Trends in macroeconomic modeling: The past quarter century

Macroeconometric model-building is now over half a century old. Econometric models of the Dutch economy were first built by Tinbergen in the middle of the 1930s, and Tinbergen's first model of the U.S. economy predates the outbreak of hostilities in the Second World War. Nevertheless, macroeconometric models have continued to evolve, sometimes in a spectacular manner, during the past quarter century. In this paper, we take a broad picture and examine eight trends that seem to us particularly important in explaining developments in this field. These trends are: (1), large scale modeling; (2), use of a team approach; (3), development of an institutional framework independent of the model's originator or originators; (4), the continuing revolution in computing technology; (5), increasing theoretical sophistication; (6), increased orientation towards forecasting uses; (7), increased orientation towards policy issues; and (8), internationalization of macroeconometric model-building. Finally, we explore some prospective developments in this field, up to the end of the current century.