The Workforce Paradox: Do Extreme Natural Disasters Accelerate or Undermine Labour Productivity?
Description
Using Jamaica as a testing ground, this study applies synthetic control methods to decouple and estimate the causal, short-run and long-run dynamic effects of Hurricane Gilbert, which made landfall in 1988, on the country’s labour productivity using aggregated data. The findings reveal a non-linear impact: labour productivity increased by an average of 14.6% in the short run (2 to 4 years post-Gilbert) but declined significantly in the long run by an average of 33.70%. On average, labour productivity would have been approximately 21.1% higher without Gilbert and its aftermath. Moreover, Gilbert exerted a more harmful effect on the output than the employment channel, whereas exchange rate depreciation, increases in remittance inflows and stronger migration pull factors are adverse moderators. Despite the current donor retreat from providing foreign assistance, the findings suggest external aid is crucial in the immediate aftermath of a natural disaster to stave off labour productivity reduction.
Files
Steps to reproduce
We collated an annual country-level panel data set from 1980 to 2021. The outcome and covariates are from The World Bank, The Conference Board, and Penn World Tables version 9.1. We collect disaster data for all countries from EM-DAT (2019) and the GDIS database. The Replication folder contains the readme file, datasets and well-detailed STATA and Python codes to facilitate easy reproduction of the research results and figures.
Institutions
- University of the West Indies at Mona
- University of Calgary