A lack of evidence to convince policymakers holds back progress on grassroots innovation in agriculture, say Brigid Letty and Martin Bell.
Grassroots innovation has long been seen by its advocates as playing an important role in smallholder agriculture. Governments have been urged to make much stronger policy commitments to supporting it, in parallel with the support commonly given to innovation based on formally organised research and development (R&D).
But such advocacy has achieved little, not least because it rests on an inadequate evidence base on the impacts and developmental benefits of grassroots innovation as a complement to conventional modes of agricultural innovation.
Without this evidence, it is unlikely that any government will support grassroots innovation.
Innovation by smallholders
The livelihoods of two-thirds of the world's poor rely to some degree on small farms of less than two hectares, which are usually located on marginal land with limited resources and poor infrastructure. The smallholders depend increasingly on innovation as they face accelerating economic and environmental change.
The limitations of conventional R&D-led innovation, which seeks to develop technologies 'on behalf of', or even 'with', smallholder farmers, are becoming clear. Externally developed technologies and practices are often not appropriate, and uptake is frequently very poor.
Grassroots innovation — which in agriculture is substantially controlled and implemented by smallholder farmers — can complement conventional R&D approaches. This is the case especially for innovation processes that combine knowledge from smallholder farmers, other stakeholders, and formal R&D.
A different form of grassroots innovation, known as local innovation, is driven by smallholders without the contribution of external knowledge.
Advocates of participatory (multi-stakeholder) grassroots innovation point to advantages over conventional approaches. These include higher rates of initial uptake of locally appropriate technologies and faster adoption of those that are taken up; bolstered farmers' skills and capabilities to experiment and innovate; and greater relevance to the needs of disadvantaged people.
There are also benefits for conventional R&D: improved farmer feedback for researchers, and lower research costs as farmers take up some of the roles traditionally reserved for researchers. And often, by accelerating the process of developing technology — and not developing technology that is never adopted — grassroots innovation results in less wastage of resources.
But there seems to be inadequate evidence to support these claims.
There are many small studies of grassroots innovation projects, and numerous evaluations for internal management purposes. But these have not been undertaken in a way that allows a cross-analysis of cases to provide common indicators that describe key features of the innovation process.
The way information has been collected does not permit comparative analysis — either between different types of grassroots innovation, or between these and more conventional R&D.
At the same time, national surveys of agricultural R&D, or of science, technology and innovation activities more broadly, do not capture participatory grassroots innovation occurring in smallholder agriculture –— its inputs or outputs, the organisational and institutional arrangements within which it occurs, or socio-economic impacts such as improved capacity of farmers to solve their own challenges through innovation.
There are a few systematic evaluations suggesting that grassroots innovation has advantages over conventional R&D in some circumstances. But the evidence base on the significance of those advantages, or about the circumstances where they are likely to be particularly important, is not adequate.
Coordinating the description and evaluation of grassroots innovation projects using a common set of indicators, is one way to build the evidence base. This information could be compiled by staff at non-governmental organisations (NGOs), and other stakeholders supporting these projects, with a view to engaging with external policy debate, and not just for internal project management.
A collaborative effort by NGOs could analyse existing data on a regular basis.
Such data could cover the history and initial conditions of projects; the actors involved; the inputs to the innovation process including the cost of any institutional strengthening; outputs in terms of technical, social and organisational innovation; and socio-economic impacts such as strengthening of farmers' skills and capabilities.
Our analysis of examples from South Africa, such as the case of a farmer experimenting with a method of producing potatoes, highlights the benefits of grassroots innovation and the type of 'non-conventional' indicators (labour-saving ability, for example) that could be developed. 
Another way of strengthening the evidence base, which has not been used to date, is to extend the scope of organisational, provincial or national surveys of formal R&D and innovation activities to include components that are explicitly concerned with supporting and enhancing grassroots innovation.
Some combination of these information-generating activities should be a substantial step towards developing an evidence base and indicators about the benefits of investing in grassroots innovation.
The challenge is not only developing new mechanisms to generate and publish the kinds of information needed to influence policy, but also to support the right kinds of analysis. We can then begin to build a strong case for committing human and financial resources to support such activities.
Brigid Letty is an agricultural development specialist with the Institute of Natural Resources in South Africa. Martin Bell is emeritus professor at the Science and Technology Policy Research Unit (SPRU) at the University of Sussex in Brighton, United Kingdom. Brigid can be contacted at firstname.lastname@example.org
 Letty, B. et al. An exploration of agricultural grassroots innovation in South Africa and implications for innovation indicator development [668kB]. Working Paper 2012-023, United Nations University–Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT) (2012)