Why do farmers sell immediately after harvest when prices are lowest?

Last registered on May 25, 2022


Trial Information

General Information

Why do farmers sell immediately after harvest when prices are lowest?
Initial registration date
May 09, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
May 09, 2022, 8:25 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
May 25, 2022, 9:46 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.


There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information

Primary Investigator


Other Primary Investigator(s)

PI Affiliation
IFPRI Kampala
PI Affiliation
IFPRI Lilongwe
PI Affiliation
PI Affiliation
IFPRI Lilongwe

Additional Trial Information

In development
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
It is often observed that smallholder farmers sell most—if not all—of their marketable surplus or cash crops immediately after the harvest to itinerant traders at the farm gate. Selling immediately after the harvest is not optimal. Thin and poorly integrated markets mean that immediately post harvest, prices in excess supply areas drop. Later, during the lean season when some of the farmers run out of stock, prices have recovered, or even increase further since farmers start to buy back. This leads to the “sell low buy high” puzzle (Stephens and Barrett, 2011; Burke et al., 2018). In addition to high supply immediately post harvest, agricultural commodities are often not yet in optimal condition. For instance, in the case of maize, fresh grains are generally not dry enough, requiring further processing and leading to increased risk of rot by the trader. Often, this is used by buyers as a reason to further drive down the price paid to the farmer. In this study, we zoom in on three potential behavioural explanations why farmers seemingly sell at sub-optimal time. One potential explanation is situated at the household expenditure side, and assumes that households face challenges in accurately predictive future expenditures. Such budget neglect leads farmer to sell more early on and save too little for later in the year. A second potential explanation is situates at the household income side. Here the assumption is that farmers face cognitive challenges in making inter-temporal cost benefit calculations (Drexler et al., 2014) and fail to commit to certain thresholds (Ashraf et al., 2006; Duflo et al., 2011).
External Link(s)

Registration Citation

Brian, Dillon et al. 2022. "Why do farmers sell immediately after harvest when prices are lowest?." AEA RCT Registry. May 25. https://doi.org/10.1257/rct.9410
Experimental Details


A first intervention tests budget neglect and comprises of an intervention that takes the farmer through a detailed budgeting exercise. The budget exercise will involve three components. A first component uses recall to provide a first approximation of what will be necessary in the future. A second component consists of segmentation, which involves defining categories of expenditures for cognitive ease. Finally, we will look at a range of risks, which involve expenses that are not certain but may materialize. We try as much as possible to attach objective probabilities to these risks and also incorporate this in the budget.
In a second treatment, we will develop, together with the farmer, a detailed plan of how much the farmer will sell over the coming year (per month or per quarter). For each sales event, the farmer will also be asked to commit to a minimum price. This will be done on a special form that farmers can than hang up in their house.
In a third treatment we will develop an intervention where we provide farmers with historical (5 years) monthly price movements for maize, soybean and groundnuts. The price data will be for the market that is closest to the treated household. We will show price movements, but also provide summary statistics like minimum price, maximum price and average price to summarize price distribution data in a few easy to understand figures (Hanna et al., 2014; Drexler et al., 2014). Price data will be obtained from the Malawian Agricultural Commodity Exchange (ACE).
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Primary outcomes in this study include stocks of ground nuts, maize and soybean held by the farmers and how they evolve over time. As there is a particular focus on marketing behaviour, we will also collect detailed information on sales made, including quantities sold, prices received, who was sold to, who made the decision and what were proceeds used for. As such, questions during follow up and endline on market participation will be similar to the recall data that was collected during baseline. Further down the impact pathway, we compare welfare, both subjective and through consumption expenditure (last month), between treatment and control households. However, detailed consumption expenditure data will only be collected during endline in March 2023 to avoid priming the budget neglect treatment to farmers in the control group.
Primary Outcomes (explanation)
To investigate impact pathways, we will also include a range of questions related to expenditure, and how easy it was for farmers. For instance, did treated households have less issues in meeting expenditures for eg. fertilizer or improved seed for the next season? Furthermore, we include a module on price expectations, which will be useful to see how expectations influence eventual prices obtained, and how interventions affect the relation between expectations and behaviour.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We use a parallel design with 4 treatment arms (one pure control and three treatments).
Experimental Design Details
Not available
Randomization Method
Randomization done in office by a computer
Randomization Unit
individual - but clustered in villages (120)
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
0 (individual level)
Sample size: planned number of observations
Sample size (or number of clusters) by treatment arms
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
IRB Approval Date
IRB Approval Number
IRB Name
IRB Approval Date
IRB Approval Number
Analysis Plan

Analysis Plan Documents

Why do farmers sell immediately after harvest when prices are lowest? A pre-analysis plan

MD5: 471031c691c54047a06c2a68c5f03912

SHA1: 1d34147a7af00ef1b676a47e093f6d09bc95305b

Uploaded At: May 25, 2022