The Kenyan agricultural sector which directly contribute 26% to the GDP, and provide formal and informal employment in the rural areas, has been facing diverse challenges making it weak and uncompetitive. The challenges include non-adoption of improved technologies, weak linkages and interaction between stakeholders, poor infrastructure and unfair competition from open market operations, among others (Kirsten, et. al., 2009). Approaches used in the past to share ‘best bet options’ with the farming communities failed to reduce the gap between on-farm and expected optimal yields (World Bank, 2006). These have ranged from linear technology transfer, farming systems to farmer participatory methodologies. Currently, the innovation systems based approach that is operationalized through agricultural innovation platforms, is in use and it seeks to blend different knowledge sources in a process of co-innovation among multiple stakeholders to enable learning, faster uptake and wider impact. It fosters interactions amongst stakeholders around a common interest with basic reference to value chains (Makini et. al., 2013). The interactions within the platforms results in enhanced innovation capacity amongst stakeholders leading to development of technical, social-organizational and institutional innovations (World Bank, 2006; Kimenye and McEwan, 2014). This is a major deviation from past approaches that focused on technologies and ignored the social and institutional environment surrounding the technologies (Hawkins et. al., 2009; Hounkonnou et. al., 2012). The new approach introduced the involvement of various partners/actors to enhance utilization of agricultural innovations for desired impact and recognition of the importance of responding to consumers and market preferences. Many organizations have embraced the use of innovation platforms in different parts of the country hence there is need to understand their impacts on the livelihoods of the smallholders in order to draw lessons and make recommendations on how to strengthen and increase the effectiveness of this methodology. Impact assessments on innovations have mainly been done to assess the economic efficiency of the innovation and the efficiency analysis (ex-post) combined with surplus approach and is the most common method used. Other methods include the livelihood, comprehensive and multidimensional approaches which go beyond the economic approach to measure goals such as food security, environmental protection and poverty reduction. The assessment methods used have thus been diverse and these also include the cost benefit analysis, economic surplus approach, comprehensive approach, livelihood approach and various combinations of these methods. Anadajeyesekeram et. al. (2007), writing on the impact of science on African agriculture and food security, stated that no single technique or method is sufficient to adequately address impact assessment hence the most appropriate approaches should be a mixture of both qualitative and quantitative methods and active participation of the beneficiaries. In a review of 28 impact assessment studies drawn from all over the world, Barrientos-Fuentes and Berg (2013) concluded that most impact assessments are ex-post although currently ex-ante assessments are increasingly being conducted. They also concluded that besides the direct effects of the innovations, the impact assessments also should bring out the social, institutional, economic and environmental effects. In their study, using the economic surplus model on the impact of sorghum research and development in Zimbabwe, Anadajeyesekeram et al., (2007) concluded that while there was a positive rate of return of 25% on sorghum SV2 variety, omission of administration costs, overhead and depreciation costs and the extended benefit flow period significantly affected the rate of return. In a similar study in Zambia, the estimated rate of return on investment in sorghum research and complementary services ranged from 12% to 19% (Chissi, 2007). La Rovere et al. (2008) used the livelihood approach to measure the impact of new varieties of maize in Mexico and Nepal which is an approach that considers the impact of innovations on farmers’ livelihoods thus shifting the focus from the innovations to people’s livelihoods in various dimensions. In their study, they concluded that there is a likelihood of over or underestimating the impacts if done too early or too late. This was demonstrated by the high impact estimated when the same assessment was conducted immediately after the end of a maize project in Oaxaca and the low impact when repeated much later. Conversely, higher impacts were recorded in a silos project when assessed much later due to farmer to farmer diffusion and presence of partners (La Rovere et al., 2008). Another study on impact of public investment in maize research in Kenya by Karanja (2007), using a two linear regression method concluded that improvement in maize yield and expansion of maize area were due to an increase in research and extension expenditures, spread of hybrid seed, seed programme, use of fertilizer and higher maize producer prices. Various other studies have been conducted in Kenya (Bolo and Makini, 2011; Kavoi et. al., 2013; Kimenye and McEwann, 2011) although their focus was on stakeholder dynamics with none considering the broad outlook on their impacts to the livelihoods of the stakeholders. However, there is still very little research published on the impact assessment of innovation platforms. Most evaluation reports use single case studies to evaluate the impact of a given innovation platform. The current study sought to analyse the livelihood impact of innovation platforms on small holders in the study areas based on the previous studies that identified successful Innovation Platforms in Eastern and Western Kenya.

Published as FARA Research Results Vol 2 (6).