Abstract
The availability of detailed and precise data on poverty at a low level of spatial aggregation is important when pursuing an effective cohesion policy. In Poland, this type of information is gathered during household surveys conducted by Statistics Poland and is made available at country, region, and selected socio-economic group level. Direct estimates relating to domains not included in a survey are burdened with a serious estimation error. In a situation of a limited (or in extreme cases zero) sample size, an estimation becomes possible through the application of small area estimation methods – indirect estimation. These techniques use variables which are strongly correlated with the researched phenomenon and which come from a census or from an administrative register. The aim of the study discussed in the article is to estimate two indicators: the rate of poverty and the depth of poverty at a poviat level, with the application of the Empirical Bayes (EB) method. The first indicator provides information on the scale of the phenomenon and the other one on its intensity, and so they constitute complementary measures of poverty. The study used data from the European Union Statistics on Income and Living Conditions of 2011 and the National Census of Population and Housing 2011. Information about the scale and intensity of poverty at the poviat level was obtained through the adaptation of the EB method based on the linear mixed model and Monte Carlo simulations. The indicators estimated this way allow for an assessment of the diversity of poverty at a local level. In addition, they are more precise and consistent with administrative registers in comparison to direct estimation results.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have