THREE ESSAYS ASSESSING THE ECONOMIC IMPLICATIONS OF HEAT STRESS IN LABOR
This dissertation assesses three aspects of the economic implications of heat stress-related labor-capacity losses. Given that low-income countries around the tropics are at most risk, our analyses focus on these and the vulnerable households within them. First, we consider the optimal allocation of labor for small-scale agricultural households. We build an agricultural household that takes into consideration that these households will be affected by heat stress as producers, consumers, and workers simultaneously. Using this model and a sample of households from Pakistan, we determine that for most households it would be optimal to increase their supply of family labor to agricultural self-employment. However, if work preferences are also affected, even modestly, then decreased supply of family labor to agriculture would be observed.
Next, we turn to country-level welfare losses across the globe focusing on the role of trade in mitigating or exacerbating these. We consider nine West African economies and determine which benefit from international trade, which are made worse-off, and we fully delineate the factors and channels that determine this. Broadly, we find that net exporters of agricultural commodities will benefit via global price changes, and conversely net importers will be made worse off by global price changes. However, countries that experience especially large labor capacity losses in their export sectors can also see loss-mitigating effects from trade as their export prices rise more sharply that the global average. An alternative perspective shows that some countries are affected more by their own heat stress-related productivity losses, while others are affected more due to global changes.
Lastly, we consider the poverty impacts of heat stress-induced labor capacity losses in West Africa. Using a macro model, we determine changes in real incomes of households near poverty in seven West African countries, then use household microsimulations to determine poverty impacts. We find that poverty impacts are heterogenous in direction and magnitude across household-types and countries. In five of the seven countries, poverty headcounts increase, ranging from 1.5% in Cote d’Ivoire to 7.8% in Nigeria. In two countries, there is either little change or a decrease in poverty: in Cameroon poverty increases by 0.6% and in Guinea it decreases by 1.7%. The key channel behind this heterogeneity is how loss of labor productivity affects relative returns to factors of production. Returns to unskilled agricultural labor can increase due to increased demand for this labor to dampen losses of agricultural output.