Methodological Problems

Everyone knows that in econometric model building for a long time now there has been a tendency towards gigantism, in which one of the recommedantions for a model was its size (measured by the number of equations). This tendency still continues, but the careful observer will now have realized that another idea has been gaining ground, namely that macrodynamic medium-term econometric models, small in size (from one to a few dozen equations) and embodying considerably more economic theory than is usually possible, are complementary, or even preferable, to large models for the purposes of policy analysis and simulations: see, for example, Bergstrom and Wymer, 1976; Wymer, 1976, 1979; Reserve Bank of Australia, 1977; Comraittee on Policy Optimisation, 1978; Deleau et al., 1980, 1982; Masson, Rose and Selody, 1980; Richard, 1980; Gandolfo, 1981; Bergstrom, 1982; Kirkpatrick, 1983; Gandolfo and Padoan, 1982, 1983; and the references therein.