I noted that I might have further thoughts on this text; here they are.
1. Just to put a finer point on something from the first post: the nub of the issue for Keynes, driving the divergence between profitability and technical efficiencies, is not “complex production” in general (nor is it money, nor is it the presence of an entrepreneur as such, which I already pointed out) but economic activity mediated by exchange. The “co-operative” unit in the co-operative economy can be large and complex in operational terms and in terms of the people and resources involved. His argument grants from the outset that there’s an agreement ensuring that each person contributes to the enterprise only if his share of the yield is greater than any loss associated with his contribution. (Obviously this isn’t always true. It’s worth bracketing that just for the moment, though.) Putting aside “miscalculation” or “stupid obstinacy” (again, this is worth doing even if it seems unrealistic), this means incentives to engage in production will generally aligned with technical efficiencies (measured in terms of preference satisfaction). The problem is that the moment you introduce exchange, there is no way to ensure that each factor of production receives a money payment equivalent to the share of output he would have received in the unified-production-unit-governed-by-agreement scenario. This is basically because when the activities of production and consumption are united in a single person (whether that is in a literal Robinson Crusoe scenario, or mediated by agreement in the “co-operative economy”), we can expect some convergence between incentives and technical efficiencies, at least as long as we are measuring the latter only in terms of subjective utilities (as the “classical theory” generally does, and which is worth entertaining as far as it goes). But the moment we introduce exchange of outputs (whether mediated by money or not), there is simply no reason to expect this alignment to persist for any given production unit in the economy. I think is the core argument Keynes is making, although he illustrates it in various specific ways. In essence, as soon as production and consumption are no longer directly weighed against each other, for any given instance of productive activity, by and within single persons, the entire rosy prediction of the classical theory — economy-wide efficient outcomes resulting from putatively uncoordinated economic activity — comes apart.
2. I think this argument is both powerful and important. It goes to the very heart of the “classical theory,” which largely forms the basis of the common-sense commitment to ‘free markets’ that so many reformers and activists are wrestling with today (on whatever level–whether they are trying to make policy or doing direct organizing, etc.) It also is worth putting aside the issue of unfair initial conditions, which is such a common point of criticism of market ideology (both in more technical debates and in broader public and political discourse) to consider what happens in Keynes’ picture. Because the point is that everything unfair about the economy need not be traced to unfair initial conditions. Instead, the inherent interruptability of the correspondence that exists in the simple, “co-operative economy” case in an economy governed in part by exchange means that there is potential for exploitation and power-perpetuation at almost every turn! That is on the one hand pessimistic compared to the rosy classical picture, but it simply means that our legal & social planning of markets and economic activity [which–separate point–is inevitable anyway] has to counteract these tendencies by channeling both economic coordination and competition in fair, egalitarian directions. But it’s helpful to eliminate the idea that doing so consciously will ipso facto lead us away from technically efficient outcomes.
3. I also think it’s worth putting aside temporarily the question of measuring technical efficiencies in subjective utility terms. We know that there is a frequent slippage from subjective utility to money/price when “the classical theory” is applied to policy discussion, and that this probably occurs selectively. (I think for example we should have a place for the objective costs of labor effort when we think about market governance.) But it is also worth really thinking about the limitations of the “classical theory” even when limited to its own terms (terms which are not irrelevant in thinking about things like fairness, either).
4. Finally, as a postscript, I think it’s unfortunate that a decent portion of the current dissatisfaction with the economic style as currently practiced seems to valorize earlier versions of the very “classical theory” (in, e.g., Smith) that Keynes trained his sights on in this text. The issue is not how formalized the idea of self-regulating markets has become since Smith and his immediate successors (though certainly, formalism serves to obscure the actual theory), but rather the idea itself. Indeed, the statistical methods used by many contemporary economists doing empirical research are innocuous or positively helpful compared to much of the rhetoric of classical political economy. Let’s not get distracted about what the problems are!
Sanjukta Paul 