Food consumption, paternalism and economic policy

The thesis consists of a summary and four papers, concerned with food consumption, behavior associated with overconsumption of food and analysis of the economic policy reforms designed to improve health. Paper [I] estimates a hedonic price model on breakfast cereal, crisp bread and potato product data. The purpose is to examine the marginal implicit prices for food characteristics associated with health. A trade-off exists between health and taste. For instance, sugar, salt and fat are tasty but can be unhealthy if overconsumed; whereas fiber is unhealthy if underconsumed. If the marginal implicit price for sugar is negative, consumers value health over its taste. Our results are the marginal implicit price for sugar is negative for breakfast cereals and crisp bread—consumers value health over the taste of sugar. For salt, we find the opposite—a positive marginal implicit price, suggesting people value its taste over health. For fat, we find a negative marginal implicit price of fat in breakfast cereals and potato products containing salt, whereas we find a positive marginal implicit price of fat in hard bread and potato products that contain no salt. For the one healthy characteristic, fiber, we find a negative marginal implicit price in breakfast cereals and a positive implicit price in hard bread. Paper [II] uses a general equilibrium model to derive the optimal policy if people overconsume unhealthy food due to self-control problems. Individuals lacking self-control have a preference for immediate gratification, at the expense of future health. We show the optimal policy to help individuals with self-control problems to behave rationally is a combination of subsidies for the health capital stock and the physical capital stock. Paper [III] estimates a demand system for grain consumption based on household panel data and detailed product characteristics, and simulate the effect on grain consumption of economic policy reforms designed to encourage a healthier grain diet. Our results imply it is more cost-efficient to subsidize the fiber content than to subsidize products rich in fiber given the goal to increase the fiber intake of the average Swedish household. Our results also imply subsidies alone give rise to an increase in fiber, and to other unhealthy nutrients. Also, subsidies alone have negative effects on the budget. We therefore simulate the effect of policy reforms in which the subsidies are funded either by taxes on the content of unhealthy nutrients or by taxes on products that are overconsumed. Our results suggest that price instruments need to be substantial to change consumption. For instance, removing the VAT on products rich in fiber has little effect on consumption. Paper [IV] explores habit persistence in breakfast cereal purchases. To perform the analysis, we use a mixed multinomial logit model, on household panel data on breakfast cereal purchases. If habit persistence in consumption is strong, short and long-run responses to policy reforms will differ. Our results are breakfast cereal purchases are strongly associated with habit persistence. Our results also imply preferences for breakfast cereals are heterogeneous over households and the strength of habit persistence is similar over educational and income groups