During the last forty years consumer demand analysis has evolved toward a system wide approach. Since Working's work (1943) on Engel curves, the "demand-system" models which have been mainly used by economists are the Linear Expenditure System proposed by Stone (1954) (e.g., Andrikopoulos, Brox and Carvahlo, 1984). These models generally assume the hypothesis of intertemporal separability, allowing current demands to be formulated in function of contemporaneous exogenous variables (i.e. prices and expenditure), but not in terms of lagged and lead regressors. By the way, despite extensive empirical evidence, the question is whether the intertemporal separability hypothesis is justified in the analysis of food demand. The purpose of this paper is to test the intertemporal separability hypothesis on food demand in Italy for the period 1960-1993. For this purpose, the model to be estimated is the "simple non additive preferences" (SNAP) structure for intertemporal preferences proposed by Browning (1991), whìch generated a demand system dependent on prices lagged by and lead by one period, as well as, current prices. The econometric estimation of the SNAP demand System gives empirical evidence for whether or not Italian food consumers reject the intertemporal separability hypothesis, or likely; whether or not consumers considered lead and lagged exogenous variables to be relevant for their purchasing decisions.
