Alkoholkonsumtion och dess orsaker

Sammanfattning: The purpose of this analysis is to use regression models to find out what factors influence the average consumption of alcohol on a national level. The data which is being used come from the WHO and the OECD. The analysis is restricted to countries that are members of the OECD. In order to analyze the material the following types of regression models are investigated: a basic linear model, a logistic model and a model based on panel data. The models are evaluated according to a few criteria: we look to include only significant parameters if possible, obtain an R^2 value as high as possible when this measurement is applicable, and we want the assumptions concerning the residuals to be fulfilled. In the end we reach the conclusion that the material used for the analysis could explain the consumption of alcohol only to a small degree. The few significant results we get, however, are in agreement with the findings of previous research. If the country has a monopoly of the selling of alcohol, the amount that people drink on average decreases, and drinking seems to increase in economically good times.