Stage-Specific Labor Demand Divergence

In Simple Words

Imagine a farmer preparing a field for planting before knowing if the rainy season will be good or bad. Since planting happens early, weather insurance does not change how many workers are needed to sow seeds. However, once the rains fall and crops grow, insured farmers who planted high-yield crops need significantly more hands to harvest the abundant crops compared to uninsured farmers.

Research question

Agricultural production involves distinct sequential stages (planting versus harvest), but empirical risk models often aggregate labor inputs, obscuring whether weather insurance alters ex-ante planting investments or ex-post harvest decisions.

Participants

N = 1,468 landowning cultivator households (>0.5 acres) across 42 rural villages in Andhra Pradesh, Uttar Pradesh, and Tamil Nadu.

Study Design

Clustered RCT evaluating harvest-stage labor days versus planting-stage labor days (sowing, soil preparation, transplanting) as a placebo test using village fixed-effects panel regressions.

Results

Insurance interaction with rainfall deviation increases harvest labor demand by 0.654 days per mm/day (t-stat = 2.39), while planting-stage labor demand shows no statistically significant response (0.459 days per mm/day, t-stat = 1.41; and 0.191 in full specification, t-stat = 0.48).

Takeaway

Evaluate agricultural labor policies separately across planting and harvest stages to prevent mistaking harvest-stage labor demand shifts for planting-stage employment changes.

Read alongside the research

Plain-English glossary

Original explanations, informed by the references below. Examples are illustrations, not data from the studies above. The research text stays unchanged.

Randomized controlled trial

A study that uses chance, like drawing lots, to decide who receives each option being compared. Researchers compare the groups’ outcomes to estimate the effect of those options.

ExampleA lottery assigns volunteers to a training programme or a comparison group. Researchers then compare their outcomes.

GET-IT glossary · Randomized study ↗Read more about randomized controlled trial →

Statistically significant

A label used when a result passes a chosen statistical threshold, often p < 0.05. It does not show how large or useful an effect is. A result that is not statistically significant does not prove there is no effect.

ExampleA study reports p = 0.03 against a 0.05 threshold. It passes that test, but the effect could still be too small to matter.

American Statistical Association · Statement on p-values ↗Read more about statistically significant →

Related findings

• Cultivator Labor Demand Sensitivity →Development

  Think of a farmer playing a game with two paths: a safe path that gives a small guaranteed meal, and a risky path that gives a massive feast if it rains but nothing if it dries up. Offering weather insurance acts like a safety net: if the risky crops fail due to drought, the insurance company pays out cash.

• Anticipatory Human Capital and Child Nutrition Investment →Development

  Imagine planting two types of fruit trees in a garden where fruit tree A sells for high prices in the future market while tree B sells for nothing. Gardeners naturally water, fertilize, and prune tree A far more diligently long before it bears fruit because they anticipate harvesting a lucrative crop years down the road.

• Asset Accumulation and Savings Persistence →Development

  Think of a short-term public works program like a heavy rainstorm over a dry farm. If a farmer lets the water run off, the soil dries up as soon as the storm ends.

• Asset Wealth and Cash Liquidity Decoupling →Development

  Imagine a rancher who owns ten valuable oxen but has no paper cash in their wallet. Even though they are wealthy on paper, they purchase smaller quantities of commercial seed because selling a working ox to buy a bag of seed is impractical.