2020-2022: Replication Project

European farmers are risk-averse and want to avoid losses

Our new replication study finds that farmers across eleven European farming systems avoid risks. Farmers also perceive losses stronger than gains and distort small probabilities. Our study contributes to the credibility revolution in economics by showing that risk aversion, loss aversion, and probability weighting are present in many farming systems.

Our motivation – studying risks in agriculture

Farming is a business full of uncertainty. Risk is not only about whether it will rain enough or whether a pest will ruin crops. Farmers also have little control on the price they pay for inputs or on how much they will receive for their produce, not to mention the risk of policy makers changing laws and regulation. To understand the agricultural sector, economists want to set up models which account for these risks, and they also want these models to include a good understanding of how much farmers would like to take or avoid risks. A first step for this is to know how risk loving or risk averse farmers are. As there are many ways to measure risk attitudes, another challenge is to make sure that the same method is applied to all farmers to get comparable results. 

Our team of 28 researchers from a network on the promotion of economic experiments for policy-making has recently studied risk attitudes of 1,430 farmers from eleven European countries. We studied arable farmers in Austria, Germany, the Netherlands, and Sweden; wine growers in Croatia; potato farmers in Northern France; organic farmers in the North-West of France; olive growers in Apulia (Southern Italy) and Andalusia (Spain); young farmers in Slovenia; and farmers of different specializations in Poland.

The method 

We replicate a study with a well-known method that uses risky gambles, in the form of lotteries, to learn about farmers’ risk preferences. This approach provides a general understanding of farmers’ risk attitudes that can serve as an input for policy simulation models. However, abstract lotteries might not perfectly capture risk preferences, as these might differ by context (for example a person may have different attitudes towards risk in farming compared to the attitudes towards personal health risks such as smoking).

With the lottery we implemented we can see whether farmers shy away from risks (risk aversion), care more about losses than gains (loss aversion), and think that improbable events happen more often than they actually do (probability weighting).

Our results

In the figure below, risk aversion describes the tendency to prefer less variable outcomes over more variable ones, even if the average outcome of the latter is equal or higher. This would be reflected by farmers preferring to receive 200 Euro for sure than to flip a coin and receive 100 Euro for heads and 300 Euro for tails. 

Loss aversion means that risks are assessed differently for losses and gains. Losses induce more pain in absolute value than gains of equal sizes. For instance, and typically, a 200 Euro gain leads to an increase in well-being while a loss of the same size leads to around a twofold decrease in well-being. 

People also tend to give more importance to events that are very improbable. An event that would happen only once in a thousand times (i.e., the objective probability is 0.1%) is, for example, perceived to happen once in a hundred times (i.e., the subjective probability is 1%). This is important to measure as it indicates why some farmers may be worried about floods or hail, which may not occur as often as they expect. We estimate a parameter that reflects the degree to which this happens, the smaller the value of this parameter, the more distorted the perception of small probabilities.

The graph presents an average of a previous study with French farmers (BJR 2014) and estimates for all eleven newly collected samples. 

A presentation summarizing the project findings