Obesity has nearly tripled since 1975, becoming one of the world’s leading risks of death. Almost a fifth of French adults are obese. The TSE Health Center supports a multidisciplinary approach to such problems that require careful analysis of the interplay of individual, corporate and societal choices about healthcare, nutrition, and our living environment. In the spirit of this ambitious project, a new paper by TSE’s Céline Bonnet and Valérie Orozco examines the impact of food, sports and health amenities on obesity in France.
How can economists help to address the obesity epidemic?
Obesity’s underlying factors are related to individuals’ choices – about diet and exercise in particular – and their genetic, epigenetic and metabolic profiles. External factors, or the living environment to which the individual is exposed, also play an important role. Economists study the way that choices made by individuals or firms are constrained by their social, spatial, and organizational interactions, which might have consequences for the prevalence of obesity. The consequences of the environment then need to be addressed by public policies.
How do you investigate the health impact of the living environment?
Our paper analyzes how obesity rates are linked to where people live, and their access to food, sports, and health amenities. Regional habits and culture may influence consumption patterns. The location of amenities may also induce differences in accessibility and generate spatial spillover effects. To investigate the impact of the living environment on lifestyle, consumption patterns and health status, it is important to account for spatial links between geographical areas.
We provide an original database that merges information on the health status and food consumption of French households with living environment data about food, sports and health amenities for 281 spatial divisions in France. We also provide control variables to account for income, education level, age and employment status. We then analyze the association between obesity and the living environment in France using a spatial econometric framework.
What picture of French spatial and consumption disparities emerges?
Obesity is less prevalent in the western region, Bretagne, and the southeast region than it is in the north and center west regions. The per capita daily total caloric intake is 3303 kcal on average and generally higher in non-urban areas. The average intake from fresh fruit and vegetables is 252 kcal, with a large difference between north and south: individuals in the southeast, who are closer to agricultural areas, consume more fruit and vegetables. We also find spatial disparities in food prices.
On average, there are approximately 0.2 supermarkets and 0.02 fast-food restaurants per 1,000 people. Both tend to be more accessible in cities. There is spatial heterogeneity in the supermarket ratio, but it is relatively small. Driving time to a supermarket from communes without supermarkets is shorter in the northwest. For health amenities, 77.2% of the population had at least one doctor in their commune. There are, on average, 3.2 sports amenities per 1,000 inhabitants. The north and southeast have fewer sports facilities.
We find a positive relationship between the sports facilities ratio in a given area and the average ratio in neighboring areas. In other words, in the case of sports amenities, each area is similar to its neighbors. For other variables, we observe no spatial autocorrelation for the general practitioner rate, a positive spatial pattern for the access costs to food amenities, and a negative spatial pattern for both food amenities.
What might explain these spatial patterns?
History matters as much as comparative advantages in explaining such patterns. For example, regional food habits – more fish is eaten in the West, more fruit in the South – contribute to spatial sorting in health status and food expenditure. Individual preferences for regional amenities, such as a preference for sunny locations within the elderly population, can produce the same results.
The organization of markets and the links between demand and supply can also explain spatial correlations. For example, highly educated and wealthy people are concentrated in urban areas, where density allows them to share indivisible goods and facilities, and to gain from variety, better labor market matches, and learning mechanisms. For sales and distribution activities, proximity to final consumers and competition are key considerations. For instance, competition may inhibit supermarkets and fast-food restaurants from being located near to areas with high ratios for these amenities. Land prices are another factor. Supermarkets also tend to be concentrated in suburbs.
What are the key policy implications of this research?
As expected, we find that the more calories people eat within an area, the higher the local rate of obesity. Similarly, higher fruit and vegetable consumption lowers the obesity rate. Our results also show that the more fast-food restaurants there are per inhabitant, the higher the obesity rate. This effect is particularly strong in suburbs. In contrast, we found no effect from supermarkets. Public authorities should therefore continue directing consumers toward healthier diets, both by improving the nutritional quality of food purchases and by regulating the location of new fast-food restaurants.
We have also demonstrated that the spatial organization of the French territory implies significant spatial spillover effects that could affect the obesity rate. For instance, higher ratios of sports amenities and general practitioners in neighboring divisions can significantly increase the local obesity rate. A partial explanation is that the more sports licenses in the neighboring area, the fewer opportunities to obtain a license locally, which increases local obesity.
Medical ‘deserts’ in rural areas are a significant problem in France. To address geographical inequalities, our evidence of spatial spillovers underlines the need to consider a wider context than just the immediate local environment. In light of the variation in size and direction of the effects in our study, the scale of interventions might also differ according to the type of amenity. Personalized monitoring and personalized health and nutritional messages are among the potential policy tools.
Céline and Valérie’s paper, ‘Spatial spillovers, living environment and obesity in France: Evidence from a spatial econometric framework’, coauthored by Cécile Détang-Dessendre and Elodie Rouvière, is published in ‘Social Science & Medicine’.
