Estimating the Distributional Effects of Imputed Rents in Greece

The usual practice in empirical distributional studies is to use distributions of disposable income. However, a householdís command over resources is determined not only by its spending power over commodities it can buy but also on resources available to the household members either in the form of private non-cash incomes or in the form of in-kind welfare state provisions. In most societies, the most important private non-cash income component is the imputed rent enjoyed by the home-owning households. From this point of view, Greece is a very interesting case. Not only is home ownership widespread, but it is also most prevalent amongst the poorest segments of the population, mainly in the form of outright home-ownership. The present paper derives estimates of imputed rents using two alternative approaches (based, respectively, on hedonic regression techniques and self-assessed imputed rents) and analyses the distributional effects of the inclusion of these estimates in the concept of resources used in distributional analysis, using the micro-data of the 2004/5 Household Budget Survey.. The results show that, irrespective of the imputation method employed, measured inequality and poverty decline markedly after the inclusion of imputed rents in the concept of resources. The main group whose relative position improves after the change in the concept of resources are the elderly, while the ìlosersî can be found mostly amongst households with children.