Can We Nudge Farmers into Saving Water?

Agriculture and water use

Agriculture is the main user of water in Europe with more than 50% of total water use. In Southern and Eastern countries, water scarcity is a severe challenge which is expected to become even worse in the context of climate change.

With water uses for irrigation already larger than rainfall in a large share of the European territory, finding ways to grow crops using less water is a major challenge.

While agronomists try to find ways to make crops less dependent on water, in this project, we try to curb water use by farmers by using nudges to alter farmers' behavior.

We paired with CACG, a public utility selling irrigation water to farmers, in order to test the effectiveness of nudges to decrease farmers' water consumption using a randomized experiment.

Nudging for green behaviors

Nudges try to leverage psychological mechanisms to change people's behaviour. In recent years, nudges have appeared as a credible alternative to regulation and taxation as a way to make consumers adopt greener practices. Social comparison nudges, that send consumers information about how their own consumption compares to that of their neighbors, have been shown to be effective at decreasing both electricity and water consumption by households. The effectiveness of social comparison nudges to influence the behavior of professional economic agents such as farmers had not been studied prior to our work.

Smart meters

Smart meters enable us to measure farmer's water consumption on a daily basis and to send to farmers information on how their consumption compares to that of their neighbours.

Text-messages

An automated platform for sengind text-messages enables us to send social comparison information weekly to farmers enrolled in our experiment.

A randomized experiment in South-West France

We pair with CACG to send weekly text messages containing information about their on and their neuighbours consumption to a random sample of farmers equipped with a smart-meter.

Randomly selecting the farmers enables us to ensure that the only systematic between the farmers receiving the nudge and the farmers not receiving it (a.k.a. the control group) is the nudge itself. The randomized experiment is the Gold Standard for inferring the impact of an intervention, such as the effectiveness of a drug.

We sent weekly messages over all the summer to farmers enrolled in the experiment and equipped with smart-meters. Farmers in the treatment group received the following message tailored to their own and their neighbors true consumption:

"Hello Mr X. Water conservation is important for your watershed. Please keep on optimizing your irrigation.

On 12/09, you have consumed 45% of your water quota. CONGRATULATIONS! The irrigating farmers of your watershed have used on average 50% of their quota."

Farmers in the control group received a weekly control message in order to avoid placebo effects due to receiving a message related to water use:

"Hello Mr X. Water conservation is important for your watershed. Please keep on optimizing your irrigation."

Encouraging results

A decrease in average consumption

After tracking the consumption in the control group, the consumption in the treated group that received the nudge decreased relative to the control group around mid-august.

Overall, by the end of the irrigation season, cumulative consumption decreased by a few percentage points in the treatment group relative to the control group, wich suggests a modest effect of the nudge.

Unfortunately, our sample size is insufficient to ensure that this decrease is due to the program and not due to sampling error.

Contrasted effects

The proportion of farmers using no water at all has increased among farmers receiving the nudge relative to the control group. This counter-effect, called a boomerang effect in the literature, probably stems from the nudge triggering some farmers to start consuming water.

We also find that the proportion of farmers consuming more than 80% opf their allocated quota has decreased in the treatment group.

Taken together, these effects suggest that our nudge has concentrated farmers' water consumption towards the average.

Nudges and irrigation: which prospects?

Overall, the results from our experiment suggest that nudges can alter farmers' behavior. Three important developments are needed:

Measure the effect of nudges more precisely. Indeed, the size of our experiment (roughly 200 farmers enrolled in both treatment and control groups) is not enough to deliver enough precision to conclude with certainty as to the magnitude of the impacts. Much larger samples of farmers equipped with smart-meters are needed to achieve required precision levels.
Get rid of boomerang effects. Boomerang effects, in our case the nudge triggering water consumption by farmers who would not have consumed water otherwise, is a serious concern. Avoiding such boomerang effects could be done by not sending messages to consumers with zero consumption, or by adding a congratulations message, or a smiley.
Increase the value of the message to farmers. Maybe adding information abour weather forecasts or plant needs, or agronomic advice, would make the message more effective.

In order to know more

To go further, you can consult the published scientific paper reporting these results or its working paper version.

Page updated

Google Sites

Report abuse