How to Get it Wrong in Rural Uzbekistan: An Ethnographic Critique of Household Survey Categories

Post-Soviet transitions have prompted a search for new policy tools and methods of data collection. The shift from universal welfare provision under the Soviet system to targeted assistance and poverty monitoring has stimulated a new interest in the measurement of living standards and poverty lines.

This has promoted the use of quantitative techniques and sample surveys (household surveys, in particular) as privileged tools for the collection of policy-relevant information. This paper contends that survey techniques have particular limitations as research tools in an environment where local-level case studies are scarce and where a host of new socio-economic processes are creating fundamental shifts in the landscape of social provision, redistribution and employment. These limitations are illustrated by drawing upon a household survey conducted by the author in four villages from two regions (Andijan and Kashkadaryain) in Uzbekistan between October 1997 and August 1998. The ambiguities surrounding five basic concepts—household, employment, access to land, income and expenditure—are discussed in detail, as are the changes in their contents and meanings in the context of transition. The illustrations provided suggest that households identified on the basis of village records (the most commonly used sampling frame) do not necessarily correspond to self-contained budgeting and consumption units. The links between co-residence, budget control and household divisions of labour are complex and fluid. There is also a growing disjuncture between “official” occupations and wages and what people actually do to make a living. This disjuncture is reflected in the reporting of jobs and incomes in ways that make an evaluation of both employment status and household finances quite problematic. Likewise, the conditions of access to land are changing as agrarian reform policies unfold. Despite the fact that land privatization measures in Uzbekistan have been slow compared to neighbouring countries, even the piecemeal commoditization of land through the expansion of leasehold markets is creating new pressures on smallholder production. Finally, the reporting of household incomes and expenditures may be influenced not only by rapidly changing conditions of employment and social provision, but also by the fact that they may reflect different kinds of logic depending on the conditions of the region under study. For example, the tendency to calculate outlays such as animals or produce with reference to their market value was much less pronounced in a region with a lower overall degree of monetization. It may be that the expansion of a market economy produces a homogenizing effect on the deployment of these categories. The conclusion suggests that a more contextually sensitive approach to the mechanisms that generate new forms of vulnerability and the use of qualitative and longitudinal methodologies are essential to an adequate monitoring of further changes.