Christian Dudel
;
Jan Marvin Garbuszus
; ;

Matching as non-parametric preprocessing for the estimation of equivalence scales (replication data)

Empirically analyzing household behavior usually relies on informal data preprocessing. That is, before the estimation, observations are preselected to obtain a sufficiently homogeneous subset of data. In the context of estimating equivalence scales for household income, we use matching techniques and balance checking at this initial stage. This can be interpreted as a non-parametric approach to preprocessing data that formalizes informal procedures. We illustrate this using German micro-data on household expenditure, showing that matching leads to results which are more stable with respect to model specification and is especially useful when applied to specific subgroups, such as low-income households. The files provided here contain the code (in "R") which is needed to replicate our analyses.

Data and Resources

Citation

Dudel, Christian; Garbuszus, Jan Marvin; Ott, Notburga; Werding, Martin (2017): Matching as non-parametric preprocessing for the estimation of equivalence scales (replication data). Version: 1. Journal of Economics and Statistics. Dataset. http://dx.doi.org/10.15456/jbnst.2017136.073635