Stata-Module for Nopo Decompositions

This Stata .ado implements a matching-based decomposition analysis of differences in outcomes between two groups following Ñopo (2008) and some convenient postestimation features. Allowed matching types are (coarsened) exact matching, multivariate-distance matching, and propensity-score matching, all of which are internally using kmatch (Jann, 2020).

Installation

To download and install the module from the Statistical Software Components (SSC):

ssc install nopo

Alternatively, install the current version from GitHub (GitHub files might contain the latest updates that have not been currently published in the SSC):

net install nopo, from("https://raw.githubusercontent.com/mhamjediers/nopo_decomposition/main/")

Suggested citation:

Sprengholz, Maximilian & Maik Hamjediers, 2024. "NOPO: Stata module to perform Nopo (2008) matching decompositions," Statistical Software Components S459289, Boston College Department of Economics.

Documentation

Please consider the help file provided with the package via help nopo and the methdological details documented in this file.

Bug Reports

Please either open an issue here or drop us a mail.

Version Log

• 1.0.1: Fixed Stata 15 error due to unallowed direct acces to special matrices such as e(b)

References

Jann, B. (2020). KMATCH: Stata module module for multivariate-distance and propensity-score matching, including entropy balancing, inverse probability weighting, (coarsened) exact matching, and regression adjustment. Statistical Software Components. https://ideas.repec.org//c/boc/bocode/s458346.html

Ñopo, H. (2008). Matching as a Tool to Decompose Wage Gaps. The Review of Economics and Statistics, 90(2), 290–299. https://doi.org/10/b6tqwq