Bart Minten
International Food Policy Research Institute (IFPRI)
K.A.S. Murshid
Bangladesh Institute for Development Studies
Thomas Reardon
Michigan State University, IFPRI, and Renmin University of China
Food quality changes, Implications, Rice value chain, Bangladesh
Socio-economic and Policy
Performance
The purpose of the study is to understand how quality impacts the rice value chain in Bangladesh from rural areas to the capital, Dhaka, which, along with its metropolitan area, is home to approximately13 million people.7 To get at this information, two types of activities were organized. Interviews were conducted with key informants in the value chain in October 2009. That information was used to design questionnaires for each level in the value chain. These questionnaires were then fielded at the end of 2009 (November and December). The implemented instruments included surveys upstream in the supply chain with rice-producing households and villages, midstream with wholesalers and millers, and downstream with retailers. Upstream in the supply chain, the district of Noagoan was selected to represent rural production areas because it is an important rice-producing district, supplying rice to the capital. The district is located about 200 kilometers north of Dhaka. The village and household survey was set up as follows. The two most important paddy-producing thana (subdistricts) in the district of Noagoan were selected. In each thana, the villages were stratified based on total rice produced in the village. Then five villages from each thana were randomly selected-two from the stratum of high-producing villages, two from the stratum of medium-producing villages, and one from the stratum of low-producing villages. In each selected village, a village questionnaire was implemented and a census of households was conducted to enumerate the paddy producers. The sampling of the rice farmers was done in a two-step process. First, rice farmers were ranked by rice area cultivated from the biggest to the smallest. A cut-off was set when 50% of the total rice area in the village was reached. Second, 11 households were then randomly selected from each group, reflecting their respective share in the rice value chain. 8 Thus, 220 paddy-farming households were selected in total. 9 A survey focusing on rice production and marketing practices was then implemented. Midstream in the supply chain, the wholesaler survey was set up as follows. First, interviews were conducted with village traders and other rural off-wholesale-market traders who buy from households in that village or from other traders (17 in total). Second, 43 traders were interviewed from the local rural wholesale market in the selected district. Third, 30 urban wholesale traders were interviewed in Dhaka, half from each of the two most important rice wholesale markets in the city, Badamtoli and Krishi markets. These traders were randomly selected after a census was done. Also in midstream, a list of all the millers in the district of Noagoan was obtained. A stratified random selection of 20 millers was done (8 with automatic mills, 5 with semiautomatic mills, and 7 with small mills)10 and detailed surveys were conducted, focusing on information related to purchases, sales, and milling patterns. These different technology strata for mills were selected given the presumed importance in the processing of different varieties by the larger mills, which are usually automatic or semiautomatic. For example, Murshid (2011) found that the increasing availability and affordability of better-quality rice in the market has become possible because of the increasing availability and use of modernized mills. Downstream in the supply chain, a retail survey was conducted in Dhaka, covering both traditional and modern retailers. First, five thana were randomly selected in different parts of Dhaka (north, east, west, south, and central). In each thana, a census was done of all retail markets, and two markets were randomly selected. A census of all rice retailers was done for each market, and 12 traders were then randomly selected and interviewed. A total of 120 traditional retailers were thus interviewed. Second, 20 surveys were conducted with modern retailers. In each thana that was selected for the traditional retail survey, a census of modern retail stores was conducted and 4 were randomly selected.11 A price survey was then implemented for both types of retailer.
Selected Paper prepared for presentation at the International Association of Agricultural Economists
(IAAE) Triennial Conference, Foz do IguaƧu, Brazil, 18-24 August, 2012.
Journal