Claudius Gräbner and Jakob Kapeller
Any discussion of the micro–macro link in heterodox economics entails two main questions. The first question is relevant for social sciences in general and asks for the correct or adequate treatment of aggregates and aggregation in social theory. Any answer to this general question incorporates a series of diverse philosophical viewpoints, including: ontological claims, for example, whether social and economic aggregates exist; epistemological questions, for example, regarding the role played by aggregates and aggregation in economic theory; and methodological aspects, for example, how to adequately model processes of aggregation. Given that economics faces myriad problems of aggregation—as in the case of market interaction, macroeconomic aggregates, or interpersonal coordination and contracting—the quest to provide suitable theoretical tools to adequately address aggregates and aggregation is of special interest to economists of different persuasions.
The second major question is more specific and asks for similarities and differences in the treatment of aggregates and aggregation among heterodox economists. From a traditional viewpoint one might question the idea that there is something like a consistent vision of the micro– macro link in heterodoxy, since different interpretations of the micro–macro link have been attributed to various strands of heterodox research. While some heterodox economists may prioritize either micro- or macro-level analysis, others emphasize the meso level as a decisive intermediate layer between the more traditional approaches focusing on either microeconomic or macroeconomic aspects.
Against this backdrop, this chapter introduces a unifying heterodox approach to the micro– macro link in economics. The first section emphasizes the analytical problems that may arise from popular misunderstandings about the relationship of individual and aggregate levels. This discussion illustrates why a thorough understanding of aggregation and aggregates in science is necessary. In the second section we show that the different heterodox approaches to the micro– macro link in economics are not only consistent, but complementary to each other and allow for a concise treatment of the micro–macro link in economics based on a set of shared fundamental principles. The third section embeds the heterodox economic view on aggregation in a ‘system-ist’ framework and demonstrates that heterodox economic theory and research practice can be substantiated and summarized by a more general, philosophical perspective on aggregates and aggregation in science. We argue that such a general philosophical framework opens possibilities for advancing heterodox theory, because it allows scholars from different heterodox starting points to relate their theories to each other via a consistent meta-language. The last section offers some concluding thoughts.
Scientific endeavor often deals with the relation between aggregate entities, like a family, a nation or a firm, and their individual constituents—family members, citizens, or employees. In disentangling this relationship between the ‘whole’ and its ‘parts,’ errors may occur, which can be understood as compositional fallacies. Such compositional fallacies arise from either a wrong treatment of aggregation or a wrong treatment of aggregates, and may lead to a deficient understanding of the whole as well as its parts. This section provides illustrations of four typical compositional fallacies and their conceptual sources, in order to develop a basic understanding of the problems usually associated with the micro–macro link in the social sciences. For a more complete exposition of the particular fallacies, see Kapeller (2015).
One typical error related to aggregation is underestimating the role of relations across individuals. Aggregates contain not only individual entities but also a corresponding set of relations, which tie their individual nodes together and create a certain structure. Taking these relations into account is crucial for an adequate understanding of the constitution of the aggregate entity. For example, a central concern in feminist economics is to take individuals and their mutual relations instead of households as the starting point for economic analysis. Otherwise the important ‘aggregation procedure’ from individual to household preferences and the underlying relational structure among individuals that explains the often detrimental position of women would be neglected (Drèze & Sen 1989). Taking the household as a fundamental economic actor, therefore, contributes to the exclusion of gender issues from economic analysis. This example illustrates that aggregation problems are central to economic analysis and can appear in rather early stages of theorizing as the behavior of some aggregate social system, like a family or a nation, strongly depends on its internal structure, that is, its relational setup.
Most standard economic models, both mainstream and heterodox, ignore relational setups entirely and determine aggregate behavior by some simple procedure of summing up across individual entities. The ‘simplistic fallacy’ is thereby based on a deficient understanding of aggregation, which views the ‘whole’ as no different from the ‘sum of its parts.’ Such a view conflicts with two basic observations. First, aggregates may develop properties that no individual part possesses, such as a firm’s success. Second, individuals may acquire certain properties precisely because they are part of some whole, for example, a country’s citizen. In both cases we find that ‘more is different’ (Anderson 1972), as these newly acquired attributes may be conceived as ‘emergent properties,’ meaning novel features that arise because an aggregate is constituted or sustained. For the case of families such novel properties include the possibilities of raising offspring, lending mutual support, and creating collective identities and a shared organization of common rights and duties, which may leave some family members in a dependent and potentially deprived situation compared with others. These aspects are often neglected, for instance, if households with multiple members are represented by a single utility function.
Another fallacy regarding the micro–macro link in economics is the assumption that causality across different levels only runs from individual elements to aggregate properties (‘bottom-up’), which we label the ‘hierarchical fallacy.’ Current mainstream economics follows such a routine by imposing a general “hierarchical stipulation that macro-theories require a microeconomic foundation to obtain full validity” (Rothschild 1988: 14). The economic mainstream thereby emphasizes the necessity of reductionist strategies, which explain aggregate phenomena by ‘summing up’ the behavior and properties of the parts (for example, Robbins 1932; Lucas 1976). By contrast, heterodox economists of different persuasions advocate for a multi-level approach to economic theorizing (for example, Dopfer et al. 2004; Lee 2011; King 2012), emphasizing the changing conditions and constraints for economic action on different levels and the mutual co-existence of top-down and bottom-up mechanisms.
One reason why such a ‘hierarchical stipulation’ creates more problems than it solves when theorizing about economic phenomena is that emergent properties not only arise permanently, but also feed back on their constituents, which cannot be assessed within a unidirectional framework. Consider, for example, innovation in market environments and the associated forces of ‘creative destruction’ (Schumpeter 1942) and ‘path-dependency’ (David 1985); or the continuous evolution of social routines (Hodgson & Knudsen 2004) and consumer preferences (Witt 2001). In these contexts, where the relationship between individual action and aggregate outcomes is manifold and complex, the epistemological presupposition of a ‘hierarchical stipulation’ of micro over macro acts as a dual barrier to clear understanding. First, it leads to the ‘static fallacy’ which asserts that aggregate properties can always be reduced to lower-level entities, thereby underestimating the dynamics and complexity of social interactions, and overlooking the unexpected. Second, it gives rise to the related ‘dogmatic fallacy,’ comprising the methodological claims that aggregate properties should always be reduced to (current) micro-knowledge and that higher-level mechanisms are mere residuals of individual behavior and therefore negligible. Consequentially, emergent aggregate properties are not accorded any meaningful explanatory role.1
In contrast, causally relevant top-down relationships in economics can be found in various contexts. We can think of the influence of social norms and regulation on business practices or the social mediation of consumption preferences within a certain community. In the case of the family, traditions, power, and hierarchy are of vital importance for understanding women’s deprivation (Drèze & Sen 1989). The importance of such top-down mechanisms in economics has long been emphasized by heterodox economists, some of whom would even claim to inverse the ‘hierarchical stipulation’ inherent in mainstream economics and, conversely, demand a ‘mac-roeconomic foundation’ for microeconomics (for example, King 2012).
One natural example of a heterodox research strategy following this tradition is stock-flow consistent macroeconomics (Godley & Lavoie 2007). Here, one starts with accounting identities and other stylized macroeconomic facts to study macroeconomic dynamics, rather than starting with speculative assumptions about the behavior of a ‘representative household’ (see Kirman 1992), as it is common in dynamic stochastic general equilibrium models. While such a ‘macro-founded’ approach provides an important alternative perspective on macroeconomic dynamics, it is not to be seen as superior to a micro-founded approach per se. Rather its contribution is to make macroeconomic constraints explicit in modeling, which serves as an important heuristic even when the model is eventually based on microeconomic relationships and assumptions as in so-called ‘agent-based stock-flow consistent models,’ which study the economy as a complex system with both bottom-up and top-down effects (for example, Caiani et al. 2016). Hence, the merit of taking different levels of analysis into account often lies in the enriching of perspectives on economic phenomena, which aligns well with heterodox economists’ practice of pluralist engagement.
Table 10.1 summarizes the conceptual and methodological pitfalls collected in this section and illustrates the relation between these four fallacies. In what follows, we first examine the principles that heterodox economists have developed to deal with aggregates and aggregation
Table 10.1 Compositional fallacies: an overview
| Main error | Fallacious routine | Underlying misconception | |
| The simplistic fallacy | Ignoring relations, i.e. underestimating the complexity of aggregation | Simply summing up individual properties | |
| Wrong treatment of aggregation: "the whole is nothing more than the sum of its parts." | |||
| The static fallacy | Ignoring the possibility of unexplainable novelties / irreducible properties | Any aggregate property can be reduced | |
| The dogmatic fallacy | Ignoring that higher-level mechanisms can be studied on their own | Always aim at providing bottom-up explanations | Treatment of aggregates: "wholes cannot be explanatory—they do not carry mechanisms." |
| The hierarchical fallacy | Ignoring the possibility of downward causation | Never provide top-down explanations |
in economics, before developing a general philosophical foundation suitable for summarizing heterodox approaches.
One overarching theme in heterodox economic theorizing is that the consideration of social wholes is important for understanding socio-economic processes and outcomes. This view implies that wholes are more than a mere sum of their parts, since they exhibit non-trivial properties and carry effects of various sorts which cannot be gleaned from looking solely to their constituent parts. However, this idea has also been subject to different interpretations and applications within heterodox economics, leading to a series of distinct vantage points on the role of aggregates and aggregation in economics. Some scholars focus on the explanatory role of top-down mechanisms, while others try to escape the simplistic fallacy by building particularly sophisticated micro-founded models. They are united in treating the micro–macro link as a complex relationship that deserves theoretical attention, because social and economic aggregates may constitute novel objects or, at least, come with novel features, which may have specific real world consequences. We now turn to four fundamental perspectives that have played a decisive role in heterodox treatments of the micro–macro link.
One main implication of the idea that wholes do make a difference is to consider the spatial and temporal variance of social configurations in order to identify distinct realms of economic activity. Such distinctions may refer to historical differences (for example, ‘medieval feudalism’ versus ‘twentieth century welfare state capitalism’), spatial variations (for example, ‘core’ and ‘periphery’) or distinct spheres of economic activities (for example, ‘competitive market societies’ versus ‘sub-sistence communities’). These distinctions are deemed important since the course and effects of economic activities depend on their social and historical circumstances.
Such differentiations are often found in classical political economy. Consider, for instance, John Stuart Mill’s distinction between the sphere of production and the sphere of distribution. For Mill, the decisive difference between these two economic realms is that while the former is constrained by nature, the latter is solely shaped by human institutions (Mill 1848: II.1.1–2). Consequently, different laws and assumptions apply in these contexts.
In Mill’s account, humankind is clearly subject to ‘macrofoundations’ in the form of environmental, historical, and societal forces, as they largely define the constraints and modes of economic activity. In this regard many heterodox economists argue that holistic factors, like culture or institutions, are important for explaining social phenomena and allow for top-down effects or downward causation within their economic theorizing. In a bold and overarching interpretation, this view may extend to the claim that social and economic conditions completely determine actions, fate, and feelings of individuals. Such a view of socio-economic determinism is often associated with Marx’s concept of ‘historical materialism,’ although such an interpretation does not do justice to the original Marxian account.2 A more accurate account would probably compare the heterodox approach to that of a physicist studying the behavior of a single element (for example, the behavior of a comet entering the solar system or the pressure in some gas-container in a lab) by taking full account of the surrounding system (for example, the composition of the solar system or the size of the container) to correctly anticipate the impact of the latter (Andersen et al. 2000). This more modest attitude is key to understanding a variety of heterodox ideas, from ecological economists’ emphasis on absolute constraints (Georgescu-Roegen 1971) to Keynes’ definition of economics as “the art of choosing models which are relevant to the contemporary world” (Keynes ([1938] 1973: 296).
A second application of the general idea that social wholes make a difference focuses on the interrelatedness of individuals. Attention is devoted to the relations between individuals, and the corresponding impact of other people’s attitudes and behavior on an individual’s economic thought and action.
Interactions among agents as well as between structure and agency are of prime interest to heterodox economists, including: the analysis of interactions and relations across individual agents and of preference formation in the context of social emulation (following Marx or Veblen); the emergence of routines in organizations (Nelson & Winter 1982); and questions of social identity and the evolution of cooperation (Axelrod 1984; Bowles & Gintis 2011). Such a perspective naturally takes relations seriously and allows for agents of different influence and power, thereby serving as a guide for theorizing about self-reinforcing effects (Merton 1968), path-dependency (David 1985; Arthur 1989), wealth concentration, power structures, and elites (Rothschild 1971) as well as other forms of cumulative advantage (Myrdal 1957). One important implication of this reasoning is the conceptualization of an economy as a circular flow, where one person’s expenditure adds to another person’s income. This circular flow views monetary transactions as fundamental interactions, which constitute mutual interdependencies among single economic actors. The consequences of these interdependencies are a major theme in heterodox economics.
This focus on the role of relations avoids the simplistic fallacy by understanding social wholes as constituted by a set of individual nodes and their corresponding relational setup. Social relations play a twofold role. First, they serve as a transmission belt for cultural norms, institutional conventions, established hierarchies, shared goals, and aspirations within a social whole. Second, relations serve as a means for understanding how individual actions might influence social wholes and, hence, provide a lens for assessing social change and novelty.
The role of social relations as a transmission belt features prominently in heterodox economic thought and can be traced back to Karl Marx and Friedrich Engels ([1845] 2004: 122), who speak of individuals as an “ensemble of the social relations”; Thorstein Veblen (1899), who emphasized the social formation of consumer preferences; or Karl Polanyi (1944), who coined the term ‘embeddedness’ to highlight how individual economic action is always embedded in a certain socio-historical context. From a dynamic perspective this view can also be used to analyze questions of social mobility, where relational structures serve as a means for preserving social hierarchies as in Bowles & Gintis (1975) or Bourdieu (1998), who studied the role of educational systems for stratification in the US and France.
The second major feature of social relations, which allows active agents to influence aggregate properties, also has a prominent role in heterodox thinking and is exemplified by conceptions such as Schumpeter’s entrepreneur (1934) and Keynes’ animal spirits (1936). Both emphasize that some individual decisions are of great impact for future developments. Against this backdrop it comes as no surprise that active agency plays an important role in heterodox approaches to economic cooperation and trust, path creation and path dependence, and institutional design (for example, Hirschmann 1970).
This dual character of social relations allows for top-down as well as bottom-up effects and thereby captures the fact that individual agents and social structure are mutually interdependent. This point was most explicitly taken up by Mark Granovetter (1985), who refined the concept of embeddedness. Granovetter distinguished between over-socialized and under-socialized conceptions of individuals, attributing the latter to neoclassical and new institutional economics, while the former can be found in holistic approaches to social and economic analysis.
Both conceptions eventually posit an atomistic conception of individuals devoid of any relational embedding. For the over-socialized individuals, action has already been completely determined by social forces as a whole and quite independently of any specific relational setup, while under-socialized individuals do not have any significant relations to others. Granovetter (1985: 487) sees the embeddedness perspective as a conceptual alternative, wherein
[a]ctors do not behave or decide as atoms outside a social context, nor do they adhere slavishly to a script written for them by the particular intersection of social categories that they happen to occupy. Their attempts at purposive action are instead embedded in concrete, ongoing systems of social relations.
In this view, issues of trust and sympathy affect all interpersonal relations, even those, which are constituted as pure economic relations of exchange. The economic implications of this reasoning are non-trivial, as they concern industrial structure, trust levels and bargaining processes as well as the level of economic performance. A classic example is given by the high-tech sector, where clusters of coordination and cooperation are particularly common. These clusters are characterized by regular interactions among the involved suppliers, developers, and customers, which lead to a quasi-integration of different steps throughout the supply chain, although theses steps are carried out by formally independent organizations. The longer such relations exist, the more do they ‘outgrow’ the market and become insensitive to market signals such as ‘prices’ (Elsner et al. 2015).
However, Granovetter (1985) stresses that social embeddedness is not only a source of trust, stability, and cooperation, but can also lead to exploitation, disorder, and conflict. Hence, Granovetter’s approach does not allow for general predictions aside from the claim that “network structures matter,” since outcomes eventually depend on the overall network structure. For example, determining whether a system is vulnerable to particular interest groups who work against general interests, thereby depleting stability and trust, depends on the concrete case at hand.
In mainstream economic accounts such structural properties largely remain implicit; for instance, most Walrasian general equilibrium models do not account for networks explicitly, but implicitly assume the system to be structured as a bipartite star network, as illustrated in Figure 10.1(a). In this setup agents are not directly related to each other, but rather connected indirectly via a central auctioneer who has direct relationships with all agents and, hence, resides in the network’s center. A change of the network structure has non-trivial effects. An otherwise identical model economy characterized by a ring network, depicted in Figure 10.1(b), exhibits very different distributional characteristics and price patterns to the star network, implicit in the conventional Arrow–Debreu economy (Albin & Foley 1992).
Real networks are, of course, neither rings nor stars. Network analysis has made impressive progress since the 1990s and found that most empirical networks look in some ways similar to Figure 10.1(c). There are few nodes with many connections, and many nodes with few connections. Furthermore, nodes are organized into different clusters.3 The economic implications of this structure is an important avenue for future research.
But can we make reasonable predictions about the meso or macro level of the economy, given a precise description of the micro components as well as their relations? Notwithstanding the obvious merits of such an approach, the next section explains why a general affirmative answer to this question cannot be provided.
Another aspect of a heterodox perspective on the micro–macro link, which is strongly intertwined with incorporating relations into social and economic analysis, is the notion that novel objects or properties constantly arise in the course of social interaction. This specific aspect of social wholes—the fact they are carriers and transmitters of genuine novelty—is at the heart of this subsection.
Figure 10.1 Three different social networks Notes: (a) represents the theoretical Arrow–Debreu Economy with the Walrasian auctioneer in the center; (b) is a ring, capturing dense neighborhood structures; (c) shows a scale free network. Source: Author’s own illustration.
Economic systems regularly produce novel features that are not predictable from past data. The emergence of novelty can be most intuitively illustrated for the case of innovation, which was a key element in the work of Joseph Schumpeter. Schumpeter ([1934] 2011: 66) distinguishes between five types of innovation, such as the development of a new good, the introduction of products of higher quality, or alternative methods of production.
All these innovations may represent or bring forth novelties that were not existent in an economic system before. They may carry new mechanisms that fundamentally change the functioning of the economic system as a whole. For example, the advent of contemporary globalization not only came with cheaper import goods, but also introduced a new mechanism—the race for national competitiveness—which puts different countries in a competition for serving the interests of powerful transnational corporations. The invention of digital computers led to new markets, new goods, and new lifestyles, which continue to influence our society via novel mechanisms in various ways. While innovations are often a creative recombination of existing ideas, neither the exact way of recombination nor its consequences for society are a priori predictable.
This fact motivates heterodox economists to further investigate the micro–macro link to gain a deeper understanding of this non-predictability. One main pillar in this context is the development of arguments on fundamental uncertainty in economic action, which focuses on the role of crucial decisions in investment whose effects are very hard or impossible to anticipate. These circumstances give scope to alternative economic motives beyond conventional utility maximization such as routines and rules, individual vision and passion (as in Schumpeter’s entrepreneur), or inherited instincts (as exemplified by Veblen’s ‘instinct of workmanship’).
Another venue of work in this context aims for a refined conception of ‘meso’ in economic analysis (Dopfer et al. 2004; Elsner & Schwardt 2014). Proponents argue for a ‘micro–meso– macro’ framework as a substitute for the conventional micro–macro dichotomy. In such a framework the economic agents represent the micro level of the economy and are heterogeneous as they carry different rules. A rule and all its actualizations constitute a meso unit, which is seen as a key element in evolutionary economic investigation, since the interaction of rules is understood as a main driver of economic change on the macro level. Assume, for instance, that creditors and debtors in a given economy mutually adapt their crediting and borrowing behavior to each other. In such a setup increased risk taking on the side of creditors, who are prone to forget or ignore past turbulences, would be mirrored by increased borrowing by debtors leading to potentially unsustainable levels of debt. Hence, we can reach the classic Minskyan result that stability breeds instability (Minsky 1986) by employing a simple model of rule convergence on credit markets.
The obvious advantage of such an analytical framework is that it facilitates a focus on economic change and thus greater understanding of the source of the unpredictability of real novelty within the economy. One more specific expression of this rather general claim is provided by the theory of path dependence, originating from the seminal papers on technological lock-in by David (1985) and Arthur (1989). We can disentangle path-dependent processes into three different phases (Sydow et al. 2009; Dobusch & Kapeller 2013).
The first phase, path creation, characterizes a situation of contingency. Events happening at this stage are usually “outside the ex-ante knowledge of the observer” (Arthur 1989: 118). They are nevertheless important because these events characterize the initial conditions for the second phase, where an ergodic dynamic process, characterized by positive feedback effects and subsequent, causally linked events, leads to the dominance of one particular standard. The positive feedback may stem from different forms of network effects based on increasing returns, preferential attachment, learning and coordination effects, complementarity, or the convergence of expectations. The last phase, the resulting lock-in, reflects the resilience of the dominating standard against change. Thus, while it is almost impossible to predict the diffusion process ex ante, it only becomes possible to identify the dominant technology after one has entered the second phase of the path-dependent process.
Hence, path-dependency theory focuses on the mechanisms underlying the introduction of novelties and the creation and persistence of social standards of different forms, including social norms, organizational rules, business practices, and technological requirements. It provides a theoretical rationale for the emergence of novelties and explicates the difficulties in predicting such novelties. At the same time path-dependency theory is silent on the effects brought forth by such novelties, which often represent controversial questions in heterodox economics. A prime example is given by the effect of the adoption and diffusion of new innovations on the level of employment; while some innovations indeed function as labor-saving devices (as in the standard Keynesian approach), others may increase employment due to the creation of additional demand induced by novel products or improvements in product-quality and versatility (as in the Schumpeterian approach, see Witt 2001). Which effect eventually dominates in the face of general technological progress or a specific innovation is, hence, a question which can be hardly answered ex ante.
These arguments imply that ex ante predictions are often difficult or impossible, since the emergence and effects of novelties can hardly be fully anticipated. Nevertheless, the relevant trajectory can of course be explained ex post, as we can trace how a specific successful innovation diffused into society and how it affects their members.
Finally, our fourth perspective on social wholes and their role in economic theorizing relates to the normative question of economic welfare in the context of aggregation. Mandeville ([1714] 1962) advanced the view that ‘private vice’ in the form of egocentric instrumental rationality leads to ‘public benefits,’ that is, the maximization of social welfare. This view is deeply inscribed in modern mainstream economics, especially in the two fundamental theorems of welfare economics. While many heterodox economists could surely accept that the Mandeville case is a possible state of affairs, they also tend to critically examine the necessary conditions for such a result. Classical examples in this context include rationality traps—for example, if I can improve my view in the theater by standing up, will there also be a collective improvement if everyone follows this rationale?—and the ‘tragedy of the commons’ which describes the unsustainable usage a public good in the absence of a suitable mode of social coordination (Ostrom 1990). More formally, such cases can be expressed in the form of a prisoner’s dilemma, which illustrates the core property of rationality traps and tragedies of the commons, namely that myopic individually rational actions will lead to the worst possible aggregate outcome. This relationship is the main reason why some heterodox economists consider a prisoner’s dilemma as one archetype for heterodox economic modeling (Elsner et al. 2015). Since the welfare aspects of social organization are a general topos of heterodox economic theorizing, we find variants of this argument in several heterodox traditions: ecological economists’ emphasis on collective good problems (such as climate change); Marxian perspectives on power and conflict; and evolutionary and institutional economists’ focus on the role of social norms, conventions, and law in resolving social dilemmas.
While the label ‘systemism’ might seem new, one can be assured that the practice of system-ism is far from something completely novel. We introduce systemism to provide a full-fledged philosophical concept, which encompasses the basic heterodox arguments on the micro–macro link in economics. The development of systemism owes mainly to the works of Mario Bunge, philosopher and polymath, who aimed to transgress the traditional dichotomy between individualism and holism, which he perceived as an outdated hindrance to social research and epistemological debate.
Bunge cites a variety of examples for what he conceives as a ‘systemist’ social research. Interestingly, within these passages the names of heterodox economists come in definitely non-random abundance. Among others, Bunge mentions John Maynard Keynes and Wassily Leontief (Bunge 2004: 187), Max Weber, Joseph A. Schumpeter, Thorstein B. Veblen, or K. William Kapp (Bunge 1999: 92–93). More recent examples of heterodox approaches compatible with a system-ist perspective include understanding economics as the study of the social provisioning process (Jo 2011) and evolutionary economists’ focus on the meso level of economic activity (Dopfer et al. 2004). In sum, these observations suggest that heterodox economic approaches are salient candidates for illustrating a systemist approach to social and economic issues and, conversely, systemism serves as a natural candidate for substantiating established practices in heterodox economic research from an epistemological viewpoint.
Systemism is built upon the fundamental twin concept of systems and mechanisms, where the latter are situated within or between the former. Thereby any object or entity in systemist analysis is considered either as a system itself or as a component of a system (Bunge 1996).
A system is composed by a set of nodes or components (its composition ) with a particular relational setup (a system’s structure or organization ) situated within a certain environment. The interrelatedness of agents not only contributes to the constitution of a specific system, but gives rise to a variety of ‘ontological novelties’: that is, some features that the whole possesses, but its components lack ( global properties, like a nation’s culture or a firm’s success); or some features that components acquire precisely because they are part of some system ( relational properties, like being a creditor, a wife, or an employee; see Bunge 1996). The concept of a system can therefore be applied on several levels. For example, a family is a system consisting of different members with particular relations to each other; at the same time, it is part of a community system within which it has several relations to other components of the community. The resulting levels take the form of a hierarchy of sub- and super-systems, which serve as a basic ontological framework. Such a hierarchical understanding of reality has been insinuated by several heterodox approaches, in particular in the work of Herbert Simon (1962), who provides an evolutionary explanation for the predominance of hierarchy in complex social systems.
The second fundamental aspect in systemism is a focus on mechanisms, meaning law-like relationships, and the attempt to provide mechanism-based explanations of social phenomena (Bunge 1997). Mechanisms work within or across social systems and lead to continuous changes and stabilization of a given system. This is why we conceive of them “as a process (or sequence of states) in a concrete system, natural or social,” (Bunge 2004: 186). Thereby three rough types of mechanisms can be distinguished. First, within-level mechanisms operate within social systems, but address only one ontological layer; for example, a reduction in hourly income may induce a household to increase working hours. Second, bridging mechanisms also work within a certain social system but can take the form of agency-structure relations (a bottom-up mechanism or upward causation) or structure-agency relations (top-down mechanisms or downward causation). The former provide a theoretical alternative for the aggregation of individual behavior, going beyond a mere ‘summing up’ of individual properties by employing theoretical mechanisms for means of aggregation. Examples of such 'bridging mechanisms’ are bandwagon effects, where final outcomes depend on the sequence of individual moves; positive feedback effects, which may lead to path-dependent properties of social systems; or emulation effects, where individual behavior conforms to or is constrained by the behavior of others as in the case of rationality traps. Finally, there are mechanisms operating between a system and its environment (overlapping and surrounding systems), such as the imitation of technologies or competition among firms. One example of a truly systemist heterodox approach integrating all these relevant processes is the agent-based stock-flow consistent models, which aim to transcend both purely individualist and holist approaches.
This focus on concrete mechanisms in systemist epistemology aims to refine the standard model of scientific explanation, by uncovering the generative mechanisms underlying empirical relationships instead of simply collecting and applying these relationships. To explain the death of a person with their property of being human, and the fact that all humans eventually die, is not very insightful; rather we should try to identify the concrete mechanisms that have led to the state of affairs to eventually arrive at more general theories (Bunge 1997: 425).4
Note that systemism is not a theory, but rather an ontological and epistemological heuristic, like “a viewpoint, or a strategy for designing research projects whose aim is to discover some of the features of systems of a particular kind” (Bunge 2004: 191). Considering this fact and the fundamental aspects of systemist models, we suggest that systemism is a well-suited philosophical framework to structure heterodox theorizing on the micro–macro link as outlined above. Based on these considerations, the next section explores the relation between heterodox economic arguments and the systemic framework.
Bunge’s concept of systemism does not only provide a suitable philosophical framework for heterodox theorizing on the micro–macro link, but also offers an intuitive way to express and conceptualize theoretical considerations on micro–macro interactions. The following examples illustrate this aspect from a practical perspective.
Figure 10.2 Income inequality, labor supply, and economic development in a systemist framework drawing on Bowles & Park (2005) Source: Author’s own illustration, based on Bowles & Park (2005).
The first illustration is provided by Bowles & Park (2005), who use the Veblenian concept of social emulation to explain the allegedly counterintuitive relationship between rising inequality and increasing working hours (Figure 10.2). Due to social emulation of preferences, a higher level of income inequality induces an increase in consumption aspirations across households. In order to live up to these aspirations, a (sizable) subset of these households may increase their working effort, which leads to an increase of average working hours. A possible extension of this argument is that this increase in the supply of labor reduces the bargaining power of workers, leading to lower wages and further increasing income inequality, resulting in a path-dependent downward spiral.
Figure 10.3 The evolution of a traffic convention based on Hodgson & Knudsen (2004) in a systemist framework Source: Author’s own illustration, based on Hodgson & Knudsen (2004).
Our second illustration considers the emergence and evolution of social conventions. Hodgson & Knudsen (2004) discuss an agent-based model where drivers are forced to decide whether to drive on the left or on the right side of a street. They study the conditions required to guarantee the emergence of a stable convention. While their major finding is that habit-formation is a probable vantage point for the emergence of conventions, the illustration in Figure 10.3 extends the underlying argument by illustrating the emergence of conventions in a systemist framework considering both bottom-up and top-down effects.
These examples show that Bunge’s concept of systemism is far away from a methodological straight-jacket. On the contrary, the schematic approach utilized in these examples aims at illustrating how this approach can be employed to facilitate conceptual thinking and the crafting of ontologically sensible theoretical frameworks on the basis of a solid epistemological foundation. It further provides a useful meta-language that enables the effective comparison of different approaches towards the micro–macro link in heterodox economics and to exploit the potential complementarities among these different approaches (see also Dobusch & Kapeller 2012).
The complex relationship between different ontological levels has received considerable attention in heterodox theorizing. This has led to a number of important independent contributions on the role of aggregates and the issue of aggregation in social research, which often allow heterodox economists to circumvent typical fallacies of aggregation identified in the second section of this chapter.
In this chapter we demonstrate that the central pillars of different heterodox conceptions of the micro–macro link are not only complementary, but can also be subsumed under a common philosophical umbrella labeled ‘systemism.’ This umbrella is a useful device helping to explore the commonalities and differences in various approaches to the micro–macro link in heterodox economics. In many cases such comparisons will facilitate ‘heterodox convergence,’ as in the case of evolutionary-institutional and social economics (Elsner 2014), while in other cases the relationship may take the form of productive disagreement. Such an endeavor requires a common denominator for the different approaches enabling mutual communication and comparability. We argue this missing component is exactly what systemism can supply.
We would like to thank Alessandro Caiani, Wolfram Elsner, Volker Gadenne, Iciar Dominguez Lacasa, and the editors for their valuable comments on earlier drafts of this chapter. All remaining errors are our own.
1 The proposition that any aggregate property can be reduced is objectionable on two levels. First, one could argue that such an undertaking is often principally infeasible, because the relevant initial data is never completely known and the associated set of calculations might lack a determinable solution. Sometimes it is even asserted that some specific higher-level activities like human consciousness (Chalmers 2006) or conscious agents (Popper 1979: 292) bear no direct mechanistic relationships with lower-level activities. Second, one could argue that such an undertaking would be practically infeasible since issues become too complicated given the limitation of our current empirical and theoretical knowledge (see Simon 1962).
2 Consider Marx’s claim that “men make their own history, but they do not make it as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given and transmitted from the past” (Marx 1852: 15).
3 One should be more precise than the space of this book chapter allows. See Newman (2003) for an introduction considering most recent advances in the literature.
4 Note that Bunge (1997) uses this argument not only to question the conventional covering law model, but also to criticize other concepts, like hermeneutics or Occam’s razor, which neglect the main task to bring forth new and testable hypotheses.
Albin, P. & Foley, D. 1992. ‘Decentralized, dispersed exchange without an auctioneer.’ Journal of Economic Behavior & Organization, 18 (1): 27–51.
Andersen, P.B., Emmeche, C., Finnemann, N.O., & Christiansen, P.V. (eds.) 2000. Downward Causation. Aarhus, DK: Aarhus University Press.
Anderson, P.W. 1972. ‘More is different.’ Science, 177 (4047): 393–396.
Arthur, B. 1989. ‘Competing technologies, increasing returns, and lock-in by historical events.’ The Economic Journal, 99 (394): 116–131.
Axelrod, R. 1984. The Evolution of Cooperation. New York: Basic Books.
Bourdieu, P. 1998. Practical Reason: On the Theory of Action. Stanford, CA: Stanford University Press.
Bowles, S. & Gintis, H. 1975. Schooling in Capitalist America. New York: Basic Books.
Bowles, S. & Gintis, H. 2011. A Cooperative Species. Princeton, NJ: Princeton University Press.
Bowles, S. & Park, Y. 2005. ‘Emulation, inequality, and work hours: was Thorsten Veblen right?’ The Economic Journal, 115 (507): F397–F412.
Bunge, M. 1996. Finding Philosophy in Social Science. New Haven, CT: Yale University Press.
Bunge, M. 1997. ‘Mechanism and explanation.’ Philosophy of the Social Sciences, 27 (4): 410–465.
Bunge, M. [1998] 1999. Social Science under Debate. Toronto: University of Toronto Press.
Bunge, M. 2004. ‘How does it work: the search for explanatory mechanisms.’ Philosophy of the Social Sciences, 34 (2): 182–210.
Caiani, A., Godin, A., Caverzasi, E., Gallegati, M., Kinsella, S., & Stiglitz, J. 2016. ‘Agent based-stock flow consistent macroeconomics: Towards a benchmark model.’ Journal of Economic Dynamics & Control, 69: 375–408.
Chalmers, D.J. 2006. ‘Strong and weak emergence,’ in: P. Clayton & P. Davies (eds.), The Re-emergence of Emergence. Cambridge, UK: Cambridge University Press, 244–254.
David, P.A. 1985. ‘Clio and the economics of QWERTY.’ American Economic Review, 75 (2): 332–337.
Dobusch, L. & Kapeller, J. 2012. ‘Heterodox united vs. mainstream city? Sketching a framework for interested pluralism in economics.’ Journal of Economic Issues, 46 (4): 1035–1057.
Dobusch, L. & Kapeller, J. 2013. ‘Breaking new paths: theory and method in path dependence research.’ Schmalenbach Business Review, 65 (3): 288–311.
Dopfer, K., Foster, J., & Potts, J. 2004. ‘Micro–meso–macro.’ Journal of Evolutionary Economics, 14 (3): 263–279.
Drèze, J. & Sen, A. 1989. Hunger and Public Action. Oxford: Clarendon Press.
Elsner, W. 2014. ‘Social economics and evolutionary institutionalism today.’ Forum for Social Economics, published online. Available from http://dx.doi.org/10.1080/07360932.2014.964744 [Accessed February 29, 2016]
Elsner, W. & Schwardt, H. 2014. ‘Trust and arena size: expectations, institutions, and general trust, and critical population and group sizes.’ Journal of Institutional Economics, 10 (1): 107–134.
Elsner, W., Heinrich, T., & Schwardt, H. 2015. The Microeconomics of Complex Economies. Amsterdam: Academic Press.
Georgescu-Roegen, N. 1971. The Entropy Law and the Economic Process. Cambridge, MA: Harvard University Press.
Godley, W. & Lavoie, M. 2007. Monetary Economics. New York: Palgrave Macmillan.
Granovetter, M. 1985. ‘Economic action and social structure: the problem of embeddedness.’ American Journal of Sociology, 91 (3): 481–510.
Hirschmann, A.O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge, MA: Harvard University Press.
Hodgson, G.M. & Knudsen, T. 2004. ‘The complex evolution of a simple traffic convention: the functions and implications of habit.’ Journal of Economic Behavior & Organization, 54 (1): 19–47.
Jo, T.-H. 2011. ‘Social provisioning process and socio-economic modeling.’ American Journal of Economics and Sociology, 70 (5): 1094–1116.
Kapeller, J. 2015. ‘Beyond foundations: systemism in economic thinking.’ in: T.-H. Jo & Z. Todorova (eds.), Advancing the Frontiers in Heterodox Economics: Essays in Honor of Frederic S. Lee. London: Routledge, 115–134.
Keynes, J.M. [1936] 2007. The General Theory of Employment, Interest and Money. London: Macmillan.
Keynes, J.M. [1938] 1973. ‘Letter to Roy Harrod,’ in: Collected Works of J.M. Keynes, Volume XIV. London: Macmillan, 295–300.
King, J.E. 2012. The Microfoundations Delusion: Metaphor and Dogma in the History of Macroeconomics. Cheltenham, UK: Edward Elgar.
Kirman, A. 1992. ‘What or whom does the representative individual represent?’ Journal of Economic Perspectives 6(2): 117–136.
Lee, F.S. 2011. ‘Heterodox microeconomics and the foundation of heterodox macroeconomics.’ Economia Informa, 367: 6–20.
Lucas, R.E. 1976. ‘Econometric policy evaluation: a critique.’ Carnegie–Rochester Conference Series on Public Policy, 1 (1): 19–46.
Mandeville, B. [1714] 1962. The Fable of the Bees or Private Vices, Publick Benefits. New York: Capricorn Books.
Marx, K. [1852] 2008. The Eighteenth Brumaire of Louis Napoleon. Rockville, MD: Wildside Press.
Marx, K. & Engels, F. [1845] 2004. ‘Theses on Feuerbach,’ in: The German Ideology. New York: International Publishers, 121–123.
Merton, R.K. 1968. ‘The Matthew effect in science.’ Science, 159 (3810): 56–63.
Mill, J.S. [1848] 1909. Principles of Political Economy with Some of Their Applications to Social Philosophy. London: Longmans, Green and Co.
Minsky, H.P. 1986. Stabilizing an Unstable Economy. New Haven, CT: Yale University Press.
Myrdal, G. 1957. Economic Theory and Underdeveloped Regions. London: Gerald Duckworth.
Nelson, R.R. & Winter, S.G. [1982] 1996. An Evolutionary Theory of Economic Change. Cambridge, MA: Belknap Press.
Newman, M.E.J. 2003. ‘The structure and function of complex networks.’ SIAMReview, 45 (2): 167–256.
Ostrom, E. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge, UK: Cambridge University Press.
Polanyi, K. [1944] 2001. The Great Transformation: The Political and Economic Origins of Our Time. Beacon Press.
Popper, K.R. 1979. Objective Knowledge: An Evolutionary Approach. Oxford: Clarendon Press.
Robbins, L. 1932. An Essay on the Nature and Significance of Economic Science. London: Macmillan.
Rothschild, K.W. 1971. Power in Economics. Harmondsworth, UK: Penguin.
Rothschild, K.W. 1988. ‘Micro-foundations, ad hocery, and Keynesian theory.’ Atlantic Economic Journal, 16 (2): 12–21.
Schelling, T.C. 1978. Micromotives and Macrobehavior. New York: Norton and Company.
Schumpeter, J.A. [1934] 2011. The Theory of Economic Development, 2nd edn. New Brunswick: Transaction Publishers.
Schumpeter, J.A. [1942] 1975. Capitalism, Socialism and Democracy. New York: Harper and Row.
Simon, H.A. 1962. ‘The architecture of complexity.’ Proceedings of the American Philosophical Society, 106 (6): 467–482.
Sydow, J., Schreyögg, G., & Koch, J. 2009. ‘Organizational path dependence: opening the black box.’ Academy of Management Review, 34 (4): 689–709.
Witt, U. 2001. ‘Learning to consume—a theory of wants and the growth of demand.’ Journal of Evolutionary Economics, 11 (1): 23–36.
Veblen, T.B. [1899] 1979. The Theory of the Leisure Class. London: Allen & Unwin.