A model to "make decisions and take actions"

The 1960s were the glorious age of large-scale, "economy-wide" macroeconometric models, which enabled forecasts and simulations of the quantitative effects of certain policy changes within systems of simultaneous equations. This paper is concerned (1) with the form and character of the kind of knowledge these models brought about; and (2) with the ways in which these models shaped the practices, objects, and the very notion of macroeconomic planning. The concepts of bricolage and infrastructure help to account for the continuous reciprocal shaping of mathematical techniques, empirical data, bits of economic theory, institutional arrangements, the aims and hopes of economic planning, images and visions of "the economy", and the practical requirements and affordances of computing technologies. The paper traces the movements and changes of the multi-sector growth model, constructed by the Norwegian economist and communist Leif Johansen, from its publication in 1960 to its implementation in national macroeconomic planning and its wider circulation from the mid-1970s on. In Norway, it became one component of a "system of models," which, far from simply being a passive "tool for decision making," structured bureaucratic procedures, and formatted the basic concepts of policies. Both in and outside Norway, it showed a certain resilience: when Keynesian macroeconometric models encountered widespread criticism, the multi-sector growth model was further remodeled, adapted to new scenarios, and built into easily applicable software packages. Its basic design, however, remained compact and consolidated economic key assumptions. Under the label of "computable general equilibrium models," it prepares the ground for a whole variety of policies until today - regardless of their political frameworks.