Introduction: Policies to promote demand-driven smallholder agriculture and improved urban food marketing system performance in Sub-Saharan Africa (SSA) need to be informed by careful food demand analysis, especially given the rapid rate of urbanization in many SSA countries. Governments, donors, and other policymakers require an up-to-date understanding of urban consumption patterns because these are among the main drivers of many of the opportunities available to small-scale farmers and because such information can help identify key leverage points to improve urban marketing system performance. It is also important to understand better the extent of, and constraints to, urban agriculture for household consumption as well as for cash income generation. Up-to-date information on urban consumers’ food production and consumption behavior in Zambia is lacking and thus, the Central Statistical Office (CSO) conducted the Zambia Urban Consumption Survey (UCS) in August 2007 and February 2008 in cities of Lusaka, Kitwe, Kasama and Mansa. This was done in collaboration with the Ministry of Agriculture and Cooperatives and the Zambia Food Security Research Project (FSRP). Lusaka and Kitwe are metropolitan cities (Lusaka is the country’s capital city and Kitwe is the biggest city on the Copperbelt) while Mansa and Kasama can be referred to as rural cities situated in the northern parts of the country where cassava is a very important staple food. Though Mansa and Kasama are Provincial capital cities (Luapula and Northern Province respectively) they have relatively less industrial activities with agriculture playing a more prominent role than in the more urbanized cities of Lusaka and Kitwe. The primary objective of this study is to develop a detailed understanding of the food and other consumption and expenditure behavior of households in key urban areas of Zambia. Key aspects of this behavior analyzed are consumer food budget or expenditure shares across different food groups and specific food items. Analysis also examines urban agriculture as well as the market share of different types of retail food outlets, such as open air markets, street vendors, shops, supermarkets, and other retail formats. This report covers the general findings of all sections of the survey instrument used in the study, presenting results using tables and figures. Interpretation of the findings is covered in bullet point form with some explanations/discussions where necessary. Using this basic survey information, additional studies on detailed topic and value-chain (such as staples, horticulture, etc) issues will further analyze findings and draw conclusions about urban production and consumption behavior, and related marketing policy issues.
Methodology: The survey was designed to cover 140 Standard Enumeration Areas (SEAs) across the 8 strata that were defined to cover areas and households in Lusaka, Kitwe, Kasama and Mansa. This corresponds to a probability sample of about 2, 800 non-institutionalized private households residing in the target urban areas. This sample is urban area wise efficient and is expected to yield reliable estimates at urban area and stratum levels. No national estimates were to be generated from the data. In order to improve the quality of the data as well as capture seasonality of expenditure and consumption, the survey was done in two phases: the first phased covered the six month period between August to January, 2007, and the second phase covered the six months from February to July 2008. During these same periods, prices of selected commodities and selected units of measure were also collected for use in further analyses, such as converting
consumption from expenditure to actual physical quantities for estimating price elasticity of demand. Data collection was conducted by way of personal interviews using 1 semistructured questionnaire to collect general consumption data pertaining to the household being enumerated. In addition to the household data collection instrument, a listing form was initially used to list all households in the selected SEA. The same panel of households visited in the first phase was followed during the second round of the survey. The number of households was over sampled in the first round to accommodate for the possibility of noncontact households in the second round. The data from the UCS survey was entered in CSPro computer application and cleaned and analyzed using the Statistical Package for Social Sciences (SPSS) software.