10 décembre 2018
Ecrit par WID.world

New paper and Stata command on correcting survey data using tax data (WID.world Working Paper 2018/12)

This new paper by Thomas Blanchet, Ignacio Flores and Marc Morgan introduces a new method to adjust household surveys using tax data.

Researchers and interested users can easily implement the correction with a single line of code using the new Stata command that that is presented in the paper. The command and its documentation can be installed from the Statistical Software Components (SSC) archive by simply typing “ssc install bfmcorr” in Stata.

Tax data show that household surveys generally fail to properly capture the top of the income distribution, and therefore need to be adjusted to estimate inequality correctly. To date, there is no consensus on how to approach this problem. The paper introduces a method to combine both datasets that has several advantages over previous ones: it is consistent with standard survey calibration methods; it has explicit probabilistic foundations; it preserves the representativeness of socio-demographic variables; and it preserves the microdata structure of the survey. The procedure is illustrated by applications in five countries, covering both developed and less-developed contexts.