Final Report Summary - MAPFAN (Microeconomic Analysis of Prices, Food and Nutrition)
Diet-related chronic diseases are a major public health concern. Addressing this concern is a key government policy objective. Poor diet is one important cause, and a number of policies are aimed at improving diet. The impact of these policies on diet, and thus on health outcomes, will depend crucially on how consumers adapt their food purchasing behaviour in response to the policy, and on how firms in turn respond in terms of the prices they set and the foods they offer.
In this project we have studied consumer food purchasing behaviour and the behaviour of firms in the food industry. We have modelled and provided evidence on how consumer choose among many differentiated products (e.g. what potato chip product to buy) depends on the prices they face, the nutritional attributes of products and the consumers’ exposure to product level advertising. We have extended this class of discrete choice demand models by adding flexibility in how purchase behaviour varies with total expenditure and in the curvature of demand. This allow these class of models to better predict the impact of counterfactual policy (e.g. changes in taxation) and better predict the distributional impacts of policy change. We have also studied demand at the grocery basket level, looking at what explains cross country differences in purchasing behaviour and at how consumers adjusted their spending over the Great Recession. On the supply side, we have modelled how firms that have market power choose their strategies over prices and advertising. Modelling both sides of the market has allowed us to consider policy changes (e.g. restricting advertising, introducing new taxes) taking into account the response of both consumers and firms.
We have developed a novel approach to modelling demand that nests commonly used models in characteristics space and allows for the main features of classical demand systems over continuous quantities of many goods. The model relaxes the weak separability assumption that is typically used in demand estimation, by creating an interaction between products through the characteristics they supply. We use the model to study demand for food and nutrients across three countries - the UK, France and the US. We show that consumers in the three countries purchase very different food baskets, in part due to differences in prices and nutritional characteristics, but importantly also due to inherent preferences and eating habits.
We have analysed the likely impact of a number of different proposed policy reforms targeting alcohol consumption, sugar consumption, fat consumption, diet and exercise and the effects of income support and other economic policies on nutritional wellbeing over the recession. Our research has often highlighted the potentially perverse effects of some of the policies, and has helped to inform and influence the public debate.
In this project we have studied consumer food purchasing behaviour and the behaviour of firms in the food industry. We have modelled and provided evidence on how consumer choose among many differentiated products (e.g. what potato chip product to buy) depends on the prices they face, the nutritional attributes of products and the consumers’ exposure to product level advertising. We have extended this class of discrete choice demand models by adding flexibility in how purchase behaviour varies with total expenditure and in the curvature of demand. This allow these class of models to better predict the impact of counterfactual policy (e.g. changes in taxation) and better predict the distributional impacts of policy change. We have also studied demand at the grocery basket level, looking at what explains cross country differences in purchasing behaviour and at how consumers adjusted their spending over the Great Recession. On the supply side, we have modelled how firms that have market power choose their strategies over prices and advertising. Modelling both sides of the market has allowed us to consider policy changes (e.g. restricting advertising, introducing new taxes) taking into account the response of both consumers and firms.
We have developed a novel approach to modelling demand that nests commonly used models in characteristics space and allows for the main features of classical demand systems over continuous quantities of many goods. The model relaxes the weak separability assumption that is typically used in demand estimation, by creating an interaction between products through the characteristics they supply. We use the model to study demand for food and nutrients across three countries - the UK, France and the US. We show that consumers in the three countries purchase very different food baskets, in part due to differences in prices and nutritional characteristics, but importantly also due to inherent preferences and eating habits.
We have analysed the likely impact of a number of different proposed policy reforms targeting alcohol consumption, sugar consumption, fat consumption, diet and exercise and the effects of income support and other economic policies on nutritional wellbeing over the recession. Our research has often highlighted the potentially perverse effects of some of the policies, and has helped to inform and influence the public debate.