Que devrions-nous entendre par « politique de croissance » ?

For various reasons ? empirical, technical and casual ? modern (neo-classical) growth theory has centered its attention on steady-state exponential growth. When the models are intented to serve as a guide to policy, the tacit presumption is that the goal of growth policy is to increase the long-term growth rate. It seems more natural to define as « growth policy » anything that permanently lifts the cyclically-corrected trend path of the model economy, even if it only adds a constant percentage to a pre-existing path. The deeper point is that once growth theory is freed of the need to generate exponential paths and ways to tilt them, the way is open to a more discriminating choice of assumptions, governed more by empirical validity and less by artificial convenience. If there is a general lesson for macroeconomic theory in these considerations, it is not that convenient simplifications should be avoided. It is rather that convenient simplifications have to be chosen with care. JEL Classification: E60, O11, O40, O43.