Photo: / Roger Reuver CC BY

Modelling a shift to sustainable diets

Peer pressure was found to be the most important driver for people to adopt a plant-based diet, when researchers integrated psychological theories into a computer model.

A massive shift towards more plant-based diets has in recent years been identified as an important element in reducing greenhouse gas emissions and competition for land. So far, most research has focused on what a sustainable diet would look like. Few studies have explored if and how food cultures with less meat could become widespread on a global scale.

In a recent paper, Sibel Eker et al. use integrated assessment modelling (IAM) to analyse the likely uptake of vegan, vegetarian and flexitarian diets in a larger population, and the environmental benefits that would follow. IAM is a widely used tool for estimating the potential of environmental policy options, since it explores the connections between society and the natural world. What is innovative about this study is the integration of theories from behavioural sciences beyond simple economic factors. Eker et al. use two psychological theories:

  • The Theory of Planned Behaviour, which has been widely used in environmental contexts and focuses on the role of social norms in shaping behaviour.
  • The Protection Motivation Theory, which is widely used in health contexts and focuses on individual risk.

In the case of eating less animal products, the first theory implies that if many of your peers have adopted such a diet you are also more inclined to make a change. The second theory instead focuses on the importance of perceived risks. In this case the health risks associated with a high intake of meat and dairy and the perceived risks from climate-related events. Both theories recognise that individuals who experience obstacles to change are less likely to take action, referred to as low self-efficacy. A schematic illustration of how the researchers think these factors will affect a shift to a vegetarian diet is presented in figure 1.

Figure 1: Conceptual framework of the diet change model. An illustration of the behavioural framework underlying the diet change model. The arrows represent a causal relation between two factors, and the polarity of an arrow indicates whether the relation is positive or negative.


The team modelled a differentiated population taking into account gender, age and educational level. The population was then divided into two groups: those who followed a meat-based diet and those who had adopted a vegetarian diet. To add to the complexity, the researchers also tried four different combinations of diets. In a reference scenario, meat-eaters ate a meat-based reference diet and the vegetarians adopted a lacto-ovo vegetarian diet. In three other scenarios meat-eaters either shifted to healthy eating guidelines (scenarios 1 and 2) or a flexitarian diet (scenario 3) by 2050, while vegetarians kept the lacto-ovo diet (scenario 1) or shifted to a vegan diet (scenario 2 and 3).

To deal with the high range of uncertainty for the variables used in the model, the researchers uses a Monte-Carlo approach. They simulated the model 10,000 times with different combinations of parameter values within the uncertainty range.

Social norms among young people (ages 15–44) was the parameter that contributed to most variance in the output. This supports earlier findings that values endorsed by the peer group have more influence on whether people make a change than actual health and climate risks.

The second most influential parameter is the self-efficacy parameter among females, which is assumed to be higher than that of males. Self-efficacy among those who have completed secondary education also ranked fairly high. It is assumed to be higher for people with higher education, but it is still an important factor, since the former constitute the largest demographic group by educational attainment. This suggests that it is important to strengthen people’s perceptions that their changing eating habits will have an impact.

When it comes to reducing climate impact, it seems that it is more effective to persuade a majority to move from regular meat-eating to a flexitarian diet than for a smaller group to abandon meat altogether. A population of 90 per cent of flexitarians and 10 per cent vegans (scenario 3) will contribute to less greenhouse gases than 50 per cent regular meat-eaters and 50 per cent lacto-ovo vegetarians (reference scenario) (see figure 2).

Figure 2: The climate impact of diet change scenarios in 2050, the results of 10,000 model simulations for 2050 for the percentage of vegetarians/vegans in the total population.

The researchers write that, to their knowledge, this is the first model that couples climate, diet and behaviour. One of the limitations is that it is based on psychological theories and data that are drawn largely from studies of Western industrialised nations. To further develop the model, they plan to collect more data from sources such as social media, as well as focusing on specific cases where cultural values and traditions also play an important role in whether people are willing to adapt their behaviour or not.

Kajsa Pira

IIASA press release 22 July 2019:
Modelling the drivers of a widespread shift to sustainable diets



In this issue