Learning, agricultural technologies and rural livelihoods in Uganda

This thesis contributes to understanding of the learning mechanisms in information dissemination programs, agricultural technologies adoption process, effects of adoption of modern technologies and rural livelihood strategies and dynamics. The study used mixed methods to answer the research questions focusing on four districts: Tororo and Iganga in eastern region and Hoima and Masindi districts in mid-western region in Uganda.

Chapter one provides the context of the thesis, general background of each chapter, the main objective and research questions and methodology used in the thesis.

Chapter two addresses the learning process and its effects of the adoption of improved maize and bean varieties in Uganda. More specifically, the concept of social, environmental and didactic learning was used to examine the use of demonstration plots to promote new varieties, role and interaction of actors (farmers, demo hosts, commercial actors) and the effect on the information dissemination and uptake of improved seeds. We find that the information dissemination, learning and uptake of improved seeds process were hindered constraints such as uncooperative demo hosts, poorly maintained demonstration plots, not considering farmers’ needs and limited interactions among farmers, demo hosts and seed company officials. Social distance and established relationships are important in information dissemination process. Adoption of improved technologies was hindered by inaccessibility, unaffordability, and unavailability. This thesis shows that imbalance of social, environmental, and didactic learning hinders the learning process and adoption of improved technologies.

Chapter three looks at adoption dynamics, reasons for disadoption and crop management practices in both improved and local maize varieties. Survey results show that the decision to adopt is not permanent; farmers shift from adoption to disadoption between years. High cost and lack of funds to buy the inputs were reported as reasons for not using the improved seeds in all the years. Household heads with primary education and above, males and wealthier were more likely to adopt the improved maize varieties. Land and labour had positive effect on the likelihood of adopting. Availability of family labour is required during peak seasons especially for adopters because improved management practices are required when planting improved varieties: row planting, number of weeding, harvesting.

Physical capital is positively influenced the size of land under improved seeds. Mechanical labour availability (oxen) and seed affordability is necessary for the extent of adoption. Gender of the household head does not influence the extent of adoption as it does the decision to adopt, therefore accessibility is necessary for adoption. The proportion of land allocated to improved maize seeds is rarely affected by the coefficients except the land size. Gender variable did not influence disadoption; the female headed households were more likely to disadopt compared to their male counterparts. Most households did not allocate all the maize area to improved varieties.

Chapter four addresses the effects of the adoption of improved maize varieties on production and income in Uganda. As indicated above farmers recycle certified seed over a varied period of time. Considering only certified seed bought in agro-dealer’s shops, adoption of improved varieties increased the yields and crop income. The increase in yields could be as a result that farmers practice improved management practices because yields decreased over time due to weather variability. Although, the yields increased, this did not translate to increased maize income per hectare. There was no significant increase in maize income per hectare. Surprisingly, despite this result the crop income increased significantly among the adopters. Adopters may have allocated their best land and management practices and additional labour to other crops complementing the effect of certified seed. The higher crop income indicates that adopting farmers cultivate other crops for income.

Chapter five investigates the livelihood strategies and dynamics among the smallholder farmers in Uganda. We look at factors that influence the choice of a strategy, their outcomes, changes in livelihood strategies and crop production between the strategies. All households engaged in annual farming. Our findings identified annual crops farming as the main strategy. Other strategies include self-employment, casual employment, salaried employment and perennial farming. The distribution of the livelihood strategies did not change much in the second year. Perennial/annual strategy cultivated the greatest number of crops and allocated more land to crop production in both years. Maize, beans, cassava, sweet potatoes, bananas were the most grown crops. However, the use of improved varieties in maize and bean production is low.

There were constraints associated with participation in other livelihood strategies compared to annual farming strategy. Age of the household head, education and wealth may have contributed to the choice of livelihood strategy. Higher literacy levels, knowledge and skills are required for non-farm employment. Dependency ratio influenced participation in perennial/annual farming and self-employment/annual negatively and positively respectively. Location of the households is important in determining the choice of livelihood strategies.

The outcomes (incomes) vary depending on the different strategies with non-farm strategies generating higher incomes. Salaried employment/annual was the most remunerative in 2013/14 while casual employment/annual topped and salaried employment/annual came second topped in 2015/16. Annual farming strategy was the least remunerative despite allocating most of their land and labour to the activity. Although, the distribution of livelihood strategies was relatively stable, individual households switched strategies. Farm strategies were the most stable strategies although they were least remunerative. Shifting of households to a more remunerative non-farm strategy requires more resources, skills, and opportunity. Switching strategies may involve re-allocation of resources between the strategies or adopting/abandoning an activity.

In the synthesis, I highlight the significance of the main findings, how the studied themes are interrelated and contribution of the study to the literature of learning, agricultural technologies and livelihood strategies. These findings have an implication on how learning takes place and how these affects the technology adoption process. Most importantly, adoption dynamics will contribute to understanding more about adoption process which is not a one decision. Another major finding that emerged in this study is that despite agriculture being important for rural households in Uganda, it is the least remunerative compared to other livelihood strategies.