A group of people walking by the Golden Temple | By Wong She Fei, Adobe.
The Mittal Institute’s Delhi Office has long been focused on fostering collaborations – both in-country and beyond. They recently had a visit from S.V. Subramanian, a Professor of Population Health and Geography at Harvard University, and chair of the Faculty Advisory Group for the Center for Geographic Analysis at Harvard University. He is also the Principal Investigator of the India Policy Insights program, based out of the Geographic Insights lab at Harvard. This interdisciplinary research lab uses fine-grained data to bring important insights and accountability to policy actions. Professor Subramanian is based at the Mittal Institute’s Delhi office while he works on the program, where he is building collaborations with stakeholders, including the government of India. He is also exploring how he can collaborate with The Mittal Institute on his current and upcoming projects. He explained more about the program during a recent interview with the Mittal Institute.
S.V. Subramanian, Principal Investigator of the India Policy Insights program.
Mittal Institute: Thanks so much for speaking with us, Professor Subramanian! Can you share with our community the motivations behind forming India Policy Insights? What do you endeavor to accomplish, and who do you collaborate with?
S.V Subramanian: India Policy Insights is based out of the Geographic Insights lab at Harvard. It is an interdisciplinary research lab based at the Harvard Center for Population and Development Studies and the Harvard Center of Geographic Analysis. Our aim is to improve precision public policy, public financing, and governance in India related to indicators of population health and development. By leveraging novel geospatial and statistical methodologies, a major aim of the project is to collate and analyze population health and development data geocoded to multiple micro and macro public policy units and make robust predictions at each level.
Another aim of the project is to foster evidence-based policy discussion, formulation, and action by disseminating this targeted data to the public domain and engaging with multiple stakeholders. Having accurate and up-to-date data on population health and development publicly available for all elected representatives can improve political leadership and accountability in India.
Korea University is also our collaborative partner and my colleague, Dr. Rockli Kim, is a co-investigator on this project. In India, our collaborators are NITI Aayog, the International Institute for Population Sciences and the Lal Bahadur Shastri National Academy of Administration.
Mittal Institute: What are your personal research interests, and what is the broad scope of the project?
S.V Subramanian: India has launched many programs and faces many challenges across a variety of sectors. I am personally interested in sectors that relate to health, nutrition, population, and development. Currently there is a strong push towards making informed, data-based decisions. The India Policy Insights project has led to the creation of a comprehensive, cross-cutting, data science platform which incorporates approximately 7,500 different policy indicators across sectors. The platform allows users to visualize and analyze the status of different indicators, making it easier to prioritize areas for intervention.
The project will have a comprehensive dashboard at four geographic scales: districts, parliamentary constituencies, assembly constituencies, and eventually, villages. View an example of a previous data-driven dashboard here. The idea behind the platform is to empower district collectors in the field, by enabling them to use data for their day-to-day decision making; for example, deciding which village or subdistrict requires intervention. In addition to this data being of academic interest, these indicators are also being monitored by the government.
Mittal Institute: You mentioned India faces challenges across sectors – is one of these availability of data? Can you walk us through the current status of data in India?
S.V Subramanian: Although there is plenty of information available at the state and district level, data is not readily available below the district level. Any public policy provision depends not only on how efficiently you can deliver a program, but also on how equitably you can deliver it. Both would involve some degree of prioritization.
For example, if there are 1,000 villages in a district, then the district collector would need to know how to prioritize his efforts to a subset of those villages – and our project aims to help them identify those villages. India Policy Insights also focuses on sub-districts and assembly constituencies besides districts and parliamentary constituencies. Currently in India there are 700 districts, 4,000 assembly constituencies, and half a million villages. Since people live in villages and not in districts, we are trying to gather data at that level.
Mittal Institute: What challenges and opportunities does the collection of data present?
S.V Subramanian: The India Policy Insights project is continuously producing value-added data and we recently published a paper on the use of machine learning models (view a past PNAS paper here). We often curate data from census variables to develop prediction models. By supplementing National Family Health Survey data with census information, we can curate data for many of India’s villages. All the value-added data is produced using a combination of statistical modelling, machine learning and geographical information systems. When collating data, it is important to differentiate between prevalence and headcount, as policy is not always about data but about optimization. The absolute burden increases in areas of low prevalence even if the area is densely populated. To better identify areas that require prioritization, we created a new metric called the Prevalence Headcount Metric.
S (“Subu”) V Subramanian is a Professor of Population Health and Geography at Harvard University, and chair of the Faculty Advisory Group for the Center for Geographic Analysis at Harvard University. He is a Primary Faculty in the Department of Social and Behavioral Sciences at the Harvard T.H. Chan School of Public Health, a Core Faculty of the Harvard Center for Population and Development Studies, and a Faculty Affiliate of the Harvard Department of Sociology. He is also an Honorary Senior Fellow of the National Institution for Transforming India (NITI) Aayog, Government of India and the PI of the Geographic Insights Lab. Prof. Subramanian was the Founding Director of Graduate Studies for the interdisciplinary PhD program in Population Health Sciences at Harvard. He received his under- and post-graduate training at the University of Delhi, and completed his PhD in geography at the University of Portsmouth. He has published over 700 articles, book chapters, and books in the broad field of population health and well-being and in applied multilevel statistical methods.