ABSTRACT
Before turning to the role of MORKMON in the policy-making process and discussing the linkages between model development and policy, it must be emphasized that MORKMON is not the only model available for policy analysis in the Bank. In particular since the mid-1980s a proliferation of models occurred, which was at least partly induced by the severe cr iticisms to which macroeconometric models had been exposed. Moreover, models are not the only research input in the policymaking process. A substantial part of the resources for research in the Bank is allocated to individual empirical or theoretical studies. At the Nederlandsche Bank, at least five different model types currently exist, each serving different purposes. For short-and medium-term projections and policy analysis, structural macroeconometric models form the most important category. Next to MORKMON, a new quarterly multi-country model for the EU, named EUROMON, has been constructed recently (De Bondt et al., 1997), and will be further developed in the near future. In addition, the CPB’s policy model FKSEC and NIESR’s world model NiGEM have been operational for some years now. Since the early 1980s leading indicator models have been developed and used to produce monthly forecasts of industrial output, cyclical turning points, and real GDP growth in the Netherlands and other industrial countries (Fase and Bikker, 1985; Bikker and De Haan, 1988; Berk and Bikker, 1995; Van Rooij and Stokman, 1996). As from 1993, similar models have been constructed for the purpose of inflation forecasting (Bikker, 1993a; Bijsterbosch and Hebbink, 1996). The third type of model available for policy analysis is a stochastic dynamic general equilibrium model calibrated for the EU-area (Bolt and Folkertsma, 1997; Folkertsma, 1999). This model is in the tradition of monetary equilibrium models which are based on optimal intertemporal behaviour of agents and incorporate inflation tax and liquidity effects in the spirit of Cooley and Hansen (1989) and Christiano (1991). Because of optimizing behaviour and rational expectations, dynamic equilibrium models are less prone to the Lucas critique. The model’s main purpose is to provide a theoretically-sound framework for investigating the long-term (welfare) effects of policy changes or shocks to the economy. VAR models form the fourth class of models, comprising Bayesian VARs as well as structural VARs. In the early 1990s output growth forecasts obtained from a VAR, which focused on the interaction between German and Dutch key economic variables, provided a point of reference for the outlook based on MORKMON (Bikker, 1993b). However, in general, VARs were not used for forecasting purposes but merely served to study the empirical regularities of the process of monetary transmission (Boeschoten et al., 1994; Garretsen and Swank, 1997; Vlaar and Schuberth, 1999; De Bondt, 2000), and to measure the monetary component of inflation in both the Netherlands and Europe (Fase and Folkertsma, 1997). The list of model types available for policy analysis also includes a generational accounting model for the Netherlands (Hebbink, 1997). This model highlights the implications of an ageing population and the effects of today’s budgetary policy stance for future generations.
