The demand for processed convenience foods and ready-to-eat foods has increased globally in recent decades with increases in household income and urbanization. This increase in demand for processed convenience foods has led to an expanded supply of these foods, but little is known about these suppliers, their business choices, and the viability of their enterprises. In this dissertation, I first seek to understand how West African households change their food consumption in response to changes in income and food prices. Then, I provide a characterization of small and medium-sized food processing enterprises (SMEs) in West Africa, analyzing their performance and efficiency to provide insight into the characteristics of successful food processing SMEs. Understanding how changes in household income and food prices affect dietary composition is key to improving food and nutrition security. Utilizing Mali’s 2018 Harmonized Survey on Households Living Standards, Chapter 1 estimates demand for food by food group and by processing level through the calculation of price and expenditure elasticities for 3,847 rural households and 2,745 urban households. A two-stage Working-Leser and Quadratic Almost Ideal Demand System (QUAIDS) model is employed to calculate elasticities for rural and urban households of different income levels. Following the estimation of the demand system, I estimate the amount a household would need to be compensated to restore their original utility under hypothetical price shocks to different food groups. Findings indicate that both rural and urban Malians are increasing their food consumption, in value terms, as they increase their total household expenditure. Within a household’s food budget, the consumption in value terms of animal products, fruits, and vegetables increases as household food expenditure increases. The budget share of cereals declines as expenditure increases but remains relatively high across rural and urban households. Additionally, households demand more processed foods as their incomes grow. Simulated price shocks to cereals and animal products have stronger impacts on household utility than any other food groups. Policymakers should focus on supporting and expanding sustainable food supply chains, particularly for cereals and animal products.Shifting focus to the supply side, in Chapter 2, I utilize a unique dataset of 320 processed food vendors in 82 open-air markets across Senegal in 2021 to examine what factors contribute to a market food vendor’s decision to process cowpea and the quantity they process. In addition to being locally produced, cowpea is a nutritious food source that can be processed for convenience. I provide a characterization of market processed food vendors, then employ a double hurdle model to examine how various regional, sociodemographic, market, business, and product portfolio characteristics affect the probability of processing cowpea and the expected value of cowpea processed by a vendor each week, allowing the factors that affect the participation decision to differ from those that affect the intensity of participation. I find that vendors of processed food products in open-air markets in Senegal are mostly women. The results indicate that vendors in rural markets are more likely to process cowpea and to process more, on average, than urban market vendors. Results from postestimation analysis indicate that nearly two-thirds of the processors most likely to process cowpea operate out of physical structures compared to only a quarter of the processors least likely to process any cowpea. This could be related to the risk of pest infestation for improperly stored cowpea grain. Overall, this study highlights the importance of cowpea processing as a channel for entrepreneurial women to start small businesses and earn income to support their families, warranting further research and investment into this sector.In Chapter 3, I describe the second-stage grain processing small and medium enterprises (SMEs) across urban areas of Senegal, estimate their technical efficiency (TE), and examine the factors that contribute to higher levels of technical efficiency using a Stochastic Frontier Analysis. I employ a dataset collected in 2018 under the Agricultural Policy Support Program that includes information on 552 grain processing street vendors and 200 semi-industrial enterprises. I find that women own and operate the majority of both of these types of grain processing SMEs. Additionally, both types of enterprises rely heavily on manual processing methods, exhibiting low levels of adoption of mechanized processing technologies. I estimate efficiencies separately for both groups of food processors and find the mean TE score of semi-industrial enterprises to be 0.642 and the mean TE score of street vendors to be 0.637. I find that street vendors relying on shared resources are less efficient. I also find that semi-industrial processors are more efficient if they are connected to formal networks through membership in a processor’s organization, providing evidence that policies that improve access to and provide support for processor’s organizations may effectively reduce inefficiency of second-stage grain processing SMEs.
Read
