On smallholder farm and farmer diversity

Feeding the world sustainably is one of today’s greatest challenges. The urgency to produce more food from a dwindling resource base calls for a sustainable intensification of agricultural production. Globally, 83% of all farm systems are smallholder farms, whose productivity could increase through the adoption of improved agricultural technologies and techniques. However, smallholders are diverse in their features, constraints and opportunities and so are their possible pathways for sustainable intensification. These pathways are made up of sequential decisions for change. Decisions in smallholder farm systems, e.g. on land and labour allocation, are often a matter of negotiation since resources are shared at household level. Therefore, when aiming to understand, anticipate or evaluate resource allocation decisions of smallholders, information is needed on individual interests and household-level decision-making dynamics. In this thesis, I address the question of how inter- and intra-household differences in Northern Ghana shape smallholder farm decisions. Chapter 1 outlines the problem statement, the specific research questions as well as the research context. In Chapter 2, I characterize local farm systems diversity to determine farm type specific constraints and opportunities for agricultural innovation. I do so, by using the multivariate statistical techniques of principle component analysis and cluster analysis using farm household data (n=80). I determined six farm types, stratified according to household, labour, land use, livestock and income variables: two types of high resource endowment (HRE), two types of medium resource endowment (MRE) and two types of low resource endowment (LRE). The HRE types were oriented towards non-farm activities or crop sales, the MRE types derived their income mainly from on-farm activities and the LRE types were generally oriented towards subsistence. Each farm type was associated to different constraints and opportunities, ranging from composting and better post-harvest storage (LRE), the procurement of donkeys for transportation and tillage (MRE) to better cattle manure management and crop diversification (HRE). Chapter 3 compares the etic, statistical typology of Chapter 2 to an emic, participatory typology. The latter resulted into a classification of farmers rather than farms i.e. grouping household or community member types (household heads, wives, sons, landless) rather than entire households. The joint application of statistical and participatory approaches provided different but complementary perspectives, allowing a multi-dimensional analysis of farm and farmer diversity. Chapter 4 operationalizes the insights into the local horizontal (farm) and vertical (farmer) diversity for a nuanced impact assessment of five project-proposed technology packages. I assessed the performance of the technology packages per farm type (LRE, MRE and HRE) and per region (Northern Region, Upper East Region and Upper West Region of Ghana). For the performance assessment I used the whole-farm model FarmDESIGN as well as a weighted scoring technique to systematically capture farmer evaluations. I then compared model results with farmer realities and found that women were more positive about the packages than men, since men heavily penalized extra costs and labour, translating into a greater congruence of model results with the male evaluation. LRE farms were projected to benefit most in relative and least in absolute terms from an adoption of the packages. I also explored alternative farm designs and found that the most promising configurations were hard to attain due to high cost and labour requirements for their implementation. Based on the encountered intra-household differences during the technology evaluation, I decided to take a deeper look at decision-making dynamics in local farm households. Chapter 5 hence examines intra-household dynamics and trade-offs in land allocation decisions of smallholder farmers, by applying concepts of economics, socio-psychology and physics. I revealed conflicting interests and a mismatch between ‘ascribed power’ and ‘exerted power’ suggesting that social power may be deployed, overruled or withheld. Power may be withheld if investments and risks, associated with a negotiation, outweigh the expected utility. Individual and household-level utilities furthermore exposed the social unacceptability of many technically promising land allocation options. I conclude that technical options should be evaluated ex-ante for their likelihood of acceptance and social implications to ensure their basic viability and sustainability. In Chapter 6, I report on methods and findings of a serious game that simulated an actual household-level negotiation between the male household head, a wife and the eldest son of a hypothetical local farm household. I used social network analysis to quantify interactions during the negotiation. While the household head was the key decision maker acting as a strategic gatekeeper in a funnel-like process, the wife and the son also had a significant influence on the household-level negotiation outcome. Model-based analysis showed that the household-level outcome was more profitable as well as agro-biologically and nutritionally more diverse and productive as compared to the household heads’ suggestion. In line with my hypothesis in Chapter 5, power was observed to be actively deployed, withheld or passively overruled depending on the decision domain and process dynamics. I observed an integrative negotiation style, resulting into high levels of satisfaction with the negotiation process and outcome by all parties, who unanimously reported a high level of similarity between simulated and real-life negotiations. Chapter 7 briefly responds to each research question and elaborates on the comprehensive insights of this thesis, including overall lessons learnt on intra-household decision-making dynamics and a matrix of local farm and farmer characteristics. I discuss the transferability of my methods and findings as well as their contribution to the debate on women empowerment in agriculture. I furthermore reflect on agricultural systems research at the interface between linear and complex systems thinking. I conclude that, in order to effectively support local smallholder farmers, R4D projects are well advised to assess possibly competing interests around any proposed change. Pathways for sustainable intensification are made up of sequential decisions for change, spanning over different decision domains that are administered by different household or community members. A systematic overview of local farm and farmer characteristics as well as participatory inquiries help to understand possible decision-making dynamics, providing a solid basis to formulate or adjust a projects’ theory of change and theory of scaling. Finally, it will be the sum of local changes and their synergetic effects that will add up to the global change that is required to sustainably feed the world.