Transient Dynamics of the COVID Lockdown on India’s Production Network

In the wake of the COVID-19 pandemic, numerous governments imposed productionrestrictions on various sectors of the economy. Prima facia there is reason to believe thatthe cost of these restrictions may be greater than their simple sum. This is because quantityconstraints on production percolate through the supply chain forcing some sectors to reduceoutput because of the non-availability of inputs. This paper develop an agent-based modelof a supply chain to study the impact of the COVID lockdown. The model is capable ofgenerating out-of-equilibrium dynamics in response to shocks like the COVID lockdowns.The model generates two counterintuitive phenomena in response to the lockdowns. Thefirst of which is that the supply of output of some sectors increases during and immediatelyafter the lockdown. And the second is that recovery after the relaxation of the lockdownentails the overshooting of GDP above its normal levels. Furthermore, the size of theovershooting depends on the stickiness of prices. These counterintuitive phenomena are intimatelyrelated to the network interaction between firms as buyers and sellers of intermediateinputs. We calibrate the model to the Indian economy to illustrate both the increase in theoutput of some sectors and the overshooting of GDP in recovery. We also use the model todecompose the total effect of the lockdown into ‘direct effect’ and ‘network effect’, wherebythe ‘network effect’ measure that portion of the reduction in the output of a sector which isbrought about by the lockdown impose on other sectors. We use this decomposition toempirically estimate the ‘network effect’ for each sector of the Indian economy. We findthat there is sizeable heterogeneity among sectors in how their network position amplifiesthe quantity constraints imposed on sectors distantly related to them as buyers-sellers ofintermediate inputs.