The title says that this is a book on the microfoundations of evolutionary economics, but it is just as well a book on the microfoundations of post-Keynesian economics. In the preface written by Taniguchi, it is pointed out that he and his co-authors decided not to include “post-Keynesian economics in the title of this book, because we believe that evolutionary economics is more comprehensive than post-Keynesian economics and the first comprises the latter as a part when a suitable analytical framework is given” (p. x). The aim of the authors is to provide theoretical foundations to evolutionary economics (and thus post-Keynesian economics) that are “as fine and logically sure as Arrow and Debreu’s model of competitive equilibrium” (p. vii). Their goal is to show why it is that a market economy “works”, that is, how a fully decentralized input-output production economy will not be driven to chaos, and thus will converge to a stable quantity structure, despite agents behaving in a realistic way and having only access to local information, and despite prices remaining fully fixed. Their hope is that these alternative foundations will convince more economists, mainstream or even some heterodox ones, to abandon approaches based on constrained optimization, demand-determined (flexible) prices, and totally unrealistic assumptions, such as the tâtonnement process, the centralization of information, the lack of inventories, and the “all at once” solution.
The book is named after evolutionary economics but offers a selective interpretation of what this stands for. It builds on the concept of satisficing but does not deal with innovation, technical change, or technical progress—standard subjects of evolutionary economics as often understood. Rather the authors tackle what they consider to be the microfoundations of evolutionary economics. They are keen to develop evolutionary economics because they observe that change is everywhere—in economic behavior, knowledge, commodities, technology, or institutions—and hence because “an evolutionary point-of-view is the best way to understand the economy and its development” (p. 2). However, from the standpoint of Shiozawa and his colleagues, “evolutionary economics lacked theoretical foundations: no theory of value, no theory of behavior, no proper tool of analysis, and no proof of how an economy works” (p. 1).Footnote 1 The authors also express frustration at their perception that “evolutionary economics pretends to criticize neoclassical mainstream economics, but in many of its arguments, implicitly or explicitly, it has imported the reasoning and the results of neoclassical economics” (p. 2). Their intent is to provide evolutionary economics with “foundations defining the nature of human economic behaviors that is consistent with its own distinct worldview” (p. 2). Their understanding of evolutionary economics is based on the complexity and instrumental behavior defined in the work of Herbert Simon and his associates. On this, I find much similarities with the views expressed by some evolutionary economists such as Dosi and Egidi (1991).
Besides the preface, the book contains seven chapters. The first two chapters are authored by Shiozawa. They provide the methodological foundations of the approach advocated by the three authors. The next three chapters are written by Morioka and explain how quantities in a multi-sectoral model can adjust to changes in demand despite fixed prices, thanks to the presence of appropriate amounts of inventories. The first of the last two chapters, authored by Taniguchi, runs simulations dealing with this question, while the last chapter tackles the special case of financial markets. I will not comment on these last two chapters.
Chapter 1 mainly deals with the limited capacity of economic agents, first to gather appropriate information, and then to treat this information and arrive at an enlightened decision. As a consequence, agents never arrive at optimal solutions, and hence, it is always possible to improve upon these solutions through evolution, that is, through retention, mutation and selection. Shiozawa thus adopts the satisficing principle along the lines of Herbert Simon, and follows the tracks of March and Simon, Cyert and March, and Nelson and Winter. Shiozawa insists that it is not our knowledge of laws or theories that allows avoiding chaos in the economic system, rather it is the mode of interaction between agents that does so, surely a lesson that can also be drawn from agent-based models. In particular, the existence of inventories, trade credit, and money balances contribute to insuring that local disruptions do not spread to the whole economy. For Shiozawa, process or sequential analysis is the proper way to integrate real (historical) time. He argues that one cannot use future variables when constructing a model, as he further argues that expectations (mistaken or not) “cannot play such an important role” (p. 43). Shiozawa notes the influences of both Keynes’s weight of evidence and Heiner’s (1983) routinized behavior in believing that the consequences of mistakes are tamed by gradual adjustments.
The second chapter outlines a large number of postulates that are at the heart of the approach advocated in the book. Shiozawa rejects instrumentalism and relies instead on realism and plausibility. Among the postulates, we find the following: exchanges are made in a monetary economy; there is a production of commodities by means of commodities; there is a finite set of production techniques (fixed-coefficient production and not a continuous infinite set of choices); prices are set by firms on the basis of normal-cost pricing and not by markets; prices only change when unit costs change and not when demand or the proportion of demand attributed to each product changes; production takes time and so production cannot immediately respond to changes in demand and thus firms must rely on stocks of inventories, both with respect to their intermediate inputs and their produced outputs; there is excess capacity, in the sense that there is always room to produce more; there is a pool of workers that can be employed, and their relative wage structure is relatively stable.
As can already be guessed from the above, a fundamental principle is the separation of price and quantity adjustments. Quantities reflect changes in demand. Prices provide the criterion to judge if a new production technique is better than previously existing ones. Prices are regulated by production costs, that is, the minimum cost for a given markup structure. What we have here is a classical or Sraffian theory of prices, slightly reinterpreted. The reinterpretation is based on three features: there is no uniform profit rate; the normal price, based on the normal unit cost, is the actual price, that is, cost-of-production prices are not prices achieved only in the long term; fixed capital is not associated with joint production. Shiozawa considers that such joint production accounting is not practiced in actual accounting and thus rejects it on realistic grounds: “Thus we can treat the depreciation allowance of the machine as a production cost of the single main product… and allocate the machine cost among the products of the normal volume” (p. 94). In this manner, “the unit cost and the product cost remain constant even when the operation rate is not the normal rate” (p. 94).
Shiozawa (2016) provides more information on what he calls the revival of the classical theory of values or a cost-of-production theory of value, as he relates it the Oxford studies on pricing and full-cost pricing, making use of Frederic Lee’s (1998) book Post Keynesian Price Theory, also cited in Taniguchi’s preface, and rejecting the concept of the demand curve, as did Lee, in particular, since prices seldom change. Shiozawa further argues that Ricardo only mentioned market prices because in his days, the most important industrial product was cotton textile, the major input ingredient, which however was an agricultural product subjected to the law of supply and demand. This is not the case anymore, so that there is no need for the study of the long-run convergence of market prices towards production prices. Normal prices based on standard unit costs are actual prices, and hence, “the new theory … can analyze short-or medium-period questions” (p. 133). All of this is line with my own interpretation of how Sraffian prices ought to be understood (Lavoie 1992, 2013). They should be reinterpreted as full-cost prices, where “the price of each commodity is simply determined by its production cost – measured at the normal level of utilization of capacity – plus a target rate of return” (Boggio 1990, p. 47).
The rest of the chapter focuses on the inventory strategy, which Shiozawa considers to be his ultimate postulate, with inventories being some target ratio of a moving average of sales. Inventories will play a key role in the quantity adjustments of the following three chapters. Shiozawa criticizes post-Keynesians for omitting inventories, and also for omitting the interconnections of the different production sectors, thus, themselves relying on a representative firm, just like their New Keynesian foes. It is true that besides Godley and Cripps (1983), Godley (1999) and Godley and Lavoie (2007, chapters 8–11), post-Keynesians usually assume the existence of a period which is long enough for production to respond fully to demand changes. Besides what can be found in agent-based models (ABM), for instance, those of Caiani et al. (2016) and Seppecher et al. (2018), few other heterodox authors take inventories into account: only two sets of names come to my mind: Duménil and Lévy (1993) in one set, and Chiarella et al. (2005) in the other. However, these authors analyze one-sector models. They cannot take into consideration the complex network of the production of commodities by means of commodities tied to the sequence outlined by Shiozawa: production decisions and orders are made at time t; firms get their inputs and produce their goods at time t + 1; the outputs become available and are sold at time t + 2.
The third chapter describes the basic theory of quantity adjustment. For Morioka, capitalism is by nature a “seller’s market”—a demand-constrained economy—as opposed to Kornaï’s supply-constrained socialist economy. He links this claim to Sraffa’s 1926 statement that firms are not limited by rising unit costs but rather by the difficulty of attracting more customers. As argued in Godley and Lavoie (2007), smooth adjustments in a capitalist economy require the existence of buffers—here stocks of inventories, especially product inventories. Changes in these inventories allow a decoupling between production and demand, since demand cannot be known with certainty before production occurs. As a support for this claim, Morioka provides short citations of Hicks and Stiglitz and a long citation by Simon (1991) according to whom adjustment through quantities is the major coordination and allocation tool in the real world. Because of the interdependencies already alluded to, all firms must hold inventories for smooth adjustments to proceed. Morioka recalls the one-sector inventory dynamic analyses of Lundberg, where stability depends on the marginal propensity to consume and the target inventory to sales ratio (as in Sraffian supermultiplier growth models where it depends on the marginal propensity to consume and the GDP share of investment). He also recalls other previous efforts at assessing stability, especially when the Lundberg framework is combined to an input-output structure.
The fourth chapter is Morioka’s own contribution to the stability issue of these quantity adjustments at fixed prices within an input-output structure, based on the time sequence outlined in chapter 2 (forecast and orders, production, delivery, and sales). It is the theoretical core of the book as the analysis is purely mathematical. Here, I only mention his main results, under the assumption that production is equal or higher than the breakeven point given by the markup, and hence that firms do not go bankrupt.Footnote 2 First, stability does not depend on errors in forecast; second, it depends on the degree of interdependence found in the input-output matrix; third, it depends on the size of the inventory buffers; and finally it depends on the way demand is being forecasted. The more smoothing there is, for instance, the greater the number of periods considered with simple moving averaging, the more likely the stability of the adjustment process.Footnote 3 Static expectations (which depend only on the most recent result) will thus destabilize the system. Morioka points out that he followed the strictures suggested by Simon: the actions and the forecasts of the agents are based on data that are easily obtainable, and the computations that are assumed to be made by these agents require little capability. What is striking in all of this is that, in contrast to the ABM studies mentioned above which rely on price tuning and which follow the same strictures, with enough smoothing the adjustment to a change in demand is a converging process despite the complete lack of reaction of prices to the evolution of sales and inventories.
Chapter 5 extends the model in a few directions. Whereas in the previous chapter the firms belonging to a given sector were assumed to be homogeneous, so that the analysis could be conducted at the sector level, here, the production techniques of these firms are different, their prices are different, and their decisions “on production and raw material orders are mutually independent” (p. 267). There is thus a link with agent-based modeling. Morioka also shows that partial inventory adjustment (as can be found also in Godley and Lavoie 2007) can replace smoothing of expected demand. In other words, even if expectations of sales are simply equal to the sales of the previous period, a partial adjustment of inventories towards the target inventories to sales ratio will help to stabilize the adjustment process. Morioka believes that such a partial adjustment would be justified by rising marginal costs, but I believe radical uncertainty would be a much better explanation for this observed phenomenon. Morioka also shows that the adjustment is more likely to be completed successfully if the growth rate in demand is smaller than the target inventory to sales ratio. Thus large changes in demand cannot be accommodated easily. Morioka concludes that the likelihood of the (conditional) short-run stability of quantity adjustments is now fairly well established. However, this “does not exclude mid- and long-term instabilities in the economy” (p. 289).
The message being conveyed by the book should by now be clear: the results achieved by the authors, which were previously only available in Japanese, constitute a great breakthrough—an achievement of paramount importance as the authors say—for the analysis of the modern industrial economy (the financial sector requires a completely different story). The authors claim that their results are comparable with those of Arrow-Debreu but obtained within the completely realistic framework of a production economy with agents disposing only of local information and frugal capabilities. Prices are not scarcity indices: changes in prices are not signaling changes in demand relative to supply. They are a tool to assess the best production method, as improvements or new combinations will be reflected by a decrease in unit costs. The information conveying changes in demand is transmitted and assessed through quantity signals—sales and changes in inventories—but not through prices. A key achievement of the authors is the demonstration that a multi-sectoral economy with produced inputs, where production takes time, can adjust to variations in demand through the realistic decisions of managers to alter quantities without any change in prices—something that previously was not thought to be possible.
The book certainly provides a welcome integration of evolutionary, Sraffian and post-Keynesian economics. It provides a new research avenue for those scholars dedicated to agent-based modeling who, with very few exceptions such as Seppecher et al. (2018), have yet to combine the input-output interdependence to the ABM stock-flow consistent model framework. The only drawback of the achieved results, as briefly mentioned at the end of chapter 5, is the lack of aggregate demand feedback. While the production of the inputs depend on the forecasted demand for outputs, the actual final demand (consumer demand and investment demand) in the models were themselves exogenous or “independent”. In other words, an increase in production, and hence in wages and profits, has no feedback effect on realized demand for final output in the next period. Given the achievements of the authors in dealing with changes in inventories and with what happens when production is not equal to demand—questions left aside by nearly all post-Keynesian or heterodox authors—I feel that it would be unjust to press this issue. Stock-flow consistent agent-based models and hence simulations may be the only way out to deal with these macroeconomic feedback effects.
On the methodological front, the book is certainly close to the methodology advocated by post-Keynesians and heterodox economists at large. In the past, I have emphasized that four key presuppositions are realism, reasonable rationality, production, and organicism. The first three features are underlined by Shiozawa. If one reinterprets organicism to mean something beyond macroeconomic paradoxes or emerging properties, taking it to mean in addition the circularity of production and hence the interdependence between sectors and the complexity of their networks, then the methodology underlying the evolutionary foundations proposed in this book are fully in accord with those of post-Keynesian economics. As the authors say a number of times, their foundations are akin to the study of effective demand at the level of the firm; what remains to be incorporated is macroeconomic effective demand.
REPE readers may find some comfort in knowing that Shiozawa makes similar critiques with regard to the theoretical foundations of post-Keynesian economics. See https://www.researchgate.net/post/Does_Post_Keynesian_Economics_need_no_theoretical_foundations
In a previous chapter (p. 38), Shiozawa insists that Schumpeter’s creative destruction only makes sense as long as not too many defaults or destructions occur.
In equation (4.18) that describes exponential smoothing, the past actual and expected values should be inverted.
Boggio L (1990) The dynamic stability of production prices: a synthetic discussion of models and results. Polit Econ: Stud Surplus Approach 6(1–2):47–57
Caiani A, Godin A, Caverzasi E, Gallegati M, Kinsella S, Stiglitz JE (2016) Agent based-stock flow consistent macroeconomics: towards a benchmark model. J Econ Dyn Control 69:375–408
Chiarella C, Flaschel P, Franke R (2005) Foundations for a disequilibrium theory of the business cycle: quantitative analysis and qualitative assessment. Cambridge University Press, Cambridge
Dosi G, Egidi M (1991) Substantive and procedural rationality: an exploration of economic behaviours in changing environments. J Evol Econ 1(2):145–168
Duménil G, Lévy D (1993) The economics of the profit rate: competition, crises and historical tendencies in capitalism. Edward Elgar, Aldershot
Godley W (1999) Money and credit in a Keynesian model of income determination. Camb J Econ 23(2):393–411
Godley W, Cripps F (1983) Macroeconomics. Fontana, London
Godley W, Lavoie M (2007) Monetary economics: an integrated approach to credit, money, income, production and wealth. Palgrave Macmillan, Basingstoke
Heiner RA (1983) The origin of predictable behavior. Am Econ Rev 73(4):560–595
Lavoie M (1992) Towards a new research programme for post-Keynesianism and neo-Ricardianism. Rev Political Econ 4(1):37–79
Lavoie M (2013) Sraffians, other post-Keynesians, and the controversy over centres of gravitation. In: Levreto ES, Palumbo A, Stirati A (eds) Sraffa and the reconstruction of economic theory: volume three. Palgrave Macmillan, Basingstoke, pp 34–54
Lee FS (1998) Post Keynesian price theory. Cambridge University Press, Cambridge
Seppecher P, Salle IL, Lavoie M (2018) What drives markups? Evolutionary pricing in an agent-based stock-flow consistent macroeconomic model. Ind Corp Chang 27(6):1045–1067
Shiozawa Y (2016) The revival of the classical theory of values. In: Yokokawa N, Yagi K, Uemura H, Westra R (eds) The rejuvenation of political economy. Routledge, London, pp 153–172
Simon HA (1991) Organizations and markets. J Econ Perspect 5(2):25–44
About this article
Cite this article
Lavoie, M. Shiozawa, Yoshinori; Morioka, Masashi; Taniguchi, Kasuhisa: Microfoundations of Evolutionary Economics. Rev Evol Polit Econ 1, 265–270 (2020). https://doi.org/10.1007/s43253-020-00004-5