In the Eye of the Storm Firms and Capital Destruction in India

Abstract

This paper examines the response of firms to capital destruction, using a new measure of firm exposure to tropical storms as a negative exogenous shock on firms’ capital stock. Drawing on a panel of Indian manufacturing firms between 1995 and 2006, we establish that, depending on their strength, storms destroy up to 75.3% of the fixed assets of the median firm (in terms of its productivity and industry performance). We quantify the response of firm sales within and across industries and find effects akin to Schumpeterian creative destruction, where surviving firms build back better. Within an industry, the sales of less productive firms decrease disproportionately more, while across industries capital destruction leads to a shift in sales towards more performing industries. This build-back better effect is driven by firms active in multiple industries and, to a large extent, by shifts in the firm-level production mix within a firm’s active set of industries. Finally, while there is no evidence that firms adjust by investing in new industry lines, firms tend to abandon production in industries that exhibit lower comparative advantage.

The Impact of Cyclones on Local Economic Growth: Evidence from Local Projections

“…where w dh is the maximum wind speed (in knots) observed at district d's centroid during cyclone h. To compute w dh we use the values given by storms' best tracks from the National Oceanic and Atmospheric Administration (NOAA) Tropical Prediction Center and feed them to Deppermann (1947) wind field model, as we explain in more details in Appendix A. w max represents the maximum wind speed in the sample and the number 33 corresponds to the Saffir-Simpson scale threshold above which winds qualify as tropical storm. This threshold is reasonable for developing countries (Pelli et al, 2022). 8 x dh ∈ (0, 1), with 0 indicating the absence of winds above the threshold and 1 indicating that the district was exposed to maximum wind speeds.…”

“…While in India a threshold of 33 knots is high enough for winds to cause serious damages on infrastructures (e.g. Pelli et al, 2022), studies in the US have used higher thresholds (e.g. Emanuel, 2011).…”