Four methodological prescriptions for empirical research

I do not wish to address “methodology” with a capital M in the sense of the meaning of life, of probability, and the definitions of truth, understanding, and progress. If we do not share roughly common notions and concepts from the outset, nothing I can write in the space of this essay will help resolve that state. Instead, given that the primary objective of econometric modelling is to provide an “explanation” for actual economic behaviour (which can include descriptive and predictive aims as well), I seek to communicate the broad framework of my approach and some of the principles and practical procedures it entails.

Over a period of about 20 years of analysing empirical phenomena, I have noted down four golden prescriptions for those who seek to study data:

I. THINK BRILLIANTLY. This helps greatly, especially if you can think of the right answer at the start of a study! Then econometrics is only needed to confirm your brilliance, calibrate the model, and demonstrate that it indeed passes all the tests.

II. BE INFINITELY CREATIVE. Be assured that this is an almost perfect substitute for brilliant thinking, enabling one to invent truly excellent models en route and so achieve essentially the same end state.