Capital as a Factor of Production

by K.H. Hennings
Article in the New Palgrave: A Dictionary of Economics;
edited by John Eatwell, Murray Milgate, Peter Newman;
London, New York, Tokyo: The Macmillan Press Limited, The Stockton Press, Maruzen Company Limited; 1987.
Vol. 1, pp 327 - 333.

Capital as a Factor of Production. The role played by capital in production has frequently been in dispute: "When economists reach agreement on the theory of capital they will shortly reach agreement on everything else' (Bliss, 1975, p. vii). Disagreements are due as much to divergent definitions, or uses, of the term 'capital' as to different views about what should be considered a factor of production. But above all there have been differing views about whether, and in what sense, capital can be said to be productive. In particular, there has been disagreement about whether it can be said that a more capital-intensive production method is more productive than a less capital intensive one. Preclassical, classical, neoclassical and neo-neoclassical economic theory have given different answers to these questions. These will be considered below, but the discussion will be confined to the role of capital as a factor of production. It should be noted in particular that the problem why capital earns its owner an income depends as much on the social institution of ownership and the institutional organization of production as on the role capital plays in production. It is only the latter, in a sense technical, problem which will be addressed here.

TERMINOLOGY. Capital goods are produced commodities which are required for production no matter how much or how little they are subject to wear and tear. A stock (at a point of time; see Fisher, 1906) of different capital goods is a capital; this concept is to be taken in a vector sense. As long as they are required in production, all capital goods can be valued, even when they are not traded on markets, as many of them are. Because of their heterogeneity, different capital goods cannot be aggregated, but their values can. A capital value is therefore the sum of the capital values of those capital goods which constitute a capital. Note that this is a book-keeping term, which depends on the valuation of the capital goods involved; the capital value can change although there is no change in the stock of capital goods. The term money capital will be used in a similar sense, but with a somewhat different connotation: it denotes the sum of money necessary to buy a specified stock of capital goods. Real counterparts in a scalar sense to a given capital value or money value can be constructed in principle (Hicks, 1974, p. 151), but not in an unambiguous manner.

PRODUCTION: BASIC NOTIONS. Production is the transformation of inputs into outputs. Inputs are those things which need to be increased in order to obtain more output by the same method of production, where the latter is defined as a blueprint which details what inputs are required when and in which proportions to produce a unit bundle of outputs. As there may be more than one method to produce the same unit bundle of outputs, a production process is defined as a particular method of production to produce a particular unit bundle of outputs. A production process always uses inputs in fixed proportions; variable proportions are represented by different production processes. If there exist various different production processes with which the same unit bundle of outputs can be produced, they will differ in the proportions in which they use various inputs; but in general it will not be possible to compare them from a purely technical point of view. Different production processes are comparable only if their costs are computed and related to the value of the outputs obtained. In general, however, any ordering obtained in this way need not be unique: two different production processes may have the same unit costs. Moreover, if the prices of inputs change a given ordering need not be preserved. Such difficulties affect the choice between different production processes; they do not, however, affect the role of capital in production, or its status as a factor of production.

Production typically is roundabout, i.e. proceeds in stages: what is produced as output in one production process is used as an input (alongside others) in another. If all these intermediate products (outputs which are used as inputs) are specific in the sense that they have only one possible use, all production processes required to produce a particular bundle of outputs can be strung together into a sequence of production processes. Consolidating all stages, one can view the sequence as transforming 'primary' inputs into 'final' outputs. Here primary inputs are these which are not produced within the sequence of production processes, if indeed they can be produced at all; final outputs are those which are not used, or used up, within the sequence.

Not all intermediate products are specific in the sense that they have only one possible use. In this case all interlocking sequences can be combined into a production system which again can be viewed as transforming primary inputs into final outputs. Without loss of generality one can assume that such a production system comprises all production processes in operation in an economy. Consolidating them amounts to adopting a 'black box' view of production. Disregarding the internal structure of the production system and of the production processes which constitute it, one links directly primary inputs to final outputs, and disregards all inputs produced and used, or used up, within the production system. The advantage of this procedure is that it reduces the number of inputs to be considered.

The definition of what is a primary input, or a final output, depends on the level of aggregation as well as the nature of the production processes involved. Production on a barren island will require many inputs as primary ones which are intermediate products in a production system comprising all production processes operating in a continent rich in resources. Similarly the final outputs produced by the island economy's production system may be confined to what are intermediate products in the production system of a continent.

By definition, an increase in output can only be obtained by an increase in inputs in fixed proportions. From this one can infer that all required inputs together are productive, and have a non-negative marginal production. This cannot, however, be inferred for any single input. This can only be done if either there are at least two different production processes for the production of the same unit bundle of outputs because then it is possible to calculate the marginal net value product of an input (Bliss, 1975, ch. 5); or if there are alternative uses for all inputs in production processes which produce other unit bundles of outputs (Uzawa, 1958). Only when there exist only one production process for a particular unit bundle of outputs and there are no alternative uses for some of the inputs it requires is it impossible to calculate their marginal contribution to the outputs obtained individually; it is of course still possible to calculate their contribution as a group of inputs.

FACTORS OF PRODUCTION. In modern usage, all primary inputs can be called 'factors of production'. Conventionally, however, primary inputs are considered, following Senior (1836), the services of agents or stocks, and the term 'factor of production' is reserved for the latter. If they are the services of natural agents or human beings, they are called 'original factors of production'; they are called simply 'factors of production' if they also include the services of stocks of durable commodities. Factors of production can therefore be defined as those agents or durable stocks the services of which are primary inputs in production processes. Factors of production are productive and have a non-negative marginal product if their services are productive and have a non-negative marginal product. The definition of factors of production just given is reasonably precise as far as natural agents and human beings are concerned. Land an labour have been considered factors of production at least since Petty (1662). Land was often understood, if tacitly, to include all beneficial powers of nature; the term 'natural agents' was introduced by Senior (1836). In preclassical theory durable stocks were called simply 'stocks' (see, e.g., Barbon, 1690), but usage of the term was often confined to trade and commerce. When production came to be seen as the dominant economic activity, produced means of production, considered as a factor of production, came to be called 'capital'. This term had been in use for a long time (see Hohoff, 1918-19; Salin, 1930; Assel, 1953), but now acquired a new meaning, thus inviting confusion and controversy. It will be useful, therefore, to trace historically the use made of that term, and the notions attached to it.

PRECLASSICAL THEORIES OF CAPITAL AND PRODUCTION. There is very little about production and its relation to capital in economic writings before the mid-18th century. Barbon (1690) provides an early, but singular, instance of an analysis in which a surplus is seen to arise from the use of what he calls a 'stock' (of capital goods) in trade as well as in the production of commodities and a sum of money. But he did distinguish, as had Barbon, between profits from 'stock' and interest on money (1752, p. 313), thus separating the investment of money from the productive use of 'stock', e.g. capital goods, although he is none too clear about the latter.

The Physiocrats were probably the first to develop a clear view of production and the role of capital in it. But they did not use the term 'capital'. Cantillon (1755) strongly emphasized the need for accumulated sums of money required to buy stocks of goods in which to trade, or with which to produce. But he called them 'funds' not 'capital'. Thus he speaks of the farmer who needs to have enough funds (assed de fond) to conduct his business. Quesnay used the term 'advances' (avances) in a similar way in the sense of money capital. Behind his usage is a clearly drawn picture of agricultural production which uses land and labour to produce output, and needs money capital to finance the lag between the expenditure on inputs and the sale of the output obtained. Probably deliberately, Quesnay eschewed the term 'capital'. Where he used it (1766b), he spoke explicitly of money capital (capital d'argent), but conceived it as invested in buildings, implements, stores of grain, cattle, and so on (1766a, pp. 172-3). These, however, he clearly conceived as productive. Moreover, his argument centres on the idea that larger advances would permit more productive production methods to be used (see Eltis, 1984, chs. 1 and 2).

Turgot (1770) was the first to develop a specific theory of capital as a factor in production when, possibly under the influence of Hume's ideas, he generalized Quesnay's theory. Quesnay had shown that advances where necessary for agricultural production. Turgot, in an attempt to develop Quesnay's theory of a society dominated by agriculture into a theory of commercial society, places commerce and manufacture on an equal footing with agricultural production, and emphasized that advances are required in all branches of economic activity. Such advances are paid out of capital, which is defined as 'accumulated values' (1770 , § LVIII). If account is taken of the various degrees of risks involved, the rates of return on all possible investments are equalized by competition between the owners of the various capitals (Turgot uses the plural, capitaux) such that the rate of interest can 'be regarded as a kind of thermometer of the abundance of scarcity of capitals in a Nation, and of the extent of the enterprises of all kinds in which it may engage' (1770, § LXXXIX). At the same time, Turgot argues emphatically that some return on all these kinds of investment is necessary in oder to keep production on the same level; if the rate of return were lowered, capitals would be withdrawn, and production could not be kept on the same level as before (1770, § XCVI). Thus to Turgot 'capitals' are money capital. Money capital is required because production is roundabout and thus needs capital goods as well as original factors of production. Like Quesnay, Turgot assumed that larger amounts of money capital make possible higher levels of production. One might be inclined to argue that therefore money capital, i.e. advances, are productive; but although Turgot is not entirely clear on this point it seems that he considered not so much advances as the capital goods which represent them as productive.

THE CLASSICAL THEORY OF CAPITAL AND PRODUCTION. The classical view of the role of capital in production was worked out by Adam Smith. He began by emphasizing the division of labour, but then switched to a detailed consideration. 'Of the Nature, Accumulation, and Employment of Stock' (1776, book II) in which he effectively adopted the theory put forward by Quesnay and Turgot. His attempts to integrate these two approaches were not entirely successful (Bowley, 1976); although the division of labour retained its status as a device which enhances the productivity of labour in classical economic theory, the emphasis was shifted to the accumulation of capital as the prime force making for growth. This was of course linked to the idea that production needs advances, and the proposition that labour was the more productive the larger these advances. Smith also changed the emphasis in another respect: he formally defined 'capital' as that part of a persons's stock of commodities which is expected to yield an income. Smith described its function as assisting labour in production: fixed capital (machines, buildings, land improvements, and 'aquired and useful abilities') 'facilitates' labour by increasing its effectiveness; circulating capital (money, raw materials, goods in process and goods in stock) 'abridges' by providing (material) advances.

This distinction is ambiguous, but characteristic for Smith's position. Fixed capital, he argued, yields an income, i.e. is productive, by being used 'without changing masters': while circulating capital needs to be either given up (in trade) or be destroyed (in production) in order to be productive (1776, pp. 279-83). What is considered are capital goods; but only money capital can circulate in the way Smith described their circulation. The two approaches can be reconciled; but the way in which Smith expressed himself invited confusion between money capital on the one hand, and capital in the sense of capital goods on the other. In fact, Smith needed both concepts. James Mill (1821), Rae (1834) and other classical writers often used the term 'instrument' when emphasizing that they meant capital goods, and continued to speak of capital in the sense of money capital. Money capital played indeed an important role in classical enonomic thought for it permitted classical writers to argue, in a rather loose way, that production methods were the more productive, the more money capital they required. It is for this reason that Hicks (1974) called them 'Fundists'. At the same time, however, they also considered the role of capital goods in production processes (Sraffa, 1960), and thus maintained a 'real capital doctrine' (Corry, 1962, p. 18).

The view that capital assists labour was attacked by Lauderdale (1804), who pointed out that capital could, and frequently did, supplant labour when circulating capital was substituted for fixed capital. This initiated the debate on the 'machinery question', and confirmed the role of capital as a factor of production: what can supplant a factor of production surely must be considered as belonging to the same species.

Smith had separated a person's stock of commodities into durable consumer goods and capital by requiring that the latter be expected to earn an income. This led to many attempts to show that not only capital goods used in production are expected to yield an income (i.e. Hermann, 1832, or Menger, 1888). These discussions often confused the role of money capital in investment processes with the role of capital goods in production processes, and contributed to the survival of the concept of money capital as a factor of production referred to above.

The view that production requires advances in the form of capital goods was so dominant that the role of fixed capital was often pushed into the background. Thus Ricardo spoke of production as 'the united application of labour, machinery, and capital' (1817, p. 5), thus equating capital with circulating capital. As Smith had subsumed the consumer goods required for the maintenance of labour under circulating capital stocks, this particular part of the total stock of commodities in an economy acquired, under the name of the 'wage fund', a pivotal role in all discussions of the role of capital in production. Following a precedent set by Smith, the wage fund was seen to be derived from, and increased by, saving, i.e. non-consumption or 'abstinence', as Senior (1836) was to call it. Destined to supply the consumption goods required as advances while production processes continue, the concept was used as a theory of wage determination on the assumption that the wages fund was given at least in the short run and thus determines the wage level when workers compete freely for employment.

In spite of all the attention Smith gave to the accumulation of capital as a factor making for economic growth, he reserved a special role for human labour as the prime factor of production, especially in those passages in which he set out his conjectural history. This emphasis, which is clearly based on the view that production requires advances, remained a feature of the classical theory of capital, and was a mainstay of the labour theory of value as developed by Ricardo and others. It is symptomatic that from this point of view the use of 'machinery and other fixed and durable capital' was considered no more than an (admittedly considerable) modification of the labour theory of value by Ricardo (1818, p. 30). More radical writers, such as Hodgskin (1827) emphasized the notion of capital goods as 'stored-up' labour (i.e. outputs produced by past labour) that had been worked out by James Mill (1821) and Ricardo (1817) and on its basis denied fixed capital the status as a factor of production.

The special role Smith has reserved for labour did not prevent him from juxtaposing labour, stock and land to parallel wages, profits and rent (1776, p. 69). This juxtaposition was elaborated into a strict parallelism between factors of production and their earnings by Say (1814) which became generally accepted by the middle of the 19th century. Thus when J.S. Mill (1848) summarized the classical theory of capital into his four propositions, he still adhered to the view that production required advances in the form of capital goods. But when he comes to discuss the laws of increase of factors of production, he treats them on an equal footing (though in exactly the order Smith had listed them: and not the land-labour-caital oder which Say had made familiar). At the same time, however, Mill often gives the impression that he means money capital when he speaks of 'capital', especially in those passages in which he argues that competition will establish a uniform rate of return on capital because capital will be transferred from one industry to another.

In a similar way Marx (1867) used the term capital to mean both a stock of commodities, and a sum of values. In addition, Marx insisted that capital goods are capital only in a capitalistic society, and thus used the term also to describe a particular organization of production in society.

Finally, the view that production requires advances in the form of capital goods which Smith had expounded, and which most classical writers accepted, was developed by a few of them into a theory which strongly emphasized the time element in production. There are some traces of this in Ricardo (1817), especially in his recognition that all the difficulties he encountered in his theory of value are due to the temporal aspect of production processes. The view was worked out in detail by Rae (1834), by Longfield (1834), and also by Senior (1836). Their work foreshadows one aspect of the neoclassical theory of capital.

Classical economic theory considered three factors of production: land, labour, and capital. Each had its own dimension: land was a stock, labour a flow, and capital was money capital in the form of a stock of capital goods. In the original conception their standing was not equal: labour worked on land with the help of capital. Hence the capital intensity of production mattered: the more money capital was invested, the more productive was labour in its efforts to work up the bounty of nature into consumable output. These notions were not, however, made very precise: that was left for neoclassical theorists. Thünen's early discussion of the marginal productivity of capital (1850) remained an exception.

THE NEOCLASSICAL THEORY OF CAPITAL AND PRODUCTION. Neoclassical economic theory was not a coherent construct: up to the 1930s there were different versions of neoclassical theory as far as the treatment of capital as a factor of production is concerned (Stigler, 1941).

Perhaps the most contentious version was the Austrian one as worked out by Böhm-Bawerk (1889). To some extent it had been foreshadowed by Jevons (1871), even though Jevons had little to say about production. But there is a clear picture in Jevons of the necessity of money capital which is 'invested' in the form of advances in time-consuming production processes. What is more, Jevons formulated, very much ad hoc, a temporal production function which postulated that there are diminishing marginal returns to the length of investment of such advances: and used it to derive the marginal product of an extension of that length, which clearly is a measure for the capital intensity of production.

Böhm-Bawerk, by contrast, consciously and explicitly developed a theory of production. It very much follows classical lines: production requires time, and hence needs advances in the form of capital goods. Capital goods are seen as produced means of production, and at the same time as stored-up land-and-labour, even though they derive their value not, as the classics had maintained, from the fact that they represent land and labour services spent in the past: but from their prospective usefulness in the production of future output. Nevertheless Böhm-Bawerk emphatically denied that capital goods can be productive, and insisted that only the production processes which they make possible are productive. Although this could have meant that the notion of productiveness was transferred from factors of production to production processes, Böhm-Bawerk did not take this step. He seems to begin by saying that only land and labour should be called productive, and ends by postulating something very much like a productivity of the length of the period of production. As in Jevons (1871), this view is based on a temporal production function in which the degree of roundaboutness of production processes is explicitly taken as a measure for the capital intensity of the production processes in operation. Böhm-Bawerk attempted to overcome in this manner the difficulty of deriving any such measure from diverse sets of capital-goods. The roundaboutness of production processes was turned into a variable which was chosen by profit-maximizing entrepreneurs subject to a given amount of money capital.

The relationship of this construction to classical economic thought is obvious. Nevertheless Böhm-Bawerk's attempt to provide a temporal theory of production based on the notion of capital as a derived factor of production, or intermediate good, turned out to be very contentious. The theory of interest which he had been built upon it was turned into what became the standard (neoclassical) theory of interest by Fisher (1907, 1930) - but only after it had been cut loose from its production-theoretic underpinnings: and after Fisher had substituted instead an analysis of investment opportunities based on the concept of money capital. Various attempts to reformulate Böhm-Bawerk's theory of the role of capital in production (Wicksell, 1893; Strigl, 1934; Hayek, 1940) generated much debate, but did not manage to rescue it.

The Austrian theory of capital is much more traditional than other versions of neoclassical theory which gave up the 'advances' view of capital. Thus Wecksteed (1894) placed all factors of production on an equal footing, including all kinds of capital goods, and postulated that 'The Product being a function of the factors of production we have P = f(a,b,c, ...)' (1894, p. 4) without even mentioning whether production takes time or not. Being considered akin to any other input in this respect, capital goods are of course productive; but nothing can be said about the capital intensity of production. Marshall (1890) argued in a similar way, although he kept to the classical tradition by reserving a place for money capital alongside the capital goods used in production. Taking up a distinction first made, it seems, by Menger (1888, p. 44), Marshall distinguished between capital goods which earn quasi-rents, and money capital which earns interest. In essence this is the distinction between production and investment: capital goods are used in production, and if used productively, earn quasi-rents; money capital is invested, and if invested successfully, earns interest. Clark (1899) equally rejected the advances view of production. In his view, production did not require advances once production processes were properly set up, or synchronized. As in Wicksteed, capital is a factor of production on an equal footing with land or labour. At the same time, Clark separated clearly between material capital goods or produced means of production, on the one hand, and capital as a quantum of productive wealth' (1899, p 119), measured in money, which is invested in capital goods. Although Clark calls this 'a material entity' (1899, p. 119), his 'capital' is money capital, just as it was in Marshall (or Menger for that matter). Knight(1933) continued in this vein, but emphasized money capital, considered as a 'material entity', so much that capital goods were almost lost sight of. As a result, 'capital' came to be seen more and more as a homogeneous mass which was created by saving decisions, which could be invested in one industry and transferred to another, which was productive in the sense that it has a non-negative marginal product if used properly, and which guaranteed higher productivity if employed in larger amounts in relation to other factors of production. Not surprisingly, this conception was attacked by the heirs to the Austrian tradition in capital theory, especially Hayek (1936, 1940), as a 'mythology of capital'. But their own position was so much bound up with the deprecated notion of a period of production that Knight's conception (1933, 1934, 1935, 1936) became the dominant doctrine.

The notion of capital as a 'material entity' was formulated rigorously by Pigou, who provided a sophisticated definition of a capital stock, consisting of heterogeneous capital goods, which 'is capable of maintaining its quantity while altering its form' (1935, p. 239). This was possible only by making some rather strong assumptions on the way the capital stock was maintained. Thus Pigou assumed, among other things, that any item of a constant capital stock that needs to be replaced is replaced by another capital good yielding equal quasi-rents at the time of replacement. Later changes in quasi-rents are disregarded. While such assuptions may be objected to, they do allow in principle to give precise meaning to the notion of a capital stock as a changing 'material entity' without aggregating heterogeneous capital goods, i.e. without negating its quality as a vector.

Walras (1874-7) and Pareto (1909) treated capital very much as Wicksteed had done: as yet another factor of production in a production system which was fully synchronized and which was not in need of advances. As they used production functions and thus assumed, as Wicksteed had done, that there always exist many production processes for the production of the same unit bundle of outputs, the productivity of capital goods was no problem for them. But because they espoused the black box view of production they somewhat lost sight of the internal structure of production, and hence of the character of capital goods as produced means of production: capital goods are in their conceptual scheme simply part of the endowment which economic agents use to maximize their satisfaction. Moreover they could not form a notion of the capital intensity of production as thy had no way of aggregating capital goods in an unambiguous manner.

Wicksell, finally, in his later treatment of the matter (1901) attempted to provide a synthesis of neoclassical capital theory by combining the general equilibrium framework of Walras and Pareto with the Austrian view of production as a time-consuming process. This led him to emphazise capital goods and their productivity. But when he came to close his system he took refuge, as Böhm-Bawerk had done, to the idea of a given fund of money capital. The importance of a given fund of money capital which acted as a constraint on entrepreneurial choices between different degrees of roundaboutness of production processes was also emphasized by Schumpeter (1911) and Cassel (1918).

Neoclassical economists have in common that they attempted to formulate a theory of production; but they differed in their conceptions (Hennings 1985). Böhm-Bawerk and those who followed him made an attempt to formulate more precisely what they saw as the gist of the classical theory of the role of capital in production: but their efforts were not generally accepted. All other neoclassical writers exept Wicksell jettisoned the advances view of capital, and were in consequence faced with the necessity of formulating a measure for the capital intensity of production if they wished to uphold the proposition that more capital intensive methods of production were more productive. Wicksteed as well as Walras and Pareto did not do so, and simply refrained from making such statements. Marshall, Clark, and Knight in one way or another attempted to solve the problem by taking refuge to a concept of capital which is in essence a notion of money capital, and which cannot unambiguously serve for that purpose. Only Pigou formulated an unambiguous concept of capital as a changeable 'material entity'.

THE NEO-NEOCLASSICAL THEORY OF CAPITAL IN PRODUCTION. The neo-neoclassical view of the role of capital in production is based on the work of Viner (1930), Stackelberg (1932), Schneider (1934) and others, who worked out the theory of production as well as a theory of production costs, and the syntheses later provided by Hicks (1939) and Samuelson (1947) of the various neoclassical theories on the basis of the Walras-Pareto theory of general equilibrium (Arrow and Hahn, 1971). Originally strongly microeconomic in nature, capital goods held an stage. But as this theory was essentially static, little thought was given to dynamic considerations that arise if concrete capital goods are shifted from one industry to another. Where such problems came up, refuge was taken in the Clark-Knight conception of capital as a fairly homogeneous and amorphous mass which could take on different forms. With the growth of macroeconomic one-sector thinking - Hicks (1932) is one of the earliest examples in this part of economic theory - this conception was more and more resorted to. It received the seal of approval in Samuelson's textbook (1948), and in numerous empirical studies based on the macroeconomic production function first proposed by Cobb and Douglas (1928). It was of course realized that capital consisted of capital goodss: abut their aggregation into a more or less homogeneous aggregate was considered an index number problem which could be solved in principle as well as in practice. It was against these notions that opposition arose in the 1950s and 1960s.

RECENT DEBATES. As Joan Robinson (1954, 1956) pointed out, the Clark-Knight concept of capital cannot serve in a macroeconomic production function à la Cobb-Douglas because it is essentially a monetary measure. Surprisingly, this contention engendered a major debate in capital-theory. Essentially two answers were given to Robinson's objection. On the one had it was argued that one should search for appropriate indices that can be used to aggregate heterogenous capital goods into a scalar measure (Campernowne, 1954).

This created a specialist literature on aggregation problems which demonstrates that in general conditions for consistent aggregation are rather restrictive, although in many cases appropriate indices exist (Green, 1964). On the other hand, it was argued that macroeconomic analysis should be abandoned in favour of microeconomic ones if heterogeneity (which after all exists in land and labour as well as in capital goods) is the issue (Swan, 1962).

In the course of the debates referred to above it was demonstrated that the value paradoxes Joan Robinson had pointed out may invalidate the idea that different production processes can be brought into a continuous ordering which corresponds to their respective capital intensities. While this point was eventually accepted, its importance is still under dispute (see Harcourt (1972) and Blaug (1974) for summaries and evaluations from divergent points of view). To some, such demonstrations completely invalidate neoclassical and in particular neo-neoclassical economic theory, because both are considered to be founded on the idea that marginal products of factors of production need to be calculated on the basis of technical data alone. Others accept such demonstrations as exceptions to a general rule. What is sometimes lost sight of in these assessments is the fact that reswitching of produciton processes, capital revaluations, Wicksell effects, et hoc genus omne do not invalidate all propositions in capital theory (whether neoclassical or not). One can well do without capital in the sense of capital value (i.e. as a scalar magnitude) for some purposes (see, e.g. Nuti, 1970). Moreover, it should be appreciated that Robinson's objections do not apply to Pigou's notion of capital as a changable 'material entity' even though it is not at all obvious that such a concept would serve well in macro-economic production function.

Another attack on neoclassical capital theory was made by Garegnani (1960, 1970, and 1976). The gist of his argument seems to be that the Walrasian model of general equilibrium, if properly extended to include the production of capital goods, cannot generate equilibrium as well as a unique rate of return on all capital goods for all possible initial endowments. As Garegnani has not specified the dynamic adjustment processes he envisages, hos claim is difficult to adjudicate. Nor is it clear in what respect, if any, it invalidates received notions of the role of capital in production processes. Recent debates (Hahn, 1982; Dum&e;nil and L&e;vy, 1985) have not thrown much light on these issues.

CONCLUSION. Capital always consists of heterogeneous capital goods; indeed it is useful precisely because goods are heterogeneous and specific in the sense that they cannot be used for all purposes. Attempts to represent them by some kind of aggregate are useful only if they preserve this aspect of capital goods. In classical economic theory the notion of advances was used as such an aggregate, although in a rather loose fashion, with an awareness of the heterogeneity of the capital goods that assisted labour in time-consuming production. Austrian neoclassical economic theory attempted unsuccessfully to make this notion more precise in the form of a temporal theory of production. Non-Austrian neoclassical and neo-neoclassical economic theory sacrificed the heterogeneity of capital goods together with the time element in production, and developed an atemporal theory of production on the basis of a concept of capital value, or money capital. Yet, as Wecksell pointed out (1901, p. 149), the valuation of capital goods in terms of prospective output is a 'theoretical anomaly'; it is nevertheless appropriate in view of their character as produced means of production. It is not surprising, therefore, that anomalies result when such concepts are used. The alternative is obviously to analyse the role of capital goods in a framework which admits their heterogeneity and permits them to be used for different purposes, i.e. in a general equilibrium framework. Such analysis have so far been mainly confined to stationary states. Some of the essential characteristics of capital goods, however, such as their specificity, are of importance only in non-stationary states. Much remains to be done, therefore, before the role of capital and of capital goods as factors of production can be said to be completely elucidated.

K.H. Hennings

BIBLIOGRAPHY

Arrow, K.J. and Hahn, F.H. 1971. General Competitive Analysis. San Francisco: Holden Day; Edinburgh: Oliver & Boyd.
Assel, H.G. 1953. Der Kapitalbegriff und die Kapitallehre bis zum Beginn der Neuzeit. Wirtschaft und Gesellung. Festschrift für Hans Proesler zu seinem 65. Geburtstag. Erlangen.
[Barbon, J.] 1690. A Discourse on Trade. By N.B.M.D. London: Milbourn for the Author.
Blaug, M. 1974. The Cambridge Revolution: Success or Failure? London: Institute of Economic Affairs. Bliss, C.J. 1975. Capital Theory and the Distribution of Income. Amsterdam and Oxford: North-Holland Publishing Company. Innsbruck: Wagner. Trans. as The Positive Theory of Capital, London: Macmillan, 1891.
Bowley, M. 1975. Some aspects of the treatment of capital in 'The Wealth of Nations'. In Essays on Adam Smith, ed. A.S. Skinner and T. Wilson, Oxford: Clarendon Press.
Cantillon, R. 1755. Essai sur la nature du commerce en general. Londres: Gyles. Edited with an English translation by Henry Higgs, London: Cass, 1959.
Cassel. G. 1918. Theoretische Sozialökonomie. Leipzig: Deichert. Trans. as Theory of Social Economy, London: Fischer Unwin, 1923.
Chapernowne, D.G. 1954. The production function and the theory of capital: a comment. Review of Economic Studies 21, 112-35.
Clark, J.B. 1899. The Distribution of Wealth. New York: Macmillan.
Cobb.C.W. and Douglas, P.H. 1928. A theory of production. American Economic Review 18, March 139-65.
Corry, B.A. 1962. Money, Saving and Investment in English Economics 1800-1850. London Macmillan.
Duménil, G. and Lévy, D. 1985. The classicals and the neoclassicals: a rejoinder to Frank Hahn. Cambridge Journal of Economics 9, 327-45.
Eltis. W. 1984. The Classical Theory of Economic Growth. London: Macmillan.
Fisher, I. 1906. The Nature of Capital and Income. New York: Macmillan.
Fisher, I. 1907. The Rate of Interest. New York: Macmillan.
Fisher, I. 1930. The Theory of Interest. New York: Macmillan.
Garegnani, P. 1960. Il capitale nelle teorie della distribuzione. Milano: Guiffrè. Trans. as Le capital dans les théories de la répartition, Paris: Presses Universitaires de Grenoble et François Maspero, 1980.
Garegnani, P. 1970. Heterogeneous capital, the production function and the theory of distribution. Review of Economic Studies 37, 407-36.
Garegnani, P. 1976. On a change in the notion of equilibrium in recent work on value and distribution. In Essays in Modern Capital Theory, ed. M. Brown, K. Sato and P. Zarembka, Amsterdam, New York, Oxford: North-Holland.
Green, H.A.J. 1964. Aggregation in Economic Analysis. Princeton: Princeton University Press.
Hahn, F.H. 1982. The neo-Ricardians. Cambridge Journal of Economics 6. 353-74.
Harcourt, G.C. 1972. Some Cambridge Controversies in the Theory of Capital. Cambridge: Cambridge University Press.
Hayek, F. A. von. 1936. The mythology of capital. Quarterly Journal of Economics 50, February, 199-228.
Hayek, F.A. von. 1940. The Pure Theory of Capital. London: Routledge & Kegan Paul.
Hennings, K.H. 1985. The exchange paradigm and the theory of production and distribution. In: Foundations of Economics, ed. M. Baranzini and R. Scazzieri, Oxford: Blackwell.
Hermann, F.B.W. 1832. Staatswirtschaftliche Untersuchungen. Munich: Weber.
Hicks, J.R. 1932. The Theory of Wages. London: Macmillan. 1963. Hicks, J.R. 1939. Value and Capital. Oxford: Clarendon Press. Hicks, J.R. 1974. Capital controversies, ancient and modern. American Economic Review 64, May, 307-16. Reprinted in J.R. Hicks, Economic Perspectives, Oxford: Clarendon Press, 1977.
[Hodgskin, T.] 1827. Popular Political Economy. London: Tait and Wait.
Hohoff, W. 1819-19. Zur Geschichte des Wortes und Begriffes 'Kapital'. Vierteljahrshefte für Sozial- und Wirtschaftsgeschichte 14, 554-74 and 15, 281-310.
Hume, D. 1752. Of Interest. Political Discourses. Edinburgh: Fleming. Reprinted in D. Hume, Essays Moral, Political and Literary, Oxford: Oxford University Press, 1963.
Jevons, W. S. 1871. The Theory of Political Economy. London: Macmillan.
Knight. F.H. 1933. Capitalistic production, time and the rate of return. In Economic Essays in Honour of Gustav Cassel, London: George Allen & Unwin.
Knight, F.H. 1934. Capital, time, and the interest rate. Economica, NS 1, August, 257-86.
Knight, F.H. 1935. Professor Hayek and the theory of investment. Economic Journal 45, March, 75-94.
Knight, F.H. 1936. The quantity of capital and the rate of interest. Journal of Political Economy 44, August 433-63, and October, 612-42.
Lauderdale, J. Maitland, 8th Earl of. 1804. An Inquiry into the Nature and Origin of Public Wealth. Edinburgh: Constable.
Longfield, M. 1834. Lectures on Political Economy. Dublin: Milliken.
Marshall, A. 1890. Principles of Economics. London: Macmillan.
Marx, K. 1867-94. Das Kapital, 3 vols, Hamburg: Meisner. Trans. as Capital, 3. vols, Chicago: Kerr, 1906-9.
Menger, C. 1888. Zur Theorie des Kapitals. Jahrbücher für Mationalökonomie und Statistik, NF 17(51), 1-49.
Mill, J. 1821. Elements of Political Economy. London: Baldwin, Cradock and Joy.
Mill, J. S. 1848. Principles of Political Economy. London: Longmans, Green.
Nuti, D.M. 1970. Capitalism, socialism, and steady growth. Economic Journal 80, March, 32-57.
Pareto, V. 1909, Manuel d'économie politique. Paris: Giard & Brière. Trans. as Manual of Political Economy, London: Macmillan, 1972.
Petty, W. 1662. A Treatise of Taxes and Contributions. London: Brooke. Reprinted in The Economic Writings of Sir William Petty, ed. C.H. Hull, Cambridge: Cambridge University Press, 1899, Vol. 1.
Pigou, A.C. 1935. Net income and capital depletion. Economic Journal 45, June, 235-41.
[Quesnay, F.] 1766a. (Premier) Dialogue entre Mr. H. et Mr. N. Journal d'Agriculture, du Commerce et des Finances, June, 61-109. Reprinted in Physicrates, ed. E. Daire, Paris: Guillaumin, 1846.
[Quesnay, f.] 1766b. Observations sur l'intérèt de l'argent (par M. Niasque). Journal d'Argiculture, du Commerce et des Finances, June, 151-71.
Rae, J. 1834. Statement of Some New Principles of the Subject of Political Economy. Boston: Hilliard Gray & Co.
Ricardo, D. 1817. On the Principles of Political Economy and Taxation. London: Murray. In The Works and Correspondence of David Ricardo, ed. P. Sraffa, Cambridge: Cambridge University Press 1951, vol. 1.
Robinson, J. 1954. The production function and the theory of capital. Review of Economic Studies 21, 81-106. Reprinted in Collected Economic Papers of Joan Robinson, Vol. II, Oxford: Blackwell.
Robinson, J. 1956. The Accumulation of Capital. London: Macmillan.
Salin, E. 1930. Kapitalbegriff und Kapitallehre von der Antike zu den Physiokraten Vierteljahrsschrift für Sozial- und Wirtschaftsgeschichte 23, 401-40.
Samuelson, P.A. 1947. Foundations of Economic Analysis. Harvard Economic Studies 80, Cambridge, Mass.: Harvard University Press.
Samuelson, P.A. 1948. Economics: An Introductory Analysis. New York: McGraw Hill.
Say, J.B. 1814. Traité d'économie politique. 2nd edn, Paris: Deterville.
Schneider, E. 1934. Theorie der Produktion. Vienna: Springer.
Schumpeter, J.A. 1911. Theorie der wirtschaftlichen Entwicklung. Leipzig: Duncker & Humblot.
Senior, N.W. 1836. (An Outline of the Science of) Political Economy. London: Griffin.
Smith, A. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. London: Strahan and Cadell. Ed. R. H. Campbell and A.S. Skinner with W.B. Todd, The Glasgow Edition of the Works and Correspondence of Adam Smith, Vol. II, Oxford: Clarendon Press, 1976.
Sraffa, P. 1960. Production of Commodities by Means of Commodities. Cambridge: Cambridge University Press.
Stackelberg, H. von. 1932. Grundlagen einer reinen Kostentheorie. Vienna: Springer.
Stigler, G.J. 1941. Production and Distribution Theories. New York: Macmillan.
Strigl, R. von 1934. Kapital und Produktion. Vienna: Springer.
Swan, T.W. 1956. Economic growth and capital accumulation. Economic Record 32, 334-61.
Thünen, J.H. von 1850. Der isolierte Staat. Theil II, 1. Abteilung: Der naturgemässe Arbeitslohn und dessen Verhältnis zum Zinsfuss und zur Landrente. Rostock: Leopold. Trans. by B.W. Dempsey as The Frontier Wage, Chicago: Loyola University Press, 1960.
Turgot, A.R. J. 1770. Réflexions sur la formation et la distribution des richesses. Ephémérides du Citoyen, November and December 1769, January 1770. Trans. as 'Reflextions on the Formation and Distribution of Wealth' in Turgot on Progress, Sociology and Economics, ed. R.L. Meek, Cambridge: Cambridge University Press, 1973.
Uzawa, H. 1958. A note on the Menger-Wieser theory of imputation. Zeitschrift f¨r Nationalökonomie 18, 318-34. Viner, J. 1930. Cost curves and supply curves. Zeitschrift für Nationalökonomie 3, 23-46.
Walras, L. 1874-7. Eléments d'économie politique pure. Lausanne: Corbaz. Trans. as Elements of Pure Economics. London: George Allen 1954.
Wicksell, K. 1893. Über Wert, Kapital und Rente. Jena: Fischer. Trans. as Value, Capital and Rent, London: George Allen & Unwin, 1954.
Wicksell, K. 1901. Föreläsningar i Nationalekonomi, Första Delen: Teoretisk Nationalekonomi, Lund: Berlingska Boktryckeriet. Trans. as Lectures on Political Economy, Vol. I: General Theory, London: Routledge, 1934.
Wicksteed, P.H. 1894. An Essay on the Co-ordination of the Laws of Distribution, London: Macmillan.

Source:
The New Palgrave: A Dictionary of Economics.
edited by John Eatwell, Murray Milgate, Peter Newman, London, New York, Tokyo:The Macmillan Press Limited, The Stockton Press, Maruzen Company Limited; 1987. Vol. 1, pp 327 - 333. Taken from the Volume in the Public Library Amsterdam, Prinsengracht 587.