摘要：This report analyzes the effects of grain market reform and food subsidy elimination in Eastern and Southern Africa on access to food for low-income consumers. The report also assesses the potential to use "self-targeted commodities" to improve vulnerable groups access to food though market development strategies and food aid programs. While much research has been devoted to understanding how producers and traders would respond to reform of staple food markets, relatively little is known about the potential or actual responses by consumers. The report presents recent findings from six household-level surveys in urban areas of Zimbabwe, Kenya, Zambia, and Mozambique between 1991 and 1994. Secondary data from South Africa and Malawi are also presented where available. The report highlights seven conclusions: (1) Consumers subsidies on refined maize meal in Kenya, Zambia, Zimbabwe, Mozambique, and South Africa have not necessarily promoted food security, because they have entrenched a relatively high-cost marketing system and impeded the development of lower-cost channels from developing. (2) The conventional wisdom of rigid urban preferences for refined maize meal was greatly exaggerated by policy restrictions under the controlled marketing systems and subsidies on refined maize meal. (3) Since the removal of refined meal subsidies and controls on maize movement, retail prices of hammer-milled whole meal have ranged from 55% to 80% those of refined meal manufactured by large-scale millers. (4) In all countries surveyed, there is an inverse relationship between whole meal consumption and household income, and a positive relationship between refined meal consumption and household income. (5) New investment in hammer milling has increased rapidly since market reform. (6) Small-scale mills have a higher labor-to-investment cost ratio and labor-to-output ratio than large-scale milling. (7) Yellow maize, which is typically available at a price discount relative to white maize, also appears to be a "self-targeted" food.
