Which Progress for Poverty Studies can we expect from new large Data Sources?
Journal of Contextual Economics – Schmollers Jahrbuch, Vol. 128 (2008), Iss. 1: pp. 65–73
Large data sources would allow us to test the impact of neighborhood characteristics, such as poverty rates, on the attitudes and behavior ot residents. The article explores the feasibility of existing large data sets for such a purpose. Unfortunately, none of the three sets reviewed, the Microcensus, the ALLBUS and the SOEP, allows for such multi-level analyses, because data cannot be regionalized due to data protection or insufficient sample size. To overcome these problems in a limited sense, it is suggested to pursue a “puzzle strategy” to combine data from different existing data sets.