In his stimulating Nobel Prize Lecture, Paul Samuelson made the important point that there were interesting macrodynamic systems which could, in no sensible sense, be associated with maximization problems. The paradigmatic example he chose was the multiplier-accelerator system. In this paper I...