The next step in the logical chain at the heart of post-Keynesian theory is an account of how the economy comes to settle at an equilibrium that is dictated by how much producers can expect to sell, and has no inherent automatic mechanism to ensure that that equilibrium either be at full employment of resources or tend towards it over time. This is where the theory departs most drastically from the framework of general equilibrium, which largely represents what counts as mainstream economics nowadays. In order to present it fully, we must first understand the implications of Say's law, which underpins most mainstream characterizations of equilibrium. We then need to understand the conditions under which Keynes first came to reject Say's law, which is the version of the principle of effective demand that is known as Marshallian in post-Keynesian theory, as it relies heavily on the theoretical framework developed by Marshall. But the notion of the economy settling at an under-employment equilibrium because of constraints of demand was also developed, separately, by Kalecki, starting from what could be called a Marxian standpoint. The difference between the two approaches is said primarily (but not exclusively) to be in the form of aggregate supply, which is assumed (Lavoie 2014: 262). Again, proceeding in chronological order of publication, we turn first to the Marshallian (or Keynesian) formulation of the rejection of Say's law.

Say's law

There are different versions of this proposition, but the one that most synthetically represents it is that supply creates its own demand, and that therefore there cannot be situations of excess supply. A corollary, then, is that there can never be a condition of involuntary unemployment of factors of production (labour in particular). The possibility of demand falling short of supply had been the object of an epistolary debate between Robert Malthus, who had suggested such a possibility, and David Ricardo, who instead denied that such a condition could ever arise. The most complete restatement of Say's law is given by John Stuart Mill, however, and can be represented as follows.