Brief Biography

This gives a brief account of my intellectual journey in the field of econometrics, and how it led to the research I have done. This also permits classification of my research by the themes I have been pursuing.

For an account of my intellectual journey in Islamic Economics, see link

For a personal biography, see link.

A Bayesian Start: During Graduate School at Stanford, I learned and accepted the Bayesian Rationality Argument. I became a "rabid Bayesian" (quoting Jim Berger, who writes this). However several puzzles could not be resolved by this approach. Decision Theory shows that all admissible procedures are approximately Bayes and vice versa, and I thought this was a way to prove Bayesianity to the world, and did a lot of research on this. After my Ph.D. from Stanford, I was planning to write a text on the decision theoretic approach to Econometrics before George Judge beat me to it. In the meantime, I had been working on the Stein Paradox and trying to understand it. Eventually I did understand it and decided to write a textbook explaining it. This turned into Statistical Foundations for Econometric Techniques (SFET). Upto this point I had accepted the idea of a division between "Applied" and "Theoretical" and assumed that specialization and fragmentation of knowledge is the best way to advance knowledge. As an afterthought, I decided to add some realistic examples to SFET, to illustrate the theoretical tools in context of real data. This led to a revolution in my thinking about econometrics.

Applying theoretical models and estimators to real world data involved a collection of problems which no one appears to have thought about. Much later I discovered David Freedman, and a few others, have thought about these problems and come up with pessimistic conclusions. Most of our techniques do not work in real world situations. My more recent research is devoted to trying to map out the Limits of Inference: what can and cannot be done safely with statistical methods.

HOW TO DECEIVE WITH STATISTICS:

While the above shows weaknesses of conventional methodology, a more important theme is the deliberate use opf these weaknesses to prove whatever you want. As Disraeli said, there are "Lies, Damned Lies, and Statistics." In this subpage, we collect material on how statistics/econometrics can be used to deceive.