DFL Decomposition in Stata

DFL Decomposition in Stata

With two partners, I tackled the gender wealth gap in Germany. My contribution was the bulk of the empirical method, and roughly half of the coding, as well as proofreading. As required, running the main .do file automatically calls the others, so that the project can be replicated with a single click.

Lessons learned

This course was rather unique. Rather than learn methods, we had an introduction to the data set and were expected to replicate a paper with no codes given. We pulled and merged from dozens of data files, selected variables, and pored over documentation for over a month before touching the empirical analysis.

The experience represented a number of firsts for me. In addition to being my first proper programming project, it was my introduction to big data (30+ waves of panel data, generally over 10,000 individuals, with hundreds of economic, social, and health variables). Files and our paper can be found here: https://github.com/jimbodonahue/GSOEP

Note that the data set is unavailable for privacy reasons.