MPI toolbox for Stata now available on SSC

After a considerable period of development mpitb, a toolbox to estimate and analyse multidimensional poverty indices (MPI) for Stata, is finally available on the Statistical Software Components (SSC) archive. You can download and install the toolbox by issuing ssc install mpitb in Stata.

The toolbox emerged from my work on the global MPI and various research projects. I frequently found myself to code solutions for the same problems again and again. As a consequence key features of the toolbox include

• the estimation of several indices of the Alkire-Foster framework (e.g., the adjusted headcount ratio, headcount ratio and intensity)
• the estimation of censored and uncensored headcount ratios as well as absolute and percentage contributions to the MPI for all indicators
• disaggregation of all indices by subgroups
• estimation of all indices for alternative parameter choices (e.g., weights, poverty cutoffs)
• estimation of changes over time is supported out of the box for many indices

Moreover, the toolbox comes with a streamlined workflow, several convenience functions and a few low-level tools for programmers.

The currently available documentation comprises several resources. As a quick start I suggest to take a look at the slides of a recent talk I gave at a Stata user conference. For more in-depth information, consult the backing paper and the comprehensive help files shipped with the package. You can access these help files from Stata by issuing help mpitb (assuming the package is installed). The Stata package also comes with auxiliary files, which may be optionally downloaded during the installation process. These files allow you to replicate the examples provided in paper and slides. Together slides, paper and help files already provide a good overview. Yet, many more things can be done with mpitb, which is why I’m planning to compile a more comprehensive documentation geared towards user interests.

Despite the considerable development period mpitb is neither free of bugs, nor are all features already implemented. Indeed, the principal structure as a toolbox seeks to support the implementation of novel forms of analyses for multidimensional poverty in the future. The toolbox is also distributed under version control at gitlab. Please report bugs, feature requests, etc, using the issue tracker of the repository. For questions and other requests, please use the service desk to avoid your mail slipping through the cracks. Any kind of feedback is always welcome.