Quality of statistical match of household budget survey and SILC for Turkey

This paper presents the quality analysis of the statistical matching conducted for a research study on household consumption behavior, household indebtedness, and inequality for Turkey. The match has been done for four years (2005, 2008, 2009, and 2012) of Household Budget Surveys (HBS) and the Survey for Income and Living Conditions (SILC). The aim of the statistical matching is to transfer household expenditure data from the HBS to the SILC to create synthetic data sets that have information on household consumption expenditures as well as household income and indebtedness. We are following the methodology of constrained statistical matching, using estimated propensity scores developed in Kum and Masterson (2010) to produce the synthetic data sets that we need. The analysis shows that the match is of high quality.