Assessing the impact of hermetic storage on marketing behavior and livelihoods of rural households in Ethiopia

Last registered on December 13, 2017

View Trial History

Pre-Trial

Trial Information

General Information

Title

Assessing the impact of hermetic storage on marketing behavior and livelihoods of rural households in Ethiopia

RCT ID

AEARCTR-0002635

Initial registration date

December 12, 2017

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

December 13, 2017, 3:07 PM EST

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

Locations

Country

Ethiopia

Region

East Wollega

Primary Investigator

Name

Hugo De Groote

Affiliation

International Maize and Wheat Improvement Centre (CIMMYT)

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Bart Minten

PI Affiliation

IFPRI

Contact Investigator

Additional Trial Information

Status

In development

Start date

2017-12-14

End date

2018-12-31

Keywords

Agriculture

Additional Keywords

maize, storage, hermetic bags, PICS bags, marketing

JEL code(s)

O13, Q13, Q18

Secondary IDs

Abstract

Introduction: To lift rural household out of poverty, improved market links and integration are. Most rural household produce most of their own food, during one agricultural season per year, so they need to store it to bridge the lean season and protect them from price fluctuations. Recent new technologies, in particular hermetic bags, offer new possibilities as they reduce losses from insects and fungi to almost none. The major benefit of the hermetic bags are, however, allowing farmers to store longer, as the risk of storage losses is removed. In the short run, they can store their cereals longer, and sell or consume them later in the season. In the longer run, they can invest more and increase the production of more susceptible but productive crops such as maize. By increasing production and selling later at a better price, they can substantially improve their income, food security and livelihoods. Therefore, this study is designed to evaluate the impact of hermetic storage, first on the decrease of storage loss, but especially on behavioral change.
Study design: Eight hundred seventy farm households will be selected randomly in one district of East Wellega, an administrative zone in West Ethiopia with high levels of storage losses and high maize producton; 435 households will be given 3 hermetic bags, to store about 300 kg of maize, and provided with training for proper use, the other 435 will serve as control. Data on storage, marketing, storage losses, income and food security will be collected at baseline, right after harvest, and after the end of the storage period, 6-7 months later, over two years.
Expected results. In the first year of using hermetic bags, we expect rural households to increase their consumption, postpone their marketing, and increase their income. In the second year, we expect an increase in area and production, investment in more hermetic bags, and increased income and consumption. In the long run, we expect increased market integration, with farmers increasingly producing for the market.

External Link(s)

Registration Citation

Citation

De Groote, Hugo and Bart Minten. 2017. "Assessing the impact of hermetic storage on marketing behavior and livelihoods of rural households in Ethiopia." AEA RCT Registry. December 13. https://doi.org/10.1257/rct.2635-1.0

Former Citation

Sponsors & Partners

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

Request Information

Experimental Details

Interventions

Intervention(s)

The main intervention is to provide at least three hermetic bags (capacity 90-100kg each), free of charge. SG2000 staff will provide instruction on the use of the bags during group demonstrations and discussions. The training will provide an explicit discussion on how bags can help household improve food security by reducing losses, but also income by allowing sales at a later time, and by increasing the area and production of hybrid maize. During the baseline survey, treatment households will provided with an extra information module on the hermetic bags, including technical details as well as strategic options. During subsequent surveys, treatment farmers will be able to ask further technical questions on the technology.

Intervention Start Date

2017-12-14

Intervention End Date

2018-10-31

Primary Outcomes

Primary Outcomes (end points)

- maize storage losses (%),
- length of storage of maize (days)
- income maize sales (Ethiopian birr)
- food security (Months of Adequate Household Food Provisioning (MAHFP), household food diversity score, household food insecurity access scale (HFIAS) score)
- maize production (kg/year)

Primary Outcomes (explanation)

- maize storage losses (%, measured through count and weigh method or farmer’s estimate) is expected to decrease thanks to hermetic bags which control storage insects
- length of storage of maize (time from harvest to last sales or consumption): because of reduced losses, farmers can store longer.
- Income from maize sales: as stored grain is well protected, farmers can store longer and sell at a better price
- food security: as rural household have less storage losses and higher income, the food consumption, diversity and security increases.
- maize production: as farmers’ income from maize production increases, they can invest more inputs (land, labor, fertilizer, ...) in maize production

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Overview: The main design is a randomized controlled trial, using a stratified simple random sampling design. One district (woreda) will be selected purposely, in which households will be selected randomly. The criteria for selection are:
- The household head and the spouce did not attendedany field demonstration on hermetic bags in the previous years.
- Household has access to land for crop cultivation, and produced at least 300 kg during the last harvest.
- Household head has provided fully documented informed consent.

Based on a household survey executed by IFPRI study, the areas with highest losses in maize fall in the South West maize growing areas in the country, in particular the adminstrative zone East Wollega.
Based on agricultural statistics, key informants and secondary information, one or two woredas will be selected with high maize production and high lossess, accessible, close to but not including SG2000 woredas.
Within each woreda, a set of neighbouring kebeles (lowest adminstrative unit) will be selected. With kebeles as strata, farmers with selected randomly, and evenly divided over treatment and control. To study the effect of sampling density, different kebeles

Experimental Design Details

Randomization Method

The kebeles in the woreda will be determined, with their number of households. The total sample size (870) will be distributed proportionately to the number of households in each kebele. Randomization will be done in the field, after obtaining the list of households of each kebele (smallest administrtative unit). The households will be numbered, each will be assigned a random number, then they will be sorted by that random number. The top first ranked households, up to the number required for that kebele will be selected. Every other household will be assingned to the treatment group, the rest to the control.

Randomization Unit

The unit of randomization is the household. Selection will be stratified over kebeles.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

No clusters

Sample size: planned number of observations

870 households

Sample size (or number of clusters) by treatment arms

870 households

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

To detect a minimum difference of 0.2 standard deviations in a primary outcome with 80% power and a 0.05 significance level, 393 households are needed per arm. This was rounded to 435 households per arm or 870 households in total to allow for a 10% loss to follow-up.

Supporting Documents and Materials

IRB