Effect of social information on farmers' irrigation decisions.

Last registered on July 13, 2017

View Trial History

Pre-Trial

Trial Information

General Information

Title

Effect of social information on farmers' irrigation decisions.

RCT ID

AEARCTR-0002283

Initial registration date

July 07, 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

July 12, 2017, 8:37 PM EDT

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

Last updated

July 13, 2017, 3:11 AM EDT

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

Locations

Country

France

Region

Primary Investigator

Name

Philippe LE COENT

Affiliation

INRA

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Julie Subervie

PI Affiliation

INRA-LAMETA

Contact Investigator

PI Name

Arnaud Reynaud

PI Affiliation

INRA-TSE

Contact Investigator

PI Name

Sylvain Chabé-Ferret

PI Affiliation

INRA-TSE

Contact Investigator

Additional Trial Information

Status

On going

Start date

2017-06-01

End date

2017-12-01

Keywords

Agriculture, Environment & Energy

Additional Keywords

JEL code(s)

Secondary IDs

Abstract

Social information has been shown to influence a diversity of human decisions : household energy and water consumption, charity giving, student alcohol use... These behavior changes however entail limited monetary costs or even reduce costs for the decision maker. Does the effect of social information stand in the context of costly decisions?
Irrigation is a key input to ensure optimal crop yield in many contexts throughout the world. Yet, the reduction of water use by farmers is a key challenge considering the current and expected modification of rainfall patterns and the resulting reduction of water resources available for irrigation. However, reducing water use may potentially incur significant costs for farmers by reducing yields. In this experiment, we intend to test whether providing social information may contribute to the reduction farmers' water use for irrigation. The potential effect of peer behavior relies on the assumption that farmers have upward biased beliefs on the behavior of others. Correcting these biases may incline farmers to reduce water use for two reasons: i) to conform to the social norm and/or ii) because they may be reassured by the fact that others are reasonable and are not over-consuming water.
Our experiment takes the opportunity of the installation of irrigation smart-meters in 3 watersheds in the South West of France. These smart-meters communicate daily to the Irrigation Scheme Manager (ISM) the amount of water used by farmers. The principle of the experiment will therefore be to test the effect of providing weekly social information on water use coming from these smart-meters on individual farmers' irrigation decisions.

External Link(s)

Registration Citation

Citation

Chabé-Ferret, Sylvain et al. 2017. "Effect of social information on farmers' irrigation decisions.." AEA RCT Registry. July 13. https://doi.org/10.1257/rct.2283-2.0

Former Citation

Chabé-Ferret, Sylvain et al. 2017. "Effect of social information on farmers' irrigation decisions.." AEA RCT Registry. July 13. https://www.socialscienceregistry.org/trials/2283/history/19389

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 principle of the intervention is to provide information coming from smart-meters to farmers on i) their level of water use and ii) the level of water use of other farmers of the same watershed. This information will be provided weekly via mobile texts by the Irrigation Scheme Manager. Farmers without smart-meters will only receive information on the water use of others. Water use in cubic meters is not an appropriate indicator because smart-meters can be associated to a different amount of land and different type of crops. Each smart-meter is however associated with a water quota. We therefore considered that the relevant and comparable indicator of water use is the percentage of attainment of the water quota. It will be used in the communication with farmers and as the indicator of water ised/

Intervention Start Date

2017-07-10

Intervention End Date

2017-09-30

Primary Outcomes

Primary Outcomes (end points)

As mentioned in the "intervention" section, the main indicator of water use in this experiment is the percentage of attainment of the water quota (%Quota). The outcome which will be examined is the weekly evolution of % Quota (for farmers with smart-meters) as well as the % Quota at the end of the intervention.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

In this experiment different RCTs are carried out. In both RCTs, treatment and control group receive weekly texts by mobile.
The first RCT is carried out with farmers that are equipped with a smart-meter. The control group receives a standard message requesting farmers to limit their water use. The treated group receives weekly information on his personal attainment of the irrigation quota (individual %Quota) and the average attainment of the quota of farmers of his watershed (social information on % Quota).
The second RCT is carried out with farmers that are equipped with standard water-meters. The control group receives a standard message requesting farmers to limit their water use. The treated group receives weekly information only on the average attainment of the quota of farmers of his watershed (social information on % Quota coming from farmers equipped with smart-meters).

Experimental Design Details

Randomization Method

In both RCTs, the sample is stratified by farmer type (collective or individual structure), watershed, size of the quota and percentage of attainment of the quota in 2016. Treatments are randomly assigned in each strata (28 in the first RCT and 23 in the second one) in office with a computer.

Randomization Unit

Individual farmer randomization is used in both RCTs.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

RCT with smart meters: 200 farmers. RCT with traditional water-meters: 261 farmers. There is no cluster in this study.

Sample size: planned number of observations

RCT with smart metters: 200 farmers. RCT with traditional water-meters: 261 farmers

Sample size (or number of clusters) by treatment arms

RCT with smart metters: 200 farmers. RCT with traditional water-meters: 261 farmers. There is no cluster in this study.

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

The regression with fixed effects on strata reveals a standard error of .0339 for the nudge. The minimal detectable effect on the total water consumption is therefore 0.0952, which represents 28% of the average water consumption (in % of the quota). This power may be improved considering that weekly consumption will be subsequently analyzed.

Supporting Documents and Materials

IRB