Addiction and Alcohol consumption at the household and individual level: theory and applications

Non disponibile

In modern societies, it is a common opinion to consider alcohol consumption as a social problem, especially when consumed habitually in large quantities or when binge drinking is a relevant phenomenon. The negative effects to the individuals and the society are numerous and cover several fields. Numerous examples could be given, as liver diseases, car crashes, violence, and so on. These consequences of alcohol consumption have long been studied in medical and sociological fields, but, in our opinion, still not enough has been done from the economic perspective. If policy makers pursue the objective of reducing alcohol consumption, or at least its negative affects, economists should give them accurate instruments to properly evaluate and predict the consequences of such policies, including benefits or losses in terms of social welfare. To accomplish this duty, a firts step consists in classifying properly the negative effects of alcohol consumption, allowing to focus on specific objectives and to find specific indications for each class of issues. Our proposal is to distinguish between individual effects, household effects and social effects. The individual effects are those direct effects which consumption of alcoholic beverages brings to the drinker itsef. These effects include general health diseases, liver diseases, emargination, car crashes consequences, working problem, lack of concentration and so on. From an economic point of view, several of these aspects can be synthesized by the loss of utility due to the addiction status of the consumer. In economics, the usual tool for analyzing addiction is the Rational Addiction model (RA) by Becker & Murphy (1988). This model represents an important milestone in the modelling of addiction since it started an important empirical literature on addiction. The use of this model has been criticised by several authors both because of some simplyfing assumptions and because applications of this model to real data have been often conducted without taking into account some peculiar statistical issues (Gruber & Köszegi, 2001; Gul & Pesendorfer, 2001; Lowenstein et al. , 2003; Auld & Grootendorst, 2004; Bernheim & Rangel, 2004; Wangen, 2004). The model, however, is still widely used to empirically test for the presence of addiction, mainly because of the simplicity of the linear Euler equation which has to be estimated. In this work an estimate of the Rational Addiction model is conducted using a series of Italian cross sectional data on households expenditure on alcoholic beverages. Using these data, which allows to analize individual expenditure on alcohol, the hypothesis of addiction is rejected. Even though this result may be driven by the fact that if alcohol is consumed in moderate quantities it is unlikely to cause addiction, we believe that perplexities and critics around the RA model have to be taken properly into account befor making statements of absence of addiction to alcohol in the Italian society. This is one of the reasons for which we propose an extended model, which could possibly overcome some shortages of the RA model and allow for more robust results. Household effects arise when alcohol consumption not only causes problems to the consumer itself, but also to its family (household level externalities). Typical examples of these externalities are violence, serenity deprivation, the diminished economic resources left for the household members and so on. In particular, in this work we focus on the intra-household distribution of resources to verify whether households with a heavy drinker show a different behavior in distributing resources between household’s members. In Chapter 6 we apply the Collective Choice model (Apps & Rees, 1988; Chiappori, 1988, 1992) to Italian households and find that when a relatively large amount of alcohol is consumed, the distribution of resources within the household is shifted from the average distribution, and that this difference is greater for low income households...