How a certain economic reality is made, strengthened, and spread? This article examines the problem through the situation of the French General Planning Commissariat in the late 1960s, which at that time held a key place in the political-administrative area. However, given the force accorded to the market and the external constraint, this institution became increasingly questioned. In this context, a new macroeconomic model called "FIFI", developed by a group of economists at the INSEE (Institut national de la statistique et des études économiques), introduced some innovations regarding the formulation of economic policies and their bases. The model has been designed to provide a picture of the French economy in 1975 if no decisions were taken, in order to identify future problems which could arise from the trends and to determine the solutions which could be recommended. As it is an instrument that attributes causality (between various entities such as wages, unemployment or external constraint), the FIFI model tends to displace the type of imputations on which the participants of the Plan rely, especially trade unions. Hence, these economic metrologies not only produce a reality, made up of numbers and relatively independent; they also destabilize the political forms that are explicitly based on values.