I wrote a panicky blog post last week about the questionable absence of statistical software in my econometrics class. Today, I discovered my intuition (and the course syllabus) misled me into believing something so utterly outrageous. The good news is that my econometrics course will integrate a statistical package into the coursework. The bad news is that my econometrics course will integrate EViews into the coursework.

I have to admit something is better than nothing as far as software in econometrics goes. But why we're using EViews is a mystery to me. As I understand it, most academic research economists use either Stata or R, while researchers in private industry tend to use SAS or R. So, by learning EViews, I'm essentially learning a dead language. It's like learning Sanskrit, but without the redeeming esoteric and historical qualities.

Furthermore, and more importantly, EViews does not allow users to customize and extend the software as readily or as easily as Stata or R. I've never tried to extend EViews, but from everything I've read, it's much less extensible as Stata or R. Both those packages also have thriving user communities, each of which has produced hundreds of useful add-on packages and plug-ins.

The one potentially redeeming quality of EViews is its sugar-coated user interface. I haven't used Stata enough to really comment on its user interface. But I will concede that anyone uncomfortable working in the command line or with object-oriented programming would probably have trouble diving into R.

In most cases, I would praise the package with the superior user interface and more gradual learning curve, and bash the package that takes some time to master. When it comes to more professionally-oriented software, however, I put much more emphasis on flexibility, extensibility and robustness. Professional software can't omit important features for the sake of simplicity — if a researcher needs a feature, the feature needs to either exist or be easily added.

Still, I do feel better (if marginally so) about my econometric course. The idea of being confined to datasets of five observations was frightening.