Estimación de sistemas de ecuaciones de gastos y demanda

Abstract

Severe limitations to the use of the classical theory of consumer behavior as a basis for empirical analysis stem from the large number of parameters demand equations contain, from the difficulty of defining quantities for many categories of expenditures, from the general lack of time series data and from the occurrence of multicollinearity among price series. Restricting the classical model to one where decision making is made step wisely by preliminary budgeting over time periods and categories of items provides a mean of alleviating these limitations and of bridging the gap between theory and empirical analysis. The method proposed, permits the estimation of expenditure functions for major categories of items and of demand equations. It makes use of both cross-section and time series data to limit the requirements of these last. It enables to make forecasts step wisely, from consumption to expenditure levels on major categories of items, and to quantities demanded. An application is made with Argentine data.