Sample design considerations for informal sector surveys

1 The survey units An informal sector survey is concerned with economic units of special types namely householdbased and other small-scale establishments. Exactly what type of units are in the scope of the survey is a substantive issue, dependent on circumstances and data requirements, and is not considered here. However, certain common features can be noted. Typically, the survey is confined to non-agricultural units, but covers diverse activities. The units of interest lack formal or legal separation from the household as an economic entity, are owned and operated by persons alone or in partnership as self-employment activity, without employing regular employees. In developing countries in particular, the establishment is typically located in the operator’s home, in temporary premises, or without a fixed place of business. In addition, the population of interest may include micro-enterprises which, while employing one or more regular employees, operate on a scale below a certain level. In so far as the economic units of interest are household-based and have a one-to-one correspondence with households, the informal sector survey design is likely to be similar to that of a household survey. However, there are many features which distinguish such surveys from population-based household surveys. Fundamental differences arise from the fact that we are dealing with a population of units which tend to be very heterogeneous, unevenly distributed in the population, and also less stable than typical households. Furthermore, the precision requirements, and possibly even the measurement objectives, may differ from one category of units (type, size, sector of economic activity) to another. All these factors contribute to increased complexity of the survey design, including sample selection and estimation procedures. While such complexity cannot be avoided, it remains highly desirable that it be kept to a minimum. Even more important is the need to avoid complicated procedures at the implementation stage .