According to the World Health Organization (WHO), an individual is deemed to be overweight if his or her body mass index (BMI, the ratio between his weight in kilograms and his height squared) is higher than 25; above 27, his risks of disease increase significantly and he is obese if his BMI exceeds 30. According to this indicator, the prevalence of overweight in the French adult population rose from 29.7% to 37.5% between 1990 and 2002 (OECD Health data for 2005). To understand this increase in overweight, which is also observed in all other developed countries, economists have first of all focused on the effects of lower food prices, which induced a rise in calorie intake and a trend decline in physical activity at work, responsible lower calorie expenditure. These long-term disturbances to the energy balance of individuals are the source of the obesity "epidemic" we see today. The word "epidemic" may suggest that being overweight or obesity is a transmissible condition. This is of course not the case from a strictly biological point of view. Recent research has nonetheless focused on the social transmission of overweight and obesity via social interactions between friends, or social norms concerning corpulence.
To study the impact of these factors, the INRA scientist used data from the INSEE survey on "Household living conditions in 2001", which recorded the declarations of more than 5000 individuals concerning their habitual weight*, their ideal weight, their actual height and certain attitudes towards food. The researcher supposed that individual representations of ideal weight were generated from several sources of information: in the first place, social norms, the principal focus of his research, and secondly habitual weight, because individuals representations are also governed by a "reality principle", a cognitive strategy that allows individuals to preserve their wellbeing by adapting their ideal to reality. Regarding this latter point, it was thus noted that 40% of the individuals surveyed considered that their ideal weight was their habitual weight. The statistical model built on this basis thus postulated that social norms and habitual corpulence jointly determine individual representations of ideal body shape which, in turn, affect dietary behaviour.
Measuring social norms
To quantify social norms regarding corpulence, the author posed two hypotheses. Firstly, the social reference group was that to which an individual was assigned (because of his age, gender, financial and cultural assets). Secondly, the specifications were shared expectations on how members of the group should behave or the attributes they must own (in this case, a particular corpulence).
The ideal BMI of each individual contains information on the specifications regarding corpulence that prevail in the social group to which he is assigned; the social norm is thus measured by the ideal BMI of the group.
Social norms affect the individual representations of women, while only the real BMI determines those of men
The results show that for women, a 1% increase in the social norm (the mean ideal BMI of the assigned group) induced an increase in the ideal individual BMI of about 0.5%. For men, an identical variation in the social norm had no effect on the ideal individual BMI. This difference in gender is consistent with results in the international literature on the impact of beauty stereotypes, to which women are the most sensitive. It may partly explain why female obesity is increasing more rapidly than male obesity (+64% between 1997 and 2006 versus +40% for men). In women, the impact of environmental changes (lower prices, advertising pressure, etc.) is so strong that they induce modifications to social norms that amplify the effects of these changes.
In addition, a 1% variation in the habitual BMI induced an 0.5% increase in the ideal BMI for women and an 0.8% increase for men, which means that men adapt their representations of ideal body shape to the reality of their own body more easily than women.
Social norms modulate the effect of public policies by modifying individual representations, notably amongst women
The research also showed that individual representations of ideal body shape were very strongly correlated to certain nutritional attitudes. Individuals who considered they were overweight with respect to their ideal weight tended to avoid fatty and sweet foods to a greater extent, consuming more low-fat or low-sugar products, and were more frequent in declaring that they consume a balanced diet and less alcohol.
Policies to subsidise fruit and vegetable prices, which would cause a 1% reduction in the real BMI, would reduce the ideal BMI of an individual who wishes to lose weight by a minimum of 0.5% amongst women and 0.8% amongst men. In the case of women, because all individuals in the group would simultaneously modify their representations, the mean ideal BMI for the group would also be modified downwards. Overall, to the direct effect of -0.5% would be added a second reduction in the ideal BMI of each woman of -0.5%, caused indirectly by a change to the norm. This second effect might not be observed in men. This example shows that social norms can modulate the effects of public policies through a modification to individual body representations.
The researcher also considered the impact of other types of social interactions, such as network effects on adolescents or within a couple, in order to understand why the social norms on corpulence observed in the USA are evolving upwards, despite active policies to combat obesity. Analysing and understanding social interactions in the context of diet provides important input to both the development and the evaluation of public policies to combat obesity.
.................................................................
*habitual weight, without being precisely the true weight measured accurately on day 0, is a reference weight that is "borne in mind".
.............................................................
Reference :
- Etilé, F. (2007). "Social norms, ideal body weight and food attitudes". Health Economics, vol. 16, n° 9, pp 945-966.
Pour en savoir plus :
http://www.inra.fr/internet/Departements/ESR/publications/iss/
N° 1 - March 2008 - Normes sociales de corpulence et politiques nutritionnelles.
|
Search :
