An Agent-Based Model to Simulate Meat Consumption Behaviour of Consumers in Britain.

The current rate of production and consumption of meat poses a problem both to peoples' health and to the environment. This work aims to develop a simulation of peoples' meat consumption in Britain using agent-based modelling. The agents represent individual consumers. The key variables that characterise agents include sex, age, monthly income, perception of the living cost, and concerns about the impact of meat on the environment, health, and animal welfare. A process of peer influence is modelled with respect to the agents' concerns. Influence spreads across two eating networks (i.e. co-workers and household members) depending on the time of day, day of the week, and agents' employment status. Data from a representative sample of British consumers is used to empirically ground the model. Different experiments are run simulating interventions of the application of social marketing campaigns and a rise in price of meat. The main outcome is the mean weekly consumption of meat per consumer. A secondary outcome is the likelihood of eating meat. Analyses are run on the overall artificial population and by subgroups. The model succeeded in reproducing observed consumption patterns. Different sizes of effect on consumption emerged depending on the application of a social marketing strategy or a price increase. A price increase had a greater effect than environmental and animalwelfare campaigns, while a health campaign had a larger impact on consumers' behaviour than the other campaigns. An environmental campaign targeted at consumers concerned about the environment produced a boomerang effect increasing the consumption in the population rather than reducing it. The results of the simulation experiments are mainly consistent with the literature on food consumption providing support for future models of public strategies to reduce meat consumption.