Aftalion’s Multi-layered Structural Analysis

Aftalion’s analysis of structural transformation is closely related to his approach to identifying relevant units of analysis in the economic system, and of identifying policy measures consistent with the specific causal structure of transformation relevant to any given historical context. His work on periodic overproduction crises is clear evidence of a research methodology that avoids concentration of attention on aggregate magnitudes or atomistic units by focusing on industrial sectors and intersectoral relationships.Footnote ¹⁸ Intermediate levels of aggregation are central to Aftalion’s heuristic and characterize his approach to economic dynamics. The movement of any economic system, through time is described in terms of dynamics involving a large but finite number of sectors to be identified by certain common features that keep them distinct from other sectors in the economy.

This approach leads to what Jean Lhomme, in his essay on Aftalion’s heuristic, describes as an attempt to formulate “a positive theory” through “successive approximations”: “three tentative explanations are attempted […], of which each one will bring theory increasingly close to reality.”Footnote ¹⁹ At the first stage of approximation, Aftalion’s overproduction theory of economic crises can explain “what takes place at the two poles of each periodic cycle, at the point of crisis and at the beginning of expansion”—that is, the fact that “prices fall as a result of excessive production. They rise as a result of insufficient production.”Footnote ²⁰ However, overproduction theory in itself does not explain an important characterizing feature of modern periodic crises—that is, “the duration of [price rises and falls], the duration of the phases of the [economic] rhythm.”Footnote ²¹ For lack of sufficient production cannot account for the persistence of the price rise “during the three or four years of prosperity,” as it would be unreasonable “to believe in continuing underproduction while production itself is continually increased.”Footnote ²² For this phenomenon to be explained, one must go beyond the dynamics of general production and focus on the intensity and sequencing of fluctuations in particular industries (the second stage of approximation). In Aftalion’s view, moving to this new set of phenomena allows a refocusing of the analysis on “special fluctuations” whose relevance is missed at the macroeconomic level and that could be important in providing an explanation for the duration of cycles and their internal structure.Footnote ²³ The switch from a macroeconomic approach (fluctuations of general production) to a sectoral approach (special fluctuations in particular industries) is an essential step in explaining the structural transformation of the economic system during business cycles:

Yet, the special fluctuations in fixed capital industries does not provide a comprehensive explanation for the intensity of business fluctuations in the economic system as a whole. To achieve that, the analysis of cycle dynamics must switch back to the consideration of a macroeconomic adjustment mechanism (the third stage of approximation). In this case, it is no longer fluctuations of production levels that are brought into focus, but fluctuations of aggregate demand working through changes in the absolute and relative magnitudes of different types of income (wages, profits, rents): “[e]ach fall in the price of a commodity brings about a fall in the incomes and thus in the demand of those who are employed in its production. It is thus a factor determining the fall in prices of other commodities [… ] Correspondingly, each price rise is a factor contributing to the possibility of a general price rise.”Footnote ²⁵ In short, each stage of approximation is relevant not only as a step in the process of theory construction, but also as a device essential to understanding a specific layer of the causal process under consideration. Different phenomena can be explained depending on which causal layer is considered. Aftalion practices a heuristic that moves back and forth between macroeconomic and sectoral levels of aggregation by identifying causal factors that can only be singled out by focussing on a specific aggregation level but ultimately produce a cumulative effect. This plural heuristic is a distinctive feature of Aftalion’s research strategy and finds applications in a variety of fields, and today may be applicable to industrial and infrastructural policy to support decarbonization, which are considered, respectively, in the fourth and fifth sections below.

Aftalion’s Theory and the Structural Business Cycles Tradition

Aftalion’s theory belongs to the structural business cycles tradition of Mikhaylo Ivanovich Tugan-Baranovsky, Arthur Spiethoff, Mentor Bouniatian, and Dennis Holme Robertson.Footnote ²⁶ However, his work differs significantly from the other contributions in that tradition. The key difference is that these contributions concentrate on characterizing and sequencing interactions between disaggregate production structures and macroeconomic fluctuations without explicitly theorizing dynamic structural interactions that either preserve or transform the industrial structures so described. The economists mentioned above are principally concerned with measuring these interactions, but generally offer no theoretical approach that can be mobilized to understand or manage structural transformation across the multiple causal layers characterizing each industrial structure.

Aftalion’s distinctive features within the structural business cycles tradition are (i) the consideration of the different layers of interdependence within each network of productive activities, (ii) the consideration of the relationship between layers of interdependence and the timing of transformation (so that transformations at a given layer of interdependence may presuppose, or be conducive to, transformation at a different layer), and (iii) a policy framework that brings into focus the relationship between systemic policy objectives and the differentiated dynamics of individual industrial sectors. Aftalion’s relationship to the structural business cycles literature should be understood by paying close attention to his sophisticated approach to causality in economics (sequential and multilayered). This approach also situates Aftalion’s contribution to Accelerator Theory within the multilayered and historical framework of his overall theoretical conception, avoiding the current consensus narrative, which places Aftalion in a lineage that sees the “Accelerator” as the reciprocal (and thus a supply-side version) of the “Multiplier” found in Keynes and the Cambridge Keynesian Tradition.Footnote ²⁷ Both multi-layered causality and Aftalion’s analysis of the relationship between investment and aggregate demand are characterizing features of a theory that is “capable of addressing the macroeconomic dimension of fluctuations and crises without losing track of the sectoral differences and of the propagation mechanism of uneven dynamics from one sector to another.”Footnote ²⁸

In short, and differently from other contributions to the structural business cycles literature, Aftalion’s approach involves a flexible utilization of fundamental units of analysis, which may be different depending on the problem at hand. It also involves a flexible approach to levels of causality and types of explanation. Industrial sectors may empirically be identified in a plurality of ways, and this may in turn lead to a plurality of intersecting and not mutually exclusive explanations of dynamic processes. For example, some industrial sectors, such as steelmaking, might be characterized by their supply chains, which include backward linkages to iron ore extraction and energy production, as well as by forward linkages created by the use of steel in the fabrication of automobiles or alternatively in industrial equipment, which is an input into other manufacturing sectors. They are thus vulnerable both to supply shocks from backward linkages and demand shocks from forward linkages. Therefore, alternative sectoral specifications are likely be associated with different patterns of interdependence between sectors and different transmission mechanisms of economic triggers or policy decisions throughout the economic system. Aftalion’s acknowledgement of multiple, and not mutually exclusive, ways of identifying the relevant sectors of the economy can be closely associated with his interest in economic structures as flexible patterns of interdependencies between sectors. In Aftalion’s framework, the choice of units may determine the relevant pattern of interdependence, so that multiple and partially overlapping causal layers can be detected depending on which units are chosen. The implications of this point of view are far reaching, as any given economic system would be conducive to manifold specifications of relevant units (manifold specifications of its structure). Plurality in causality is inherent to Aftalion’s structural analysis and makes his approach to dynamics especially interested in the specific overlap of causal layers that characterizes each historical context.

Monetary Policy in a Multilayered Causal Framework

Aftalion’s analysis of monetary policy and criticism of central bank monetarism is a response to the monetary disturbances of the 1920s and 1930 and addresses those disturbances in terms of the theoretical framework he had developed when investigating the recurrent crises of an industrial economy.Footnote ²⁹ In Aftalion’s view, monetary policy can produce effects through multiple transmission channels and the structure of any given economic system determines which transmission channel would be the most influential under specific conditions. This makes it extremely difficult to undertake “a general action on prices through credit policy.”Footnote ³⁰ One important reason for this difficulty is the time-differentiated responses of different industries to the phases of the business cycle:

In his account, different sectors of the economy are likely to respond in different ways, or at least with different time lags, to any given macroeconomic intervention. However, there is broad coincidence of business cycle phases across different industries, due to “the participation one is generally observing in practice of a great number of industries to the alternating cyclical movements, so that we may refer to periods of general expansion or contraction.”Footnote ³² This situation suggests that macroeconomic interventions should not be excluded, while highlighting the need of a prudential approach to monetary and credit policy and to macroeconomic policy in general. Macro policies should be sensitive to context and timing for they may otherwise backfire and lead to outcomes opposite to the wanted ones. The early phases of expansion are especially vulnerable in this respect, seeing that, at that time, the general tendency is not yet fully established, and signs of monetary restriction may thwart the upward movement of the economy.Footnote ³³

Monetary and credit policy work through distinct transmission channels due to the heterogeneity of industrial sectors, and primarily to the lack of synchronization in their responses to dynamic impulses. For this reason, Aftalion envisages substituting the “new banking policy” (nouvelle politique bancaire), which is what became known as central bank monetarism, with a banking policy which he describes as “subtle and nuanced” (subtile et nuancée). The aim of this banking policy should be to provide specialized credit—that is, to provide credit facilities “by following the special conditions of each one of the industries under consideration.”Footnote ³⁴ Provision of liquidity would in this case be adapted to the differentiated liquidity needs of different sectors, and macroeconomic policy would aim at coordinating, through liquidity provision, the heterogeneous dynamics of industrial sectors in the business cycle. This open-ended causal mechanism is at the core of Aftalion’s criticism of the “new banking policy,” which demands from central banks that they “manage the economic activity of a country” so as to “‘direct the money’ in such a way as to achieve the stability of its purchasing power, the stability of the general price level.”Footnote ³⁵ This policy makes the general price level into a measure of the purchasing power of money independently of the absolute and relative movements of individual prices:

To conclude, Aftalion’s discussion of monetary policy is consistent with his emphasis on structural transformation as a fundamental constitutive feature of capitalism. His critique of the aggregate price level as a meaningful policy target derives from his view that monetary and credit policy should reflect actual and/or desired patterns of transformation within the network of interrelated activities rather than mirroring the dynamics of macroeconomic variables whose economic meaning can be entirely different depending on the direction followed by changes in the internal composition of aggregates. Aftalion’s consideration of the links between structural transformation, layers of causation, and the means and ends of economic policy provides a unique economic policy framework in times when structural transformation objectives (such as decarbonization) require means capable of triggering and sustaining a pervasive transformation of the structures of interrelatedness in the economy.

Structural Transformations in the Era of Decarbonization

In the third decade of the twenty-first century, there is an emerging consensus that the scale of resource consumption which enabled global economic growth in the post-war period and the associated negative environmental externalities (chiefly but not exclusively greenhouse gas emissions, air pollution, and biodiversity loss) threatens planetary boundaries. As James K. Galbraith, among others, has recently observed, classical political economy was well aware of the natural limits of the productive capacities of the primary sector; others have argued for decades for an incorporation of resource economics into multisectoral models, largely to no avail.Footnote ³⁷

Neoclassical climate change economics, as established by 2018 Nobel Laureate William Nordhaus, represents both a significant achievement and a flawed attempt to understand and model the interactions among the macro-economy, the energy system, and the climate system. In his Nobel lecture and in subsequent publications, Nordhaus argued that a global temperature rise of 4°C would be optimal, balancing equally the trade-offs between the costs of climate change mitigation and those of adaptation, even as the 2016 Paris Agreement had committed to limiting global median temperature rise to 1.5°C.Footnote ³⁸ Nordhaus’ approach, classified as an aggregate cost-benefit model, posits that climate change would only directly affect certain sectors of the economy, and that the macro-economy as a whole would not be affected by the propagation of intersectoral shocks, even if it would be by aggregate supply and demand shocks. Nordhaus’ original Dynamically Integrated Climate-Economic (DICE) model was consistent with the neoclassical consensus and, therefore, amenable to adoption as the underpinning of global climate governance via integrated assessment models (IAM) that incorporate both natural and social sciences.Footnote ³⁹ Although Regionally-Integrated (RICE) models with their regional extensions improved upon DICE models by consideration of spatial interactions, they still have no place for structural dynamics at the sectoral level.Footnote ⁴⁰ This is especially alarming as integrated assessment modeling consists not only of the damage models discussed above, which explore aggregate cost-benefits, but also IAM employs process-based models to establish optimal decarbonization pathways.Footnote ⁴¹ The absence of intersectoral interdependencies in damage models is alarming, but to omit them from the consideration of climate change mitigation scenarios and associated shared socioeconomic pathways is troubling as it assumes structural economic change to be exogenous rather than an outcome of deep decarbonization itself.

Subsequent academic literature has confirmed that Nordhaus’ approach has led to significant underestimation of the economic costs of adaptation to climate change, chronically unreliable carbon pricing, and increasingly doubtful estimates of the wider societal costs involved in missing key climate targets.Footnote ⁴² Yet it is not immediately evident how to improve these models within the existing framework, and the 2020 rework of the damage models performed by the Potsdam Institute for Climate Impact Research (PIK), although better grounded in climate science and with more realistic social discounting, made no fundamental changes to the economic modelling approach.Footnote ⁴³ Grubb et al. are concerned to improve these models by introducing dynamic aspects, and they identify “induced innovation” (encompassing endogenous innovation between high and low carbon technologies, economies of scale, and knowledge diffusion), frictions (“inertia”), and path dependencies as the “few, key dynamic assumptions that affect optimal abatement.”Footnote ⁴⁴ Although they construct their “dynamic features of emitting systems” somewhat narrowly and maintain an aggregate approach, this represents a significant step forward and appears more tractable than agent-based client models developed by some evolutionary economists, which are nevertheless helpful in establishing the limitations of market-based solutions (carbon taxes, green subsidies) in inducing green transitions and thereby in providing support for the role of industrial policy.Footnote ⁴⁵ What evolutionary approaches cannot offer is a heuristic for understanding the mutual dependencies, indirect effects, and feedbacks associated with industrial policy that is designed to accelerate specific sectoral decarbonization pathways.

While formal approaches to intersectoral modeling of climate change risks are rare in the literature, initiatives such as the Intersectoral Model Impact Comparison Project (ISIMIP) were promising in their attempts to model the consequences of intersectoral interdependencies for magnifying or mitigating climate damage.Footnote ⁴⁶ Unfortunately, to the extent this approach has achieved traction at all, it mainly used to model trade-offs in the primary sector (agriculture, land use), where data quality is higher due to the near-universal presence of agricultural subsidy regimes. These models are not more widely employed, although the earth-systems science literature notes ISIMIP as one approach that respects the interdependencies of climate risks.Footnote ⁴⁷ Perhaps not surprisingly, the greatest support for these approaches comes from climate scientists, to whom the logic of material interdependencies is obvious, and the greatest resistance comes from adherents to Nordhausian climate change economics and its assumptions that market-based solutions are axiomatically sufficient. As resources like the Carbon Emissions Accounts & Datasets (CEADS) develop and become more widely used in the estimation of optimal sectoral decarbonization pathways, revisiting intersectoral interdependencies between them will be possible, thereby providing policy support for coordination of sectorally-targeted industrial policy.Footnote ⁴⁸

A more intersectorally nuanced approach to climate modeling would also yield better monetary policy recommendations for the decarbonization era. Quantitative easing (QE) has been the standard policy approach to macro management since the 2007 subprime crisis, first in the US and the UK, then also in the Eurozone, culminating in the role that QE played in managing the exogenous shock of the COVID-19 pandemic. This approach has entailed exclusive reliance on monetary instruments in a macroeconomic framework. Recent debates have highlighted the dangers of relying exclusively on quantitative easing through aggregate monetary policies; but concerns that QE could produce inflationary pressures have only recently received empirical support and have yet to be demonstrated as a global phenomenon.Footnote ⁴⁹

Meanwhile other voices have questioned the climate impacts of recent QE, highlighting the extent to which such interventions retard decarbonization efforts, by prolonging existing economic structures and effectively subsidizing older industrial sectors in order to maintain full employment.Footnote ⁵⁰ Proponents of the view that classic QE is unhelpful to decarbonization have instead promoted “green” quantitative easing, as if establishing the principle that central banks should preferentially buy green bonds would be sufficient to ensure the effectiveness of green QE as long as there is adequate transparency and accountability.Footnote ⁵¹ This is specialized credit policy through the backdoor, but falls short of what is required to effect deep decarbonization, as it merely classifies industries as “green” or “brown” and does not consider how to manage the intersectoral dynamics required to achieve the traverse, let alone the intersectoral trade-offs, of having done so. A reconsideration of Albert Aftalion’s intellectual project suggests that more rigorous approaches to harnessing a specialized credit policy to support multi-sectoral decarbonization are possible, but the implementation of such a regime requires central banks to take seriously the intersectoral interdependencies of embedded carbon flows in the economy, not only in the energy and transport sectors, but also in the built environment and in agriculture and mineral extraction. Recent methodological innovations in environmentally-extended input-output analysis have established that tracking these flows is possible, but time lags in the production of reliable data are a major impediment to making such analysis routine.Footnote ⁵² Rebuilding and expanding the capacities of state and other public actors to make evidence-based policy at the sectoral level is urgently needed to deliver deep decarbonization, in terms of guiding both specialized credit policy by central banks and coordinated sectoral support in infrastructural and industrial policy. This would also be the fullest realization of Aftalion’s intellectual project.

A Political Economy of Transformations in Capitalism

In addition to its relevance for understanding a new growth regime inaugurated by a global push for decarbonization, Aftalion’s multi-layered analysis of industrial dynamics has significant implications for the morphology of capitalism under structural transformation. In Aftalion’s view, industrial interdependencies generate conditions for evolutionary trajectories that must be satisfied independently of the institutional context. In other words, institutions and policies must be embedded in a wider framework of which material structures are a fundamental constitutive element.Footnote ⁵³ This approach finds a remarkable expression in Aftalion’s belief that the alternating phases of expansion and contraction characterizing capitalist institutions are a direct consequence of the technical conditions of production technology, and would therefore be effective also in a different institutional context (as in a socialist economy): “[m]y principal thesis is that the chief responsibility for cyclical fluctuations should be assigned to one of the characteristics of modern industrial techniques, namely, the long period required for the production of fixed capital.”Footnote ⁵⁴ Conversely, Aftalion’s approach lends itself to the view that, independently of the institutional context in which transformations take place, certain conditions regarding the interrelatedness and time sequencing of structural change need to be satisfied for any transformational trajectory to be successful. Aftalion’s conceptualization also helps explain why economic development cannot be reduced to a matter of setting the institutional framework and expecting structural transformation to flow from it.

There are several implications of Aftalion’s line of reasoning. First, it is possible to read Aftalion as pre-figuring Luigi Pasinetti’s “Separation Theorem,” distinguishing “natural” from “institutional” levels of analysis, as there is a sense that incomes arising from capital, land, and interest would exist under socialism, because they are “natural” (that is, technologically determined) rather than the result of a particular property regime.Footnote ⁵⁵ Secondly, it becomes necessary to disentangle the “natural” from the institutional aspects of transitional paths. In this light, Aftalion distinguishes between the “natural” (that is, structural) and the institutional meanings of distributional categories such as profits, rents, and forms of interest:

The above argument leads Aftalion to distinguish between the “natural” profits, interests, and rents arising from the objective (mainly technological) conditions of production and the actual distribution of those incomes to particular individuals or social groups according to the existing institutional set up.Footnote ⁵⁷ This point of view calls attention to the distributional dynamics associated with the transformation of economic structures; and it suggests the need for distinguishing between the distributional changes required by the structural transformation and those changes that reflect existing institutional arrangements and the relative influence of social groups, and which are not necessarily aligned to the former.Footnote ⁵⁸ This provides an alternative view to the one implicit in Peter Hall’s formulation of growth regimes, which takes an institutionalist approach to both material and social realities. Given the material constraints at the base of the climate crisis, Aftalion’s approach offers a particularly helpful heuristic for understanding the challenges and opportunities associated with the era of decarbonization. Lastly, Aftalion’s approach suggests that economic policies successful in driving a structural transformation as a systemic policy objective presuppose the identification of the relevant units of analysis, patterns of interdependence, and patterns of sequential causality required for the transformation to take place. This framework may allow several degrees of freedom to policy actions, which may accordingly adjust to specific institutional and political contexts. However, no transformation objective, including that of deep decarbonization, would be feasible unless its structural prerequisites are first identified and met.Footnote ⁵⁹ Aftalion’s contribution provides essential building blocks to the construction of a historical and structural political economy rooted in the multi-layered structures and multiple time horizons involved in structural dynamics.