Permanent Income and Adaptive Expectations: Estimation of a EU Consumption Function

      Our paper estimates an aggregate consumption function using Milton Friedman’s permanent income hypothesis for the16 members of the European Union that participate in the European Monetary Union (euro-zone). The research employs quarterly data covering the period 1995-2009 for the EU-16. Permanent income and permanent consumption are not observable variables. In order to test whether the permanent income hypothesis is supported by the EU-16 data, we combine the permanent income hypothesis with the adaptive expectations model. In this way we can relate the hypothesis to observable variables such as the actual income and the actual consumption. We were expecting the combined model to be supported by the data. We employ multiple regression analysis using OLS as well as several econometric tests (including stationarity testing) to test the validity of the model. After conducting the tests we are led to a log-linear model that still combines the two theories. The log-linear estimated model is better supported by the data than the linear one. It also provides implications for the short-term and long-term elasticities of consumption. However, the adaptive expectations coefficient seems to be relatively unimportant in determining consumption.  Our paper is partially motivated by Manitsaris 2006, who used the same model with an annual dataset to estimate a consumption function for the EU-15.  Our findings on the elasticity of consumption are partially (short term elasticity of consumption) inconsistent with those of Manitsaris 2006.