Interesting discussions on RCTs

I was recently at the Centre for the Study of African Economies annual conference in Oxford. I was actually even presenting a paper but that is something for another day - let us just say that the paper was weak beforehand, and felt even weaker after. You know you;re on a road to nowhere when the discussant points out that "it is not clear exactly what the contribution is here".

In any case, while I missed the last day plenary discussions, I did manage to catch them here, where the third discussion on RCTs and structural models is particularly interesting, actually creates some discussion and debate and kind of raises passions (for a conference plenary session anyway...). Actually, about half-way through there was music to my ears when Francis Teal bemoans the use of RCTs and the rejection of simple descriptive statistics and the use of cross-sectional data - he should have been in my presentation to stand up for my simplicity. Mind you, David Mckenzie doesn't seem quite so impressed, and neither was he for my talk...

I'd never come across Glenn Harrison I must confess, but he makes a pretty clear case for using all methods of analysis, and gives a useful defence of theoretical models. Unfortunately I am either too lazy or stupid to understand the details of any of these theoretical models so have grown to dislike, and at times disdain them. However, after a talk like that which even talks up the much maligned CGEs, there is not much room for ignoring the importance of theory, and of being humble about what we do and don't know. In fact, I already blogged about my recent epiphany regarding theory in economics....

Anyway, some interesting discussions to behold.