Targeting Social Assistance in a Transition Economy: The Mahallas in Uzbekistan [antikvár]

Executive SummaryCountries in transition from the planned system have had to re-consider how best to target public resources on those in need. New risks to the population have appeared, such as open unemployment. The state's tax revenue and hence ability to provide comprehensive safety nets is often reduced. And systems of state transfers inherited from the communist period have frequently proven inappropriate or have been difficult to adapt. A common hole in the pre-reform safety net was a social assistance scheme providing support of "last resort", in part due to an official denial that poverty could exist under communism.The paper illustrates several issues involved in designing new social assistance schemes in transition economies, including (i) the definition of household resources and hence "need", (ii) the choice of agent to administer the scheme, and (iii) the reaction of the populace, the potential clients, to targeted social assistance.The example used is a scheme introduced in 1994 in the Central Asian republic of Uzbekistan. This country is the third largest former Soviet republic, with well over 20 million inhabitants, and it was one of the poorest repubUcs during the Soviet era. The new social assistance scheme, administered by traditional local community groups, the "Mahallas", is intended to be of particular benefit to families with children. At least 1 in 10 households received support from the scheme in 1997. The Mahalla assistance scheme provides an interesting example of a highly decentralized and flexible system of benefit targeting, with lessons for both other transition economies and elsewhere. While the scheme contains clear guidelines to the Mahallas, there is no formal set of necessary or sufficient conditions for benefit.The scheme is investigated using a household survey of living conditions in Uzbekistan carried out in 1995. The survey contains a variety of indicators of resources and need, including income, durable goods ownership, agricultural assets, employment status, and the nutritional status of children. It also collected information on households' knowledge of the Mahalla assistance scheme, on their applications for benefit, and on awards made, thus allowing the process of receiving benefit to be broken down into its various stages.Analysis of the survey data shows that the Mahalla scheme does deliver benefit much more frequently to less well off households than to better-off households. (This is the case both when household welfare is measured by income alone and when it is summarized by a range of household characteristics.) In this sense, the results are a positive demonstration of the potential