Model-Selection Inference for Causal Impact of Clusters and Collaboration on MSMEs in India

Do agglomeration-based spillovers impact firms more than the technical know-how obtained through inter-firm collaboration? Quantifying the effect of these treatments on firm performance can be valuable for policy-makers as well as managers/entrepreneurs. I observe the universe of Indian MSMEs inside an industrial cluster but with no collaboration (Treatment Group 1), those in collaboration with other firms for technical know-how but outside a cluster (Treatment Group 2) and those outside cluster with no collaboration (Control Group). Selection of firms into these treatments and sub-sequent performance of the firm may be simultaneously driven by observable factors. To address selection bias and overcome model mis-specifcation, I use two data-driven, model-selection methods, developed in Belloni et al. (2013) and Chernozhukov et al.(2015), to estimate causal impact of the treatments on GVA of ˝rms. The results suggest that ATE of cluster and collaboration is nearly equal at 30%. I conclude by offering policy implications of the results.



Samarth Gupta
Dec 2020

Macro