Since the ESB crisis, various studies have focused on the determinants of individual perception of Food Safety--such behavior differing widely according to the country. In this context, different authors put forward the notion of trust as being a determining factor for defining individual positions in an uncertain and complex "debated universe", as in all environmental disputes (Boisvert & Vivien 2005). Consumer factors that are known correlates of perception of Food Safety are many and varied, including demographic variables such as age, sex, education, income level, and psychological variables such as perceived risk, personal utility, product perceptions, personality and attitudes... However, on the methodological level, these studies have merely tried to highlight either individual or contextual determinants for the formation of individual preferences The studies in experimental economics have essentially focused on individual explanations, whereas, various studies based on cross-national surveys have only sought to understand the national and institutional determinants. Unfortunately, there are two types of interpretative errors corresponding to a confusion of the level of analysis: The first is an ecological error (Robinson, 1950) which takes place when an analysis (a regression, for example) on groups is done, taking into account the average values of the individuals of each group; the error consists in interpreting differences between groups (contexts) as differences between individuals, i.e. between the elementary-level units. The second, an atomic error is the opposite of the preceding one: for dealing with observations of an elementary level does not allow us to extend any conclusions to a contextual level. Furthermore an institutionalist approach had to understand the societal coherence of individual behavior observed. Therefore we use methods that allow us to combine and explain individual and institutional determinants. The data used are taken from the special European survey "Eurobarometer 64.1" 2005 dealing with Risk Issues. The sample of 25000 individuals covers the population of 25 members countries of the European Union. After a map of the main consumer risk perceptions towards Food, using an innovative factor analysis approach (namely canonical correspondence analysis), the relationship between consumer characteristics and consumer perception of Food safety is examined through an institutionalist econometric model of consumption. We have to consider the existence of structural breaks and the intrinsically dynamic nature of social reality (Prattern 2005). Institutional effects and endogeneity are therefore two issues that our econometric model must deal with. The multilevel multiprocess model (combining multilevel analysis and simultaneous equations modeling) developed by Steele et al. (2007) clearly fits these methodological issues but also our theoretical framework. Contrary to other methodologies (e.g. structural equations approaches), it requires no a priori hypotheses and is robust to a variety of institutional configurations. Our work confirms the results of different studies on the importance of individual values and attitudes and of the participation in certain social networks, or "the importance of general value orientations or worldviews". While confirming the necessity of the evolutionary or historical institutionalism, our proposition, having an essentially empirical dimension, aims to be a first step in the establishment of stylized facts and the consolidation of empirical economics, necessary to scientific dialogue. This proposition is an extension of our previous papers on the opposition to GM Food in Europe (Rousselière & Rousselière 2009).