1.1 Universal versus Social-General theories and tendencies versus trends

For a start I want to have two issues out of the way. The first is that tendencies are not trends. A trend is a statistical device imposed on or/and observed from empirical figures. On the other hand tendencies might be, but need not be, causative for trends. This needs emphasising because some current mainstream economists as well as some philosophers of science (e.g. Popper 1957) mix up tendency and trend.

The second is that, in this essay, I am not concerned with so called ‘trans-historical laws’ (sometimes, confusingly, called for short ‘historical laws’), I make a distinction between (1) trans-historical universal theories (‘all human beings are mortal’), (2) historicist theories (‘feudalism necessarily develops into capitalism’) and (3) social-general theories (‘within the domain of capitalism: if prices go up demand slows down’). All natural scientific theories are in fact of a trans-historical kind (evolutionary theories might be classified separately). Some psychological and social theories might be trans-historical. By their aims e.g. Freud’s, and Maslow’s theories are trans-historical. Much of neoclassical economics is by its aim trans-historical (as against the modern neoclassical institutionalism).

Much of the ‘floor’ for the mainstream discussion of laws in the social sciences (including economics) has been set by Popper’s The Poverty of Historicism (1957). For the time being I share his queries about historicist theories in this work. However, in taking ‘trans-historical’ (i.e., universal) theories for the prototype of scientific theories generally, I think he goes much too far. Many (if not most) empirically interesting social scientific theories are of the ‘social-general’ type.

This needs emphasising, because my ‘tendency case’, discussed in Section 2, is of a social-general type theory. In his Capital Marx sets out a theory that is particular to the capitalist mode of production.³

1.2 Tendencies: laws as ‘normic laws’

In the Introduction I referred to problems of the application of a positivist methodology in economics (either in a verificationist or falsificationist variety). Since the 1980s such problems have been well documented in the economics methodology literature. Some years before that, Bhaskar (1975, 1979) provided a rigorous critique of empiricist positivism, centering the discussion on the notions of law and tendency. The kernel of this critique is rather simple.

The foremost problem lies in positivism’s bequest of the Humean concept of law, that is, that laws are constant conjunctions of events (plus some disputed contribution of mind). Though a constant conjunction of events is not always considered as a sufficient condition for a law, it is generally considered as at least a necessary condition. Related to this concept of law is the notion that laws find phenomenal expression as events or states of affairs, and that only the phenomenal is real (Bhaskar 1975, p. 64; 1979, p. 158).⁴

Thence the first principle of the positivist account of science is “the principle of empirical-invariance, viz. that laws are or depend upon empirical regularities”. From this derive theories of causality, explanation, prediction, the symmetry of prediction with explanation, the development of science, etc. The second is “The principle of instance-confirmation (or falsification), viz. that laws are confirmed (or falsified) by their instances” (Bhaskar 1979, p. 159; cf. 1975, p. 127). From this derive various theories of demarcation and scientific rationality.

The kernel of Bhaskar’s critique lies in his application of the distinction between closed and open systems. In the natural sciences (apart from astronomy) experimental situations have the character of closed systems, and it is only in such situations that a constant conjunction of events can occur. Outside it, in the open system of the “real world”, disturbing or counteracting forces operate. (Thus, for example, the law of gravity will only be related to a constant conjunction in cases where there are no disturbing factors.) Laws then must either be restricted to closed systems (whence they are not universal or general laws), or the empirical status of laws in open systems must be doubted. In the first case the question is why the empirical should be privileged in closed systems:

Thus the argument is that from the perspective of empiricism there are no universal or general laws. But even if pure positivism is inapplicable, could not a pragmatic attempt be made to apply their criteria? – as some methodologists have in fact claimed (for example Klant 1972; 1984 and Blaug 1980; 1992). This would be decisive to the extent that with the positivist methodological criteria the object of study would get reduced to, or identified with, its empirical manifestations (cf. Bhaskar 1979, p. 167).

For science to be an intelligible activity, Bhaskar argues, the invariance principle must be dispensed with.⁵ For Bhaskar the status of laws in both the natural and the social sciences is similar. They are tendencies, or as he also calls them: normic laws.⁶ The crucial point is that laws are not open system empirical regularities and that open system empirical regularities are not laws: On the one hand, counteraction both by different laws and by accidental/contingent events may prevent the phenomenal expression of normic laws. On the other hand empirical regularities may be the joint outcome of the operation of several laws, or indeed be accidental or contingent; that is there may be empirical regularities for which there is no natural or systemic necessity.

But the similar status of laws in the natural and the social sciences (naturalism) does not imply that social objects can be studied in the same way as natural objects (scientism). The point, mentioned in almost every elementary economics textbook, is that the social sciences do not have the opportunity (ontologically or because of moral objections) to experiment. The domain of social science is not more complex than that of the natural sciences: they are both open systems; ceteris paribus and probability are not the unique inventions of economists. Nevertheless, natural scientists, if their conceptual apparatus is similar, may reach agreement on the events produced in an experimental situation (on their relevance and interpretation, different schools may diverge). However, the absence of closed systems in the social sciences seems to imply that there are no decisive test situations for social scientific theories. If that is so, the conclusion to draw from this is not that there are no social laws. There are laws in both the natural and the social sciences but they are tendential. The difference between natural and social science is that the conditions for the identification of laws are different (Bhaskar 1979, p. 163).

In summary: If one identifies laws with constant conjunctions of events then there are no non-superficial general laws, in natural science or in social science. This does not imply that there can be no laws (this would only be the case if the invariance principle were taken as axiomatic); it only implies that laws (which are always tendential) are not immediately manifest in open systems.

1.3 Tendencies: powers, effects and phenomenal results

This concept of laws as tendential laws, expounded in the previous subsection, appears to be akin to that of J.S. Mill. Bhaskar (1979, p. 161) seems to deny this; of course, the philosophical foundation of such a concept may be different for them. For those methodologists writing on the issue of tendencies J.S. Mill is a common reference point. In this subsection, in order to further clarify different notions of tendency, as briefly set out in the Introduction, I will reformulate in my own words what I take to be the gist of Mill’s view on tendencies in his 1836 Essay on method (reprinted in his 1844 collection of Essays). My reformulation is restricted to some points relevant to the current discussion. For reasons that will become clear later on, I will make a strict difference between a result and an effect of a tendency. Mill does not always make this strict difference (cf. 1836, pp. 337–8). Nevertheless this is an interpretation of Mill.

In the essay Mill seems to make no distinction in ontological status between the natural and the social to the extent that in both domains certain ‘powers’ or ‘forces’ are operative that produce results (phenomenal). Some results are more complicated than others in that many different powers rather than one or a few different powers are operative so as to produce a result. If we had a full picture of the world⁷ we would have for the sum of all powers (P) and results (R):

For each result taken in isolation (e.g. the Result numbered 127) we would know its cause or causes, for example,

Now step aside from this case of a full picture. Suppose that in reality we already have grounds to know that P(i) is an operative causal power (borrowed from other sciences or ascertained via induction from within political economy), though we do not know (all of) ‘its’ results. We don’t even know if there exists at all a result (phenomenal) produced by just this one cause. Then let us take P(i) in isolation (because we don’t have a full picture and we wish to study causes one at a time – Mill 1836, p. 322). Now suppose we have information about the working of P(i). In this case we may have grounds to argue that P(i), in isolation, tends to produce or tendentially produces an effect F(j):

(where -t→ stands for tendency; or perhaps rather ‘tendential operation’)

Note that both effects and results are occurrences even if we may not be able to perceive the effect (thus there may not be an immediate empirical counterpart for an effect). Note also that there is no principal difference between the latter case of isolation (b) and the former full picture case (a). In fact the full picture case (a) should have been written as:

In the case of result R(127), in representation (b), we had three different causes P2, P7, P8 each producing a tendency towards some effect, perhaps counteracting each other, the outcome of which is result R(127).

Thus we had, for example:

In view of the ontological status of tendencies (see below), it is to be noted that for Mill effects may not be less true than results: “That which is true in the abstract, is always true in the concrete with proper allowances.” (Mill 1836, p. 326)

So far my interpretation of Mill.

In discussing tendencies and the related concepts, one may first of all make a difference between their status as either epistemological or ontological or onto-epistemological (by the latter I refer to those philosophies (of science) that principally do not want to make a separation between epistemology and ontology – e.g. Hegelian dialectics). If one makes a separation between epistemology and ontology, then the least problematic of the ‘tendency related concepts’ is that of the phenomenological empirical ‘result’. One might claim existence for the latter, even if we might not be able to perceive it in the absence of thought and theory and their cultural mediation (e.g. snow and sorts of snow, or a rate of profit and sorts of rates of profit). ‘Results’ have at least an onto-epistemological existence. In an onto-epistemological framework ‘results’ and all the other concepts – powers, tendencies and effects – are in fact all equally problematic. These problems are ‘solved’ in the presentation of their systematic connection – which is a presentation of both theoretical and methodological content. (For the purposes of this essay I will not dwell any more on this onto-epistemological position – see Reuten and Williams 1989, Part One; Smith 1990; 1993; Arthur 1993.) I will proceed by making an (as if) separation between epistemology and ontology.

Far more problematic then, is the status of the concepts of power, tendency and effect. Least problematic is the notion of ‘power’. Usually one makes ontological existence claims for ‘powers’ (forces, motives), that is, if one has at least some aim for explanation. What about tendencies and effects? It is clear that one can be an ontological realist for powers and results, whilst merely allotting an epistemological status for tendencies and effects (they are merely theoretical devices). Ruben (1982, pp. 49–56) claims that this holds for Marx. In his interpretation of Marx, tendencies are merely theoretical simplifications in the face of a lack of

In terms of the representations above, Ruben’s interpretation would mean that Marx might agree with representation (a), though not with representation (d). Nevertheless, Ruben writes, “At any given stage in our acquisition of knowledge, we may have to accept laws of tendencies as the best we can do for the present …” (p. 55).

Anticipating the next section (§2) I can merely say that Ruben’s interpretation is interesting, but that other interpretations equally fit the text. Marx (1894F, p. 318) uses, for example, the phrase of “an actual tendency of capitalist production”.

Mill indeed seems an ontological realist about tendencies and effects. (As is Bhaskar – the latter, however, conceives of effects and results as in different ontological layers.)

These issues of ontology make up half the story of the different notions of tendencies. The other half is about the proper ‘place’ for ‘tendency’ in a representation similar to (e) above. There are two main possibilities here. (Note that in my Mill representations I did not want to take sides as to these possibilities. It is even possible that Mill takes in fact a third, intermediate position, between the two ones indicated below.)

First. In representation (c) (and d) tendencies may be conceived of as the operators of the powers. Thus tendencies ‘belong’ to powers. (It is even stronger to say that, inherently powers are always tendential in character.) This may, more explicitly, be represented as:

Operating ‘through’ the Effect in a Result (on which more than one power operate – as in representation (e) – all or some of which we may not know), we have

This is what I call the power notion of tendency, or sometimes the tendency as power notion. It seems almost inevitable that this notion involves an ontological claim of a real existence of a tendency. (Or of a power being inherently tendential in character.)⁸ The Result in this case may, of course, diverge from the Effect if we had more powers operating on the result.

A second notion of tendency, tendency as expression, or tendency as outcome, allots tendency to the Result. This may be represented as:

Although this may involve an ontological claim about the existence of the Effect (F), this need not be the case. The Effect may indeed be a theoretical device (as Ruben, we have seen, interprets Marx).

Finally, with respect to Marx there is at least one further difficulty, related to the fact that Capital describes a multiple conceptual structure. When Marx uses the term ‘expression’ this is often the expression of a force (power) at that conceptual moment in his presentation, perhaps at a still abstract level. Thus we may in fact have a ‘strain’ of forces and expressions. Apparently ‘expression’ would then be rather similar to Effect. However, expressions are often ‘preceded’ by an ‘immediate’ operation of the force/power, which is rather more similar to Effect. The term for a more final empirical ‘result’ is the term ‘manifestation’. There is, of course, not much point in spelling this further out before we get to the textual analysis, but the easiest way to pre-empt this is to say that in an equivalent of representation (h) there may follow a string of Effects and Results, of which the final term is a ‘manifestation’ of a power or a string of powers.

2.1 Chapter 13, The Law Itself

This chapter covers about twenty pages (1894F, pp. 318–38). It opens with a numerical example in which at a constant rate of surplus-value (e = s/v) and a rising composition of capital (c/v), the rate of profit (r = s/(c+v) or R/(c+v)) is shown to decline. Rewriting the expression in the usual way (Marx does not do this) we have:

r = s/(c+v) = (ev)/(c+v) = e/(c/v + 1) (1)

After a comment of about one page the concept of tendency turns up. As we will see, the German text seems ambiguous as to the exact meaning of the term tendency. The Fernbach and the Untermann translations apparently take different positions here. Marx writes (Fernbach’s translation of 1981):

[*] The hypothetical series we constructed at the opening of this chapter therefore expresses the actual tendency of capitalist production {die wirkliche Tendenz}.¹⁰ With the progressive decline in the variable capital in relation to the constant capital, this tendency leads to a rising organic composition of the total capital, and the direct result of this is that the rate of surplus-value, with the level of exploitation of labour remaining the same or even rising, is expressed in a steadily falling general rate of profit [**]. (We shall show later on (ch. 14) why this fall does not present itself in such an absolute form, but rather more in the tendency to a progressive fall.) The progressive tendency for the general rate of profit to fall is thus simply {nur} the expression, peculiar to the capitalist mode of production, of the progressive development of the social productivity of labour. This does not mean that the rate of profit may not fall temporarily for other reasons as well, but … (Marx 1894F, pp. 318–19)

Thus we have a tendency (T) of the capitalist mode of production (CMP), which has an immediate effect (F) to an expression (E):

CMP[T] → c/v↑ → (F): eEr↓ (2)

The tendency seems to be the power or the operation of the power (see representations f and g in the previous subsection §1.3). The tendency [T] leads to a rising value composition of capital (c/v), the immediate effect (F) of which is that the rate of surplus-value (e) gets expressed (E) in a falling rate of profit (r).

In the translation from the German of the second sentence in the quotation above, there is an important difficulty of interpretation (see the passage from * to **) relating to one kernel of the concept of tendency (power versus expression). The German text reads:

[*] Die ⟨Der⟩ im Eingang hypothetisch aufgestellte Reihe ⟨Fall⟩ drückt also die wirkliche Tendenz der kapitalistischen Produktion aus. Diese ⟨Sie⟩ erzeugt [‘produces’ – what is it that produces: the tendency, or capitalist production?] mit der fortschreitenden relativen Abnahme des variabelen Kapitals gegen das konstante eine steigend ⟨fortwährende⟩ höhere organische Zusammensetzung des Gesamtkapitals, deren unmittelbare Folge ist, daß die Rate des Mehrwerts bei gleichbleibendem und selbst bei steigendem Exploitationsgrad der Arbeit sich in einer beständig sinkenden allgemeinen Profitrate ausdrückt. [**] (Marx 1894G, pp. 222–3; 1894M, p. 287)

The German text (diese erzeugt) leaves room for another interpretation from Fernbach’s, that is, the mode of production produces the rise in the organic composition of capital, and this is expressed in the (tendental) fall of the rate of profit. The 1909 translation by Untermann follows this interpretation:

[*] The hypothetical series drawn up at the beginning of this chapter expresses, therefore, the actual tendency of capitalist production. This mode of production produces a progressive relative decrease of the variable capital as compared to the constant capital, and consequently a continuously rising organic composition of the total capital. The immediate result of this is that the rate of surplus-value … is represented by {expressed in} a continually falling general rate of profit [**]. (Marx 1894U, pp. 212–13)

In my opinion the Untermann translation is the superior one as it fits the remainder of the quotation from ** onwards (esp. the italicised bit – see the quote from the Fernbach translation (first quotation of §2.1) which from here on is rather similar to the Untermann translation). Thus we seem to have a ‘tendency as an expression’ notion. The capitalist mode of production (CMP) inherently produces an increasing social productivity of labour (prodtt) and this gets expressed in a tendental fall of the rate of profit (r).

[CMP: prodtt↑] → c/v↑ → (F): e(E)r↓[T*] (3)

(Cf. representation h in §1.3). It remains to be seen if representation (3) is consistent with the further text of this and the next two chapters.

The larger part of the current chapter (ch. 13, the law as such) is devoted to the concomitance of a (tendential) decline in the rate of profit and a rise in the mass of profit. This concomitance is stressed over and again. We have a

double-edged law of a decline in the profit rate (r) coupled with a simultaneous increase in the absolute mass of profit (R), arising from the same reasons {causes}. (Marx 1894F, p. 326; 1894G, p. 230)

Apart from in the quote above the term tendency appears only twice in this chapter. One passage reads:

Thus the same development in the social productivity of labour is expressed, with the advance of the capitalist mode of production, on the one hand in a progressive tendency for the rate of profit to fall and on the other in a constant [beständigem; i.e continuous] growth in the absolute mass of the surplus-value or profit appropriated; so that by and large [im ganzen], the relative decline in the variable capital and profit goes together with an absolute increase in both. (Marx 1894F, p. 329; Marx 1894G, p. 233)

Again: the tendency seems the expression (now coupled, though, with a second expression). This may be represented as:

This is consistent with representation (3), that is the Tendency as Expression interpretation.¹¹

This concludes the kernel of ‘the law as such’ or, as Marx also calls it, “the general law” (e.g. 1894F, p. 339). For the next chapter (14) it is useful to somewhat further spell it out in terms of representations (3) and (4). In my reading this law is not merely about the falling rate of profit itself (its tendential fall). The general law of the CMP is the following (see the first three pages of ch. 13 from which I have quoted above):

α. The CMP brings about a (progressive) increase in the social productivity of labour (the production of absolute and relative surplus-value as expressed in the rate of surplus-value e).

β. A dominant way of realising this (α), is by increasing the rate of surplus-value concomitantly on increasing the organic composition of capital.

γ. Its (α and β) immediate effect {Folge} is in a twofold expression, that is in: (a) a fall of the general rate of profit and (b) a rise in the social mass of profit.

This may be represented as:

However, the law does not just operate in an immediate manner. (This is about α and β and γ – not merely about the latter or about the latter two.) So the law has the character of only/merely/just {nur} a tendency (cf. Marx 1894F, p. 319; 1894G, p. 223).

While in the chapter at hand the term tendency has not been used more than the three times referred to, it is nevertheless more often implicitly referred to as in e.g.:

Viewed abstractly, the rate of profit might remain the same … (Marx 1894F, p. 336; 1894G, p. 239; 1894M, p. 319)

or:

The rate of profit could even rise, if a rise in the rate of surplus-value was coupled with a significant reduction in the value of the elements of constant capital, and fixed capital in particular. In practice, however, [!] the rate of profit will fall in the long run, as we have already seen. (Marx 1894F, p. 337; 1894G, p. 240; cf. 1894M, p. 319)

This last statement “in practice … already seen” is remarkable indeed. Is is a rather definite statement about the expression of the law – or even its empirical manifestation. (I have found it peculiar considering especially the status of this chapter, i.e., prior to the theory about the counteracting causes that affect on the rate of profit.) It is interesting then to find that this particular sentence is not in Marx’s manuscript.¹²

2.2 Chapter 14, Counteracting Factors

The chapter on the ‘counteracting factors’{causes}, covers about ten pages (Marx 1894F, pp. 339–48). It opens with an empirical observation, which is followed by a passage that is crucial to the interpretation of what Marx means by a tendency.

If we consider the enormous development in the productive powers of social labour over the last thirty years {i.e., 1835–65} alone, compared with all earlier periods, and if we consider the enormous mass of fixed capital involved in the overall process of social production quite apart from machinery proper, then instead of the problem that occupied previous economists, the problem of explaining the fall in the profit rate, we have the opposite problem of explaining why this fall is not greater or faster. Counteracting influences must be at work, checking and canceling {aufheben} ⟨durchkreuzen⟩ the effect {Wirkung} of the general law and giving it¹³ simply {nur} the character of a tendency, which is why we have described the fall in the general rate of profit as a tendential fall. The most general of these factors {Ursachen} are as follows. (Marx 1894F, p. 339; 1894G, p. 242; 1894M, pp. 301–2)

This text sustains my interpretation of the general law as a tendency law (representation 5, as paricularly comprising all the elements α and β and γ).¹⁴ So we have the general law (ch. 13), which appears to be a tendential one because counteracting influences operate (ch. 14).

The causes are next commented upon under separate headings:

1. More intense exploitation of labour. Concerning an increase in the rate of surplus-value (e), we may distinguish: either such an increase concomitant on a rise in the composition of capital (c/v), with c increasing and v decreasing; or, such an increase independent of an increase of c (with c/v rising merely as a result). This section is about the latter. With e.g. increasing intensification of labour (or prolongation of the working day) one labourer works up more means of production (c), therefore e rises and, for the same amount of capital, the amount of labour decreases. Hence for a given capital less labour is being exploited more intensively. For a given capital, profit, or the mass of surplus-value, s = ev. Each of the two factors on the righthand side, if I am right, are called contrary tendencies by themselves. This point, as I will show, is important to the general interpretation of ‘the law’.

It has already been shown, moreover, and this forms the real secret of the tendential fall in the rate of profit, that the procedures for producing relative surplus-value are based, by and large, either on transforming as much as possible of a given amount of labour into surplus-value or on spending as little as possible labour in general in relation to the capital advanced; so that the same reasons {Gründe} that permit the level of exploitation of labour to increase make it impossible to exploit as much labour as before with the same total capital. ⟨The same number of labourers is being exploited more, but a decreased number of labourers is being exploited by the same capital.⟩ These are the counteracting {widerstreitenden} tendencies which, while they act to bring about a rise in the rate of surplus-value, simultaneously lead [act] to a fall in the mass of surplus-value produced by a given capital, hence a fall in the rate of profit. (Marx 1894F, p. 340; 1894G, p. 243; 1894M, p. 302)

The last sentence is puzzling. First, ‘counteracting’ seems rather: tendencies that counteract each other (rather than tendencies that act counter to an original tendency). Second, the bit after the comma is perhaps confusing: the fall in r is not the conclusion of the sentence. Rather the ‘widerstreitenden’ tendencies operate on the rate of profit in a non-uniform way.¹⁵

So it seems now that we have two influences (of the same offspring) that counteract the law; and this gives the law a tendential character.

Apart from this we see here introduced a theme that we will meet throughout this chapter, which is that tendency and counteractions (or again counteractions by themselves) are discussed in terms of one and the same offspring.

Thus we had in the previous chapter, for the total social capital, a fall in the rate of profit, together with an increase in the mass of profit (due to accumulation of capital and a social rise in variable capital). Now, looking at a given capital amount we see e rising and s falling (with v going down). For a conclusion we have:

It does not annul {aufheben} the general law. But it has the effect {er macht} that this law operates {wirkt} more as a tendency, i.e., as a law whose absolute realisation is held up, delayed and weakened by counteracting factors {gegenwirkende Umstände}. … the same factors tend {streben} both to reduce the rate of profit and to slow down the movement in this direction. (Marx 1894F, pp. 341–2; 1894G, pp. 244–5)

This quotation, in combination with the first one provided from ch. 14 (Marx 1894F, p. 339), reveals either an inconsistency or a subtle differentiation. In the earlier quotation counteracting influences were said to cancel {aufheben} the law’s operation, hence its tendency character. Now the ‘widerstreitenden’ influences do not annul {aufheben}, but merely weaken the general law, hence the latter’s operation as a tendency.

I now turn to the remaining ‘counteracting factors’ on which, in the context of this essay, I can be briefer.

2. Reduction of wages below their value (1894F, p. 342). This is a section of two sentences only.

We simply make an empirical reference to this point here, as … it has nothing to do with the general analysis of capital … It is none the less one of the most important factors in stemming [aufhalten] the tendency for the rate of profit to fall. (Marx 1894F, p. 342)

This is a contingent element, i.e., one exogenous to this law. It is characteristic for Marx to make such an empirical reference.

3. Cheapening of the elements of constant capital (1894F, pp. 342–3). This very important factor is about the price effect of productivity increase on the value of constant capital, whence the change in the ‘technical’ composition of capital is not translated in an ‘aliquot’ change in the value composition. (On the general theoretical discussion of this issue, see Moseley 1992, ch. 1.) This issue is related to

… the devaluation of existing capital. This too is a factor that steadily operates to stay {aufhalten} the fall in the rate of profit … We see here once again how the same factors {Ursachen} that produce the tendency for the rate of profit to fall also moderate the realisation of this tendency. (Marx 1894F, pp. 342–3)

4. The relative surplus population (1894F, pp. 343–4). This section is about the retardation of the rise in c/v, brought about by a rise in unemployment caused by any previous rise in c/v.

The creation of such a surplus population is inseparable from the development of labour productivity and is accelerated by it, the same development as is expressed in the decline in the rate of profit. … here again the same reasons that produce the tendential fall in the rate of profit also produce a counterweight to this tendency, which paralyses its effect {Wirkung, operation} to a greater or lesser extent. (Marx 1894F, pp. 343–4)

5. Foreign trade (1894F, pp. 344–7). The cheapening of prices by foreign trade may affect the rise in the organic composition not being translated (to the same extent) in the value composition (c/v).

2.3 Conclusions to the chapter on the counteracting factors

Marx’s conclusion is:¹⁶

We have shown in general, therefore, how the same causes that bring about a fall in the general rate of profit provoke counter-effects {Gegenwirkungen hervorrufen, i.e., call forth counteractions} that inhibit this fall, delay it and in part even paralyse it. These do not annul {aufheben} the law, but they weaken its effect {wirkung, operation}.¹⁷ … The law operates therefore simply as a tendency, whose effect {Wirkung} is decisive only under certain particular circumstances and over long periods. {So wirkt das Gesetz nur als Tendenz, dessen Wirkung nur unter bestimmten Umständen und im Verlauf langer Perioden schlagend hervortritt.}⟨dessen Wirkung nur unter bestimmten Umständen und auf lange perioden ausgedehnt schlagend hervortritt.⟩ (Marx 1894F, p. 346; 1894G, p. 249; 1894M, p. 308)

By now the possibility for another interpretation of Marx’s notion of tendency is gradually being revealed. We have seen here and before that in the German text the term ‘Wirkung’ (operation, action) is used consistently to describe the law. (In the English text this is almost consistently translated by ‘effect’.) It would perhaps go too far to reverse back to the ‘power notion of tendency’, nevertheless something like ‘operators of powers’ (see the comment just before and after representation (f) in §1.3) or at least operation of powers (causes) seems at stake.

After all there seem two tendential elements in the operation law. The first is that for various reasons internal to the law (endogenous reasons) an increase in the rate of surplus-value (e) (either that independent of c/v, or that concomitant on c/v) dominates over the effect of c/v on the rate of profit. Hence r↓[T*]. The second is that for various reasons – again internal ones – the value composition of capital (c/v) may in fact not rise. It may seem attractive then to represent the tendential character of the general law as follows:

However, for the operation of the tendential element [e↑ ↔ c/v ↑][T] these have explicitly been called counteracting tendencies in the case of s = ev (section 1 of ch. 14) only; but even here not in a clear cut way. The term ‘widerstreitenden’ tendencies may in fact refer to merely these themselves only (i.e., e and v). In the other sections Marx writes rather in terms such as counteractions, not in terms of counteracting tendencies. From this one might infer that representation (6) is wrong. We are stuck, however, with an ambivalence as to the meaning of ‘Wirkung’ (operation, action).

Finally there is the issue of the empirical manifestation: “The law operates therefore simply as a tendency; it is only under certain particular circumstances – stretched over long periods – that its operation comes to the fore in an articulate way.” (This is what I make, in translation, of Marx 1894M, p. 308.) Does this mean that this ‘tendency law’ in the long run results in a fall of the rate of profit? This is far from obvious from the quotation. It seems rather that: within a sufficiently long span of time there will always occur a constellation of circumstances for which the rate of profit will actually fall; at other constellations, however, the rate of profit might rise. (Nowhere in the text is there, to be sure, a statement about the average development in the rate of profit). For the time being, therefore, r↓[T] in representation (5) seems not operational.

2.4 Chapter 15, Development of the Law’s Internal Contradictions

This very perceptive chapter comprises about 25 pages (1894F, pp. 349–75). Again, I merely pick out the explicit references to the concept of tendency: there are only few here. Apparently, as we will see, this chapter is not very telling about the notion of tendency. Apparently, then, my current §2.4 cannot add much to the earlier conclusions. However, as I will indicate in my general conclusions, the fact that the term tendency is used so scarcely in this chapter is telling. The chapter is in 4 sections:

1. General considerations (1894F, pp. 349–55). This is a general summary of the process of production in reference to Capital I. It provides comments on Ricardo’s treatment of the issue.

2. The conflict between the extension of production and valorisation (1894F, pp. 355–9). In summary the argument in this section runs as follows. First. “As the capitalist mode of production develops, so the rate of profit falls” (p. 356) – as argued for in chapter 13. This fall would be counteracted by a decrease in the value (cheapening) of the components of capital (either variable capital ↔ increase in relative surplus-value; or constant capital) (argued for in chapter 14). This cheapening gives rise to the devaluation of the existing capital. The latter conditions the fall in the rate of profit, and delays it. Second. The mass of labour that capital can command does not depend on its value but rather on the mass of raw and ancillary materials, of machinery and elements of fixed capital, and of means of subsistence, out of which it is composed, whatever their value may be.

These (the first and second point) are two moments of the accumulation process: “sie schliessen einen Widerspruch ein, der sich in widersprechenden Tendenzen und Erscheinungen ausdrückt. Die widerstreitenden Agentien wirken gleichzeitig gegeneinander.” (Marx 1894G, p. 259; cf. 1894F, p. 357.) This may be translated as: ‘they contain a contradiction that is expressed in contradictory tendencies and phenomena. The antagonistic agencies act simultaneously in opposition to one another’. (Note that this is consistent again with the ‘tendency as expression’ notion.) We have simultaneously: impulses to increase and decrease of the working population; decrease in the rate of profit and devaluation of capital which puts a stop to this fall; development of productivity and a higher composition of capital.

These factors may at one time assert themselves side by side in space, and at another assert themselves in time one after the other; periodically “the conflict of antagonistic agencies finds vent in crises. Crises are never more than momentary violent solutions for the existing contradictions, violent eruptions that re-establish the disturbed balance for the time being.”¹⁸

In the next two sections the term ‘tendency’ does not turn up any more, apart from in a comment on prevailing economic theory (1894F, p. 366). Section 4 is entitled “Supplementary Remarks” (1894F, pp. 368–75). In the face of the empirical manifestation of the law, section 3 deserves consideration:

3. Surplus capital alongside surplus population (1894F, pp. 359–68). This section contains the presentation of the process of economic crisis in terms of e.g.: concentration, overaccumulation, and devaluation of capital; capital to lie idle or destroyed, breakdown of the credit system, stagnation {Stockung} in production, fall in wages. For our purposes the following quote is important:

The stagnation in production that has intervened prepares the ground – within the capitalist limits – for a later expansion of production. And so we go round the whole circle once again. (Marx 1894F, pp. 363–4; 1894G, p. 265)

Thus it seems that the fall in the profit rate is a periodical matter rather than a trend-like phenomenon (as quite some commentators on Marx have interpreted him). Of course, the “missing sentence” in Marx’s manuscript, referred to at the end of my §2.1, backs up my conclusion.

3.1 General conclusions to the case

Chapter 15 of Marx’s Capital III has presented the ‘tendency law’ in its cyclical expression, that is in economic crisis/stagnation. Especially the devaluation of capital and the destruction of capital in crises/stagnation is highlighted – the effect of which is a rise in the rate of profit: “And so we go round the whole circle once again.”

Why is it that in this last chapter the term tendency has been so scarcely used by Marx? Here the previous two chapters come together. If Marx’s were a ‘power notion of tendency’ then we might expect that term to have been used over and again in that chapter 15. For the ‘expression notion’ this may be different. In fact chapter 15 is about expressions throughout. However, these are the expressions of a complex. Here it is shown how the rate of profit does not fall continuously, but rather cyclically. Thus, we simply do not have the, say, unilateral profit fall of chapter 13. In chapters 14 and 15 we see why the rate of profit does not fall continuously, and why, therefore, chapter 13 must formulate a ‘tendency law’.

This backs up my first conclusion: ‘tendency’ refers only to ‘the general law’ of chapter 13 – so it is used merely in reference to that general law. There are no ‘countertendencies’. There are, however, counteracting causes. The latter have been traced, where relevant, to the same offspring of the forces behind the tendential fall in the rate of profit itself – thus the interconnection has been sought.

There are, nevertheless, two exceptions to this. One apparent and one inconsistent to my conclusion. The first has been dealt with in §2.3 (e and v as counteracting themselves). The second is the phrase ‘contradictory tendencies’ (1894F, p. 357) referred to in §2.4 under 2.

A second conclusion relates to the ambivalence as to the meaning of ‘Wirkung’ (operation, action) of ‘the law’ – pointed at in §2.3. It seems after all that the term operation is linked to the expression of the law. This is not so surprising if indeed Marx’s is not a ‘power notion of tendency’. If we have a force (non-tendential in itself, or ontologically non-tendential) which is not expressed immediately, then indeed we might well conceive of its ‘working’ as tendential. Therefore, my second conclusion is that a ‘tendency’ in the case of Marx discussed here is not a power or force. Tendency refers to the resulting expression of force (the working/operation of forces being laid down in a law) to the extent that this expression does not work out univocally, but rather in a contradictory way.

A third and tentative concluding remark concerns the ontology-epistemology question referred to in §1.3. In the text studied I have found no evidence at all that Marx’s concept of tendency is an ontological one. It would be rash to conclude from this that his must therefore be an epistemological one (cf. Ruben referred to in §1.3). On the other hand I have also no evidence from this case, that Marx’s notion of tendency is not an epistemological one.

3.2 Marxian theory and empirical research: some further and tentative concluding remarks

How might one do empirical research on the basis of tendency laws? Having considered the case above, one is bound to be left with this query. This issue deserves a full separate paper, nevertheless I will make a few tentative remarks. In order to make sense of empirical research in the light of tendency laws, I believe one is almost enforced to take a position similar to Lawson’s (1989; 1992) reworking of Kaldor’s notion of ‘stylised facts’. Although Lawson adopts philosophically a realist position to the extent that powers and tendencies have ontologically a real existence, this position is not essential to the problem of empirical research that I am considering.

A tendency for Lawson is “a power that may be exercised and yet unrealised in manifest phenomena”. Powers themselves exist “by virtue of certain enduring structures” (1989, p. 62). Laws then are defined similar as with Bhaskar and Mill (§1.2 above). Although the effects of tendencies, he writes,

will frequently be modified or hidden by the operation of irregular countervailing mechanisms and so forth, their persistency coupled with the irregular operation of the countervailing influences may allow their effects to ‘shine through’. And to the extent that any manifest phenomenon appears to reveal some degree of uniformity, generality or persistency … it would seem to provide a prima facie case for supposing that some enduring generative mechanisms are at work. … And conceptualisations of such partial regularities, of course, are the obvious candidates for representation as ‘stylized facts’. … stylized facts can provide an access to enduring things as indications of possible manifestations of the effects of (possibly a combination of) causal tendencies. (Lawson 1989, pp. 65–6)

I believe that a similar procedure is currently the best we have, that is, if we take serious the notion of tendency law without refraining from serious empirical work on their basis. Indeed, as Lawson indicates, Kaldor (after 1966) was concerned about the increasingly formalistic and sterile non-empirical economics of his days – times have not changed in this respect (see e.g. Rosenberg 1992).

I guess that anyone studying chapters 13–15 of Marx’s Capital III (the case of §2 above) cannot be but impressed by the conscientious and thorough exhibition of that theory up to minutest detail (given its methodological level of abstraction). It may also appear a very realistic theory, even for those who would not want to share its value theoretical notions. (Schumpeter, e.g., it seems, owes much to Marx’s approach – see Schumpeter 1943).

Nevertheless, that theory is insufficient and must be developed further. Marx’s Capital is indeed an unfinished project: (a) from the point of view of his own aims (Reuten 1996); (b) in terms of the unclarity of the extent of its theoretical break from classical economics (Reuten 1993); (c) in terms of the unclarity of its methodological break from either ‘analysis’ in Ricardian vein, or dialectics in Hegel’s traces (Murray 1988; Smith 1990, 1993; Arthur 1993a, 1993b). One problem, and challenge, is that the latter two aspects hang together.

Such theoretical development cannot be carried out in the absence of thorough empirical research. Indeed (restricting myself to the theory of the case of §2) I believe that the kind of empirical research as carried out by Weisskopf (1979), Wolff (1986), Moseley (1991), and Duménil and Lévy (1993; 1996) must be undertaken. Especially if these are combined with inter-comparisons and discussion as in Moseley (1991).

Of course, one can have many theoretical objections to such empirical research. Not merely because of their particular operationalisation of the theory and the many ad hoc decisions that one is bound to make in the face of poor statistics. One may foremost object to the empirical application of an incomplete theory (as indicated above). I believe that one must have reservations here. However, at the same time I am convinced that such empirical research must be carried out and cannot await theory development.

One argument for this is in the concern for the reclaiming of a ‘real-world political economy’, which comprises the explanation of stylised facts. Of course, what these latter are is not independent of a (theoretic) discourse. However, their communication and explanation affect the social discourse.

The, related, other argument is in the cross fertilisation of methodological, theoretical and empirical research. I firmly believe that methodological, theoretical and empirical research must be carried out alongside each other, short of the ideal of a really integrated triple approach.