We empirically assess the distributional impacts and welfare effects of policies to incentivize renewable electricity production for the case of Italy. We use data from the Household Budget Survey between 2000 and 2010 to estimate a demand system in which energy goods' shares of expenditure are modelled using different empirical approaches. We show that the general Exact Affine Stone Index (EASI) demand system provides more robust estimates of price elasticities of each composite good than the commonly used Almost Ideal Demand System (AIDS). The estimated coefficients are used to perform a welfare analysis of the Italian renewable electricity production incentive policy. We show that different empirical approaches give rise to significantly different estimates of price elasticities and that methodological choices are the reasons for the very high elasticities of substitutions estimated using similar data by previous contributions. We find no evidence of regressivity of the incidence of the Italian renewable incentive scheme in the period under consideration. The renewable subsidies act as a middle-class tax, with the higher welfare losses experienced by households in the second to fourth quintiles of the expenditure distribution.