Augmentation of the Samples size for the Household Budget Survey

Abstract

The sample designs for the Albanian Household Budget Survey (HBS) are based on the around 7900 households of the Census 2011 stratified for prefecture and urban/rural areas with Probability Proportional to size sampling according to the number of households inside the strata. The sample design is not optimized for the new 61 municipalities of Albania. There is a need for publication on small domains like municipalities for important indicators of this survey, at least for the consumptions of the 12 main groups of consumption.  Important indicators should be published for the municipalities if the precision is acceptable. Either spatial or temporal aggregation (pooling) are possible solutions for certain (larger) municipalities. This study aims to give solutions for augmentation on the number of PSU to reach the publication on municipalities at least on some of them and save the design effect of HBS survey in Albania.