Sunday, October 20, 2019

Abhijit Banerjee's inconsistent criteria

Abhijit Vinayak Banerjee is one of the three economists honoured with the Nobel this year for adopting Randomised Controlled Trials in attempts to alleviate poverty. These trials emanated in the pharma industry and adopting them in economic studies is a trail blazer in its own way. Just like any adoption of a new methodology in any field, this transplant of a pharma idea in the economic field has attracted a lot of scepticism. But the Nobel committee thought it fit to award the prestigious prize to the pioneers in this area namely Abhijit Banerjee, Esther Duflo and Michael Kremer.

It is an opportune time for Banerjee and Duflo to market their recently published book "Good Economics for Hard Times." Any trial aims to isolate the variables which result in a particular consequence. For example, one may try to find out which variable results in faster reduction in poverty: adoption of Universal Basic Income or incentives for investment.

These trials will yield useful results only if strict criteria are made applicable to understand the cause and effect relationship. If Banerjee's recent utterances are any indication, Banerjee does not seem to apply consistent criteria. He keeps saying that high rates of income tax do not negatively impact economic growth. He arrived at this conclusion because the American economy grew at a fast pace during Dwight Eisenhower's presidency when the marginal tax rate was vertiginous at 95% (Is it not possible that other variables counterbalanced the impact of tax rates? There was no controlled study done.)

Applying the same logic, when Banerjee was asked if the NYAY proposal was an electoral failure because the Congress party which included this in its electoral campaign was comprehensively defeated, his response was disappointingly different. With an apparent self-interest (Banerjee was one of the architects of NYAY),  Banerjee claimed that there were other variables which caused the defeat of Congress at the hustings.

No comments: