2.1 – Theory of Distribution

8Wicksell made an essential contribution to neoclassical thinking about social policy by applying marginal analysis to the distribution of factor incomes. The core of his theory of distribution, as expounded in the second part of his Lectures on Political Economy (1901), consists of two constituents. First, under the assumptions of a perfectly competitive market setting, it is demonstrated that the prices of the production factors (wages, rent and interest on capital) correspond to the latters’ marginal productivity in general equilibrium. [7] The principle of scarcity, which in classical Political Economy was confined to the determination of rent, is thus extended to determine all factor prices. Second, if factor prices equal marginal productivity, aggregate factor incomes correspond to aggregate output. However, in critique of Walras’s generalization of the theorem of “production exhaustion”, Wicksell demonstrated that this correspondence applies only under constant returns to scale. [8]

9Contrary to standard interpretations of the marginal productivity theory of distribution, Wicksell was fully aware of the possibility that factor pricing in perfect competition does not lead to socially optimal results. In his examination of Ricardo’s analysis of labour-saving technical progress, Wicksell discussed the case of innovations that reduce the marginal productivity of labour, but lead to an increase of aggregate output under full employment (1934, 133-43). This case is illustrated by figure 1, where B represents the new technique and A the older one. [9] In a competitive labour market, the displacement of labour (L_A,0 —> L_B) will reduce the wage rate (w₀ —> w₁). In Wicksell’s view, this will induce firms that use technique A to hire displaced workers and expand their production at the lower wage rate (—> L_A,1). Aggregate employment returns to its prior level, but at lower wage incomes. [10]

Figure 1

Marginal productivity, technical progress and wage formation in Wicksell (1901)

Marginal productivity, technical progress and wage formation in Wicksell (1901)

10Due to his neo-Malthusian concerns about overpopulation Wicksell never subscribed to the belief that a perfectly competitive market system would automatically produce wage incomes at (or above) the level of subsistence. Thus he went even further in his argument about the introduction of new techniques of production:

“Nor is the result any different if we assume that wages are already at the subsistence level (and cannot, according to the usual view, fall lower). In reality, wages can not only be forced below it for a little, but can remain below it indefinitely, if the labourers and their families can make up the difference by poor relief.”

2.2 – Interest and Prices

11Wicksell’s Geldzins und Güterpreise (1898), too, had its background in distributional concerns. In the introduction to this “study of the causes regulating the value of money”, Wicksell (1936, 3) stressed that inflation and deflation are “disturbance[s] to the social mechanism”, and hence “an evil in itself ”. He added that the gains of the implicit redistribution of incomes and wealth do not outweigh the losses. However, in the perspective of Wicksell’s monetary theory the dividing lines between winners and losers are not between production factors, and it is not labour per se that would be in a precarious position:

12

“A general rise in prices is, of course, to the disadvantage of all those who receive fixed money incomes, as is the case to-day with a constantly increasing number of social groups. It is also to the disadvantage of all those who derive the whole, or a large part, of their incomes by lending money capital of one kind or another… Lastly, a general rise in prices is to the disadvantage of labour, so long as it has not the power to enforce a corresponding rise in wages. But it must not be forgotten that a rise in wages may precede a rise in prices, acting as its direct cause. It will indeed appear later that this must be regarded as the most probable procedure whenever the rise in the price level is gradual and permanent, as opposed to those more fortuitous changes which are brought about by speculative buying and the like. That being so, it is not possible, without further qualification, to speak of a rise in prices as causing a general injury to labour.”

13The objective of the study was thus to define the conditions under which money could function as a stable standard of value. Wicksell, who was writing at the times of the gold standard, was far-sighted in stressing that money is a “matter of pure convention”, essentially created by way of bank lending. His theory of cumulative changes in the price level revolves around differences between the “natural rate of interest”, [12] and the market level of the banks’ lending rates. These differences tend to lead to an expansion (contraction) of credit volumes that generates aggregate excess demand (supply) and upward (downward) pressures on the price level. This gap-theory of inflation and deflation led to the conclusion that the value of money can be stabilized by way of a simple rule for monetary policy: The (central) Bank ought to react to any change in the price level by moving the market rate of interest in the same direction, until the price level stops changing. This implied that price level stability could be considered a sufficient condition of monetary equilibrium.

14However, even though distributive concerns with inflation and deflation shine through Wicksell’s lines of argument in Interest and Prices (1898), Lectures on Political Economy, vol. II (1906), and later writings on monetary issues, there is remarkably little analysis of the effects of monetary “disturbances” on factor income shares. The conclusion that workers do not generally suffer from changes in the price level per se, at least not more than other “social groups”, is largely derived from three assumptions that Wicksell makes — often with reference to empirical observation rather than analytical necessity. The first is the assumption of full employment underlying Wicksell’s analysis of interest-rate gaps and changes in the price level (1935, 195 ; 1936, 90 and 143). This is closely connected with the second assumption, the up- and downward flexibility of nominal wages. Wicksell kept this assumption throughout most of his writings with reference to workers’ lack of means of subsistence in the case of unemployment (Boianovsky 1998, 498) ; in modern parlance, he assumed a low “reservation wage level”. This argument is fully consistent with Wicksell’s views of the wage effects of labour-saving technical progress and the need for wage subsidies. The full-employment assumption, in turn, made Wicksell downplay the redistributive and other real effects of cumulative processes of inflation and deflation as non-cumulative and of secondary importance (1935, 195-97 ; 1936, 143-45). He acknowledged that inflation would lead to “forced saving”, in terms of a redistribution of purchasing power from consumers to firms, which reap windfall profits that they can reinvest. But in his theory, there was no room for unplanned investment that could change the capital stock to the extent that the “natural rate of interest” would be affected by the income redistribution following from monetary expansion.

15The third assumption that Wicksell made in rejecting the view of inflation as “a general injury to labour” was the proposition that only unanticipated changes in the price level matter (1935, 129 ; 1936, 3). All expected changes would be taken into account in wage contracts, and workers would not generally suffer from money illusion (see also Boianovsky 1998, 234-36 and 247). It is rather the unexpected changes in profits (windfall gains and losses) that create (temporary) “disturbance in the social mechanism”. They may also cause temporary shifts in the aggregate income shares of workers, but there is next to no analysis of the net effects in Wicksell’s writings. Real profits on non-financial investment tend to rise in cumulative processes of inflation, but generally at the expense of those who hoard money, or of those who receive fixed rent payments in long-term contracts or other fixed incomes (such as civil servants and pensioners). The main disturbance is thus a matter of personal income distribution, not of functional distribution. It is a redistribution of incomes and wealth within the owning classes, through the link between credit and goods markets – and not between capitalists and workers through their links in goods and labour markets. Wicksell argued that, when an interest-rate gap produces a credit expansion which leads to excess demands for goods and labour, the workers will normally be in a position to compensate themselves and thereby, in fact, get the inflationary process going.

16What about deflation? Again, if it is anticipated, as was largely the case in the early decades of the gold standard (1873-96), it should not affect real economic activity. However, if prices are falling rather rapidly, deflation may produce an overproportional contraction of credit due to a devaluation of collaterals ; and it may produce chains of bankruptcies, as the rise in the real debt burden of producers will not be foreseen in all loan contracts. [13] Deflation will then lead to some cutbacks of production and employment. In this situation, workers might accept wage reductions to preserve their job, but as prices will also fall, due to the decline in nominal asset values and the volume of extended loans, the effects on the real wage level, employment and aggregate wage share are uncertain. At any rate, Wicksell does not discuss them in greater detail.

2.3 – The Dichotomy, Austrian Capital Theory and the Wicksell Effect

17Wicksell preached the inclusion of monetary theory into a coherent body of “pure economic theory” (1934, 7), but he practiced the neoclassical dichotomy of real and monetary analysis – where “real” means “not concerned with the influence of monetary variables”. The real sphere of exchange, production and distribution was dealt with in Value, Capital and Rent (1893) and the first volume of the Lectures in Political Economy (1901) ; the monetary sphere was explored in Interest and Prices (1898) and the second volume of the Lectures (1906). It may be argued that Wicksell had good reasons to separate his analysis of the redistributive effects of labour-saving technical progress from his theory of changes in the price level. Technical progress may affect the functional distribution of factor incomes, and the latter may determine the choice of techniques (as discussed in section 2.1). Inflation and deflation tend to change the personal distribution of purchasing power along other dividing lines, not along those of the production factors. In Wicksell’s perspective, the two spheres could nevertheless be regarded as well-connected by his concept of the natural rate of interest, if this is taken to correspond to the equilibrium rate of interest on capital in Lectures I (1901, chap. II.2).

18It may, however, be argued that the link is insufficient and the analysis incomplete, if not flawed. In Wicksell’s analysis, gaps between the natural rate and the market rate of interest lead to investment-saving imbalances, but these imbalances have no significant effects on real variables. By downplaying the redistributive effects of inflation (“forced saving”) on capital formation, Wicksell ignored feedbacks from the money rate of interest to the natural rate. Thereby he missed a possible connection of his monetary theory of interest rates with his repeated, self-critical attempts to turn Böhm-Bawerk’s theory of capital and interest into a consistent theory of distribution (cf. Uhr 1960, chap. IV-VII ; Kurz 2000 ; Sandelin 1998).

19Wicksell’s admiration for Böhm-Bawerk was rooted in the latter’s solution of a conundrum in neoclassical distribution theory. If factor prices are to be explained in terms of scarcity, factor quantities must be determined independently of prices. Land and labour, the originary factors of production, can in principle be measured in terms of their own physical units of space and time. Accordingly, Wicksell developed the first stages of his theory of production and distribution (1901, chap. II.1) by taking recourse to the thought experiment of “non-capitalistic production”, in which land and labour are the only inputs. In this way, he could consistently show that, given a linear-homogenous production function and perfect competition, factor prices correspond to marginal factor productivity. In the case of capital, things are different. Unless the range of goods in the economy is reduced to strictly one single good that serves for investment and consumption at once, “social capital” (i.e. the economy’s aggregate stock of capital) needs to be measured in terms of a unit of account “extraneous to itself ” (Wicksell 1934, 149). The obvious candidate is money, the standard of value. However, “it is futile to attempt… to derive the value of capital-goods from their own cost of production or reproduction ; for in fact these costs of production include capital and interest…We should, therefore, be arguing in a circle” (ibid. — Wicksell’s emphasis). Furthermore, Wicksell (1934, 146) starts out with his theory of capital and interest by pointing out that it is “enigmatic…that the possession of capital, apparently at least, does procure something more, namely a permanent income in the form of interest, either without sacrifice of capital or while capital is constantly being replaced”.

20Referring to Ricardo, von Thünen and Jevons, Böhm-Bawerk had emphasized the importance of “the time element in production”. In combination with the intertemporal structure of consumption plans, the production period became the core concept of his Austrian theory of capital and interest. For Wicksell (1934, 172), “the real kernel of the capital concept” was this: all production requires time ; different processes of production require inputs of capital over different periods, but their rates of return must correspond to a uniform rate of interest in perfect competition. Böhm presented the rate of interest as a premium for waiting, in compensation for the marginal disutility of abstaining from the consumption of present goods. According to Wicksell, Böhm’s concept of time preference of consumption (the agio in the valuation of present over future goods) primarily determines the speed of capital accumulation, not the latter as such. In Wicksell’s view, the rate of interest is determined by the “profitability of round-about methods of production” and, hence, by the marginal productivity of capital ; time preference has merely an indirect effect on it by its influence on the size of the capital stock. Wicksell (1934, 167-95) also refuted Böhm’s concept of the average period of production, since it presupposes that the capital intensity of production is independent of the distribution of income, and since it is incompatible with compound interest calculus. On the other hand, he criticized the circular argument of Walras and his followers, who derived the rate of interest from the value of the capital goods, and the latter in turn from their cost of production that include capital and interest (see the quotation above). In the end, however, Wicksell escaped from this circularity only by the trick of assuming the value of the capital stock as given.

21Wicksell’s own approach is to determine the rate of interest on the base of the marginal products of dated inputs of land and labour. Capital is regarded as “a single coherent mass of saved-up labour and saved-up land, which is accumulated in the course of years” (1934, 150). In general equilibrium, the marginal products of the dated inputs of labour and land, valued at the goods’ prices, correspond to the rates of wages and rent, multiplied with the compound interest for the maturation terms of the respective goods. This leads to the conclusion that the marginal productivity of capital corresponds to the respective discount rate, i.e. the rate of interest (1934, 150-66).

22However, Wicksell (1934, 148-49) discovered a “curious divergence” that destroys the symmetry of the marginal productivity theory of distribution at the macroeconomic level. The explanation is again the fact –or, as Wicksell puts it: “The theoretical anomaly”– that the aggregate stock of capital can be measured only “as a sum of exchange value – whether in money or as an average of products”, and the same holds for the “social marginal productivity rate.” (1934, 149) The problem is the classical one of the absorption of net real saving by rising real wages (and/or rent) during a process of capital formation:

23

“If we consider an increase (or perhaps a decrease) in the total capital of society, then it is by no means true that the consequent increase (or decrease) in the total social product would regulate the rate of interest. In the first instance, new capital competes with the old and thereby results, in the first place, in a rise of wages and rent, possibly without causing much change in the technical composition of the product or the magnitude of the return. For this reason, interest must certainly fall ; but it need not fall to zero, or anything like it, even if the additional product of the new capital is almost nil.”

24In the final equilibrium position that is reached after the increase in capital, the rate of interest does not necessarily correspond to the social marginal productivity of capital. In fact, Wicksell proved rigorously in Value, Capital and Rent (1893 ; 1954, 147-52) that, after a rise in the capital stock, the social marginal productivity of capital must always be smaller than the rate of interest (cf. Uhr 1960, 122-26). Under the label “Wicksell effect”, which was introduced by Uhr (1951, 850-52), this result came to be used and extended to the analysis of changes in technique, which played a central role in later debates about neoclassical and alternative theories of capital.

25What remains to be noted in the context of this paper, is a certain regret that Wicksell did not combine his discussions of the (price) Wicksell effect with his monetary analysis. In terms of the dynamic effects on the distribution of incomes, the case of a monetary expansion that follows from a positive interest-rate gap might not essentially differ from the classical case of net real saving. This would, however, entail the possibility that the “natural rate” of interest undergoes a downward adjustment towards the money rate, and that the banking system could hence, inadvertently or not, influence capital accumulation, output and income distribution – a possibility that Wicksell rejected in the face of his critics (Wicksell 1935, 198-200 ; cf. Uhr 1960, 256-60). It was explored along different routes by his followers.

3.1 – Cantillon and Ricardo Effects

27Hayek (1933, 17) described his ABC theory in a nutshell as follows: “Monetary causes…start the cyclical fluctuations ; successive changes in the real structure of production…constitute those fluctuations.” He based his explanation of business cycles on Böhm-Bawerk’s theory of capital and on Wicksell’s theory of the cumulative effects of interest-rate gaps (Hayek 1933 ; 1935). Hayek thus emphasized the Austrian view that the equilibrium rate of interest, i.e. the rate that brings investment in line with “voluntary saving”, must equal the rate of time preference of consumption. He rejected, however, Wicksell’s orientation towards aggregate variables, arguing that deviations of the market rate of interest from the equilibrium rate may leave the price level unaffected, while inducing changes in the structures of prices and production. If the lending rates of the banks are lower than the equilibrium rate, this is perceived as a signal for a profitable lengthening of production processes. Firms plan investments in additional stages of production in which they construct and employ machinery and other capital goods that are expected to increase productivity. Credit demand rises and banks will inject additional money, in excess of the savings deposited with them, into the economy by lengthening their balance sheets.

28The monetary expansion disrupts the general equilibrium that Hayek categorically assumed to be the starting point of any business cycle – a point at which labour is fully employed and all other resources are fully utilized (1935, 31). The banks’ “excess lending” pushes the system out of equilibrium by helping the producers of capital goods to bid away resources from shorter processes (with less stages of production) and from the direct production of consumption goods. As the prices of capital goods rise faster than the prices of consumer goods, this bidding turns into self-reinforcing changes in the price structure, the allocation of resources and the distribution of incomes: it is a process of “forced saving”, that may or may not be accompanied by a change in the value of money (1935, 85-100):

29

“The ’depreciation’ of money in the hands of the consumer can be, and frequently will be, only relative, in the sense that those diminutions in price which would otherwise have occurred are prevented from occurring. Even this causes a part of the social dividend to be distributed to individuals who have not acquired legitimate claim to it through previous services, nor taken them over from others legitimately entitled to them. It is thus taken away from this part of the community against its will.”

30With reference to Richard Cantillon’s 18^th-century description of the detrimental effects of an “abundance of money” on prices, incomes and consumption, Hayek (1935, 8-9 and 88-100) took the position that such a violation of consumers’ intertemporal preferences is unsustainable: “It is highly improbable that individuals should put up with an unforeseen retrenchment of their real income without making an attempt to overcome it by spending more money on consumption.” (1935, 88) In the course of the monetary expansion wages will rise and contribute to a relative increase in the prices of consumer goods:

31

“These decisions will not change the amount of consumers’ goods immediately available, though it may change their distribution between individuals. But –and this is the fundamental point– it will mean a new and reversed change of the proportion between the demand for consumers’ goods and the demand for producers’ goods in favour of the former.”

32According to Hayek, this reversal in relative profitability inevitably brings about a crisis, in the course of which the investments in more roundabout production processes must be abandoned and the structures of prices and production revert to their original states. However, for the explanation of the upper turning point in the cycle, Hayek (1933, 175 ; 1935, 151-52 ; 1969, 282-83) took recourse to the assertion that the banks would have to stop the credit expansion and “malinvestments”. As this argument and the systemic necessity of a crisis were much disputed by his critics, Hayek (1942) switched from framing his argument in terms of the Cantillon effect to the “Ricardo effect”, a disproportionality of circulating and fixed capital of the sort discussed by David Ricardo in his famous “machinery chapter” (Ricardo 1821, chap. 31). [14] Hayek argued that, even if the banks did not stop the credit expansion, the lengthening of the production processes would sooner or later lead to a shortage of consumption goods that disrupts the transition to a structure of production that is more productive, but not consistent with intertemporal preferences. The basic idea was already expressed in Prices and Production:

33

“The situation would be similar to that of a people of an isolated island, if, after having partially constructed an enormous machine which was to provide them with all necessities, they found out that they had exhausted all their savings and available free capital before the new machine could turn out its product. They would then have no choice but to abandon temporarily the work on the new process and to devote all their labour to producing their daily food without any capital.”

34Even though this passage seems to imply some latter-day version of the classical wage fund doctrine, Hayek’s ABC theory is not clearly articulated in terms of functional income distribution. The redistribution of purchasing power provides a link between the monetary impulse and its transmission through the structures of prices and incomes, but the argument runs in terms of a redistribution among “individuals” rather than classes. Attempting to wed neoclassical general equilibrium theory with monetary business cycle theory, Hayek emphasized that the temporary failure of the interest-rate mechanism to match investments with intertemporal consumer preferences would lead to a distortion of the personal income distribution that is eventually corrected by the price mechanism and the ensuing crisis.

3.2 – Critique and Philosophical Turn

35Hayek (1928 ; 1929) had begun to develop his ABC theory before the onset of the Great Depression, which started with a financial market crash and the failures of numerous banks after a strong monetary expansion in the late 1920s. His prediction of an inevitable crisis seemed to be borne out by the facts and attracted great attention to the four lectures of Prices and Production, which Hayek presented at the London School of Economics in 1931. Since his theory carried the controversial message that the crisis is not dysfunctional disequilibrium, but an indispensable cure of monetary excesses and malinvestment, it was subject to much scrutiny and critique (see, e.g., Hicks 1967 ; Hagemann / Trautwein 1998). The critical comments add up to the conclusion that Hayek’s approach fails to provide a general theory of business cycles, because it is based on extremely restrictive assumptions and inconsistent arguments. In our context, suffice it to mention the main points pertaining to income distribution.

36In order to make the Cantillon effect work in the expansionary phase, it must be excluded that the profitability of investments in more roundabout processes is quickly reduced by wage increases and consumer spending. In Hayek’s vertical structure of production, in which the leading investors are assumed to spend their outlays on labour and land only (the original factors of production), and in which wages are flexible, it is not clear why such lags in consumer demand should occur. Even if it is granted that a redistribution of incomes takes place in the way described by Hayek, there is no general self-reversal inbuilt in the price mechanism, by which the original set of intertemporal preferences would assert itself in the market process. As Sraffa (1932, 48) put it rather bluntly:

37

“One class has, for a time, robbed another class of a part of their incomes ; and has saved the plunder. When the robbery comes to an end, it is clear that the victims cannot possibly consume the capital which is now well out of their reach. If they are wage-earners, who have all the time consumed every penny of their income, the have no wherewithal to expand consumption.”

38In other words, those who benefit from increased profits in the capital goods industries tend to have a lower marginal propensity to consume and might not drive up the consumer prices to the critical degree. Furthermore, if some of the investments in “unduly elongated” production processes are completed during the boom, the price rise is counteracted or at least slowed down. Even if there is a crisis, this does not necessarily imply a return to the original structure of production, as a substantial part of the newly created production capacities remains in operation, once the financial losses have been written off.

39All in all, Hayek’s ABC theory failed to gain widespread acceptance, not only because of the quick success of Keynes’s General Theory after 1936 (as Hicks 1967 points out), but also because its analysis of the distributional consequences of monetary expansion was found seriously lacking. Even though Hayek never quite gave up on the Ricardo effect (cf. Hayek 1969), he essentially turned away from attempts to extend general equilibrium theory to business cycles and money. From the mid-1930s onwards, his writings on the price system and markets as spontaneous order and discovery processes, in which tacit and dispersed knowledge is efficiently communicated, carry more of a disequilibrium flavour. In these contexts, Hayek occasionally addressed issues of income distribution in the negative. He stressed the futility to give a precise meaning to such “weasel words” as “distributive justice” –most prominently in Mirage of Social Justice (1976) and the Constitution of Liberty (1960, 99), where he stated that “the results of the individual’s efforts are necessarily unpredictable, and the question as to whether the resulting distribution of incomes is just has no meaning”. Yet, given the general thrust of the older Hayek’s arguments, it is no far-flung conclusion that he continued to consider “forced saving” to be an illegitimate transfer of claims. [15]

4.1 – Lindahl and Myrdal on Redistribution of Incomes in Cumulative Processes

41The key contributions of the Stockholm School to Wicksellian macroeconomics are Lindahl’s “Means of Monetary Policy” (1930), translated as The Rate of Interest and the Price Level in Lindahl (1939, Part II), and Myrdal’s Monetary Equilibrium (1939), originally published in Swedish in 1931 and, as part of an anthology edited by Hayek, in a German version in 1932. Like Hayek, Lindahl and Myrdal based their analyses of monetary equilibrium and macroeconomic disequilibrium on Wicksell’s concepts of interest-rate gaps and cumulative processes, extending them to the analysis of fluctuations in output and other real variables. Like Hayek, Lindahl and Myrdal rejected price-level stability as a sufficient condition for monetary equilibrium. Yet they differed from Hayek (and to some extent also Wicksell) in almost every other respect. They did not attempt to demonstrate that the original equilibrium position of the system is stable in the sense that money would be neutral in the long run. On the contrary, they criticized Wicksell’s notion of the natural rate of interest, arguing that in the intertemporal price relations of a monetary economy the loan rate is always part of the price structure (Lindahl 1939, 247-48 ; Myrdal 1939, 50-53). Hence, they put the focus on the formation of expectations in the coordination of investment and saving plans, and on different constellations of expectations, plans and capacities in the cumulative processes that develop in the case of a difference between the market rate of interest and the “normal rate”, which brings investment in line with planned saving. Myrdal reduced Wicksell’s definition of monetary equilibrium to one single condition, namely that the “money rate of interest” brings “gross real investment” in line with “saving plus total anticipated value-change of the real capital” (1939, 96). As this implies that the value of real capital equals its production costs (1939, 95-104 and 163), Myrdal actually developed an early version of Tobin’s q. Lindahl (1930 ; 1933) went even further and defined national income in Fisherian terms as a flow of discounted yields on capital. In this perspective, the market rate of interest was of key importance for the formation of income.

42Which role does the (re)distribution of incomes play in Lindahl’s and Myrdal’s analyses of cumulative processes? In Myrdal’s Monetary Equilibrium the issue is not directly addressed. Myrdal is more concerned with the examination of equilibrium conditions and the required adjustments out of equilibrium in his famous terms of ex ante and ex post (Myrdal 1939, p. 45-47 and chap. VII). Distributional aspects come into the picture merely in connection with the stickiness of prices and wages, and the stability of consumption in a downward cumulative process –but the arguments are highly interesting with regard to modern macroeconomics.

43While Wicksell considered price-level stability to be a sufficient condition for monetary equilibrium, and hence as the relevant criterion for an interestrate policy that stops cumulative processes, Myrdal (1939, 133) argued that, at first sight, monetary equilibrium (in his definition of the value of real capital equalling its production costs) would be quite independent of movements of the price level. He then added the reservation that this would only be true if the price movement were “perfectly uniform”. He held this assumption to be untenable in view of the existence of contracts with fixed interest rates, prices and wages, of monopolistic price-setting and “a general element of inertia” (1939, 134).

44

“In any general price movement, therefore, which has not been anticipated with full certainty by all those having a part in price formation, the distribution of incomes and property must necessarily change. Consequently the demand and supply functions of different commodities change and also their price relations, including the specific price relations relevant to monetary equilibrium… A monetary policy aimed to preserve the equilibrium relations must, therefore, adapt the flexible prices to the absolute level of the sticky ones.”

45This suggestion has recently been echoed by the “neo-Wicksellian” approach of the New Neoclassical Synthesis, in which Woodford (2003, 13 fn 7) refers to Myrdal as a precursor of his main policy conclusion: in the models of the new synthesis, it is basically the stickiness of prices and wages that causes deviations of output from its natural rate (the level consistent with the natural rate of interest) ; hence monetary policy must eliminate inflation to prevent price stickiness from generating real effects. However, in the “paleo-Wicksellian” approach of Myrdal (and Lindahl), price stickiness is merely a subordinate part of the problems caused by coordination failures of the interest-rate mechanism. [17] The bigger problem is the flexibility of prices and changeability of profit expectations that develop in the cumulative process. As Myrdal argued, the stickiness of wages may accentuate the cumulative process at an early stage, but it will eventually help to stabilize the economy, i.e. provide a lower bound for a cumulative downward process. This idea is developed in his account of a “tightening of credit conditions”, which depresses the values of real capital and hence profit expectations:

46

“The whole process derives its special character from the fact that the institutional circumstances are such that there is an important resistance to a fall in certain prices, especially wages. To begin with, the collective contracts run for a comparatively long period, which naturally makes the sensitiveness –e.g., variability within the business cycle– of wages small, and even at the expiration of the contracts the workers strenuously resist reduction in money wages…[T]he workers are ready to put up with greater unemployment than before. As the wage reduction cannot then, in any case, be as great as would be required for unchanged unemployment, the latter increases, production is restricted, and the total money income of the economy falls in greater proportion than that corresponding to the reduction in rates of wages.”

47In the following Myrdal stressed that, even though this process implies a redistribution of the (shrinking) aggregate income in favour of labour, the consumption of the workers and the unemployed (financed by transfers or past savings) will stabilize the economy. Together with the “consuming habits of the middle and upper classes”, dictated by “social conventions”, the effective demand for consumption goods stops the further depression of capital values, and thereby halts the downward cumulative process. The new lower-level equilibrium will be characterized by a “largely unchanged price level for consumption goods ; capital values which will be sufficiently lower to correspond to the higher interest rate, or more generally, to the tighter credit conditions ; somewhat lower wages, particularly in the capital goods industries ; some, perhaps quite considerable, unemployment, especially in the capital goods industries ; a production volume restricted generally but particularly in the capital goods industries, implying a shorter time structure of production” (Myrdal 1939, p. 169). As Myrdal wrote these lines originally in early 1932, we have here an anticipation of Keynes’s concept of “underemployment equilibrium” along the lines of Wicksellian monetary theory and (a light version of ) the Austrian theory of capital!

48However, when it comes to addressing the issue of income redistribution in cumulative processes, Lindahl (1930) was much more to the point. In his chapter on the cumulative process caused by lowering the level of interest rates, Lindahl (1939, 162-64) pointed out that the immediate effect will be “an increase in all capital values” and “a redistribution of income in favour of borrowers but against lenders”:

49

“Although saving may be decreased as a direct result of the low rate of interest, the redistribution of income in favour of entrepreneurs having debts will be an influence in the opposite direction. Hence it is difficult to say whether there will be an increase in net savings or not…The alteration in the profitability of investments of different types will, however, cause a tendency both towards an increase of stocks and towards an alteration of production in favour of longer investments.” (Lindahl 1939, p. 163-64)

50Thus far Lindahl’s argument appears to run along the lines of Hayek’s story, which was developed around the same time. However, instead of restricting himself to postulates of a self-reversal of the structural changes, Lindahl examined different scenarios under alternative assumptions about the degree of employment and capacity utilization, intersectoral labour mobility, the rigidity of the investment period and the state and adaptation of expectations. As a result, he arrived at a position towards the notion of “forced saving” that was diametrically opposed to Hayek’s. Again, wage stickiness was a key argument, but in Lindahl’s upward cumulative process it is an upward stickiness that makes workers’ demand for consumption goods increase only slowly, when capital values rise. This leads to a restriction of consumption that enables “production to become more capitalistic”.

51

“How can a lowering of the loan rate of interest, which is generally supposed to have a tendency to decrease (voluntary) saving, thus cause an increase in total saving? The solution appears to be, that…it occasions a redistribution of incomes such that those with a relatively strong disposition to save find their incomes increased, at the expense of those whose disposition to do so is relatively weak…Even workers, who on account of their relatively small incomes cannot save much, find their share of the national dividend decreased. On the other hand entrepreneurs find their incomes increased and have a strong incentive to apply it to further capital investment.”

52In the original Swedish version of 1930, Lindahl worked with the temporary equilibrium approach and emphasized that, what may be called as “forced saving…from the point of community as a whole” is “in large part quite voluntary” from the point of the individual (1939, p. 173). Going over to the dynamic perspective of sequential disequilibria, Lindahl added in the 1939 English translation that the inflationary cumulative process will yield windfall profits to the entrepreneurs. Their incomes, “as calculated ex post, will exceed their anticipations. This additional income constitutes the ’unintentional’ saving for such a period. (For this ’unintentional saving’ the term ’forced saving’ does not seem to be very appropriate)” (1939, p. 175-76). Both the voluntary, planned saving and the unintentional saving of the entrepreneurs tends to be turned into additional investment that increases the capital stock. The notion of “forced” only applies to the restriction of consumption on the side of the lenders and other groups who fail to adjust their nominal incomes to inflation.

53It is worth noting that, contrary to the classical view and prior to Keynes’s General Theory of 1936, Lindahl (1930 ; 1939, p. 175) emphasized that it is the “quantity of saving” that adjusts to “the amount of real investment”, and not vice versa: “[T]he causal element will…be the alteration in the distribution of income due to the shift in the price level.” Lindahl’s redistributive income mechanism of balancing investment and saving ex post differs, of course, from Keynes’s aggregate income mechanism, and in Lindahl’s case it is monetary policy that, by lowering the level of interest rates, may affect the capital stock and shift the equilibrium position of the system. Yet Lindahl’s transmission mechanism, too, suggests a fundamental non-neutrality of money in the long run.

54Chiodi/Velupillai (1983, 103) even argue that Lindahl’s 1930 essay “contains, in a nutshell, the whole neo-Keynesian theory of distributive shares (whether of the Kaldor version in terms of different types of income or of the Pasinetti version in terms of classes of people)”. However, Chiodi and Velupillai ignore the end of Lindahl’s story, or rather: his skepticism about the stability of the new equilibrium. Lindahl (1939, 180) concluded that, if (and only if ) people do not adapt their price expectations to ongoing inflation, the cumulative process comes to an end when the maturing of the investments increases the output of consumption goods and stops the rise in their prices. The economy will reach a new equilibrium state with a larger capital stock, a lower level of interest rates and a higher price level, compared to the initial state. If, however, people adapt their expectations to ongoing inflation, the latter will accelerate, “until the process is brought to an end by a crisis” (1939, p. 182). Lindahl deemed this to be much more realistic, but did not elaborate on the nature of the crisis and the character of the distribution conflicts that might induce it. [18]

4.2 – Political Activities

55When the Great Depression made its full impact on the Swedish economy in 1931, Sweden left the gold standard and became the first country to adopt price-level stability as the norm for monetary policy (cf. Jonung 1979). In the same year, Lindahl was appointed advisor to Riksbanken, the central bank of Sweden, and helped to construct the relevant price index. As he had argued at great length in a 1929 monograph about monetary policy rules, [19] his preferred norm for monetary policy was actually not price-level stability but Davidson’s rule, according to which the price level should move inversely to productivity, thus stabilizing nominal income. Lindahl argued that this norm would lead to more efficient risk-sharing between entrepreneurs and their creditors, the adverse output effects of negative productivity shocks on profits would be balanced by price increases, whereas positive shocks would be beneficial to wage earners and other consumers. Despite his preference for nominal income targeting, Lindahl fervently defended the official target of price-level stability, as it seemed appropriate in order to stop ongoing debt-deflation and to counteract expectations that inflation would rise strongly after the transition to floating exchange rates. Lindahl’s advisory activities at Riksbanken can thus be interpreted as a practical consequence of his Wicksellian views on money and income distribution.

56It is hard to find such a connection in the works and activities of other members of the Stockholm School, which may be somewhat surprising given their similar backgrounds and close cooperation. Practically all of them were involved in research and advisory work in the 1920s and 1930s that was related to distributional issues. [20] Lindahl, Myrdal, Johansson, Lundberg, Kock and Svennilson were investigators in the large research project on “Wages, Cost of Living and National Income in Sweden 1860-1930”, carried out under the supervision of Gösta Bagge at the University of Stockholm and funded by the Rockefeller Foundation from 1926 until 1935. Lindahl, Myrdal, Ohlin, Hammarskjöld, Johansson, Kock and Svennilson were engaged as secretaries or experts by the governmental Committee on Unemployment that worked from 1927 until 1935. After the onset of the Great Depression, Ohlin wrote several memoranda on monetary policy and unemployment for this committee, while Lindahl and Myrdal devised strategies for countercyclical fiscal policies. Yet, despite the lip service paid to the pioneering works of Wicksell, all this work was hardly concerned with his theories of money and income distribution.

57Together with his wife Alva, Myrdal published a provocative book on the “population crisis” in 1934, where the Myrdals vented –in some contrast to Wicksell several decades before– concerns about declining population growth and the negative consequences that this would bring for investment in public infrastructure and the economy in general. A fiscal redistribution of incomes and other welfare policies were at the core of their strategy to accelerate Sweden’s birth rate, then the lowest in Europe: “[S]tate intervention…would tend to equalize consumption across social classes and income groups. Put slightly differently, the program’s aim would be to eliminate –through governmental action– the incremental expense burdens associated with childbearing and child-rearing by poorer families.” (Barber 2008, p. 57) Over the next decades, Myrdal came to use the Wicksellian concept of the cumulative process in different contexts, such as the social discrimination of ethnic minorities or the underdevelopment of regions (Myrdal 1957, chap. 2). He also praised the Nordic welfare state and the principle of equality as a precondition for economic growth and concept to counteract cumulative processes of underdevelopment (Myrdal 1957, p. 39-42 ; 1984, p. 154 ; see also Barber 2008). He did not, however, apply it to monetary theory again.

58The most likely reason for the lack of further systematic work on money and income distribution in the Stockholm School is the loss of importance of monetary policy in the early 1930s. When the Great Depression deepened in 1932 and 1933, price-level stability was seen as a necessary but not sufficient condition of stabilizing investment, production and employment. In 1933, the Swedish crown was pegged to the British pound, and the exchange-rate target reduced the autonomy that Swedish interest-rate policies had enjoyed for a short time (cf. Jonung 1979). Fiscal policies became more important and turned attention away from Wicksellian concepts of interest-rate gaps and redistributive income mechanisms.

5.1 – Socialist Supply-Side Policies

60At the end of the Second World War, Myrdal, Lundberg and other economists, expected a deep post-war depression, comparable to the one that Sweden had experienced in the years after the First World War. Instead, the Swedish economy went into a phase of strong growth, due to the export demands arising from European reconstruction and the war in Korea. The boom helped to achieve full employment, but under the Bretton Woods regime of fixed exchange rates it created a distributional trilemma for the trade unions. As Meidner (1948) put it, the unions faced the choice between:

1. wage drift in the export industries, which would jeopardize solidarity among the trade unions, if other sectors failed to follow ;
2. inflation, if the unions fought for an increase in the general wage level, risking losses of jobs in the export industries, as cost spirals would make them less competitive ;
3. legitimatory conflicts within the trade union movement, if the trade unions exerted wage restraint to stabilize prices and jobs, but thereby contributed to a profit inflation in the export industries.

Figure 2

The Rehn-Meidner Model

The Rehn-Meidner Model

61In view of this trilemma, the Swedish trade union economists Rehn and Meidner fervently rejected the idea of a trade-off between full employment and price-level stability –almost two decades before the famous Phillips curve debates began to turn that way. Rehn and Meidner argued that, in the small open economy of Sweden, which operated under a regime of fixed exchange rates and centrally negotiated wage contracts, inflation would be detrimental to employment and distributional fairness. In their 1951 memorandum for LO, the Swedish trade union federation, they devised an unconventional strategy to deal with the trilemma. With hindsight, the Rehn-Meidner model can be described as a comprehensive concept that aimed at eliminating the target conflicts in the “magic pentangle” of high GDP growth, full employment, internal and external monetary stability, and a fair income distribution. The targets were combined with unconventional instruments, in accordance with the Tinbergen rule, by which the number of instruments must correspond to the number of targets, and each instrument must be clearly assigned to one target (see Figure 2). More importantly, however, the instruments were linked with each other to create a virtuous circle of economic growth and stability. Each instrument was in turn supported by at least two other instruments, even if the latter primarily served their own targets.

62The core instruments were the solidaristic wage policy, and the restrictive fiscal and active labour-market policies ; the policies of collective funding and structural adjustments naturally followed from them and complemented them. Under the slogan “equal pay for equal work”, the wage policy of LO was to close the wage gaps between comparable jobs in highly profitable and less profitable firms. This compression of wage differentials required the coordination of wage negotiations in a centralized bargaining system (which actually came into force in the mid-1950s). The benchmark for the implicit redistribution of wage gains should be an average increase of nominal wages that corresponds to average productivity growth, but gives no inflationary impulses (LO 1951, 141). This policy implied, however, wage restraint in highly productive companies, which generated “excess profits” that could create legitimatory problems for the Solidaristic Wage Policy. Hence, Rehn and Meidner pleaded for restrictive fiscal policies that contribute to a fair income distribution in two ways: Income taxation should transform excess profits into public revenues, and sales taxation should reduce inflationary pressure at times of high wage increases. The public savings that would result from these policies could be used for deficit spending in recessions, but should create a public budget surplus over the cycles. In combination with tax-subsidized investment funds under trilateral governance (and later the “general pension fund”, established in 1960), these public savings would allow “the formation of collective capital” to fund improvements in the infrastructure and other measures of structural adjustments required to make fast productivity growth consistent with full employment.

63The key to the latter was the support of the Solidaristic Wage Policy by active labour-market policies, financed out of the sources of collective funding. On the downside of the wage equalization, less productive firms should not have their profits subsidized by local wage restraint ; they should rather be allowed to exit from their markets, thus indirectly contributing to an increase in productivity levels. Displaced labour should be comprehensively taken care of by support for retraining, matching and mobility rather than “passive” policies of unemployment benefits and wage subsidies.

64In the context of this paper, three things are notable about the Rehn-Meidner core of the Swedish model. First, it was a strategy of what later came to be known as “supply-side policies”. For various reasons of political economy, redistributive policies in welfare states are frequently associated with Keynesian concepts of employment policy. By contrast, Rehn explicitly rejected Keynesian demand management, which in those days was carried out as a mix of price controls and expansionary “employment programmes”. He pleaded for a mix of restrictive demand management with selective supply-side measures, in order to keep low inflation consistent with full employment:

65

“Purchasing power should be pumped away so that the demand surface sinks. The islands of unemployment which then appear must not be flooded over again by re-raising the ’demand surface’, but instead must be removed… by special (local, marginal) measures to create extra employment in any place or industry where private enterprise does not keep the demand for labour at the same level as the supply. There should also be strong stimuli to move from places or branches with low, to those with high, demand for labour.”

66With hindsight, some of the measures may look fairly conventional, ranging under the fashionable concepts of “workfare” and “flexicurity”. Other parts of the strategy look unorthodox, such as the expansion of the public sector and collective funding of structural adjustments through high taxes. In this, Rehn and Meidner referred back to Myrdal’s recommendation of High Taxes and Low Interest-Rates (1944), which aimed at lowering the level of the equilibrium rate of interest and yielding some democratic control over the expansionary investment process –without striving to nationalize industries (a popular strategy elsewhere at the time). So, as a second point, it should be noted that the Rehn-Meidner model was a strategy to secure equitable growth, by which trade unions would support rather than resist labour-saving technical progress. But what has this strategy to do with Wicksellian views on money and income distribution? The most obvious link is the fear of inflation, most vividly expressed in Rehn’s Hate Inflation! (1957). In the 1940s and 1950s, the key instrument of controlling inflation was no longer discount rate policy, but a mix of interest-rate regulations, quantitative credit controls and fiscal measures (cf. Jonung 1994). Under the Bretton Woods regime, it appeared to be more effective to use a combination of centralized wage and restrictive fiscal policies to prevent aggregate excess demands from feeding into cumulative changes in the price level. As the final point it may thus be argued that Wicksell’s concerns about inflation as a “disturbance of the social mechanism” are echoed in the Swedish welfarestate model of the 1950s, but the macroeconomic policy regime had changed substantially (largely as a consequence of the external influences of the Great Depression and the Second World War).

5.2 – Social Engineering and Further Fears of Inflation

67Despite the Wicksellian heritage of an anti-inflation consensus, the Rehn-Meidner model was criticized for creating an inflationary bias. One of the most prominent critics was Erik Lundberg (1907-1987). As a junior member of the Stockholm School, Lundberg (1937) had extended Wicksell’s theory of cumulative processes and Lindahl’s sequence analysis of inflation to the analysis of model sequences of economic growth and cycles, which allowed him to integrate both “Keynesian oversaving” and “Hayekian overinvestment” (or malinvestment) into a common framework. As director of Konjunkturinstitutet or NIER, the national institute of economic research, from 1937 until 1955, Lundberg refined and operationalized the concept of inflationary gaps, which had originally been developed by Keynes (1940) and referred to excessive purchasing power in goods markets (cf. Hansen 1951, 60-82 ; Berg 1993). In his controversy with Rehn and Meidner, Lundberg (1952, 3) conceded that “profit inflation” was the core problem, arguing that there “can be no balance in the economy (including in this concept equality of the supply of, and demand for, labour) unless the general price level behaves so as to prevent the creation of profit inflation or deflation”. However, Lundberg doubted that the strategy of Rehn and Meidner would work, because it did not use productivity growth as external benchmark for wage bargaining. Instead it transformed the average productivity into to a dependent variable. If wage policies were calibrated to the developments in the most productive sectors and firms, they would tend to reduce investments in other sectors. These losses could not be compensated by collective funds, as the required tax revenues would not materialize for lack of inflation. If wage policies, on the other hand, were moderate in order to protect jobs, they would create a high risk of profit inflation. Lundberg (1952, 59-65) argued that restrictive fiscal policies would not efficiently deal with that risk, due to lags, distortions and the low allocative efficiency of state activities in general. He insisted that it would be more efficient to control inflation by a return to discount-rate policies that are taylored to affect profit expectations.

68Lindahl, too, wrote and lectured intensively about “the inflation problem” in the post-war era. He did not criticize the Rehn-Meidner model as such, but in his book on The Money Value Game (1957), he observed that monopolistic price-setting had become much more widespread in goods and labour markets, bedding for inflationary dynamics in every cyclical upswing. Instead of analysing cumulative processes in alternative scenarios, as he had done in the 1930s, Lindahl (1957, p. 10-16) described them as a typical three-stage process of demand expansion. The sequence starts at the end of a depression, when loan rates are low. At the first stage, investment begins to exceed planned saving, such that monetary equilibrium is disturbed ; but free capacities in the system allow for an expansion of employment, real income and saving ex post, without changing the price level. Lindahl argued that market forces would not automatically restore monetary equilibrium once full employment is attained. The imbalance of investment and saving could, at the second stage, easily translate into cumulative inflation. Lindahl defined this stage as “profit inflation”, corresponding to the Wicksellian case of unplanned saving from windfall profits (as in Lindahl 1930). The capital stock tends to grow faster than at the first stage, but the underlying redistribution of income and wealth is not sustainable. Sooner or later, “people will begin to regard the inflation as a permanent phenomenon and adjust their activities accordingly…Inflation therefore has a tendency to accelerate, especially when the workers come up with ever increasing wage demands in order to compensate themselves for the losses of real income that they suffer due to inflation” (Lindahl 1957, p. 11-12). This is “income inflation”, the third stage in Lindahl’s sequence – and his description of wage-price spirals looks like an anticipation of Milton Friedman’s acceleration theorem. Yet it should be noted that Lindahl, who had rejected Wicksell’s concept of a natural interest, did not subscribe to any notion of a (unique) “natural rate of unemployment”, because he did not consider money to be neutral in the long run. In his view, profit inflation tends to increase total saving (planned and unintentional) beyond the level of the first stage, whereas income inflation reduces total saving, as the wage-price spirals eliminate windfall profits, while uncertainty about the future value of money deters households from saving. [21] Profit inflation generates full employment, whereas income inflation, through its detrimental effects on the capital stock, can lead to “significant unemployment and general depressions” (Lindahl 1957, p. 14). At the end of Lindahl’s three-stage sequence, investment and saving may (if unaltered by monetary policy) converge at a lower level, restoring monetary equilibrium at higher rates of inflation and unemployment – a state for which Myrdal later coined the term “stagflation” (cf. Streeten 1998, p. 546).

69In the line of Swedish studies of inflation and income distribution finally the Dane Bent Hansen (1920-2002) deserves to be mentioned for his widely praised, but quickly forgotten Study in the Theory of Inflation (1951). He himself characterized it as being influenced by Lindahl, his thesis supervisor, but also by Keynes and Hicks:

70

“This study is likewise Wicksellian in its analysis of inflation, but goes further than the post-Wicksellians in that spontaneous price increases are taken into account ; and further than the post-Keynesians in that monetary excess demand in the factor-markets is also taken into account. Furthermore, repressed inflation, hitherto much neglected in purely theoretical analysis, is also considered.”

71Hansen formalized many aspects of profit and income inflation, describing the former as a disequilibrium process and the latter as a general quasi equilibrium system, in which real wages remain constant, while inflationary gaps and parallel price and wage rises persist in the goods and labour markets. Hansen’s sophisticated framework is noteworthy, because it specifies quite precisely many of the conditions and ex ante / ex post-distinctions of variables required to analyze the redistributive income mechanisms that Lindahl (1930) and Myrdal (1931) had only hinted at. Yet, setting the focus on the formal analysis of inflationary gaps, Hansen (1951) did not systematically discuss the links between monetary policy and income distribution.