Quasi-experimental Methods and Inferences: Two Studies on COVID-19 and Technology Engagement Behavior

Inference of causality in complex socio-economic systems is a challenging task. Decision-making is dynamic and inter-temporal in nature. The driving role of expectation in economic choices makes it difficult to apply the usual idea of causality in the form of a sequence of observed causes and effects. At the same time, observational data comprising economic environments and choices made suffer from endogeneity in the sense that the external environment affects optimal choices, and choices, in turn, may affect the external environment. In this talk, Professor Chakrabarti will discuss some quasi-random experimental methods that allow us to break the problem of endogeneity and make causal inferences about responses of economic agents in terms of both pecuniary and non-pecuniary factors. In particular, Professor Chakrabarti will describe two such natural experiments using Covid-19 as a shock to study how people’s behavior regarding technological engagement changed. The key takeaway from this discussion would be the nature of data and the design of the quasi-experiments that make it possible to apply such causal inference methods based on observational data.