1 Introduction
In developed countries, the patterns of the excessive weight and obesity of adults
and children have become a critical public health concern. Data from the United
States show that the prevalence of overweight and obesity began to increase around
the mid-1980s, and has continued to increase dramatically, despite the fact that tech-
nological advances in medicine have improved the general health of the population
in the past half-century. It has been estimated that 3.7 and 5.3 percent of medical
spending are attributable to overweight and obesity, respectively (Finkelstein et al.,
2003; Daviglus et al., 2004).
In European Union countries, it has been calculated that further extension of
the health care system would reach costs of about 7 percent of the total healthcare
expenditure (Commission of the European Communities, 2005). Although much
later with respect to the United States and some continental European countries,
the overweight issue has also become significant in Italy (Costa-Font et al., 2009),
where the percentage of adults classified as overweight (including the various forms
of obesity) has risen by 4 percent in the last decade (ISTAT, 2007).
However, the complex range of social and economic factors which explain the
patterns of obesity are not understood and identified, although the fact that a vast
body of studies has recently been attempting to explore its determinants (Kopelman,
2000; Lakdawalla and Philipson, 2009; Powell and Chaloupka, 2009). The research
programs quoted in the works of Cutler et al. (2003) and Bleich et al. (2008) explain
rising rates of obesity as the result of over-consumption of calories, associated with
technological changes which lower the unitary cost of food.
One strand of empirical works focuses on economic determinants to respond to the
question of whether changes in food prices translate into changes in the prevalence
of overweight and obesity. These tests provide an interesting picture of the long-run
association of food prices and overweight changes such as the use of aggregate data
cannot identify heterogeneous individual (or household) behaviour and may result
in a policy-maker receiving weak or unclear information.
Conversely, it would be interesting to obtain direct estimates of overweight on
changes in food consumption by changes in relative food prices, with a focus on the
segmented groups of a society (Schroeter et al., 2008). The simultaneous absence in
Italy of true panel data and discontinous surveys in measuring individuals weights
(and heights, i.e. body mass index) does not enable us to account for this behaviour.