Research analysis
Value crop output at constant prices
- Select crops: at least 25 growers; over 0.25% of farm area
- Cap top 2.5% of yields within each crop
- Value cleaned output at median prices; divide by cultivated area
Illustrative summary of the source-described yield valuation for 568 Kenyan households. Crop output uses cleaned yield multiplied by planted area; constant prices are median reported sale prices. No treatment estimate is shown.
See source →How
When to use
Measuring agricultural production value in rural smallholder trials is prone to extreme outlier yields and localized price fluctuations across markets. Comparing output using variable farm-gate prices can confound productivity gains with price shifts or local market friction. This method standardizes crop yield valuation across study participants using top-tail winsorization and constant median prices.
Standardize crop yield values across study participants to evaluate treatment impacts on agricultural output.
- Restrict yield analysis to crops grown by at least 25 farmers where land allocation exceeds 0.25 percent of total farm area.
- Clean reported crop outputs per acre by winsorizing individual farm yield observations at the top 2.5 percent level within each crop category.
- Calculate total farm output per crop by multiplying cleaned per-acre yields by reported planted land area.
- Obtain a standardized constant crop price by calculating the median self-reported sale price across all market categories for each crop.
- Compute total agricultural output value per acre by summing crop-level valuations across all grown crops and dividing by total cultivated land area.
Why it works
Applying a constant median crop price isolates physical output gains and crop composition changes from local market price variation.
Example
A researcher valued farm production for 568 smallholding households in Migori County, Kenya following a watchman trial. Winsorizing yields at the top 2.5% and applying constant median prices revealed that watchman provision increased total agricultural output value by 5,002 KES per acre.
About the study
Participants
568 smallholding farming households across 76 villages in Migori County, Kenya.
Study Design
Methodological data cleaning protocol (2.5% top-tail yield winsorization and constant median-price valuation) applied to endline agricultural survey data in a cluster-randomized trial.
Results
Applying this valuation protocol demonstrated a statistically significant 5,002 KES per acre increase in agricultural production value (p < 0.10) on a control mean of 30,694 KES per acre.