New Research on Poverty Traps

Come to think of it, one of the most basic question of Development Economics has to be: why does poverty persist, i.e., why do (some of) the poor stay poor?

I am very excited about some recent research I have been doing with my LSE colleagues Oriana Bandiera and Robin Burgess, and two of our students, Clare Balboni and Anton Heil, that combines theory and data from a randomised control trial from the BRAC ultra-poor project in Bangladesh to answer the following question - can a one-time push get the poor out of poverty or is it something more systemic?

In other words, we try to test between a "poverty trap" view vs a "bad fundamentals" view of poverty. According to the former, poverty is a vicious cycle that can be broken by a "push" that will help the poor to reach a better self-sustaining equilibrium. According to the latter view, the poor do reach their potential as opposed to staying trapped in a bad equilibrium, but because of bad economic fundamentals (low human capital, bad infrastructure) this potential is very limited.

Here are the slides of Oriana Bandiera's Kapuscinski Development Lecture delivered recently at the University of Barcelona where she discusses some of this work, along with her other related work on the topic. 

  

Knowledge and Man vs Machine

Growing up in Kolkata's academic and intellectual circles, I was always baffled by "erudition" being the most respected intellectual attribute. Someone who knows a lot, has read a lot of books etc was referred to in hushed tones of respect.

Yet I was drawn to ideas, creative expressions, and analytical insights and was never too keen on "knowing a lot" unless I had to, on a given topic. Somehow it felt that erudition was a means to an end, but not an end itself. The few times that someone had a big impact on me (and I have written earlier about my High School maths teacher Pinaki Mitra in this context) it was because they opened up new horizons and methods of intellectual exploration. I fell in love with mathematics because Pinaki-da's off-the-textbook ruminations convinced me that a lot of half-baked philosophical puzzles about the concept of "infinity" that I had in my muddled teenage brain could be formulated so crisply and elegantly.

A quarter century later, machine intelligence is replacing most routine tasks, including storage of facts and data as well as routine analysis. Google can beat anyone on just the capacity to store and draw upon knowledge (with due caveats about errors). Where does that leave the cult of the erudite?

Or, to put it provocatively, can a Sidhu Jyatha (the avuncular figure in the Detective Feluda stories of the Satyajit Ray who had an encyclopedic memory) compete with Google the Great?