Unmixing the Metaphors of Austrian Capital Theory
Money & Macro2

Unmixing the Metaphors of Austrian Capital Theory

The distinctiveness of Austrian capital theory (ACT) against competing theories of capital lies in (1) its process approach over time as opposed to a comparative statics approach, (2) its emphasis on relative prices as opposed to the general level of prices, and (3) its recognition of the importance of the heterogeneity of physical capital. On the strength of these analytical convictions, ACT has persisted into the 21st century as a robust, though marginal, perspective in macroeconomics.

Unfortunately, progress in the decades since its divergence from neoclassical capital theory has been sporadic at best. Severe conceptual problems have been masked by a sense that the pure axiomatic theory of capital is essentially complete, and that its conclusions are invariant the particular metaphors and paradigms used for capital. In fact much of ACT is not pure theory, and several important conclusions hinge on precisely on the simplifying metaphors used to make it operationalizable. In particular, the expositional choice to build up from a Crusoe economy commits authors – possibly inadvertently – to a point-output model of circulating capital that cannot account for important features of actual industrial processes. Though some authors in the Austrian tradition have made important advances since Hayek (1931), no subsequent work gained widespread consensus as a canonical statement until Garrison (2001), essentially a formalization of Hayek (1931). Even those authors making progress in the intervening period have been on the whole conservative, preferring to leave the edifice intact despite significant revisions to the foundation.

This paper argues that further progress will consist in developing a flow-output model of capital, where investment plans terminate in a production process rather than a consumption good. This approach is implicit in work by Lachmann, Lewin, and Cachanosky, particularly in the latter’s papers on financial duration. Nevertheless, for the most part they still lack a self-consciousness about their distinctiveness from the Hayek-Garrison point-output model, which remains standard in Austrian pedagogy. The attempt to maintain backward compatibility with the constructs used in point-output models, while understandable, retards further progress.

The following section surveys the conceptual problems that have arisen with ACT over the past century, and especially with the Austrian business cycle theory (ABCT) that largely motivates it, along with the various attempts to deal with these problems along the way. Most of these problems arise from capital theory’s pre-marginalist Ricardian heritage and the complementary decision to build up from an isolated Crusoe economy. The paper then argues that dealing with these problems has major implications for the conclusions of ABCT. In particular, if we are to preserve the analytical distinctives of ACT in light of these conceptual problems, it will be necessary to discard many of the important conclusions of ABCT, particularly the relevance of malinvestments as an explanation for business cycles and the necessary connection between boom and bust.

A Brief History and Critique of the Canonical Austrian Theory

Austrian capital theory has inherited a complement of metaphors from its origins in the work of Ricardo and Böhm-Bawerk. Its further development through Hayek and up to the present day reflect growing awareness of tensions and inconsistencies in these metaphors. The leading lights of Austrian capital theory in the 20th century – Hayek (1941), Mises (1966), Lachmann (1956), and Kirzner (1966) – have sought solutions in the more rigorous application of the subjective method, that is, by attempting to expunge metaphorical elements in favor of axiomatically valid pure theory. Of these four however, as we will see, only Lachmann has managed to break free from a point-output model, and therefore from the complement of Ricardian metaphors.

Capital in its pure theory sense refers to the productive aspect of any property claim; that portion of its value that reflects its usefulness for future consumption as distinct from immediate consumption (Lachmann 1956: xv). It is distinct from the concept of the capital good, which refers to a good whose value is derived entirely (or at least predominantly) from capital value. Provided we remain at this level of abstraction, anything we say about capital will be valid for any form of capital, whether a durable good, a good-in-progress, land, human capital, financial capital, or anything else. 1

In practice however, the capital value and the consumption value of a good are not disentaglable. To tell a relevant and operationalizable causal story requires a mid-level theory shaped by the choice of a paradigmatic capital good. This is the type of concrete good we have in mind when telling a causal story, and to which the capital aspect of all other goods can (in principle) be assimilated by analogy. Such a theory, though necessary to refer the results of capital theory to the real world, will not be axiomatically valid, and the attempted retreat to pure theory will not suffice to solve problems inherent in the application of mid-level metaphors.

Both Mises (1966) and Kirzner (1966), following Böhm-Bawerk ([1891] 1930), build up a theory of capital from a Crusoe economy. In the hands of Austrian writers, this device focuses the analysis on the pure logic of choice; to show to what extent the problems of economic interrelation can be reduced to problems inherent in single acts of choice. In the case of capital theory, the argument runs something like this: Crusoe faces a choice between a hand-to-mouth existence gathering berries, or devoting some time up front to tilling a plot of land and cultivating crops. In order to do so, however, he must determine how much time he is able to spare from direct production without starving.

Already in this stylized example we can see the main elements of classical capital theory: the increased productivity of indirect (roundabout) production, the necessity of foregoing current consumption (saving) in order to increase future consumption (investing), the fact that Crusoe must still eat while waiting for his crops to grow (the subsistence fund), and the plan as encompassing the production of a specific good (a point-output model).

Though Austrian writers have largely been aware of particular problems with generalizing these elements to an exchange economy, there has been a nearly universal faith that generalizing in this manner is possible with the scrupulous application of the subjective method. In fact, whatever problems the subjective method did solve, it could not touch other fundamental problems in the metaphors borrowed from a Crusoe economy. In order to see the relevance of these metaphors and paradigms, it will be necessary to rehearse some well-trod critical ground before offering a constructive solution.

Roundaboutness and the Circulating Capital Metaphor

Compare the following two framings of Crusoe’s capital problem:

  1. Crusoe desires to produce a meal. He can produce it directly by gathering, or spend some time on a longer production process in order to get a bigger or better meal.
  2. Crusoe desires a stream of food. He can secure one stream of food by gathering, or spend some time up front in order to secure a larger stream over time.

Both of these parables follow Kirzner’s (1966) subjectivist emphasis on planmaking. Indeed, Kirzner spends his entire analysis building on case (a) above. The crucial question, however, is what we regard the plan as encompassing. Do Crusoe’s plans terminate in the production of a good, with each new good constituting a new plan, or do they terminate in the establishment of a production process, the stream of production being considered merely as the continuation of a plan? This we will call the distinction between a point output (a) and a flow output (b) model of capital.

Though there are some cases where either interpretation will do, it is not quite an arbitrary choice. If we imagine a production process to be composed of (1) up-front costs, (2) maintenance costs over time, and (3) a flow of output over time, a point-output model is a sensible depiction of a process where maintenance costs are primary and up-front costs are negligible. Each repetition of the process can be considered as a plan in isolation. A flow-output model, on the other hand, is necessary where up-front costs dominate maintenance costs, especially where maintenance costs can be regarded as routine, and therefore part of the initial plan. With an up-front investment in something like a durable tool, each repetition of the productive process cannot be considered in isolation, and can only intelligibly be considered as a continuation of the overarching plan for which the initial investment was made.

These two models are each associated with a different paradigmatic capital good. In a point-output model, Crusoe’s capital (remembering that capital is defined as the potential for future consumption) consists in partly-finished goods. This was the basis for Böhm-Bawerk’s ([1891] 1930) capital theory, with growing trees or aging wine as the paradigmatic capital good. After a certain amount of waiting, capital goods mature into consumption goods of varying quality. In this model there is no up-front investment (or it is ignored); production is a succession of (possibly overlapping) input→waiting→output sequences. Hayek’s (1931) model added some sophistication to Böhm-Bawerk’s by accounting for a flow input, allowing for more complicated production processes where capital goods pass through multiple ‘stages of production’. Nevertheless, for Hayek, as for Böhm-Bawerk, a capital good was still fundamentally understood as an unfinished consumption good.

In a flow output model, on the other hand, the paradigmatic capital good is a tool, or a durable capital good, something distinct from the consumption good, whose planned use cannot be separated from the entire flow of output it is intended to generate, net of maintenance costs.

Laying it out this way makes it clear that point-output processes can be assimilated relatively easily to flow-output models, provided the process is repeated, but flow-output processes cannot generally be assimilated to a point-output model. Nevertheless, most authors have taken it quite for granted that anything said in terms of one can be said equally well in terms of the other. Bruno et al. (1966) argued that “there is no essential difference between the circulating-capital and [durable]-capital models.” Hayek in numerous places (1929; 1931: Preface; 1932b: §XII) insisted that the issues in circulating and durable capital models were, aside from some expository differences, identical in all essential economic respects, and even endeavored to “abstract from the existence of durable goods” (1931: ch. 2). In order to integrate with Ricardian static theory, therefore, which dealt in stocks rather than flows, early expositors of capital theory chose to assimilate durable capital processes to a circulating capital model. Subsequent authors have largely maintained this choice, even while recognizing the insufficiency of the Ricardian apparatus that motivated it in the first place.

In order to assimilate durable capital into a circulating capital model, capital would have to be regarded as in some sense an unfinished consumption good rather than a good in its own right. For Ricardo, and continuing well into the Marginalist era, this was done by considering capital to represent embodied labor time – or, more generally, “labor, nature, and time stored up” (Mises 1966: 493). This is the crucial simplifying assumption behind Ricardian economics, and the reason it sits so uncomfortably beside the Marginalist elements grafted on by Böhm-Bawerk and later Austrians.2 Durable capital, therefore, was to be thought of as the indirect application of original factors over an increased number of stages. Per Ricardian analysis, any increases in real income would redound up the supply chain and accrue only to the owners of those original factors applied at each stage of production (Mises 1966: 493; Hayek 1931).

The decision to take circulating capital rather than durable capital as the paradigmatic capital good, or (what is the same thing) the decision to employ a point-output model rather than a flow-output model, has resulted in numerous dead ends since the time of Böhm-Bawerk. The most significant was the ‘average period of production’ concept, intended to measure the average period of time original factors were ‘tied up’ in circulating capital goods before being ‘released’ in the form of consumption goods. More capitalistic production embodied more of the original factors, and thus more cumulative waiting between the start of a production process and the end.

Among Austrian writers, this collapsing and homogenization coexisted for a long time with an explicit emphasis on the heterogeneity of capital, with the latter serving to obscure the former. As Lachmann (1956: 73) argued,

On the one hand, no other economist saw more clearly than [Böhm-Bawerk] the essential heterogeneity of all capital… On the other hand, his theory is essentially an attempt to reduce this ‘complex’ to a common denominator and to measure all changes in it in the single dimension of time… [H]e failed when he tried to introduce the incongruous element of single-dimension measurement into a theory conceived in terms of heterogeneous products.

Later Austrian writers did become aware of particular difficulties in Böhm-Bawerk’s exposition. The average period of production, though used uncritically in Hayek (1929), was later called a “meaningless abstraction” in Hayek (1936) and an “empty concept” in Mises (1966: 489) – and this was the consensus among those sympathetic to Böhm-Bawerk’s project.

Nevertheless, despite the wide acknowledgement of Böhm-Bawerk’s error in this narrow respect, it was not until Hayek (1941: vi) that a major Austrian capital theorist acknowledged that the conceptual problems ran much deeper, and even there they were not solved. The basic problem was that the circulating capital metaphor, passing from Ricardo through Böhm-Bawerk and Hayek, depended implicitly on a backward-looking labor theory of value to homogenize capital and consumption goods, whether the resulting aggregate was understood as a unit of time, labor time, or capital more generically – a point noted as early as Fetter (1902) but not widely internalized until Sraffa (1960), when the roundaboutness metaphor lost currency among non-Austrians for other reasons. The tools developed with the circulating capital metaphor in mind, most obviously the average period of production but also more abstract concepts like ‘roundaboutness’, could not, in the end, be rescued.

Monetary versus Physical Units on Capital

If limitations of point-output models rendered the roundaboutness concept nonoperational, issues raised during the Cambridge Capital Controversy made it clear that capital intensity, even divorced from a notion of roundaboutness, was not conceptually viable. Capital theory could not, therefore, simply replace roundaboutness with capital intensity, as Hayek tried;3 a more fundamental reconceptualization would be necessary.

Sraffa (1960, ch. 12) showed that capital intensity could not be understood as a one-dimensional scalar. In particular, if the flow of inputs into the production of a good was not constant over time as Hayek (1931) assumed, the optimal technique might vary in unpredictable ways with the interest rate. If two production techniques could not be unambiguously ordered as to capital intensity, the simple Ricardian story, where saving and investment work through capital to determine long run income, could not be applied to actual industrial processes. Samuelson (1966) summarizes the issue:

The phenomenon of switching back at a very low interest rate to a set of techniques that had seemed viable only at a very high interest rate involves more than esoteric technicalities. It shows that the simple tale told by Jevons, Böhm-Bawerk, Wicksell, and other neoclassical writers – alleging that, as the interest rate falls in consequence of abstention from present consumption in favor of future, technology must become in some sense more “roundabout,” more “mechanized,” and “more productive” – cannot be universally valid.

Samuelson goes on to provide arithmetical examples of reswitching for both circulating and durable capital models. Even before Sraffa’s proof, however, Lachmann (1956: xiv) could claim that “most economists agree today that except under equilibrium conditions, a ‘quantity of capital’ is not a meaningful concept.” Though the main target of the reswitching debate was neoclassical capital theory, the Austrian theory – now bereft of both the roundaboutness metaphor and the capital intensity concept – was an important casualty,4 and no doubt this gave an air of obsolescence to ACT.

The issue, again, is one of aggregation. Following the classical economists, capital theory was one of the last holdouts in continuing to insist on analysis in physical terms. In this case, the purpose of the insistence was that aggregating capital in value terms makes the capital stock a function of the interest rate. In Ricardian analysis, however – formalized as the Cobb-Douglas production function of neoclassical theory – diminishing returns to capital in physical terms pin down the interest rate at the productivity of the marginal piece. To abandon physical aggregation of capital was to admit that the quantity of capital was a function of the interest rate, and not vice versa, and therefore to abandon the Ricardian project of income determination by ownership class.5 Though conservatism had carried Ricardo through the middle of the century, if capital could only be meaningfully aggregated in value terms, the edifice would have to – at last – be abandoned totally.

Toward a Coherent Austrian Capital Theory

Even if no particular reconstruction of the Austrian theory has yet surpassed the Hayek-Garrison exposition, a number of authors have made important steps toward a more analytically satisfactory ACT, even if we still lack both a full statement of the amended theory and a reevaluation of its original claims. Though the former would be a book-length task, this section takes some steps toward the latter by delineating a distinct Lachmann-Lewin-Cachanosky model from the standard Hayek-Garrison model.

The Duration Concept

One important corrective has been developed in Lewin & Cachanosky (2015, 2018) and Cachanosky & Lewin (2016, 2018), who build on Hicks (1939; 1973) to provide an alternative concept to replace the average period of production and its long succession of abortive replacements: namely, financial duration.

Duration, in its modified or Macaulay form, represents the semi-elasticity of the capitalized value of a stream of income with respect to the interest rate (see Cachanosky and Lewin 2016 for a derivation). In this respect it captures much of what the ‘average period of production’ concept was meant to: industries with a longer duration will be more sensitive to changes in interest rates, and therefore more prone to malinvestment under the guidance of misleading intertemporal price signals. The important difference is that the quantity responding to changes in the interest rate is not a (physical) quantity or aggregate of capital, whether measured in time or labor time, but capital value. As Lewin and Cachanosky (2018) point out, by making “capital intensity” a function of the interest rate rather than vice versa, the Ricardian project of determining the incomes of various classes of owners is rendered defunct, precisely the break necessary to put ACT on firmer footing.

Lewin and Cachanosky trace through several implications of the duration concept for ACT. In the first place, the time structure of production as a series of stages is an untenable holdover of the circulating capital metaphor. Contra Mises’ (1966: 333) attempt, Lewin (1999: 64) sums up the conceptual problems with arranging capital goods by order:

The formula is crucially dependent on being able to identify the stages of production. It is assumed that the process begins at stage 1 and ends at stage n. In this way any kind of “looping” (coal is used in the production of iron and vice versa), where the output of one stage becomes available as an input of an earlier stage, is ruled out. Second, if the output is a flow (as it usually is), then we must also have some way to connect inputs that occur at time periods n–t with precisely that output that arrives at time period n and separate them from those that need to be connected to outputs occurring at time periods n+j where j is an index of time periods occurring after n. In other words, if the production process is a flow input—flow output process, a set of inputs are used to produce jointly a set of outputs occurring over time and the measuring of T becomes more problematic. Similarly, we must be able to identify the amount of labor time l that is used. This obviously presumes that it is possible to reduce any labor heterogeneity to comparable terms, like efficiency units, and then to measure the number of such units supplied per period of time.

Even so, the main thrust of Lewin and Cachanosky is conservative, preferring to repair the foundation without modifying the edifice. They argue that financial duration “captures what it was that the Austrians struggled to express” (2018) and that, besides some infelicitous metaphors like the time structure of production, and some misidentification of where malinvestments would arise (2016), the basic logic of ABCT is sound. Lewin (1999), after the above quote, is unwilling to abandon the roundaboutness metaphor because “even where we have a simultaneous and perfectly synchronized production process, considerations of the time structure and the decisions related to it must still enter” (p. 66). The vision of Böhm-Bawerk’s critics of “consumption and investment tak[ing] place at the same time”, he argues, “is valid only for an economy that has reached a state of stationary equilibrium.”

In a pure theory sense, duration does capture much of the intuition behind the roundaboutness metaphor and the average period of production intended to quantify it. Nevertheless, because the term ‘roundaboutness’ suggests a plan terminating in a consumption good – a point-output model with a circulating capital paradigm – duration is not simply a drop-in replacement. Though it is possible to attach the meaning of duration to the term ‘roundaboutness’, the constellation of nonoperational metaphors surrounding the latter, and their pedagogical persistence, suggest the desirability of replacement rather than rehabilitation. Lewin and Cachanosky’s (2018, §V) “financial reformulation” of ABCT, for example, in addition to swapping out financial duration for the average period of production, employs a flow-output durable capital paradigm that departs fundamentally from the circulating capital metaphors used in the Hayek-Garrison model. Lachmann, in his corpus, largely does the same, though without an explicit duration concept. Both tell flow-output stories unselfconsciously; indeed, it is not clear that any of these authors have fully appreciated the ramifications of the different paradigm. Most commentators, both friends (e.g. Ebeling 2015) and critics (e.g. Vienneau 2008), have accordingly overlooked this essential difference:6 it is not that we have simply replaced a problematic measure with a workable one; it is that we have replaced the entire set of metaphors used to build up ACT.

Duration and Durable Capital

The flow-input-flow-output (FIFO) paradigm suggested by Lewin above entails substantially different metaphors for capital than either Hayek’s flow-input-point-output (FIPO) or Böhm-Bawerk’s point-input-point-output (PIPO) models. If the paradigmatic capital good in a PIPO model is trees, and the paradigmatic capital good in a FIPO model is a good-in-progress (say, a brick, to which is added successively larger structures), only a FIFO model commits to a paradigmatic durable capital good receiving inputs and delivering outputs synchronously. Only this way can we adequately capture the place of durable capital in production plans without the futile attempt to reduce capital to “embodied” factors. Only this way can we do justice to the synchronous quality of production without resorting to a Knightian model that assumes the reconciliation of plans in a market that a capital theory ought to explain (as averred by Hayek [1936] and Kirzner [1966]).

The important step is not simply to elevate durable capital to a first-class citizen in capital theory. So long as we regard durable capital as stored up labor time, another detour taken by original factors of production in the process of their transformation into consumption goods, “there is no essential difference between the circulating-capital and [durable]-capital models” (Bruno et al. 1966). It is not circulating capital as such that poses so many intractable problems; rather, it is the Ricardian habit of thinking of productive processes in point-output terms. A durable paradigmatic capital good will be necessary, but not sufficient, for a productive reorientation.

What would such a reorientation look like? Most importantly, the unit of analysis in capital theory will not be the capital good, but rather the investment project. Rather than resulting in a final consumption good as in point-output models, the investment process in a flow-output model is understood to result in a production process (or changes to an existing one) that itself generates a flow of goods. Rather than focusing on the time-structure of production as intermediate goods flow through it, a flow-output capital theory would focus on changes to production processes – and without the diversion to “original” factors of production. As Knight (1934, §12) pointed out, though somewhat obscurely, the original factors to which the income from a capital project is imputed are not those gifts of nature present at the beginning of time, but rather those present at the initiation of that project. That these factors were themselves produced with the aid of land and labor is irrelevant except to the extent that they were produced with the income from this project in view, in which case the plan’s boundaries would include them too.

This argument distinguishes Lachmann (1956) from Kirzner (1966), both of whom set out to analyze capital theory with the plan as the basic unit of analysis. To take the project instead as our paradigmatic plan orients the subjective method in the context of a flow-output model that was only implicit in Lachmann’s work. Once a capital structure is in place, production itself, along with capital maintenance, can be seen as routine, in the sense of being foreseen as part of the original plan.7 This allows us take seriously both the insurmountable problems with Ricardian circulating capital metaphors, without giving up on the distinctives of the Austrian process approach. To develop this line further is fundamentally Hayekian in spirit,8 even if it does not follow Hayek’s own analysis in many particulars.

A flow-output model with durable capital at the center is the more natural setting for the duration concept. It is hardly sensical to speak of the duration of a piece of circulating capital, except in a highly stylized arithmetical example. In real-world usage, it is not typically durable capital itself that has a duration, but rather investment in durable capital; a capital project. In other words, it is not the stock of capital to which the Austrian malinvestment story applies, but marginal changes to the capital structure. To imagine a factory receiving inputs and producing outputs synchronously, as Knight did, does not require the economy as a whole to be in equilibrium, nor does it require plans to be pre-reconciled. It requires only that we regard particular investments as starting from a given baseline capital structure, the result of previous investments.

To call the duration of an investment “roundaboutness” is to strain the metaphor to the breaking point. Roundaboutness, in its straightforward meaning, pertains to and suggests a circulating capital metaphor, goods in progress taking detours through a longer production process, and is therefore misleading in the context of a durable capital FIFO model. “Time is germane to [the problems of capital],” Lachmann (1956: 111) argues, not as the dimension of capital-intensity, but “as the dimension in which capital resources are turned from one mode of use to another.” Investment will be required in order to change the composition of any stock of capital, and as noted before, it is not generally possible to say whether the resulting configuration is more or less capital intensive (or roundabout) than the previous. This being the case, it should be obvious that regarding capital as an aggregate stock is a distraction from the essential relevance of time to capital theory: not that production takes time, but that investment takes time, regardless of its ultimate effect on the capital-intensity of the production process. Stages of production are irrelevant to the story except as an aspect of the general structure of complementarity within the capital structure that constrains the sorts of investments that can be made.

This change in perspective gives us a framework to appreciate not only the role of time in coordinating plans and expectations, the essential contribution of the Austrian process approach, but also the simultaneity of consumption and production once the investment period is complete, which addresses Knight’s concerns. Time is relevant because the primary goal of investment is not capital maintenance (as in the steady state), but alteration of the capital structure.9

Implications for Business Cycle Theory

One point which has been suggested by various authors on various grounds is that, in light of difficulties with the structure of production concept, “higher order” industries – i.e. those further up the supply chain – may be the wrong place to look for a boom.10 The story of expectations disappointed by the failure of complementary circulating capital goods to appear partway through the productive process only makes sense in the context of a highly stylized model of production – the sort of model Austrians tend to look askance at in other contexts. The prices most affected by low interest rates will not in general be pieces of circulating capital further up the supply chain, but investment projects of long duration, whose income streams lie furthest in the future.

More important, however, is the implication that the Austrian business cycle story – however valid its causal chains – is not necessarily a business cycle story at all once we correct the theoretical shortcomings. It is significant that though Lachmann wrote at length about capital theory throughout his career, he hardly raised the question of business cycles at all after 1941.

If we were to restate the ABCT story in light of the necessary modifications to ACT, it would be something like: rises in interest rates can cause previously viable investments-in-progress to become nonviable, if they should need to refinance in the interim. Such a story is sensical. But aside from the pervasiveness of the interest rate as a price in the economy, it is unclear why this should be of more theoretical import than any other speculative failure to anticipate future price movements. By comparison, exchange rates are a similarly pervasive price in small open economies. Any current exchange rate necessarily indicates an immediate-run balance between speculators anticipating a rise and those anticipating a fall. Any motion, therefore, is bound to disappoint a substantial fraction of speculators. And yet, these continual failures only result in a crisis under special circumstances. It is unclear why speculative failures to anticipate the movement of interest rates in particular should carry such outsize economic significance. The perception of interest rates being (in the short run) under the control of the central bank, and its connection (in the short run) with the volume of circulation, appears to have led Austrian writers to assign these particular failures far more importance than is warranted.

The connection with the volume of spending, however, warrants a more thorough discussion. The canonical Austrian story, drawing on Wicksell, offers a rather more specific proximate trigger. The central bank (or private banks with an elastic currency issue [Hayek 1933, though see White 1999]) holds interest rates below the ‘natural rate’, i.e. the rate that equilibrates saving and investment, which leads to a mispricing of the value of waiting. An investment boom results, which – being inconsistent with consumer preferences – eventually forces interest rates higher, liquidating these investments.

Notice how much of this fleshed-out story – as compared to the skeletal story in the prior paragraph – depends on nonviable concepts above. In the first place, the ‘natural rate’ concept depends crucially on analysis in physical terms (see the discussion in Hayek 1931; 1932a and Sraffa 1932), in a way that does not – in light of the difficulties above – translate readily into a monetary economy. Second, the concept of saving underlying the natural rate concept is a Ricardian economy where saving is accumulated in physical terms, with the interest rate a declining function of the size of the subsistence fund (=saving) (see Harwick 2019b). An unsustainable investment boom is one where entrepreneurs are induced to deplete the subsistence fund. However, if the Ricardian metaphor is nonoperational in a monetary economy, the natural rate of interest is also nonoperational, and whether an investment boom is unsustainable or not cannot be referred to it.

A second potential connection is between currency issue and interest rates in the short run. Quite apart from interest rate movements, a monetary expansion can generate a boom if it causes real money balances to rise before prices do. If this is not correctly perceived as a temporary boom, the short-run increase in general profitability can be imputed to the prices of capital goods, inducing overinvestment in those capital goods which will turn out to be unwarranted by the pattern of consumer demand once prices do begin to rise and choke off demand.

Whether or not rational entrepreneurs will be misled this way has been the subject of a great deal of debate. But more importantly, even supposing that there is such a “cluster of errors”, there is no reason to think that the liquidation of these investments at the end of a boom will necessarily lead to a recession. There is certainly some opportunity cost for reassigning specific capital, and real income will be to some degree lower than if the boom had not happened. But where in this story is a sudden crash? It is hard to imagine that the failure of the marginal investments previously induced by a monetary expansion would diminish real income suddenly and sufficiently enough to constitute the main explanation for recessions.

A third potential connection might answer the previous challenge by bringing in financial markets. A series of realized bad investments can ramify through financial markets, forcing financial firms to contract their issues of broad monies and financial assets (see Harwick 2019a). This can cause a decline in real money balances, slow spending until prices fall, and a rise in interest rates due to a scarcity of liquidity. In principle the liquidation of marginal investments following a boom could snowball in this way, reminiscent of an Austrian business cycle, though to my knowledge no recession has been demonstrated to have been triggered by a previous overinvestment in exactly this manner.

Even supposing such a story, however, the causal chain differs in important ways from the classical Austrian story. Most importantly, the snowball effect is critical. It is not the malinvestments being liquidated that constitutes a recession; rather, the recession forces the liquidation of more investments than would be warranted by the end of a boom. In this sense, a central bank could fruitfully prevent those further liquidations through monetary expansion at the beginning of a recession without propping up the previous nonviable investments, as both Mises and Hayek feared. The object of monetary policy, therefore, is not to set an interest rate matching saving to investment – indeed, it is not clear what that would mean outside of a Ricardian model – but to supply a quantity of money matching money demand so as to avoid the general pressure on prices that leads either to overspending or underspending.


For much of its life, and especially since Mises, Austrian capital theory has been subservient to the needs of the business cycle theory. The business cycle theory, however, is crucially reliant on just those parts of the capital theory that depart from pure theory and do not survive more rigorous specification. By using Lewin and Cachanosky’s duration concept to build a flow-input-flow-output model centered on durable capital as the paradigmatic good, we can give flesh to Lachmann’s fruitful suggestion to use the investment project, rather than the capital good, as the appropriate unit of analysis in capital theory, without abandoning the distinctive process orientation of the Austrian approach, and leave behind once and for all the vestiges of Ricardian analysis that have retarded progress for over a century now.

It remains for future work to evaluate the relative importance of ABCT-adjacent causal processes such as those suggested in the previous section. With a more theoretically satisfactory approach to the problems of capital theory, without being stymied by the variety of paradoxes that beset the capital theorists of previous generations, the difficulties inherent to applied capital theory have the potential to become much more soluble.


  1. Other uses in economics, such as ‘social capital’, are more analogies than proper instances of capital, as they do not generally have a rental or sale value.
  2. Hayek (1936) at various points both asserts that capital can only be sensibly measured in value terms and not physical, and criticizes Knight’s use of value units to homogenize capital into a fund. Yeager (1976) retains the “attractive quasi-homogeneity” of capital by retreating to pure theory and defining it quasi-tautologically as the thing (whatever that may be) which embodies waiting as a factor of production. It is not clear that this is more operationalizable than the more classical Ricardian device, or even that it suggests a paradigmatic capital good (though Yeager does seem attached to the circulating capital metaphor).
  3. The translator of Hayek (1929) from its original German brings attention to his rendering of Kapitalintensität (an obvious cognate to “capital intensity” in English, though Hayek [1939: 17] thought this rendering “somewhat too literal”) as “roundaboutness” (Ibid.: 234). This does not match Böhm-Bawerk’s ([1891] 1930) terminology, for whom “roundabout production” was translated from Productionsumwegen (“detours in production”). Hayek did, however, use “roundabout processes” in his English work (e.g. 1931).
  4. Besides their common wage-fund heritage, the affinity between Austrian and Neoclassical capital theory as antagonists to the reswitching debate can be seen in Schibuola (2014), where Garrison’s account of ABCT is reprised using the Solow growth model in place of the Hayekian triangle, with largely the same results.
  5. Oddly, the Cambridge (UK) antagonists to the debate, though correct about the basic incompatibility between marginalism and the Ricardian income-determination project, preferred to reject the former to preserve the latter.
  6. Ironically, O’Driscoll & Rizzo (1985), the most influential book explicitly advancing Lachmann’s research agenda, contains in its chapter on business cycles (ch. 8) a markedly Hayekian and pre-Lachmannite exposition of ABCT which would later be developed into Garrison (2001). Ebeling (2015), similarly, even states that Lachmann “remain[ed] true to the Böhm-Bawerkian emphasis on capital goods as intermediate goods within time-structures of production”, despite the fact that the lack of this emphasis is precisely what separates Lachmann from preceding Austrian capital theorists.
  7. Hayek (1936) in arguing against Knight’s conception of capital as intrinsically perpetual emphasizes that capital maintenance is not routine; that capital equipment is not necessarily replaced with identical items, and that active decisions are involved at each step. This seems like an overstatement. For our purposes, we can say that planned replacement with an identical item is a routine extension of the original plan, and replacement with different equipment would be a relevant change to the capital structure in our sense.
  8. Interestingly, Hayek (1936, fn. 33) suggests avoiding the use of the term ‘capital’ altogether.
  9. This simultaneity is distinct from the steady state logic Lewin criticizes; indeed, it fits comfortably within what Wagner (2010) named a neo-Mengerian approach to economics:

    Suppose that 95 percent of enterprises are operating within their execution phases, leaving five percent of enterprises at nodal positions where they are either creating or revising plans. This kind of situation would generate observations that would fit with the reasonably predictive properties of models of static equilibrium. An established furniture manufacturer that also owned its forests would confront the world in pretty much simultaneous fashion. During any year, or other time span, it would be planting trees, harvesting trees, buying and repairing equipment, and making furniture, all of it appearing to be simultaneous. . . . The source of the motion [in an economy] . . . is the five percent of enterprises not in stasis at any particular instant that are eroding the static reposes of the other enterprises.

  10. Cachanosky & Lewin (2016) argue that investments of long duration “do not correlate coherently” with what were understood by previous Austrian authors as higher-order stages of production. Salter & Luther (2016) argue that, in the context of a rational expectations model, whether the boom occurs in higher-order industries or somewhere else has little relevance to the Austrian story. Luther & Cohen (2014; 2016) discuss some of the difficulties in operationalizing the stages of production concept in the context of empirical work.


CapitalEugen Von Böhm-BawerkF.A. HayekFrank FetterIsrael KirznerLudwig LachmannLudwig Von MisesPaul SamuelsonPeter LewinRoger Garrison


Facebook Twitter Reddit StumbleUpon


  • 1

    Raymond Niles

    Oct 16, 2014 at 3:20 | Reply

    Im thinking about this. It is interesting.

    I agree with your general point that the focus must be on investment. How hard — i.e., how much investment does it require — to reconfigure production? If the change in demand is large and it requires significant amounts of investment to alter a production process, a huge price change will happen in that industry (either a bubble or a collapse).

    So, I think you are right that it is the time for investment (to alter a production process), not the time for producing goods, that matters.

    Does this mean that “roundaboutness” or a time-structure of production is no longer valid? It may be that this concept is confusing when trying to understand the role of capital in the business cycle. For conceptual clarity, one may want to jettison the Bohm-Bawerkian framework, or most of it, as you suggest. (I am still not sure about this.)

    However, it is still true that time is a valid concept when thinking about these things. For example, investment comes from savings, the size of which is a function of time. The more time there is to save, the larger the amount that can be invested. Thus, a large investment is also a more “time-consuming” process.

    From your earlier post, it is clear that bubbles form when large changes in demand confront an industry that requires a large investment to change output (e.g., housing). Such large investment is time-consuming.

    So, we may be able to rescue the concept of roundaboutness, not by applying it to capital stocks, but by applying it to the levels of investment required to alter an industry.

    However, an industry that takes a lot of investment to alter is also one with a large embedded capital structure. Consider the stock of housing, as an example. Then we are back to roundaboutness. It applies to both the stock of capital and to the quantity of investment required to alter output in that industry.

    Does this mean that Bohm-Bawerk is rehabilitated? Probably not. I think that your focus on investment as opposed to capital stock is important and may represent an improvement to ABCT.

  • 2

    Jim Caton

    Oct 16, 2014 at 22:13 | Reply

    This all fits very well with Lachmann and sits well with me, at least mostly. Time structure of production is not so significant to coordination in equilibrium. It does matter out of equilibrium where market actors must form new expectations concerning the profitability of particular ventures and must wait to see if they correctly judged market conditions and interpreted prices. Once in equilibrium, the time element is only important in that the processes are constantly rolling over in a sustainable manner. That’s boring anyhow.

    What is significant is that the structure of production must change to meet certain needs. The time element is important in that it effects profit margins of producers, but this is taken care of well enough with a run-of-the-mill financial framework. Austrian theory is special because of its emphasis on capital structure. Changes to it take time both to plan and integrate. The idea of stages of production is partly a hangover from equilibrium theorizing – otherwise profit margins make nonsense… – and as you aptly point out, Ricardo.

Leave a Reply

More Content