provide any per diem. The project also financed some small in-
vestments in materials to enable some implementation e.g. provision of
insulated boxes for transport of milk from farms to milk processing
facilities by motorbike. Key characteristics of the innovation platforms
and their districts are summarised in Table 1. Data to calculate mean
annual rainfall estimates for 2009–2013 were sourced directly from the
AGRYHYMET Regional Centre (Niamey) of the Permanent Inter-State
Committee against Drought in the Sahel (CILSS), and local human po-
pulation estimates were sourced from the LandScan database of the Oak
Ridge National Laboratory, US Department of Energy.
The design of project monitoring and evaluation required mon-
itoring and evaluation at the scale of each innovation platform. Thus
Table1
Locationsandcharacteristicsofthenineinnovationplatforms.
Numberof
participants
SiteCountryLocationAgro-ecological
zone
Meanannualrainfall
2009–2013(mm)
Estimateddistrictpopulationdensity
2000(km−2
)
EntryPointTotalMenWomen
BanforaBurkinaFaso10°38′N14°45′
W
Sudanian102280Milk:souredmilkandyoghurtproductsintoBanforaand
BoboDioulassomarkets
571355216
BarabouléBurkinaFaso14°12′N1°51′WSahelian67915Meat:sheepandgoatmeatviaregionalmarketsinto
Ouagadougoumarkets
22184
GuiderCameroon9°56′N13°57′ESudanian918100Milk:freshandsouredmilkintoNgaoundérémarkets52502
ManiChad12°43′N14°41′ESahelian55075Meat:sheepandgoatmeatintoNdjamenamarkets22220
MartapCameroon6°54′N13°03′ESudanian168220Milk:freshandsouredmilkintoMarouamarkets807010
MindifCameroon10°24′N14°26′ESahelian782150Meat:sheep,goatandcattlemeatviaGarouamarketsinto
Yaoundémarkets
18816424
TankantoSenegal12°49′N15°02′
W
Sub-humid91540Meat:sheepandgoatmeatviaregionalmarketsintoDakar
markets
25819860
ThiouBurkinaFaso13°48′N2°39′WSahelian75740Meat:sheepandgoatmeatviaregionalmarketsinto
Ouagadougoumarkets
705317
VelingaraSenegal15°00′N14°41′
W
Sahelian5278Meat:sheepandgoatmeatviaregionalmarketsintoDakar
markets
302010
A.D. Sparrow, A. Traoré

4. during 2012–2013 each innovation platform, through its facilitator,
maintained records and reported on the numbers of meetings, training
activities and change activities implemented as a group, including the
numbers of men and women involved in each “event” (Table 2). In
November 2013–January 2014, each innovation platform was asked to
self-assess its “performance” or “functionality” to date, against nine
criteria (Table 3). The objective of the self-assessment was two-fold: (1)
to enable the innovation platform members to reflect upon how it was
progressing against its expectations, in terms of both group processes
and moving towards desired outcomes, and (2) to enable APESS to
gather data on the progress of innovation platforms for its reporting to
CORAF/WECARD. All members of innovation platforms were invited to
join the self-assessment, not just the kernel of producers who are also
members of APESS. The platform self-assessment was undertaken as a
group activity, with the APESS facilitator supporting the group mem-
bers to reach a consensus score against each criterion through open
discussion. Scores were not sought from individuals.
Data on the production system, socio-economic circumstances and
household well-being of the APESS-member producers in each in-
novation platform were sourced from an APESS database that was in-
dependent of the design of the ISIAE project. APESS has adopted a
simple instrument to assess and monitor member well-being and food
security of its members, based on a household-scale survey and a
spreadsheet summary. APESS uses this instrument for monitoring and
evaluation against its overarching mandate of improving the liveli-
hoods of livestock producers through improved education and market
networking. The instrument is called the bilan simplifié (“simplified
balance sheet” or “appraisal”; Barbedette, 2013). Through a multi-day
interview, data are collated for a particular year about the household,
its members, the available labour contributions, the land at its disposal,
the productivity of the land, the costs of inputs such as fertiliser and
veterinary supplies, the household's consumption of crop and livestock
products, the sale of crop and livestock products for cash, use of labour
for off-farm income, gifts and remittances, non-production expenses
including food, healthcare and education, transport, telecommunica-
tions, etc. Barbedette (2013) provides a detailed description of the bilan
simplifié and a thorough guide for its use. The two key calculated in-
dicators on the summary spreadsheet are (1) bottom-line annual profit
or loss in dollar terms, and (2) the taux de couverture (“coverage rate”).
Taux de couverture is the number of months in the year that the
household would have been able to feed itself from its own production
(excluding off-farm labour). If the taux de couverture is > 12 months,
the household is food-secure and the farm can trade commercially and
potentially make a profit. If the taux de couverture is < 12 months, the
household is food-insecure on an annual basis. Producers are en-
couraged to undertake the bilan simplifié assessment in order to estab-
lish their household's current situation as a baseline and then, as they
engage in APESS educational activities, farmers are encouraged to re-
visit the process in order to monitor the progress (or not) in the well-
being of the household as a result of adoption of technological in-
novations or new farm management strategies. In our analysis, all 23
variables on the bilan simplifié were tested as descriptors of the type and
Table 2
Summary of activities by innovation platforms (2012–2013 inclusive).
Meetings Training activities Change implementation activities
Site Country Number Cumulative people
involved⁎
Women
involved (%)
Number Cumulative people
involved⁎
Women
involved (%)
Number Cumulative people
involved⁎
Women
involved (%)
Banfora Burkina Faso 9 60 30 2 8 10 11 18 17
Baraboulé Burkina Faso 2 2 18 1 5 0 17 9 13
Guider Cameroon 3 8 0 2 20 50 4 4 0
Mani Chad 2 11 0 0 0 0 4 138 6
Martap Cameroon 5 17 12 4 20 30 4 20 30
Mindif Cameroon 12 11 14 0 0 0 3 50 10
Tankanto Senegal 3 76 22 2 20 58 7 300 38
Thiou Burkina Faso 3 16 10 2 22 24 5 9 20
Vélingara Senegal 1 30 33 1 1 0 4 68 38
⁎
Not every person participated in every meeting, training activity or implementation activity.
Table 3
Innovation platform self-assessment criteria (late 2013-early 2014).
Criterion Short name of criterion
(used in Tables 5 and 6)
Indicator
The innovation platform is established and functional. Overall function The innovation platform has governance arrangements, including institutional
relations and decision-making, and evidence of scheduled meetings over the past
years.
The innovation platform has several major types of actors and
the participation of these.
Level of participation The innovation platform goes beyond a meeting almost exclusively of producers and
includes the active participation of representatives of the value chain and support
services.
The innovation platform has identified and is in the process of
supporting entry points suited to innovation.
Relevance of entry point Appropriate innovations were developed by the project; they are about to be tested
through project activities and thereby provide a stimulus for transformation.
The innovation platform is incorporating new knowledge
from research and through participation in research.
Research The research activities introduced or initiated by research participants or advisors
have an impact on the evolution of the platform
The innovation platform has access to good facilitation and
capacity in bargaining/brokerage.
Facilitation and
negotiation
Meetings of the innovation platform are organized by trained facilitators and there is
evidence that there has been discussion and consensus around the actions of the
platform.
The innovation platform has prospective opportunities to
achieve outcomes.
Opportunity analysis Additional financial support for the innovation platform will provide evidence
attesting to the achievement of outcomes.
The innovation platform supports explicit effects on gender. Gender Innovations and the path to the impact will benefit women.
The innovation platform invests in communication and
knowledge sharing.
Communication Knowledge is captured and shared amongst members of the innovation platform and
beyond through networks and media.
The innovation platform is supported by a functional
framework for monitoring and evaluation.
Monitoring and
evaluation
The monitoring and evaluation data collected support the desired effects and
behavioural changes.
A.D. Sparrow, A. Traoré

5. size of the production system, the socio-economic circumstances and
the household well-being of innovation platform members (Table 4).
Statistical analysis was undertaken using the innovation platform as
the unit of replication (n = 9), and thus bilan simplifié data collected at
the level of households were aggregated to the scale of the platform.
Due to the small number of innovation platforms, and the lack of nor-
mality displayed by the major of descriptors, all data summary and
analysis were by non-parametric approaches. For the bilan simplifié
data, medians for each variable were calculated amongst producer-
members within each innovation platform. Pair-wise correlations be-
tween all variables at platform scale were then calculated using the
Spearman rank correlation method.
3. Results
All nine variables of innovation-platform self-assessment are posi-
tively correlated (Table 5) i.e. the perceived effectiveness of the in-
novation platform according to any individual assessment criterion
tends to reflect effectiveness according to any other criterion, and
consequently the level of overall functionality of an innovation plat-
form is a direct expression of the level of the other eight assessment
criterion.
In contrast, self-assessed functionality of innovation platform is
generally poorly characterised by quantitative variables describing the
membership and activities of the platform (Table 6). Functionality is
correlated with the number of meetings and, to a much lesser extent,
the number of training opportunities, but is not correlated with the
Table 4
Variables describing the type and size of the production system, the socio-economic circumstances and the household well-being of innovation platform members, as extracted
from the APESS bilan simplifié database.
Variable name Description
People in household Number of people in household
Active people in household Number of people in household able to contribute farm labour
Land area Area of land (ha)
Crops - sales Value of all crops (cereal, tubers, vegetables, fruit and forage) sold into the marketplace
Crops - gifts Value of all crops (cereal, tubers, vegetables, fruit and forage) given to other households
Crops - autoconsumption Value of all crops (cereal, tubers, vegetables, fruit and forage) consumed within the household
Livestock - sales Value of all animal products (beasts, hides, meat and milk) sold into the marketplace
Livestock - gifts Value of all animal products (beasts, hides, meat and milk) given to other households
Livestock - autoconsumption Value of all animal products (beasts, hides, meat and milk) consumed within the household
Total Production Total value of farm production
Crops - expenses Value of all inputs for crop production
Crops - expenses (%) Value of all inputs for crop production (percent of total turnover)
Livestock - expenses Value of all inputs for animal production
Livestock - expenses (%) Value of all inputs for animal production (percent of total turnover)
Total production expenses Total value of farm inputs
Total production expenses (%) Total value of farm inputs (percent of total turnover)
Value of autoconsumption Total value of farm products consumed within the household
Household purchases Value of food purchased from markets
Dependence on market Value of food purchased from markets (percent of all food value)
Receipt of gifts Total value of products received from others without payment
Insertion into markets Total value of farm production sold into the marketplace (percent of turnover)
Total expenses Sum of the value of farm inputs and other household expenses
Taux de couverture Months per year equivalents during which household could feed itself on household-based production alone
Table 5
Spearman rank correlation matrix between the self-assessed scores of innovation platform functionality using the nine
criteria.
Overall function 1.00
Level of participation 0.92 1.00
Relevance of entry point 0.68 0.50 1.00
Research 0.87 0.74 0.84 1.00
Facilitation and negotiation 0.88 0.82 0.76 0.91 1.00
Opportunity analysis 0.89 0.84 0.70 0.90 0.87 1.00
Gender 0.81 0.61 0.93 0.93 0.89 0.84 1.00
Communication 0.66 0.53 0.62 0.67 0.60 0.72 0.71 1.00
Monitoring and evaluation 0.82 0.80 0.70 0.84 0.84 0.92 0.80 0.88 1.00
Overallfunction.
Relevanceofentrypoint.
Research.
Gender.
Monitoringandevaluation.
Legend: Positive > +0.75 +0.50 to 0.74
Level of correlation Negative < –0.75 –0.50 to –0.74
Facilitationandnegotiation.
Levelofparticipation.
Communication.
Opportunityanalysis.
A.D. Sparrow, A. Traoré

6. numbers of actions, the numbers of members involved in any of the
three categories of activities (meetings, training activities and actions),
and the percentage representation of women. Indeed, functionality is
(weakly) negatively correlated with actions i.e. there were fewer re-
ported actions, involving fewer people, in platforms that self-assessed to
be more functional.
Self-assessed functionality of innovation platform is strongly
correlated with a wide range of variables describing the production
system and economic circumstances of the households of producer-
members, as characterised by data from the bilans simplifiés (Table 6).
All aspects of platform functionality are positively correlated with
median size of farms, the value of crop and pasture production – for
family consumption, gifting and sale – and expenditure on crop and
pasture inputs.
Table 6
Spearman rank correlation matrix between the self-assessed scores of innovation platform functionality and the variables describing platform
activities, participant socio-economic well-being, local rainfall and district population.
Overallfunction
Levelofparticipation
Relevanceofentrypoint
Research
Facilitationandnegotiation
Opportunityanalysis.
Gender
Communication
Monitoringandevaluation
Training – number of sessions 0.63 0.57 0.13 0.36 0.49 0.26 0.31 0.17 0.24
Training – number of people 0.41 0.39 –0.26 0.05 0.14 0.06 –0.05 –0.03 –0.03
Training – proportionof women 0.52 0.41 0.05 0.33 0.28 0.19 0.19 –0.07 0.00
Meetings – number of sessions 0.84 0.73 0.71 0.72 0.70 0.83 0.75 0.50 0.66
Meetings – number of people 0.17 –0.18 0.37 0.22 0.19 –0.02 0.42 0.12 –0.08
Meetings – proportion of
women –0.06 –0.26 0.18 0.12 0.25 0.06 0.34 0.10 0.04
Activities – number of sessions –0.23 –0.14 –0.44 –0.39 –0.08 –0.34 –0.33 –0.65 –0.46
Activities – number of people –0.33 –0.59 0.27 –0.13 –0.27 –0.35 0.11 –0.10 –0.36
Activities – proportion of women –0.01 –0.27 0.00 –0.08 0.02 –0.19 0.16 0.15 –0.13
People in household 0.10 0.33 0.04 0.20 0.21 0.11 0.00 0.07 0.25
Active people in household 0.04 0.33 0.02 0.17 0.19 0.15 –0.05 –0.02 0.26
Land area 0.67 0.77 0.47 0.65 0.68 0.66 0.53 0.31 0.60
Crops – sales 0.66 0.64 0.58 0.71 0.52 0.58 0.54 0.31 0.47
Crops – gifts 0.59 0.57 0.69 0.69 0.53 0.52 0.58 0.40 0.53
Crops – autoconsumption 0.80 0.78 0.64 0.84 0.81 0.76 0.74 0.45 0.68
Livestock – sales 0.25 0.17 0.64 0.49 0.45 0.38 0.58 0.75 0.66
Livestock – gifts –0.03 0.01 0.46 0.15 0.24 –0.04 0.26 0.25 0.25
Livestock – autoconsumption –0.33 –0.39 0.28 –0.06 –0.10 –0.22 0.11 0.22 0.03
Total Production 0.15 0.10 0.63 0.39 0.34 0.20 0.47 0.56 0.48
Crops – expenses 0.60 0.66 0.63 0.70 0.57 0.59 0.53 0.30 0.55
Crops – expenses (%) –0.19 0.06 –0.14 –0.31 –0.17 0.02 –0.26 –0.09 0.06
Livestock – expenses 0.16 0.19 0.20 0.20 0.41 0.26 0.32 0.58 0.54
Livestock – expenses (%) 0.28 0.35 –0.33 0.01 0.18 0.13 –0.11 –0.16 0.00
Total production expenses 0.44 0.48 0.52 0.60 0.57 0.40 0.47 0.56 0.64
Total production expenses (%) 0.34 0.51 –0.10 0.18 0.16 0.39 –0.05 0.01 0.25
Value of autoconsumption 0.24 0.25 0.39 0.44 0.34 0.28 0.37 0.56 0.52
Household purchases 0.05 0.19 0.17 0.20 0.29 0.09 0.16 0.34 0.38
Dependence on market –0.42 –0.38 –0.75 –0.60 –0.44 –0.45 –0.58 –0.58 –0.59
Receipt of gifts 0.31 0.37 0.51 0.50 0.39 0.28 0.37 0.43 0.50
Insertion into markets 0.51 0.57 –0.10 0.26 0.34 0.35 0.11 0.42 0.44
Total expenses 0.10 0.19 0.28 0.29 0.28 0.11 0.21 0.39 0.40
Taux de couverture 0.39 0.29 0.56 0.73 0.52 0.51 0.58 0.53 0.57
Mean rainfall (mm) 0.91 0.85 0.60 0.75 0.81 0.69 0.69 0.31 0.56
Population.density 0.04 –0.12 0.36 0.32 –0.05 0.21 0.24 0.04 0.01
Legend: Positive > +0.75 +0.50 to 0.74
Level of correlation
Negative < –0.75 –0.50 to –0.74
A.D. Sparrow, A. Traoré

7. Some aspects of platform functionality are positively correlated
with level of sales of animals or animal products (meat and/or milk),
but most aspects of functionality are not correlated with aspects of
animal production (Table 6). There is a non-significant tendency for less
functional platforms to have members with more auto-consumption of
animal products in the home (Table 6).
Different variables describing platform functionality are correlated
with the bilan simplifié variables describing dependence on purchases
from the market and sales of products into the market – respectively
negative and positive correlations (Table 6).
Most aspects of platform functionality are positively correlated with
taux de couverture, the number of month equivalents per year that a
local producer-household is able to feed itself (Table 6).
Platform functionality is not correlated with local population den-
sity, but strongly positively correlated with mean annual rainfall
(Table 6). This is probably an indirect effect, expressed through crop
and pasture productivity, all aspects of which are positively correlated
with rainfall (rs = 0.57 to 0.66, Table 6).
4. Discussion
Our results show very limited support for our first hypothesis, that
functionality of innovation platforms is affected by numbers of activ-
ities and participants. There is a positive correlation only between the
functionality scores and number of meetings, not any other measure of
activity or participation. However, there is very strong support for our
second hypothesis, that functionality of innovation platforms is affected
by the production system and socio-economic circumstances of local
producer-members, with a wide range of correlations between func-
tionality and well-being indicators. There is also partial support for our
third hypothesis, that functionality of innovation platforms is affected
by regional characteristic, with a positive correlation between func-
tionality indicators and rainfall, but not with population density.
In general, according to our analyses, innovation platforms are self-
assessed by their members to be most functional in higher-rainfall
zones, where participants already have larger farms (by area), already
have diversified production systems within which they engage more in
crop and forage production (with its attendant costs for inputs), have
higher plant yields per unit area, feed forage more to animals, sell more
animal products into the marketplace, have higher overall economic
through-put, and have higher taux de couverture i.e. innovation plat-
forms are judged to be most helpful by those who already have a
modicum of success in engaging with markets for profit and have re-
latively high household well-being For these platform participants,
meetings are useful, especially if the innovation platform has small
membership, but training and formal actions (in particular) are not
useful. Perhaps members are able to conduct all their negotiations in
support of innovation efficiently and effectively during meetings, and
are spared the effort of less efficient extra-mural follow-up. In general,
few women are involved in any of these activities, even though the
members of the platform judge gender to be respected and incorporated
well i.e. there is an apparent contradiction between word and action
with regard to gender.
Innovation platforms are judged to be much less functional in low
rainfall areas, with poorer participants, who have small investments in
(and lower production of) crops and forage, limited sales of both plant
and animal products into the marketplace, high dependence on the
marketplace for purchase of their household food needs, low overall
economic through-put, and low taux de couverture i.e. the people who
most need help to increase their well-being and lift their households out
of poverty are the people who perceive innovation platforms to be less
functional. Because the self-evaluation approach was subjective and
relative rather than absolute, it is possible that these poorer platforms
members had different (and greater) expectations of their first en-
counter with an innovation platform and these expectations were not
met by the implementation of the platforms in this project. It is also
possible that poorer platform members assess their production systems
to involve more dimensions and higher levels of risk and/or that they
can access fewer resources to sufficiently reduce risk exposure (see
Ayinde et al., 2014).
It has been assumed that IAR4D and other participatory approaches
that address strengthening of innovation systems with diverse mem-
berships through building strengthened formal or informal institutions
are better for agricultural development than traditional linear models of
technology transfer (Clark, 2002, Sumberg, 2005, Hawkins et al.,
2009), and assessments of the potential for better coordination along
value chains have been conducted (e.g. Chiuri et al., 2013), but this
hypothesis has rarely been tested rigorously. Many case studies that
provide evidence of success (e.g. Adekunle et al., 2012; Sanyang et al.,
2013) are not the result of controlled experiments in which participa-
tory approaches are directly compared with a non-participatory control.
In addition, it is difficult to know whether the lack of negative case
studies is due to the absence of reporting in the literature more than the
absence of actual failures on the ground. Furthermore, success has most
often been reported in strictly narrative terms, using direct quotes from
enthusiastic participants, rather than in terms of improved quantitative
outcomes, such as absolute agricultural production or the percentage of
production that is delivered into markets.
A recent meta-analysis by Nkonya et al. (2013) provides a rigorous
analysis of the effectiveness of innovation platforms. In that study, sites
with innovation platforms are compared with sites without platforms.
They could not show any statistically significant effect, although cau-
tioned about interpretation of the results given the relative short
timeframe of implementation of the majority of the cases examined.
With a similar statistical contrast, Pamuk et al. (2014) found that in-
novation platforms increased the adoption of crop management in-
novations, but not soil water, soil fertility or post-harvest management
innovations, but also expressed caution about the relatively short
period of platform activity at times of assessment. In terms of actual
well-being outcomes, Pamuk and van Rijn (2015) used a food con-
sumption score (FCS) that combined aspects of food quantity and nu-
tritional quality, and found that FCS was correlated only with institu-
tional and human capacity-building, and not with any of several other
principle associated with “IAR4Dness”, including measures of partici-
pation and social capital.
Whilst our study has not been able to address all of these concerns
identified in the literature, we have analysed the extent of variation in
functionality amongst the nine innovation platforms established and
explored possible causes of that variation – and the results show a level
of nuance in perceived value of IAR4D that has not been reported be-
fore.
Our results resonate strongly with the findings of Olarinde et al.
(2013). This economic impact evaluation of research for development
trials in northern Nigeria and Niger found a significant benefit of IAR4D
on production and profitability of cereals, legumes and fruit/vegetable
crops compared with conventional technology transfer, once producers
fully engaged with innovation platforms. However, the potential benefit
of IAR4D was not realised because the propensity of producers to par-
ticipate in IAR4D was strongly positively correlated with household
size, availability of boreholes in villages, and number of extension visits
and participation in research in the five years prior, and negatively
correlated with residence in the Sahel relative to the higher-rainfall
Sudanian and northern Guinean savannas. The authors imply that in
their study area, past investment in extension and agricultural research
has been concentrated in areas with higher potential for commercial
agricultural, generating a reinforcing feedback that has largely ex-
cluded poorer Sahelian households.
The dichotomy between richer and poorer producers, in terms of
their ability to engage with markets and value chains, has been ob-
served elsewhere in Africa. For example, in Kenya, Krishna et al. (2004)
and Burke et al. (2007) concluded that small-scale dairy production,
with an orientation towards markets, was a strategy to stay above the
A.D. Sparrow, A. Traoré

8. poverty line rather than a strategy to escape from poverty. Indeed 80%
of the households in Kenya that had escaped from poverty attributed
their improved situation to non-agricultural businesses and employ-
ment, rather than development of their farms (Kristjanson et al., 2010),
and development of agriculture and livestock production was often a
consequence of the improved economic circumstances of the house-
hold, rather than the inverse (Harris and Orr, 2014). Even in the study
of Nkonya et al. (2013), in which innovation platforms were observed
to benefit animal production in Rwanda, one can now wonder about
cause and effect: what was cause and what was effect?. Was IAR4D
more profitable for the producers who already had substantial income
from selling meat and milk into the marketplace, as in the current
study? The dichotomy between richer and poorer producers is not
limited to animal production systems. Jayne et al. (2010) concluded
that in Kenya, Mozambique and Zambia, 50% of the maize sold into the
markets had been produced by only 1.0–2.2% of the producers, de-
pending on country. The majority of producers were caught in a vicious
cycle in which low production inhibits development of markets, which
in turn limits the capacity of smallholders to use productive agricultural
technologies in an economically sustainable manner, thereby reinfor-
cing subsistence agriculture.
5. Conclusions
Regardless of whether the correlations of innovation platform
functionality with indicators of rural household productivity and well-
being reflect different functionality in an absolute sense or different
expectation of function, and given our underlying assumption that
perceived functionality is an indicator of potential for future adoption
of innovation, our results present a challenge for IAR4D concept pro-
moters such as CORAF/WECARD, as well as for APESS, the NGO which
coordinated this project and whose members who made up a significant
number of the participants. As implemented, innovation platforms ap-
pear to be most effective for what might be termed “the rural middle
class”. Perhaps this reflects the preoccupation of IAR4D and innovation
platforms, as defined by and implemented through FARA and CORAF/
WECARD, with market value chains and market insertion. If innovation
platforms are to be more functional for, and perceived as effective by,
the poorer smallholders in the semi-arid regions, their implementation
and conception seem to require modification. Alternatively, IAR4D
should be accepted as a relevant approach for participatory develop-
ment of markets and value chains, and a different conception of par-
ticipatory development is required for poorer smallholders whose ex-
pectations are driven by their perception that they must achieve self-
sufficiency for their families (taux de couverture ≥ 12 months) before
they attempt to develop a market orientation. In essence, the assump-
tion that value chain connectivity is the key limitation to improved
productivity, and thus the appropriate entry point to initiate innovation
platforms, may be invalid in semi-arid regions – and a less prescriptive
entry point is required to catalyse participatory action research to di-
agnose the actual limiting factors in the local systems.
It is likely that smallholder producer-households that do not reliably
meet self-sufficiency are enmeshed in a complex of limiting factors that
epitomize “wicked problems”. The term “wicked problem” derives from
the social sciences and has particular significance in the global research
in economic and social development (e.g. Rittel and Webber, 1973;
Peterson, 2009). In wicked problems, a set of reinforcing feedback loops
maintain an undesirable status quo in terms of economic and/or social
conditions of concern. Changing to a more desirable situation requires
changing many drivers or system parameters simultaneously e.g. si-
multaneously changing multiple policies and institutions. For example,
Guo et al. (2013) show how lack of use of inorganic fertilisers in the
more remote parts of East Africa is an entirely rational response to
economic constraints caused by the costs of transport of fertiliser from
port to farm and of surplus grain production from farm to market, as a
result of the combination of poor transport infrastructure and cross-
border tariffs. From the perspective of IAR4D implemented through
market-oriented innovation platforms in our study, the principle focus
on building value-chain connectivity may be addressing only one of the
suite of factors that constrain households below taux de couverture of
12 months, and failing to address a range of other absolute constraints
and risk-based concerns of producers (see Ayinde et al., 2014), thereby
perpetuating confinement to a “poverty trap” (Nelson, 1956; Carter and
Barrett, 2006).
For APESS, the results are a challenge because they can be inter-
preted to suggest that APESS' key stakeholder group, livestock families
in the Sahel, may not benefit unequivocally from an IAR4D approach
focused on commercialisation of meat and milk products into the
market. The most functional of these market-orientated innovation
platforms lie within the subhumid zone of West and Central Africa, the
higher-rainfall half of the Sudanian savannas, and thus on the southern
fringes of the existing mandate and strategic focus of APESS. This leaves
APESS with a choice between (1) seeking different approaches better
suited to its current stakeholders or (2) expanding its mandate and
stakeholder group southwards into higher rainfall zones in which small-
to medium-scale agri-business for meat and milk production has greater
potential.
Acknowledgements
The research was conducted as part of a project funded by the
Australian Government's Department of Foreign Affairs through the
African Food Security Initiative (DFAT Agreement Number 57865), a
joint venture of the Commonwealth Scientific and Industrial Research
Organisation (CSIRO) and the West and Central African Council for
Agricultural Research and Development (CORAF/WECARD) based in
Dakar, Senegal. Human research ethics approval was sought and ob-
tained from the CSIRO Social Science Human Research Ethics
Committee (application 026/15). The authors acknowledge the help
and assistance provided by the staff of APESS in the Ouagadougou
headquarters, and in regional centres in Burkina Faso, Cameroon and
Senegal, as well as all the members of the nine innovation platforms
which participated in this study. Thoughtful comments received from
Yiheyis Maru, Richard Stirzaker, Monica van Wensveen and two
anonymous reviewers have been invaluable in preparing this paper.
References
Adekunle, A.A., Fatunbi, O., 2012. Approaches for setting-up multi-stakeholder platforms
for agricultural research and development. World Appl. Sci. J. 16, 981–988.
Adekunle, A.A., Ellis-Jones, J., Ajibefun, I., Nyikal, R.A., Bangali, S., Fatunbi, O., Ange, A.,
2012. Agricultural innovation in sub-Saharan Africa: experiences from multiple-sta-
keholder approaches. In: Forum for Agricultural Research in Africa (FARA), (Accra,
Ghana).
Adjei-Nsiah, S., Klerkx, L., 2016. Innovation platforms and institutional change: the case
of small-scale palm oil processing in Ghana. Cah. Agric. 25 (6), 65005.
Anttiroiko, A.V., 2016. City-as-a-platform: the rise of participatory innovation platforms
in Finnish cities. Sustainability 8 (9), 922.
Ayinde, O.E., Muchie, M., Olatinwo, K.B., Adenuga, A.H., Oyewole, A.E., 2014. Analysis
of adoption and risk perception in innovation system in Kwara state Nigeria: a case
study of improved rice. Int. J. Agric. Sci. 10, 816–822.
Barbedette, L., 2013. Ce qu'ont appris sur les exploitation familiales des membres de
l'APESS: Synthèse d'étape, 13 tendances. Association pour la Promotion de l'Elevage
au Sahel et en Savane, Ouagadougou.
Burke, W.J., Jayne, T.S., Freeman, H.A., Kristjanson, P., 2007. Factors Associated with
Farm Households' Movement into and out of Poverty in Kenya: The Rising
Importance of Livestock. (Michigan State University Development Working Paper
No. 90).
Byerlee, D., 1998. The search for a new paradigm for the development of national agri-
cultural research systems. World Dev. 26, 1049–1055.
Carter, M.R., Barrett, C.B., 2006. The economics of poverty traps and persistent poverty:
an asset-based approach. J. Dev. Stud. 42 (178–1).
Chambers, R., Pacey, A., Thrupp, L.-A. (Eds.), 1989. Farmer First: Farmer Innovation and
Agricultural Research. IT Publications, London.
Chiuri, W., Birachi, E., Buruchara, R., Adekunle, W., Fatunbi, O., Pali, P.N., Wimba, B.,
Bizosa, A., Nyamurinda, B., Nyamwaro, S.O., Habumugisha, P., Tuyisenge, J.,
Bonabana-Wabbi, J., Karume, K., Kasenge, V., Kamugisha, R., Fungo, B.,
Tumwesigye, S., Kato, E., Nkonya, E., 2013. Market access for agro-enterprise di-
versity in the Lake Kivu Pilot Learning Site of the sub-Saharan Africa Challenge
A.D. Sparrow, A. Traoré

9. Programme. Afr. J. Agric. Resour. Econ. 8, 120–134.
Clark, N.G., 2002. Innovations systems, institutional change and the new knowledge
market: implications for Third World agricultural development. Econ. Innov. New
Technol. 11, 353–368.
Guo, Z., Koo, J., Wood, S., 2013. Fertiliser Profitability in East Africa: A Spatially Explicit
Policy Analysis. Paper Presented at the Global Geospatial Conference, Addis Adaba,
4–8 November 2013.
Hall, A., 2011. Putting agricultural research into use: lessons from contested vision for
innovation. In: UNU-MERIT Working Paper Series no. 2011-076. United Nations
University-Maastricht Economic and Social Research and Training Centre on
Innovation and Technology, Maastricht, The Netherlands.
Hall, A., Janssen, W., Pehu, E., Rajalahti, R., 2006. Enhancing agricultural innovation:
how to go beyond the strengthening of research systems. World Bank,
Washington DC.
Hall, A.J., Sivamoha, M.V.K., Clark, N.G., Taylor, S., Bockett, G., 2001. Why research
partnerships really matter: innovation theory, institutional arrangements and im-
plications for developing new technology for the poor. World Dev. 29, 783–797.
Hall, A.J., Yoganand, B., Sulaiman, R.V., Raina, R.S., Prasad, C.S., Naik, G.C., Clark, N.G.
(Eds.), 2004. .), Innovations in innovation: reflections on partnership, institutions and
learning. ICRISAT, Patancheru, Andra Pradesh, India.
Hall A.J., Yoganand B., Sulaiman R.V., Raina R.S., Prasad C.S., Naik G.C. and Clark, N.G.,
(Eds.), Innovations in innovation: reflections on partnership, institutions and
learning. ICRISAT, Patancheru, Andra Pradesh, India, 2004.Harris, D., Orr, A., 2014.
Is rainfed agriculture really a pathway from poverty? Agric. Syst. 123, 84–96.
Hawkins, R., Heemskerk, W., Booth, R., Daane, J., Maatman, A., Adekunle, A.A., 2009.
Integrated Agricultural Research for Development (IAR4D): A Concept Paper for the
Forum for Agricultural Research in Africa (FARA) Sub-Saharan Africa Challenge
Programme (SSA CP). FARA, Accra, Ghana.
Hounkonnou, D., Kossou, D., Kuyper, T.W., Leeuwis, C., Nederlof, C.S., Roling, N., Sakyi-
Dawson, O., Traore, M., van Huis, A., 2012. An innovation systems approach to in-
stitutional change: smallholder development in West Africa. Agric. Syst. 108, 74–83.
Jayne, T.S., Mather, D., Mghenyi, E., 2010. Principal challenges confronting smallholder
agriculture in Sub-Saharan Africa. World Dev. 38, 1384–1398.
Krishna, A., Kristjanson, P., Radeny, M., Nindo, W., 2004. Escaping poverty and becoming
poor in 20 Kenyan villages. J. Hum. Dev. 5, 211–226.
Kristjanson, P., Mango, N., Krishna, A., Radeny, M., Johnson, N., 2010. Understanding
poverty dynamics in Kenya. J. Int. Dev. 22, 978–996.
McIntyre, B.D., Herren, H.R., Wahungu, J., Watson, R.T., 2009. Agriculture at a
Crossroads: International Assessment of Agricultural Knowledge, Science and
Technology for Development (IAASTD) Global Report. Island Press, Washington DC.
Mulema, A.A., Mazur, R.E., 2016. Motivation and participation in multi-stakeholder in-
novation platforms in the Great Lakes Region of Africa. Community Dev. J. 51,
212–228.
Nederlof, S., Wongtschowski, M., van der Lee, M. (Eds.), 2011. Putting Heads Together:
Agricultural Innovation Platforms in Practice. Bulletin Vol. 396 KIT Publishers,
Amsterdam.
Nelson, R.R., 1956. A theory of the low-level equilibrium trap in underdeveloped
economies. Am. Econ. Rev. 46, 894–908.
Nkonya, E., Kato, E., Oduol, J., Pali, P., Farrow, A., 2013. Initial impact of integrated
agricultural research for development in East and Central Africa. Afr. J. Agric. Res.
Econ. 8, 172–184.
Ojasalo, J., Tahtinen, L., 2016. Integrating open innovation platforms in public sector
decision making: empirical results from Smart City research. Technol. Innov. Manag.
Rev. 6 (12), 38–48.
Olarinde, L.O., Ayanwale, A.B., Oladunni, O.A., Kaku, S.N., Adekunle, A.A., Fatunbi, O.,
2013. The integrated agricultural research for development (IAR4D) and its impacts
on marketed crops: data analysis of the Kano-Katsina-Maradi pilot learning site. In:
Invited Paper Presented at the 4th International Conference of the African
Association of Agricultural Economists, September 22–25, 2013, (Hammamet,
Tunisia).
Pamuk, H., van Rijn, F., 2015. Heterogeneity in agricultural innovation systems' impact
on food security: evidence from sub-Saharan Africa. In: Proceedings of the 29th
International Conference on Agricultural Economists, Milan, Italy, August 9–14, pp.
2015.
Pamuk, H., Bulte, E., Adekunle, A.A., 2014. Do decentralized innovation systems promote
agricultural technology adoption? Experimental evidence from Africa. Food Policy
44, 227–236.
Peterson, H.C., 2009. Transformational supply chains and the 'wicked problem' of sus-
tainability: aligning knowledge, innovation, entrepreneurship, and leadership. J.
Chain Netw. Sci. 9, 71–82.
Pingali, P.L., 2012. Green revolution: impacts, limits, and the path ahead. PNAS 109,
12302–12308.
Pinstrup-Andersen, P., Watson II, D.D., 2011. Food Policy for Developing Countries: The
Role of Government in Global, National, and Local Food Systems. Cornell University
Press, Ithaca, NY.
Poulton, C., 2014. Democratisation and the political incentives for agricultural policy in
Africa. Dev. Policy Rev. 32, s101–s122.
Pretty, J., Toulmin, C., Williams, S., 2011. Sustainable intensification in African agri-
culture. Int. J. Agric. Sustain. 9, 5–24.
Rajalahti, R., Janseen, W., Pehu, E., 2008. Agricultural innovation systems: from diag-
nostic toward operational practices. In: Agricultural and Rural Development
Discussion Paper 38. World Bank, Washington DC, USA.
Rittel, H.W.J., Webber, M.M., 1973. Dilemmas in a general theory of planning. Policy. Sci.
4, 155–169.
Sanyang, S., Muluh, G.A., Kuiseu, J., Taonda, S.J.B., Kafando, A., Stirzaker, R., 2013.
Innovation Platform for Technology Adoption (IPTA): maize value chain in Burkina
Faso. In: van Huis, A., Röling, N., Hounkonnou, D., Youdeowei, A. (Eds.), Promoting
African Smallholder Agricultural Production Systems through Institutional
Innovation: Proceedings of the CoS-SIS International Workshop, Elmina, Ghana,
October 22–26, 2012. Wageningen University, Wageningen, pp. 19–32.
Sanyang, S., Pyburn, R., Mur, R., Audet-Bélanger, G. (Eds.), 2014. Against the Grain and
to the Roots: Maize and Cassava Innovation Platforms in West and Central Africa. LM
Publishers, Arnhem.
Sanyang, S., Taonda, S.J.B., Kuiseu, J., Coulibaly, N., Konate, L., 2016. A paradigm shift
in African agricultural research for development: The role of innovation platforms.
Int. J. Agric. Sustain. 14, 187–213.
Schut, M., Klerkx, L., Sartas, M., Lamers, D., McCampbell, M., Ogbonna, I., Kaushik, P.,
Atta-Krah, K., Leeuwis, C., 2015. Innovation platforms: experiences with their in-
stitutional embedding in agricultural research for development. Exp. Agric. 52,
537–561.
Siziba, S., Nyikahadzoi, K., Nyemeck, J.B., Diagne, A., Adewale, A., Oluwole, F., 2013.
Estimating the impact of innovation systems on maize yields: the case of IAR4D in
southern Africa. Agrekon 52, 83–100.
Sumberg, J., 2005. Systems of innovation theory and the changing architecture of agri-
cultural research in Africa. Food Policy 30, 21–41.
Thiele, G., Devaux, A., Reinoso, I., Pico, H., Montesdeoca, F., Pumisacho, M., Andrade-
Piedra, J., Velasco, C., Flores, P., Esprella, R., Thomann, A., Manrique, K., Horton, D.,
2011. Multi-stakeholder platforms for linking small farmers to value chains: evidence
from the Andes. Int. J. Agric. Sustain. 9 (3), 423–433.
Triomphe, B., Floquet, A., Kamau, G., Letty, B., Vodouhe, S.D., Ng'ang'a, T., Stevens, J.,
van der Berg, J., Selemna, N., Bridier, B., Crane, T., Almekinders, C., Waters-Bayer,
A., Hocde, H., 2013. What does an inventory of recent innovation experiences tell us
about agricultural innovation in Africa? J. Agric. Educ. Ext. 19, 311–324.
A.D. Sparrow, A. Traoré