Interest and Profit: An Empirical Assessment of the Monetary Theory of Distribution for the Euro Area

ABSTRACT

Several authors, especially those who share a Classical–Keynesian point of view, argue that the interest rate determines the rate of profit in the long run. Considering the eleven founding economies of the euro area, I find that, adjusted for the rate of growth of gross national income, there is a positive long-term relationship between the real interest rate and the net rate of profit. The results are confirmed even when I estimate a model with nominal interest rates, inflation, and a yield curve. These results imply that the European Central Bank (ECB), when deciding monetary policy, is not neutral in determining income distribution.

KEYWORDS:

• Monetary Theory of Distribution
• interest rate
• profit
• Sraffa
• central banking
• ECB

JEL CODES:

• B51
• C23
• E43
• E58

Acknowledgement

This paper was a by-product of the ‘Monetary Policy Project’ led by Prof. Louis-Philippe Rochon, to whom I am deeply grateful for his support. I would like to thank the anonymous reviewer for her helpful and constructive comments that greatly contributed to improving the final version of the paper. I would also like to thank the editors of this journal. Remaining errors are my sole responsibility.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1

By ‘normal rate of profit’ we simply mean what the theories of distribution endeavour to explain: classical and Marxian expressions such as ‘natural rate of profit’, ‘permanent rate of profit’ or ‘general rate of profit’, and marginalist expressions such as ‘natural or real rate of interest’ and ‘equilibrium or full-employment rate of interest’, never refer to actual or effective profits, but to normal profits. The later, reckoned gross of interest, correspond to the rate of return on capital which would be obtained by firms using dominant or generally accessible techniques, and producing output at levels regarded as normal at the time the capacity was installed. As actual profits always deviate from normal profits, economic theory is in substantial agreement in regarding the normal rate of profit as a magnitude which cannot be arrived at statistically and empirically. (Pivetti 1991, p. 20, emphasis added)

2 This quotation has been translated to English by Matías Vernengo http://nakedkeynesianism.blogspot.com.ar/2012/09/sraffa-and-confidence-fairy.html.

3 Determining what the exogenous variable is it is undoubtedly very complex. As suggested by Panico,

Yet, owing to the complexity of the conflicts over distributive shares, it may be impossible to say a priori whether the real wage rate or the rate of profits has to be taken as an independent variable. The specific economic, social and political conditions prevailing over a certain period will determine the outcome of the conflict between workers and capitalists over the social product, and consequently the direction of the causal links between distributive variables. (Panico 1988, p. 6)

4 To quote from Ricardo’s work:

I have already said, that long before this state of prices was become permanent, there would be no motive for accumulation; for no one accumulates but with a view to make his accumulation productive, and it is only when so employed that it operates on profits. Without a motive there could be no accumulation, and consequently such a state of prices never could take place. The farmer and manufacturer can no more live without profit, than the labourer without wages. Their motive for accumulation will diminish with every diminution of profit, and will cease altogether when their profits are so low as not to afford them an adequate compensation for their trouble, and the risk which they must necessarily encounter in employing their capital productively. (Ricardo 1821, emphasis added)

5 ‘ … for the long-term interest rate and the normal remuneration of “risk and trouble” establish, in each sphere of production, the minimum or “necessary” level below which the profit rate cannot go, over the long run, however intense one may suppose the forces of competition to be’. (Pivetti 1991, p. 29).

6 Actually, the real interest rate impacts on the rate of profit. As Ciccone (1990) points out, ‘The fundamental issue, as I understand it, concerns what, in Pivetti’s analysis, the monetary authority have as a target. What they are assumed to control, through the nominal interest rate, is precisely the real rate of interest’.

7 This phenomenon is known in the mainstream literature as the ‘Gibson Paradox’ (Sargent 1973). Referring to this, Pivetti stated that:

There is nothing ‘paradoxical’ in the positive correlation between interest and the price level, when the phenomenon is viewed in the light of Tooke’s theory: given money wages and production techniques, a rising (lowering) of prices as a result of a lasting rising (lowering) of interest rates would merely reflect the adaptation of prices to normal costs, caused by competition. (Pivetti 1998, p. 43)

Kaldor also suggests that ‘There is evidence for believing that interest costs are passed on in higher prices in much the same way as wage costs’ (Kaldor 1982, p. 63). Other authors who contributed to this idea include Tooke, Marx and Keynes (among others; see Panico 1984 and Zolea 2021). For recent empirical evidence on the Gibson Paradox, see Deleidi and Levrero (2021).

8 See Dvoskin and Feldman (2021, p. 276) for a criticism of Franke’s model.

9 See Dvoskin and Feldman (2021, p. 268) for a criticism of Ciccarone’s model.

10 This paper provides an interesting theoretical and empirical review of the literature on the topic. Some papers that try to deal with the empirical part are those by Duménil and Lévy (2001, 2004, 2007) and Shaikh (2011). However, these papers are more focused on descriptive statistics rather than econometrics techniques and are based on a Marxian theory in which the Say’s law holds. Other non-mainstream scholars who have addressed this issue include Hein and Ochsen (2000), Power, Epstein, and Abrena (2003) and Hein (2010), who have found that there is a close relationship between rentiers’ income and the real interest rate.

11 See Deleidi and Levrero (2021) who show that monetary policy is able to permanently affect long-term interest rates and that the central bank has a certain degree of freedom in setting the levels of the short-term policy rate.

12 The fact that some European countries decided to peg their currency and be part of a monetary union frees us from the problem of working with economies as if they were open economies and dealing with the nominal exchange rates that might also affect distribution (see Álvarez 2021; Dvoskin, Feldman and Ianni 2020; Gahn and Machado 2019; Ianni 2016). The case of sterling and the possible impacts on monetary policy during the 1970s is also discussed by Kaldor (1980, p. 317).

13 I have chosen the GNI instead of the GDP to include also the foreign income of residents.

14 The normal rate of profit is the expected rate of profit on newly installed equipment.

15 This is one of the key points in economic theory. The level of activity can affect the effective rate of profit (Garegnani 1992). A passage in Kaldor (1980) is useful to clarify this, as he considers that an increase in interest rates could have an effect on activity levels and thus bankrupt firms:

And where circumstances are such that the rise in interest charges cannot be passed on they eat into profits; with continually rising rates, this is bound to lead to a situation where firms become insolvent for lack of cash to pay interest on their loans, or where they have to borrow in order to pay interest on previous borrowing, a process that is sure to lead to bankruptcy. (Kaldor 1980, p. 315)

Hence, activity levels (and capacity utilization) must be taken into account when analysing the effective rate of profit.

16 This equation can be derived from the theoretical ARDL model developed in Greene (2000, p. 605).

17 The choice of the real interest rate follows a fairly simple logic. The system expressed in nominal terms says very little about relative earnings in inflationary contexts. In these contexts, it cannot be known the true distributional adjustment cannot be known if real variables are not analysed. This is why I decided to do it with the variables in real terms. However, one referee drew my attention to the fact that the objectives of monetary policy may be other than a real interest rate, e.g., as financial stability. This is why I included Appendix E where I work out the same model with the nominal interest rate, inflation and the yield curve (Models 2 and 3) and with average (ten-year) inflation (Models 4 and 5).

18 Given that the long-term real interest rate reaches negative values in some years, it is not possible to transform it into logarithms.

19 For Pedroni and Kao, I have chosen the existence of an individual intercept. For Fisher I have chosen lag intervals (1 1) and intercept (no trend) in CE – no intercept in VAR.

20 This effect is also recognized by Valle Baeza and Mendieta Muñoz (2012).

21 Normality in residuals is analysed in Appendix D. Coefficients corresponding to the short-term effects by country of the ARDL model are also available upon request.

22 This test was performed with different lags (2, 3 and 4) for robustness check. The results are available upon request. With two lags there is no causality, the p-values are greater than 0.1. Following the Toda and Yamamoto procedure I performed this test also with 3 and 4 lags. Here I show the result with 3 lags. With 4 lags the results hold. The problem when using many lags is that the test loses power (Enders 2004).