Ex-Ante Input Complementarity vs. Ex-Post Harvest Labor Decoupling
RCTClinical Trial
Agricultural production occurs in distinct stages, with ex-ante planting decisions made before weather realizations and ex-post harvest decisions made after weather outcomes are known [17, 18]. While crop insurance increases the rainfall sensitivity of ex-post harvest labor demand by encouraging ex-ante investments in weather-sensitive inputs, it has zero effect on ex-ante planting-stage labor demand [18-20].
Picture this
Imagine a baker preparing for a festival. Before knowing how many visitors will attend, the baker buys expensive high-yield cake mixes that require sunny weather to bake properly. Preparing the dough in the morning (the planting stage) takes the exact same number of helpers whether it rains or shines. However, if the weather turns out great and a huge crowd arrives, the baker needs to hire a large team of evening workers to decorate and pack all the extra cakes (the harvest stage). Index insurance changes the baker's initial choice of cake mix, but only causes labor hiring to spike during the final packing stage when weather is good.
What the evidence says
Insurance significantly increases the rainfall elasticity of harvest labor demand ($\beta = 0.654, t = 2.39$), but has no statistically significant effect on planting-stage labor demand ($\beta = 0.459, t = 1.41$ or $\beta = 0.191, t = 0.48$ in placebo specifications) [19-21].
- Who was studied
- 1,468 cultivating households owning >0.5 acres across 42 treatment villages in India [7, 21].
- How
- Village fixed-effects regression comparing ex-post harvest labor days against ex-ante planting labor days across insured and uninsured cultivators [19-25].
What to do
Distinguish between ex-ante planting labor and ex-post harvest labor when evaluating labor market impacts of agricultural risk management policies.
From the source
"In contrast to the estimates for harvest labor, the effect of realized rainfall in the village on the difference in demand for planting-stage labor across insured and uninsured cultivators is not statistically different from zero." [20]
Risk, Insurance and Wages in General Equilibrium