Static microsimulation models like POLIMOD are designed to assess the effects of current or prospective tax and benefit reforms. Since the purpose of POLIMOD is to inform policy, the accuracy of the model ouput is a key concern. Thus, a significant problem faced in the construction of such models is that the microdata required are never available for the current year (or any prospective year). Models therefore are built from data collected in previous years, which has been ëupdatedí to render it representative of the year of interest. This paper describes the method of updating the Family Expenditure Survey (FES) income and expenditure data used in POLIMOD and discusses several issues concerning the choice of updating method. It furthers our understanding of the structure behind POLIMOD and develops an answer the question: How accurate can we claim our model is and therefore how valid are our predictions and policy proposals?
