#econtwitter I understand being a researcher/economist leaves no/very little time to update software skills. However there's an ongoing shift that should not be ignored. STATA and Matlab have become old

(my) Anecdotal evidence shows that only the youngest economists (and surely not all of them) are familiar with open source popular programming languages like R and Python, while faculty 32+ years old stick to what they have been taught in their BA/MA (i.e. STATA/Matlab)

Using STATA/Matlab for research is totally fine. However I think there is an issue being neglected on the teaching side. I think we can all agree that nowadays students today should be exposed to good data analysis/ econometrics teaching.

Good teaching of econometrics includes:
a) Exposing students to the many existing methods to recover 'good' (in its broadest definition) estimators from a theory perspective
b) Showing them in practice how to analyze data and implement those methods

Consider point b). Every good econometrics course includes applied assignments. Wouldn't it make sense (from a university perspective) to expose students to the statistical software which are most required by employers these days? Otherwise...

... what's the point of buying (expensive) licences for Softwares that students won't use ever again?
Just look for the number of job listings that require STATA/Matlab vs the number of those that require Python/R in absolute values today, and in relative changes wrt recent years

Learning how to perform MSc level data analysis in R/Python does not require exceptional skills. My final advice to economists is: ask MSc students to solve assignments in R even if YOU don't know it. And ask the PhD students to learn it by themselves so that they can teach it.

The investments should be beneficial for everyone.

BA/MSc students: they learn a language that may actually help them find a (nice?better? better paid?) job

Junior/Senior economist: you 'train' a generation of PhD students with skills that may also help you in research. R/Python have such a huge potential! I personally have seen improvements in R exponentially faster than in STATA. PS: these PhDs are potential (high-skill) co-authors

PhD students: you get to learn languages very useful both in case you pursue academic path or not. By teaching them you'll learn them very well (hopefully). You are probably the ones who are gonna benefit the most of the investment

University: stop spending money for licences.

Net losers: private software developers. But unfortunately (for them) this is how the digital world is evolving. As long as open source languages become more and more popular among data scientists, the future is theirs...