Volumes 60–63, December 2012, Pages 115–121
Diagnosing the scope for innovation: Linking smallholder practices and institutional context
Diagnostic research in support of innovation
- http://www.sciencedirect.com/science/article/pii/S1573521412000334
- doi:10.1016/j.njas.2012.06.008
- Get rights and content
Abstract
This
paper analyses the diagnostic studies of this special issue to
underline their function in probing the opportunity for transformational
change and the potential of socio-technical novelties in such processes
of change. The studies document the ability of poor, illiterate farmers
to create novelties, and, sometimes, to develop the institutional and
informational capacities needed to support and disseminate the
novelties. The studies also show that it is not easy for farmers to change
‘the rules of the game’ that are encoded in routine practices, the
relationships amongst organizations, normative behaviours, informal or
formal regulations, bylaws and so on. The general methodologies of the
studies documented in this special issue are discussed and their
potential, strengths and weaknesses are indicated. The studies might not
have yielded significant policy lessons but they have provided
well-grounded insights into processes of sense-making, contextually
relevant criteria for and processes of assessment, and into the
initiation of change. They have developed sufficient initial
understanding for building and informing institutional innovation. How
successful (or not) that process has been will be analysed in later
reports from the CoS-SIS programme.
Keywords
- Action research;
- Diagnostic study;
- Socio-technical novelty;
- Institutional change
1. Introduction: the diagnostic metaphor in innovation in practice and research
Diagnosis
refers to the ancient practice of determining the nature of a disease,
malfunction or disorder; the identification of symptoms is the key
process in the analysis, from which causation is inferred. The emphasis
in medical practice is placed on ‘what is wrong’. The diagnosis of
systemic failures has been common in development studies, giving rise to
a wide variety of prescriptive remedies. Where the treatment has
appeared to produce improvement or cure, success has been claimed for
the treatment, although the pathways between intervention and outcome
often were left un-described in such studies. The ‘diagnostic study’ as
an approach to probing the opportunity for transformational change, and
of the potential of socio-technical novelties in such processes, is of
more recent origin and derives from diverse fields of knowledge. These
include studies of entrepreneurship, and policy discussion of whether
entrepreneurial ability can be developed through appropriate
interventions.
The work of
researchers such as Anil Gupta or Robert Chambers have documented the
ability of even poor, illiterate farmers to create novelty, and
sometimes also to develop the institutional and informational capacities
needed to support and spread their novelties at and beyond local level.
In this set of papers, too, farmer-developed novelties are recorded.
However, what is noteworthy is that in the process of revealing such
novelties, the studies come to probe deeper questions. Yemadje et al.,
for instance, report farmers’ role in developing a system of oil palm
fallow on the Adja plateau, Benin, under population pressure. The study
raises the question of whether the system has become an arena for
resolving or intensifying competing claims on land. Tenure issues are
involved and the issue of the balance of power between those who own
land (in a practical and not necessarily legal sense) and those who
borrow land cannot be avoided. This leads to the realization that an
analysis based on the social category ‘farmers’ may be unhelpful because
the category bundles together heterogeneous interests that are better
disentangled. Other authors in this issue make similar efforts to
disaggregate who in fact is doing what in the situation examined.
Amankwah et al., for instance, distinguish the novelties developed by
‘positive deviants’ in the small ruminant sector in northern Ghana;
Doumbia et al. analyse the development of dairying by particular farmers
and an entrepreneur, in defiance of the prevailing ‘rules of the game’
set by the Office du Niger, Mali.
However, evidence is much scarcer that farmers alone are able to change
‘the rules of the game’ that are encoded in routine practices, the
relationships amongst organizations, normative behaviours, informal or
formal regulations, bylaws and so on. The dominant ‘rules of the game’
tend to be set and maintained at other levels of governance than the
local, and by other actors, who may overtly act in their own rather than
farmers’ interests. Moreover, in any one locality the assemblage of
institutions may be made up of discordant elements that are the outcome
of the accidents of history rather than design, an assemblage that no
single actor commands or understands in its entirety. This set of
diagnostic studies represents a first step in building an analysis and
shared understanding of ‘what is wrong’ in the assemblage. CoS-SIS’
intention is that the diagnoses help to identify novel products,
technologies, and practices but also behaviours, relationships, rules,
norms and other features of the institutional landscape, that would
offer stakeholders a potential for positive change. In framing the
studies in this fashion, a number of methodological issues arise that we
consider further below.
The
studies in addition offer empirical information related to mid-range
theories about innovation processes. They thereby contribute to three
strands of practice and research. For practitioners, they offer insights
into how an appropriate institutional understanding can be developed of
a given socio-technical problem in a specific context [1].
Secondly, they address preliminary ‘what’ and ‘how’ questions that
arise in relation to the development of novelties in local processes of
system innovation [2]. Thirdly, they contribute to the stream of studies that are feeding the ‘new analytics’ [3] of governance. This paper reflects further towards the end of this paper on these three substantive issues.
The
issues addressed are not, strictly speaking, new. They were articulated
in the 1970s as fundamental and pressing agricultural development
concerns by, amongst others, Uma Lele and her team at John Hopkins
University, Baltimore, USA, and later in the World Bank [4], ORSTOM, Paris, and its colleagues throughout francophone Africa [5],
and by the Agricultural Administration Unit's worldwide network of
university- and field-based researchers, development administrators,
policy-makers and practitioners (based at the Overseas Development
Institute, London) [6], [7] and [8].
The expertise and understanding vested in the then dominant centres of
power to command the development discourse, of how governance,
technologies and institutions actually work in specific contexts and in
fine detail, have faded with the passing of the first generations of
those intimately involved in the late colonial and post-colonial
development experiences. More recent discourses, based on assumptions of
market-led or market-mediated economic growth and development
trajectories, have focused attention on aggregated economic
prescriptions for the development of agriculture and food systems and
agro-ecosystem management. Local and national institutions’ functions
and their actual functioning in a particular context, in these
narratives are occluded because they are assumed to converge over time
under compelling market imperatives. Furthermore, because in competitive
markets it is often technical and product innovation that give
enterprises an edge, over the last decade the discourse has narrowed
consideration of managed change and induced innovation still further, to
a focus on issues of profit-driven technology and product development,
promotion and market share, and associated organizational changes.
Innovation, in this perspective, is the outcome of entrepreneurial
creativity, drawing on the purposeful organization of systems of support
and information flows within and between private commercial, business
and public actors.
However,
countervailing processes, such as (the growing) consumer preference for
products with traceable non-market qualities (such as health benefits,
food safety, the social well-being of producers, environmental
sustainability), as well as persistent poverty, increasing malnutrition,
observed and forecast climate change impacts on agro-ecosystems,
adverse natural resource trends and intensifying competition for these
resources, are challenging numerous communities of research and practice
to question again the inter-relations amongst institutions and their
context [9].
By examining once more the potential for system innovation by means of
purposeful institutional innovation in particular socio-technical
contexts, a wider range of stakeholders is brought into view. The study
of innovation as constituted in multi-dimensional relational processes
reveals the possibility of multiple viable entry points and pathways for
purposeful change. In particular, the institutions in which existing
relationships are encoded and normalized, present themselves as gaps,
constraints to and opportunities for change.
2. Methodological issues
2.1. Boundary decisions
The
scope and focus of the diagnostic studies were derived from
pre-analytic choices made at national level and through priority-setting
by local stakeholders (Introduction, this issue). Those interested in
studying processes of change, and in intervening in those processes
through the way in which research is conducted, acknowledge the
importance of positioning ‘boundary-setting’ as a conscious choice that
helps to make a ‘messy problem’ tractable [10].
There is no objective way to judge if the choices made are ‘right’ or
‘wrong’; what is important is that key stakeholders are brought together
in a process in which they themselves assume joint responsibility for
the choices made. Diagnosis of a situation of interest can itself serve
as a step in the process of bringing diverse actors together to consider
and deal with the set of inter-relationships constituted in
multi-layered institutions encoded and embedded for instance in the
political rules of the game, local practices and also biophysical states
and flows. Osei-Amponsah's diagnosis focusing on the quality of palm
oil produced by local processors and Quarmine's diagnosis of the cocoa
sector, or Akpo's of the oil palm seed system, bring these constitutive
arrangements clearly into focus as the actors seek to develop novelties
in products, bylaws, regulatory norms, practices and local
organizational arrangements, and thereby to find or create the
institutional spaces for change across their respective domains.
What
binds such elements together is people's experience of the situation,
around which interactions are organized in a social space or, following
Foucault [11], a dispositif. Yemadje's diagnosis of the socio-technical dispositif
constituted by tenure arrangements, or Totin's analysis of the
inter-relationships between socio-technical novelties and institutions
in water management, suggest that the drawing together of the actors in
order to improve a situation that is perceived by the stakeholders
themselves as problematic is not necessarily a process that
spontaneously arises. This observation in the past has led to the
characterization of rural societies as locked into tradition, inert,
fatalistic or resistant to change. The evidence and analysis of CoS I,
and numerous other studies over the last five or more decades, is that
rural societies, and individuals within them, are as alert and eager for
change as anyone else when opportunities are accessible, profitable,
and realizable. What makes the difference, then? It is the hypothesis of
CoS-SIS that purposeful effort is needed to change in particular the
institutional constraints that restrain opportunity for small scale
producers, a hypothesis that we pursue in the next paragraph.
Sidibé’s
paper suggests that the techno-organizational interventions described –
that were designed to promote innovation in the sheanut sector of Mali –
can be said to be ‘successful’ only if diagnostic and practice
boundaries are narrowly drawn. The innovations that occurred may be said
to have created longer term and higher scale difficulties in so far as
they appear to have locked the innovation and innovators into
export-oriented market niches that offer only limited prospects of wide
access to the opportunity, or for effective co-ordination along the
value chain, or for buoyant market growth. Togbé et al., with respect to
the cotton sector in Benin, and Totin et al., with respect to the
inter-relation between arrangements for managing rice and water in
Benin, similarly provide evidence of ‘institutional traps’ that limit
the power of proven novelties to effect wider change, institutions that
are held in place by the prevailing pattern of interactions at and
between multiple levels. The weight of the evidence in the papers in
this issue indicate that, without a change in relationships among key
actors, socio-technical innovation at the farm level will continue to be
constrained. The contribution by Osei-Amponsah et al. explicitly draws
attention to the fact that key actors have not been included in the
dominant socio-technical arrangements impacting palm oil quality in
Ghana. In consequence of their diagnosis, in this study the design of
their innovation pathway has begun with the creation of more effective
relationships amongst a wider set of stakeholders.
By
widening the boundaries of analysis to include institutions and the
interactions among socio-technical and organizational arrangements at
and between multiple levels, this set of papers demonstrates that there
exist in each context several viable pathways for innovation, even – as
Amankwah et al. insist – under homogeneous farming system conditions.
The (potential) existence of ‘multiple pathways’ raises the issue of
whom, then, has power to govern the choices that are actually made among
pathways, an issue to which we return towards the end of this paper.
We
further see evidence in these papers of the effort it has taken each of
the researchers, most of whom were trained in the agricultural
sciences, to widen their disciplinary boundaries, to encompass and
interpret ‘the social’, and specifically ‘the institutional’ into their
thinking about technology, and into their researching practice. The
initial scoping studies [12]
led by the post-doc research associates provided for most of the
research team the first indications of the difficulties this poses, as
the researcher moves from the laboratory and experimental field, a world
that the researcher more or less controls, into a world full of
surprises. This is the first indication that ‘purposeful change’ is not a
matter of social engineering but of design of processes that allow for
learning, adaptation, and experiment.
There
is, moreover, another balancing act to perform. The ‘field of study’
here presents itself as constructed by others’ experience of reality. At
the same time, the ‘research object’ – such as an enterprise, value
chain or commodity sector examined in each of these diagnostic studies –
becomes naturalized within the frame of reference of the researcher,
thus once again, to an extent isolating the researcher from the totality
of others’ experience. The identification and characterization of
‘research objects’ that cannot be manipulated or measured under the
disciplines of experiment, places great weight on the accuracy of
observation, in a richly described context. A diagnostic study thus is a
compromise; it preserves a separation between the researcher and the
researched but the boundary acquires what Hatchuel has called a ‘degree
of porosity’ [13], a boundary that the researcher has to negotiate continually.
The
challenges are compounded as the researcher positions him- or herself
in the position of someone who, through their researching practice,
‘intervenes’ in the processes that are observed, thereby helping to
transform the context or the processes observed. This raises profound
questions of knowledge production, to which we return below.
2.2. Cases as evidence of what?
Current
interest in the gold standard of policy research has tended to displace
former interest in case studies that, by means of cross-case comparison
through time or space, might serve to test policy, practical
interventions, or theory. The gold standard has two (sometimes merged)
design features: before/after, with/without analysis of matching cases;
randomized assignment of ‘treatments’ (where the designed intervention
is regarded as a treatment) and, where sufficient numbers of such
studies allow, meta-analysis of the entire data set. CoS-SIS has chosen
not to construct a research pathway based on matching cross-case
comparisons but also not on statistical analysis of randomized
treatments. The programme argues for a middle way appropriate to the
study of ‘situated experiences embedded in specific histories and
responding to the particularities of framing conditions that were
themselves evolving throughout the period of research’ [14].
It follows that the notion that a set of diagnostic studies such as
these are merely anecdotes without collective explanatory power is
rejected a priori. Of what, though, might they claim to provide
evidence?
There are two
dimensions of innovation for which these studies provide explanatory
evidence. One has to do with innovation as a ‘process’, the other with
intervention research as a form of knowledge management and learning
that is constitutive of action [13]; we address the latter in Section 2.3.
Innovation in our view is not an event but a process and as such
requires a research methodology that can track the process over time and
space. In this perspective the diagnostic studies are but a step along a
methodological pathway (further elaborated in [1]). Hoholm and Araujo [15]
review the promises and challenges of research that seek to understand
the mechanisms and dynamics of how innovations emerge, unfold and become
institutionalized. They argue for concentration on observable practices
and actions in actual rather than assumed contexts while cautioning
that analysis of multi-actor perspectives need to be informed by
observation of the interplay among perspectives in situated actions and
practices. Togbé, Doumbia and Kpéra provide evidence of the power of
such a focus.
Hoholm and Araujo [15]
also warn that a presumption that system innovation is necessarily a
multi-level affair settles too many issues in advance of observation.
All the papers in this issue attempt a preliminary sorting of the level
or scale of inter-relationships that either open the space for or
constrain the novelties they describe. In each paper's construction of
institutional hierarchies, where the determining influences on or
drivers of change might lie is revealed as highly diverse – an important
insight for those who seek simplifying and universalizing prescriptions
that ignore the bricolage of institutions described in Yemadje's paper,
or who see innovation as an isomorphic process in a rationally ordered
world.
2.3. Intervention research as a knowledge production model
Hatchuel [13],
speaking of collective action, makes an important distinction:
‘intervention research is not a means of producing knowledge for action
but is rather a constitutive process of action’. This set of studies
straddles this distinction, with Doumbia et al.’s and Amankwah et al.’s
contributions lying more towards the ‘production of knowledge for
action’ end of the spectrum, and Osei-Amponsah et al.’s and Quarmine et
al.’s contributions being more clearly a ‘constitutive process of
action’.
This set of papers
does not aim to develop a theory of collective action but numerous
instances of collective actions are described. The study of collective
action (amongst other points of interest) directs attention to how,
specifically, in the instance observed, a technique is coupled to
efficiency, under criteria of efficiency that make sense to those using,
or proposing to use that technique. The studies in this issue provide
rich insight into the experiences by which such efficiency criteria are
formed. Quarmine et al. point for instance to the specific issues that
arise in this regard when access to information is asymmetrical, and
suggest how such asymmetry might be overcome by bringing about changes
in the relationships among the stakeholders. The papers also highlight
the challenges involved in bringing diverse actors’ appreciation of the
efficiency criteria that matter into a common frame that can be used
both to design innovation processes and assess progress.
The
preliminary diagnoses (that are intended by the CoS-SIS programme to
feed into innovation processes) taken as a whole further reveal the
beginnings of collective effort to construct and reconstruct knowledge
and relations that both organize and represent the actors’ positions and
interdependency [14].
Kpéra's ‘rich description’ of stakeholders’ perceptions of their own
and others’ interests open up to a perceptive analysis of existing
‘pathological’ inter-dependencies but also of the potential for the
collective design of arrangements that might offer greater mutual
satisfaction, based on the co-construction of knowledge. We refer here
also to Snowden [16] and to Blackmore et al. [17],
who suggest that the link between diagnosis, action, and knowledge
management is a matter of ‘probing the space’ for purposeful change
through systematic enquiry in a researching process that involves the
practitioners in shared reflection. The CoS-SIS researchers were engaged
to a varying degree, at the time the work reported here was undertaken,
as stakeholders in this kind of process. For instance, Akpo et al., as a
matter of building stakeholders’ trust in the researchers’ practice,
report the organization of ‘shared reflection’ as one of the very first
steps taken. Others, too, have sought to involve the stakeholders in the
co-production of knowledge and shared reflection, at varying points in
the processes described. Overall, the diagnostic studies reveal that
while the sciences are a necessary component in knowledge production
they are not a sufficient social practice if the purpose is
socio-technical innovation.
3. What we learn from the content and the issues that arise
We
now attempt to draw out what might be learned from the content,
organized under the three themes outlined in the introduction, that lie
close to the heart of innovation practices and studies.
The
concept of innovation directs attention to the purpose for which change
is desired and towards the actions taken to bring this about i.e.,
toward sense-making among the actors involved. Sense-making can take
many forms, including building shared conceptual frameworks, developing
or enforcing norms of behaviour, jointly analysing the messy situation
or problem at stake, and determining purposeful actions for change.
Attention is directed also towards the criteria by which technical,
managerial or organizational novelties are assessed, and how these are
assessed ‘in action’ (rather than ‘in the abstract’ against disciplinary
criteria). A related question in the complex social situations
described (which are quite distinct from the bounded enterprises in
which innovation studies frequently are conducted), where the actors
bring to bear long historical memories and fine socio-cultural
sensitivities, is who initiates the process of purposeful change, for
which reasons.
3.1. Sense-making
The
issues analysed in the diagnostic studies are not new and, as in the
oil palm seed system analysed by Akpo et al., they have been recognized
both by researchers and the stakeholders for a long time. However, the
diagnostic studies not only document the issues: they draw the contours
of the institutional arrangements that allow the problem or mess to
persist, provide evidence to the stakeholders of the seriousness of the
effects and consequences and, importantly, indicate ways forward that
are grounded in the stakeholders’ own perceptions, interests and, in
most of the cases presented here, also in their willingness to engage in
effort to transform the situation for mutual benefit. We note in
addition that although the emphasis in most of the studies is on opening
the institutional space for novelties to break out of their niche to
effect larger scale change in the institutional regime, Amankwah et al.
raise the question of when and where ‘copying’ the discoveries of
positive deviants, based on building capacity and incentives for
farmer-to-farmer learning, might be the more effective strategy.
The
analyses explicitly build on the perspectives of the smallholder
farmers, processors or pastoralists involved. They point to the lack of
effective organizations to represent and defend the interests of local
users of social, technical and natural resources vis-a-vis others in a hierarchy of institutionalized relationships. Bannerjee and Duflo [18]
are among the many scholars who provide evidence of how assumptions
about the opinions and preferences of smallholders cause aid programmes
to fail. Conventional wisdom, formed and circulated among those far from
the field, is shown to be a bad advisor. However, we do not conclude
therefore that smallholders are right or that their views and
experiences necessarily must prevail above all others’. We do insist
that by revealing that they, as all other actors, have reasons for doing
what they are doing, smallholders must be taken seriously as
knowledgeable agents of innovation.
Two
related issues can be considered. The first is that many complex issues
cannot be ‘solved’ on the basis of expertise alone but will be resolved
(if at all) by the stakeholders – who enter into multi-stakeholder
processes with partial and divergent views on the issues that concern
them. In such situations, the development of mutual understanding of the
diversity of views and interests and of the inter-dependency amongst
interests, is of crucial importance, and a necessary concomitant of any
decision by the stakeholders to act in concert to bring about change.
The second is that ‘diversity trumps expertise every time’ [19]
when messy situations are the focus of concern. Laboratory experiments
show that groups composed of diverse actors take better decisions with
respect to such situations than those homogeneously composed of experts.
Agricultural scientists concerned by the lack of uptake of
technological solutions to problems around which neat boundaries have
been drawn, need to consider if they have not, in fact, made a
categorical error in the classification of ‘the nature of the problem’.
No technology will ‘solve’ the competing claims, and the conflicts that
arise from these, documented by Kpéra et al., although, as the authors
point out, rigorously grounded technologies and scientific information
will be needed to inform the collective processes of conflict resolution
and change on which the stakeholders may have embarked.
