A macro Economy as an Emergent Ecology of Plans

A Macro Economy as an Ecology of Plans

Richard E. Wagner

Department of Economics, 3G4

George Mason University

Fairfax, VA 22030 USA

tel.: ++1-703-993-1132

fax: ++1-703-993-1133

[e-mail]

http://mason.gmu.edu/~rwagner

Abstract
Standard macro theories reflect a choice-theoretic orientation wherein aggregate variables are treated as acting directly on one another. Macro phenomena are thus reduced to the same order of simplicity as micro phenomena; macro variables differ from micro variables only by their larger size. In contrast, this paper treats the relationship between micro and macro as non-scalable. Macro phenomena emerge through micro interaction and are of a higher order of complexity than micro phenomena. Rather than reducing macro to micro through scalar multiplication, macro phenomena supervene on micro interaction and are not themselves objects of direct action. A macro economy is treated as a complex ecology of plans that constitute a non-equilibrium process of spontaneous ordering.

Keywords: micro foundations; complexity; emergence; supervention; ecology of plans; turbulence; spontaneous ordering; scale-free modeling
JEL Codes: B4, D2, D8, E1
Acknowledgements: An earlier version of this paper was presented in 2010 at James Madison University, George Mason University, and at the annual meeting of the Southern Economic Association. Several participants offered helpful comments for which I am grateful. I am particularly grateful to Barkley Rosser for offering some insightful remarks in his role as a discussant at the SEA presentation, as well as to three referees for JEBO who offered valuable commentary.
A Macro Economy as an Ecology of Plans

1. Introduction

One notable feature of Snowden and Vane’s (2005) “schools of thought” survey of macro theory is that the controversies among the theorists they examine resembles a family feud. Despite numerous particular differences among those theorists, they all operate under the shared presumption that macro phenomena are scaled-up versions of micro phenomena: macro is micro addressed in a loud voice. Where micro theory refers to demand and supply functions for individual services, macro theory refers to demand and supply functions for output as a whole. Just as a firm is characterized by a production function, so too is the aggregate of firms. Where a firm’s plan of production can be altered through shocks to supply or demand, so too can aggregate output. Macro theory looks pretty much the same as micro theory, as Axel Leijonhufvud (1973) notes in his comparison of the totems of micro and macro theories.

The distinction between macro and micro theory is generally thought to correspond to the distinction between a whole and its parts. This paper likewise treats the micro-macro relationship as one of parts-to-whole. If macro is just a scaled up version of micro, macro phenomena must be equally simple and apprehensible as micro phenomena. This paper parts company with such macro theorizing by treating the relationship between micro and macro as non-scalable, with this alternative treatment bringing several analytical possibilities into play. The relationship between micro and macro is no longer a simple matter of aggregation, recognizing also that aggregation is impossible in the presence of heterogeneity in any case (Stoker 1993) (Hartley 1997: 132-46). This paper approaches non-scalability by treating a macro economy as an ecology of plans. Micro entities form plans and act on them. A macro economy, however, is not an aggregate of plans but is an ecology of plans.

The difference between an aggregate and an ecology resides in unplanned and unintended interactions among plans that are part of the ecology and which are absent in the aggregate. An aggregate is of the same order of complexity as the entities that comprise the aggregate; an ecology is of a higher order of complexity than the entities that comprise the ecology due to interactions among the elements of the ecology. Within an ecology, the same micro units can generate different macro patterns in response to different patterns of connection among the micro units. With a macro economy as an ecology of plans, the macro characteristics of the ecology emerge through interaction among micro units and their plans of action. Macro phenomena have qualities that are not contained within each of the micro units that comprise the ecology. Leonard Read (1958) explains that no single individual knows how even to make a pencil. The ability to produce pencils is a macro-level quality that emerges out of micro-level interaction. The knowledge required to make pencils is distributed among persons and embodied in organizational practices and institutional arrangements.

Within an emergent or ecological orientation, there is no reduction of macro to micro through basing theories on averages or representative agents, thus avoiding the situation identified by Kirman (1992). Outliers have significant analytical work to do when the characteristics of the population of interacting agents cannot be reduced to a typical or representative interaction. As Kevin Hoover (2001: 74) puts it: “The macroeconomy supervenes on the microeconomy but is not reducible to it.” To theorize about macro phenomena is to theorize about objects that emerge through interaction among entities at the micro level of action. In consequence, the macro level is the arena of spontaneous ordering and unseen hands while the micro level is the locus of intentional planning and action. A macro economy is thus a complex ecology of evolving plans that constitute a non-equilibrium (Katzner 1998) process of spontaneous ordering.

This paper explores how emergent-style theorizing might be brought to bear on the macro-theoretic concern with an economy in its entirety. The paper starts by contrasting the standard choice-theoretic or Dynamic Stochastic General Equilibrium (DGSE) framework for macro with the emergent-theoretic framework that is adumbrated here. The emergent-theoretic framework replaces the postulate that observations pertain to equilibrium states with the presumption that societal orderliness is always incomplete and is subject continually to evolutionary development through the emergence of new phenomena within a complex ecology of plans. Within this alternative framework, macro phenomena emerge out of or supervene on interactions among micro units within the ecology of plans that constitute an economy. After exploring some methodological matters pertinent to this contrast between macro frameworks, the remainder of the paper proceeds by presenting a set of topics where the emergent framework leads into different analytical territory from where the choice-theoretic framework leads. This perusal of topics bears a family resemblance to Thomas Schelling’s (1978) treatment of the relation between micro-level actions and macro-level results of interaction and to Peyton Young’s (1998) examination of how interaction among individuals generates the emergence of institutional arrangements. Indeed, the emergent-theoretic framework pursued here locates institutional arrangements and not resource allocations as the prime objects of macro-theoretic analysis, similar to Nathan Rosenberg’s (1960) treatment of the Wealth of Nations as being centered on institutional arrangements and not on resource allocations.

2. A Fork in the Theoretical Road

The world appears to us as generally but not universally orderly. The reasonableness of that appearance is attested by our ability to pursue deliberate action in the world with reasonable success. Not complete success to be sure, because reality also comes with turbulence that sometimes sinks well laid plans. The perception of orderliness provides the point of departure for social theorizing, for without a sense of societal orderliness there would be no object to theorize about. That point of departure, however, presents a fork in the theoretical road concerning the character of the theoretical relation between micro parts and macro whole.

The relationship between micro and macro is the province of the micro foundations of macro theory (Janssen 1993) (Horwitz 2000). Table 1 presents a six-point contrast between choice-theoretic and emergent-theoretic frameworks for constructing a macro theory that is based on micro-theoretic foundations. There is, of course, significant variation among proponents of the DSGE framework. Where some proponents of that framework take a suitable micro foundation to be the theory of perfect competition and its Pareto efficiency, other proponents embrace a micro foundation of imperfect competition and its Pareto inefficiency. Such variation, however, is superficial and resembles a family feud in that the commonalities among the disputants dominate the differences when contrasted against the emergent-theoretic framework.

First, proponents of the choice-theoretic orientation treat macro observations as representing equilibrium states to be an analyzed through comparative statics. In contrast, the emergent-theoretic orientation is non-equilibrium in character, as illustrated by Shackle (1974) and Katzner (1998) which, it should be noted, is not a disequilibrium framework. It is rather a framework that starts from the common sense observation that life faces us with both persistence and turbulence. These are both features of an ecology of plans: at any instant many plans are continuing without change while yet other plans are being introduced, and with some of those inducing future changes, including abandonment, of presently continuing plans. Regularity in society does not require equilibrium; it requires only that the social world be intelligible, which is consistent with the continual injection of change into society through the introduction, revision, and abandonment of plans. Within a macro ecology of plans, moreover, there will generally be discernable statistical relationships among macro-level variables that are a facet of the intelligibility of the social world.

Systemic equilibrium portends placidity and not turbulence. To get turbulence from postulated equilibrium requires the insertion of exogenous shocks. Even then, however, the turbulence is purely notional because it is just a place holder within the model to indicate that a new equilibrium will be obtained and can be apprehended via comparative statics. Turbulence is brought in through a deus ex machina rather than being treated as an intelligible feature of the ecology of plans that are in play inside a society. In contrast, turbulence is a normal feature of most ecologies, including the ecology of plans that comprises a society (Louçã 1997) (Potts 2000). Within a framework of limited and divided knowledge, plans are never formed in full knowledge of all other plans and the implications of those plans for the plan in question. For the plan in question, these other plans will act as exogenous shocks. But they aren’t truly exogenous shocks because they are a normal feature of the ecology of plans. For an individual the rest of the world presents mostly a parade of exogenous shocks, but for the world as a whole those aren’t shocks but merely internally generated turbulence through collisions among plans.

To treat both orderliness and turbulence as properties of the life of the ecology of plans that constitutes a macro economy requires that we follow the emergent-theoretic branch of that analytical fork, which seeks to explain the emergence or generation of orderliness instead of postulating it as an analytical point of departure. An emergent analytical framework would seek to explain macro coordination as something that arises through interaction among non-coordinated entities,¹ and would reflect a bottom-up orientation toward macro theorizing as illustrated by De Grauwe (2010). Coordination would thus be a variable quality of societal processes and not a maintained hypothesis to guide theoretical effort, as illustrated, for instance, in the generative style of theorizing conveyed in Axtell and Epstein (1996) and in the essays collected in Epstein, ed. (2006).

Second, the choice-theoretic presumption that observations pertain to equilibrium states allows reduction of macro to micro. As Hoover (1998: 87) notes, this reduction allows “the euthanasia of macroeconomics.” Within this analytical framework, the structure of micro relationships has no work to do because the presumption of equilibrium converts that structure into mere background. By contrast, the reduction of macro to macro is impossible within the emergent-theoretic framework. Macro phenomena supervene on and emerge out of micro-level interactions, as Hoover (2001) notes in an economic context and Kim (1993) examines philosophically. In this manner the entire population of micro agents, along with the patterns by which they are connected, and not just some measure of central tendency, is relevant for the macro-level properties of the emergent order. The relation between micro and macro is neither reductionist nor holistic, as McQuade and Butos (2009) explain in their treatment of adaptive systems. Causation runs in both directions, so it is reasonable to speak of macro foundations for micro as well as speaking of micro foundations for macro (Smithin 2004) (Wagner 2010). Robinson Crusoe, for instance, has neither property rights nor engages in contractual relationships. Property and contract are macro-level phenomena that emerge though interaction among multiple Crusoes, and at the same time these emergent institutions influence the pattern of micro-level activity.

Central to the supervention of macro observations on micro interaction is recognition of distinct levels of phenomena as against treating all phenomena as residing on the same theoretical plane. Mitchel Resnick’s (1994) computational analysis of a traffic jam illustrates the analytical distinction nicely. Suppose the cars entering a highway follow the simple rule of driving as fast as they can until they are three car lengths behind the car in front of them, at which time they maintain that distance. If, within this steady stream of traffic, one car slows down momentarily, the following cars will likewise slow down and a traffic jam will form. The traffic jam is an object that is distinct from the individual cars that constitute the jam. A sequence of photographs, moreover, would show the traffic jam to be moving backwards. Yet it would not be accurate to describe the traffic jam as a gigantic car that is moving backward. No car ever moves backward. The traffic jam is a distinct object that supervenes on the cars that constitute the jam. It is this quality of emergence and supervention that comes into play in treating a macro economy as an ecology of plans. A central feature of that ecology, as any ecology, is the emergence of phenomena with qualities that are not reducible to qualities possessed by the micro units that constitute the ecology.

Third, the difference between the choice-theoretic and emergent-theoretic orientations can be summarized with reference to two distinct types of orderly social configuration: a parade and a crowd of pedestrians passing through a piazza. Both of these configurations are orderly in that people can act intelligibly within each configuration. The choice-theoretic orientation effectively treats a society as a parade. Parades vary in quality, to be sure, with some parades perhaps coming close to some image of perfection while others seem quite distant from perfection. Regardless of whether a parade is thought to be perfect or imperfect, it is appropriately reducible to a point-mass entity. Whatever number of participants a parade might have, they can be reduced to a point-mass entity that can be described by a direction and speed of movement. If one were asked to explain observed variation in the quality of different parades, the explanation would surely run in terms of such things as the organizational skill of the parade marshal, the musical and marching abilities of the participants, and the amount of time given to rehearsal.

The crowd passing through a piazza constitutes a distinctly different social configuration, one that is still generally orderly but which bears little resemblance to a parade. The pedestrian crowd is not reducible to a point-mass entity. The crowd cannot be described by direction and speed. Following a particular pedestrian through a piazza gives no information about the entirety of pedestrians in the piazza. Where the parade is an organization under the direction of a parade marshal, the people within the piazza constitute an order of organizations, with each organization pursuing its own plan of action. A pedestrian crowd is a form of spontaneously ordered network, as explored in Barabási (2002). The orderly quality of the pedestrian crowd resides in the rules and principles that govern interaction among the participants. It would be ontologically mistaken to treat the pedestrian crowd as an imperfect parade that potentially could be perfected through policy, even though it is conceivable that the orderliness of the crowd could be improved along the lines examined in Schelling (1978). It is a pedestrian crowd and not a parade that is an appropriate analogy to the orderly but turbulent social configuration that is constituted through an ecology of plans.

Fourth, the choice-theoretic framework treats macro phenomena as equally simple as micro phenomena. Treating macro as an equilibrated aggregate of micro actions is equivalent to treating a parade as an aggregation of equilibrated pedestrians. Just as a parade is a scaled-up version of a single pedestrian, a macro economy is a scaled up version of an average or representative micro unit. In contrast, the emergent-theoretic alternative recognizes that the micro-macro relationship is scale free. Within the ecology of plans framework, the move from micro to macro is a move from simple to complex phenomena sketched in Hayek (1967) and extended in new directions by Koppl (2010), Magda (2010), and Rosser (2010). While the DSGE reduction of macro to micro is associated initially with Léon Walras, it is worth noting that Walras (1954: 380-81) raised briefly the possibility of treating the macro level as an ecology of plans, only to turn away due to analytical complexity. He raised this possibility by contrasting his formulation of an annual market of pre-coordinated activity with what he called a continuous market, which he described as resembling a lake where the water is agitated by the wind, in contrast to the placidity of the water in the annual market. Walras’s continual market allowed variable turbulence as a systemic feature of micro-level interaction, only he abandoned this insight in favor of the analytical closure that he thought equilibrium modeling offered.

Fifth, one facet of the macro-level complexity that accompanies the emergent-theoretic orientation is a treatment of polities as likewise engulfed in complex processes of interaction. Where political economy within the choice-theoretic framework treats polities as unified entities that intervene into market processes (Drazen 2000), an emergent-theoretic political economy treats polities as complex processes of competitive interaction among the units that comprise a polity (Wagner 2007). If the micro-macro distinction were a simple matter of scale, micro and macro phenomenon would be equally capable of guidance by planning. After all, individual action is the province of planning, so by extension the macro level would likewise be the province of planning. Within the ecology of plans that pertains to the emergent-theoretic framework, systemic planning faces limits that do not appear within the choice-theoretic framework. The choice-theoretic framework conceptualizes policy as polity acting on economy, with both entities treated in point-mass fashion as when one billiard ball strikes another. Within the emergent-theoretic framework, neither polity nor economy is reasonably reducible to a point-mass entity; moreover, the relationship among those entities is knotted and not separated.

Sixth, the choice-theoretic framework grounded in simplicity makes policy appear to be a simple matter of selecting among equilibriums through comparative statics. The end of analysis is placed on prediction as the instrument of policy advocacy. Within the emergent-theoretic framework, by contrast, the end of analysis is placed on explanation and understanding. As Kochugovindan and Vriend (1998: 64) summarize: “… because emergent phenomena are inherently unpredictable to some extent, the study of complex adaptive systems confirms that the true objective of science is not prediction but understanding.” A parade is susceptible directly to control through policy, as a parade marshal can change the direction or pace of march in an instant. Doing this is not possible within a piazza because the people who are passing through it have their own objectives which often will be maintained despite the contrary efforts of policy. Once polity is recognized as an order of competitive organizations, moreover, the problem of policy comes to be recognized as more of an institutional and constitutional matter than a matter of systemic planning and control.

3. Micro and Macro: A Bi-planar Framework for an Ecology of Plans

This paper pursues a bi-planar framework for exploring the relationship between micro parts and the macro entirety within an ecology of plans. Alternatively, Potts and Morrison (2007) use a tri-planar framework that distinguishes among micro, meso, and macro levels, with the meso-level containing institutions which reside on the macro level in the bi-planar framework. Regardless of the number of planes that might comprise an emergence-based framework, the relation between micro action and the macro resultants of micro action requires explanation in terms of planes where one plane contains activities that are projected onto a higher plane through such entities as statistics, expectations, conventions, institutions, and ideologies. There are individuals who act and there are social products of interaction among those acting individuals, and with those products shaping, channeling, and constraining individual action. In this respect, James Coleman (1990), in his treatment of theorizing about interdependence among actors in society, distinguishes between the level of action and the level of system. With respect to this framework, he notes that “the only action takes place at the level of individual actors, and the ‘system level’ exists solely as emergent properties characterizing the system of action as a whole. It is only in this sense that there is behavior of the system as a whole (p. 28, Coleman’s italics).”

Figure 1 illustrates a bi-planar framework for exploring a macro economy as an ecology of plans that emerges out of interaction at the micro level. The upper part of Figure 1 illustrates three possible observations in Phillips-curve space to make contact with choice-theoretic macro theorizing. These observations could be thought of as representing three different levels of aggregate demand corresponding to different rates of inflation and output, with a and c denoting deviations from the natural rate of output represented by b. To be sure, the various macro theories differ in how they account for those observations, but the central point in any case is that the object of theoretical attention for choice-theoretic macro theorizing is captured by those macro-level variables. While Figure 1 uses a Phillips curve portrayal of macro-level theory, such other macro-level portrayals as aggregate supply and demand or money supply and demand could be substituted. Whatever the particular type of macro relationship brought under examination, the emergent framework would treat it as supervening on micro-based network of actions and relationships.

Those micro relationships are portrayed in the lower part of Figure 1 as a network of acting entities of two types: the circles denote market-based entities and the triangles denote polity-based entities. The lightning bolts pointing from the micro level of action toward the macro level of statistics, projection, and ideology shows that the system level is real but is not the level where action occurs. Action takes place only on the micro level, with macro-level observations supervening on micro-level interaction. The aphorism “think globally but act locally” has the macro-micro relationship almost right. The only emendation required to get the aphorism exactly right is to note that there is no option to local action, for any subsequent global impact is an emergent product of micro-level actions and interactions. This doesn’t mean that the macro level is inert, not by any means. The macro-level is the home of such holistic objects as institutions, statistics, projections of future circumstances, and ideological beliefs and presumptions. Such macro phenomena can influence micro-level action as instances of downward causation, as explored by Lewis (2010) and Pryor (2008). For instance, a prolonged depression might influence beliefs about the stability properties of a market economy, changing in turn the subsequent pattern of micro-level interaction in the direction of greater public involvement in economic activity, especially when, as Kirman (2010) notes, people are to a significant extent influenced by those with whom they associate rather than acting independently of one another. This influence of macro patterns on micro actions is no different from the ability of macro-level projections of an ageing population with much wealth and leisure time inducing micro-level action in the form of enterprises that construct facilities to accommodate the greater use of recreational vehicles that might be thought to be in the offing.

Figure 1 pertains to a bottom-up formulation of the relationship between micro and macro, similar to De Grauwe (2010). There is, however, a significant difference between these formulations. De Grauwe pursues a bottom-up framework in order to develop an empirically more accurate choice-theoretic model of business cycles, and does so by varying the stock of information that agents possess. De Grauwe thus seeks to construct an alternative path to the construction of direct relationships among macro-level variables as these are reflected in national income accounts. In contrast, the cardinal presumption of the ecology-of-plans framework is that there is no direct causal relationship between micro-level action and macro-level resultants, even though empirical regularities might exist over some temporal interval, because the micro-macro connection is intermediated through societal structure.

Equilibrium-centered macro theory can, of course, give an account of interdependence among economic activities. Indeed, such an account is perhaps the prime virtue of this theoretical framework. What it can’t do, however, is give an account of turbulence that arises through inconsistencies among plans because no action is presumed to take place until all plans are mutually consistent. All plans are pre-reconciled within the equilibrium framework, just as the actions of the members of a parade are pre-reconciled. The alternative to the equilibrium framework is to treat the ecology of plans as an emergent process where macro-level objects supervene on micro-level interaction. Any relation among macro-level variables is thus intermediated through interaction among entities at the micro level.

Any plan has the structure wherein a decision to commit to a course of action is undertaken in advance of the market determination of the value of that action. Cost is borne prior to production, with the receipts generated by that production coming later (Buchanan 1969). Plans are initiated on the basis of ex ante projections or beliefs, with the ex post value of those plans ascertained only later and taking the form of windfall gains and losses. Equilibrium theory, however, neuters the distinction between ex ante and ex post, due to the stipulation that observations pertain to equilibrium states. For the distinction between ex ante and ex post to have significant analytical work to do, some such analytical framework as an ecology of plans is necessary because the distinction between ex ante and ex post can appear only in the absence of pre-coordination among plans. People may well lay plans by making efficient use of public information about plans now underway, but there will be much private information pertinent to the formation, revision, and abandonment of plans that will continually enter the ecology, leading to continual divergences between the ex ante vision on which a plan was undertaken and the ex post reality that subsequently was confronted.

The ecology of plans framework also eliminates the distinction between long run and short run. This venerable distinction is relevant for an individual plan, as it corresponds to the subsequent operation of a plan in the face of some divergence between the long run anticipation on which a plan was based and the short run judgment whether to continue with a plan that presently is not matching that long run anticipation. For an ecology of plans, however, the distinction between short run and long run is irrelevant. Within that ecology, there will be plans that are just getting underway while there are other plans that are in the process of being abandoned. The distinction between long run and short run is pertinent to individual plans but is irrelevant for the ecology of plans, for that ecology is just an arena into which plans are inserted to take wings or not.

What results from this analytical effort is a form of spontaneous order macro theory, as illustrated by Howitt and Clower (2000), Leijonhufvud (1981), Shackle (1974, and Witt (1997). Micro theory would be the domain of intentional action; macro theory would be the domain of emergent phenomena, spontaneous order, and unintended consequences. This line of analysis would reassert the sense of the distinction between what is seen and what is unseen. The micro domain of action pertains to what is seen and intentional. The macro domain of spontaneous order and emergence pertains to what is not part of anyone’s direct intention, but rather reflects interaction among participants along with phenomena that emerge through interaction. With respect to micro foundations, this effort points toward emergent-theoretic foundations for macro theory, as distinct from choice-theoretic foundations.

The challenge embraced here is to analyze the generation of orderliness cum turbulence within the ecology of plans that constitute a macro economy and not to compare the properties of some equilibrium arrangement of plans against some postulated Paretian standard. With respect to what is denoted as policy, moreover, the entities of state are likewise members of the crowd and nothing like a parade marshal. Within an ecology of plans, new plans continually are being created while existing plans sometimes are being revised or even allowed to die. The interconnection among plans in this ecology is a source of turbulence, not as an exogenous shock but as a systemic feature of what is a living even if not a sentient organism.

The actors shown in Figure 1 comprise a network that contains two types of actors and possesses a particular pattern of connection among those actors. With respect to the types of actors, the circles denote market-based entities and the triangles denote polity-based entities. Both types of entity exist and operate on the action plane, for there is nowhere else they can reside and operate. Those entities have somewhat different rules of action, which is significant for some of the turbulence that can emerge within the ecology of plans. While all entities are guided by a search for gain, that search can play out differently as between market-based and polity-based entities. Still, the preference for gain over loss is a universal characteristic of all acting entities, as Buchanan (1969) explains, even though this universal characteristic can be intermediated differently within different types of organizations.

Figure 1 shows both market-based and polity-based enterprises acting on the same plane, and with each being sources for the generation of macro-level data. It is well recognized that profit motivation differs from bureaucratic motivation, even though both entities pursue gain over loss. For instance, both each party in a dispute between market-based entities has incentive to settle the dispute without going to trial because they are residual claimants on their litigation expenses. It is different for a dispute between a commercial and a political entity because a political entity is not a residual claimant on litigation expenses. Those expenses, however, can serve as a form of investment in seeking higher office if a trial has the ability of trials to attract favorable public attention. The commercial calculus of profit-and-loss would be replaced by an alternative though related calculus of political gain. While an ecology of plans is naturally turbulent due to the operation of ex ante-ex post divergences, turbulence would also arise through institutional incongruity within the ecology, as noted by Ludwig Lachmann (1971) in his treatment of Max Weber’s institutional economics.