Blind Data

“In God we trust. Everybody else must bring data.”- NR Narayana Murthy, Chief Mentor, Infosys Technologies

The stand-out session for me at the recently concluded MDG summit was the one organized by the World Bank and titled, “Impact Evaluation”. The panel included Michael Kremer, Gates professor at Harvard University, who was literally the only person from academia that I met during the entire week at the Summit. The rest of the panel included folks from the World Bank, and was chaired by the very impressive Michael Anderson, Director-General of Policy and International Relations at DFID. The subject of discussion- data, and how it drives policies today.

Each expert presented a point of view from their respective fields of interest. The central theme was the relevance and importance of generating data to drive policy, rather than assumptions and “common sense”. Of course, after one year at Harvard, this manner of thinking wasn’t new to me. There is such a strong emphasis on data that one may be forgiven for the RCTs-for-parachutes jibe that has become part of the lore of research humor. Sitting back, though, it’s quite incredible how radically different this manner of thinking is to me, compared to what I was exposed to in medical school.

India isn’t really the hotbed of research. For all our prowess in math, and chauvinistic claims of having gifted the world the concept of zero, the sad fact is that, as a people, we do not think in terms of numbers. This, however, I have learned, is not a problem restricted to India, or even the subcontinent. It is a problem in Africa, a problem in South America, and definitely a problem in Asia. In fact, it might be interesting to study the relationship between per capita income and the reliance on data in determining policy within a country. My guess is that there will indeed be a correlation, although I’m not so sure about which would be the chicken and which would be the egg.

In the clinical world, the practice of eminence-based medicine rather than evidence-based medicine is the pejorative phrase that’s used to describe the philosophy of “I’m the boss, and I told you so” in designing patient care. Protocols and standard guidelines get thrown out the window while the heuristic practice of policy replaces them. Policy-makers and decision-takers routinely make their choices based on their personal experiences, irrespective of what the published literature suggests. This gives rise to statements like, “In my experience, this works.” Of course, the fallacy of their sample size of one could be pointed out to them, but for somebody with less training in statistical methods than a 3 year old, the argument would be a lost cause.

It’s time this changed. It’s time students were taught to start thinking in terms of numbers and hard evidence. I’ve previously blogged about my attempts to rectify the anomaly by setting up IMPULSE. The enthusiasm of the uptake of this effective intervention shows great promise for replication on a larger-scale, and as a community, the development world would do very well to sell the idea to schools and colleges in the developing world. This would enable us to develop the in-country expertise of generating data required to formulate sound policy, without having to resort to the expert advice of expensive consultants. Indeed, local capacity building is the most effective means of cost-cutting in international development.

Data is the final leveler. Whether you conduct your research in the finest labs of Harvard, or in the most trying circumstances on the field, in India, as long as your data and methodology is sound, you have the upper hand. It’s very encouraging that the world has now started to think robustly in terms of using data as the prime determinant of policy. The chair of the World Bank panel remarked with satisfaction that he was surprised that not a single member of his audience stood up, and claimed that data should be second to on-field experience in determining policy, as he expected. Mr. Director-General, I believe our trends are promising, but your sample was skewed. New York City is a world away from the chaos that is Mumbai. You should know; you work in New Delhi.