8 Money, credit and inflation

This chapter summarises the value analysis developed previously, through a critical review of Marxian contributions to the theories of money, credit and inflation. It is divided into four sections. The first briefly reviews Marx’s theory of money and credit, including the forms and functions of money. The second explains the relationship between money and prices of production, and develops the possibility of valueless paper currency in the Marxian system. The third critically analyses three Marxian theories of inflation, and indicates how they can be developed further. The fourth section concludes this chapter.

8.1 Labour and money

Marx’s derivation of money from commodity exchange, in chapter 1 of Capital 1, is neither an historical explanation of the origin of money, nor a purely logical derivation of the concept of money from the commodity (see section 1.1). Marx’s analysis presumes, first, that money and exchange are inseparable and, second, that exchange was marginal to pre-capitalist societies, while money also had ritual, ceremonial, and customary uses in these societies.¹ Money emerges historically out of the interaction of commodity owners with one another, especially exchange between different communities.² Therefore, money implies a minimum level of regularity and complexity in exchange, but it does not require either generalised barter or that most of the output is for exchange.

Marx’s analysis of the forms of value, culminating with the money form, shows that money has an essence, to monopolise exchangeability and to be the universal equivalent. Simply put, money can be exchanged for any commodity, whereas commodities do not generally exchange for each other. In addition to this, Marx’s analysis implies that the functions of money follow from its essence, rather than the converse as is presumed by neoclassical and post-Keynesian theory:

Marx’s approach to money implies that “what money does follows from what money is”: because money monopolizes exchangeability, it also measures value, facilitates exchange, settles debt, and so on. Seen in this way, there is order and internal cohesion to the functions of money no arbitrariness.³

The social and historical determinations of money, including its essence and functions, demonstrate that money is a social relation that derives from the form of articulation between commodity producers. Commodity exchange develops fully only under capitalism, when exchange becomes generalised and impersonal and the ritual and ceremonial aspects of money are largely irrelevant.

Marx distinguishes between money as money and money as capital.⁴ Money as money is the measure of value and the means of circulation, and it fulfils three functions, means of payment, store of value, and international money. Money as capital is money advanced for the production or transfer of surplus value. Money as capital is closely related to money as money, because the forms and functions of money are identical, and money may fulfil both roles simultaneously. For example, in the payment or expenditure of wages, what is C-M-C for the workers is M-C-M’ for the capitalists. Let us now look briefly into the functions and forms of money.

At the highest level of abstraction, money measures commodity values through a simple comparison of their RSNLT with its own, and expresses the result in the units of the standard of prices.⁵ If, for example, the RSNLT of the monetary unit (say, the ounce of gold⁶) is thirty minutes (0.5 hours of socially necessary labour, hSNL), and £1 is its monetary name, the standard of prices is Ω=£1. In this case, the monetary equivalent of labour (MEL, m) is:

(8.1)

This value of m implies that one hour of abstract labour creates a value of £2. It follows that the price of a commodity i produced in five hours (λⁱ = 5hSNL) is:

(8.2)

The determination of prices does not require the actual comparison of each commodity with money; therefore, as a measure of value money is merely ideal money.⁷ It was shown in section 5.3 that the price form expresses commodity values and allows differences between values and prices. There has been much controversy about the relationship between money, value and price in the labour theory of value, especially between Marxian and Ricardian interpretations (see section 2.1).⁸

A less known but equally interesting controversy involved Marx and the ‘Ricardian socialist’ proponents of ‘labour-money’, especially John Gray, John Bray, Alfred Darimon and Pierre-Joseph Proudhon.⁹ This controversy illustrates the importance of properly understanding the function of measure of value.

Gray’s is the best argued case for paper labour-money. His proposed monetary reform derives from the belief that labour alone bestows value and, therefore, that labour should be the measure of values. He argued that the use of valuable commodities (e.g., gold) as money was problematic for two reasons. First, Gray argues that the supply of the money commodity could never increase as rapidly as the supply of all other commodities put together. Therefore, it would be generally impossible to sell the entire output, production would be chronically below potential, and there would be unemployment and deprivation because of the defects of the monetary system.¹⁰ This argument is clearly wrong, because Gray implicitly assumes that the velocity of circulation is always necessarily one, that turnover periods are identical for all capitals, and that hoards and credit are not available. Second, and more interestingly, the Ricardian socialists believed that money veils exchanges and allows the workers to be exploited by the capitalists, and the debtors exploited by the creditors. However, if the workers were paid in labour-money their participation in social production would be ascertained correctly, and they would be able to draw commodities with an equivalent value from the whole of that produce.¹¹

Marx derided the labour-money idea in the Grundrisse and elsewhere, for two reasons.¹² First, if the ‘just price’ paid for the commodities were determined by the concrete labour time necessary to produce them, the economy would fall into disarray as the producers tried to make their commodities more ‘valuable’ by working less intensely. This nonsense stems from the implicit assumption that the normalisation of labours may be avoided, and that their homogenisation can be reduced into an identity between individual labour time and money (see chapter 5).

Second, if the ‘just price’ were based upon RSNLT, however determined, productivity growth would reduce RSNLTs and lead to the appreciation of the labour-money (deflation). This unwarranted outcome would benefit the cursed creditors, reduce investment and delay technical progress. Moreover,

The time-chit, representing average labour time, would never correspond to or be convertible into actual labour time; i.e. the amount of labour time objectified in a commodity would never command a quantity of labour time equal to itself, and vice versa, but would command, rather, either more or less, just as at present every oscillation of market values expresses itself in a rise or fall of the gold or silver prices of commodities.¹³

These difficulties derive from Gray’s inability to understand the necessity of synchronisation and homogenisation of labour in commodity production. When commodities are sold, money is the means of circulation (exchange).¹⁴ Again at the highest level of abstraction, in exchanges C-M-C’ the commodity owners at all times possess the same value, alternating between the original commodity, the money commodity and the newly purchased commodity. This presumption captures the essence of simple (equivalent) exchange. However, the use of gold coins causes their wear and tear (and might encourage the clipping of coins), implying that commodities generally exchange for coins worth less than their face value. The continuity of exchanges under these circumstances shows that circulating money is merely a representative or symbol of value. Symbols of money such as inconvertible paper may, therefore, perform the same service as pure gold:

in this process which continually makes money pass from hand to hand, it only needs to lead a symbolic existence. Its functional existence so to speak absorbs its material existence. Since it is a transiently objectified reflection of the prices of commodities, its serves only as a symbol of itself, and can therefore be replaced by another symbol. One thing is necessary, however: the symbol of money must have its own objective social validity. The paper acquires this by its forced currency.¹⁵

Let us now see how Marx analyses the functions of store of value, means of payment and world money.

Money functions as a store of value when it is hoarded. Hoarding is often justified in the literature because of individual preferences or uncertainty. Although these factors can play an important role in the formation of hoards, there are structural reasons why money hoards are formed in the course of capitalist production.¹⁶ The most important reason is that production involves regular expenditures that are generally disconnected from the accrual of sales revenue. Producers must also accumulate reserves in order to meet unforeseen expenses, maintain and replace fixed capital, expand the output, pay dividends, offset price fluctuations and so on. These idle reserves are normally deposited in the banking system, and they form the basis of the bank reserves. Development of the credit system reduces each capitalist’s hoarding needs, because the hoards of the capitalist class are available to borrow.¹⁷ Bank loans facilitate the realisation of long-term and large-scale investment projects; however, they also facilitate speculative activity and, more broadly, increase the likelihood that localised disturbances (accumulation of inventories, price changes, technical innovations, etc.) will spread and trigger economic crises.

Money functions as means of payment when it settles transactions undertaken previously, and cancels a promise to pay. This is particularly important in commercial credit, that finances the sale of produced commodities, and banking credit, that finances new production.

Finally, the functions of money are performed in the international arena by world money, that is value in pure form and a crystallisation of abstract labour recognised across the globe. National currencies must be convertible into world money in order to allow domestic commodities to be exchanged for foreign goods, and to facilitate the inclusion of domestic labour into the international system of production.

Adjustment of the quantity of money to the needs of circulation is a complex process involving all functions of money.¹⁸ In a simple commodity money system, it was shown above that the value of the money commodity plays an essential role in the determination of prices (see section 8.2). However, at a more complex level of analysis the quantity and velocity of money are important determinants of the expression of value as price.

Marx rejected the quantity theory of money (QTM) in the case of gold because, for him, the quantity of circulating money changes in order to realise the value produced. These changes happen primarily through hoarding and dishoarding, the output of the gold-mining sector, international bullion flows and changes in the velocity of money. For example, if the output grows the additional money necessary for its circulation will be made available through the above channels; alternatively, if the gold stock increases (with all else constant), the additional gold will be hoarded or velocity will decline. In contrast with the QTM (and Ricardo), prices remain unchanged in both cases.¹⁹

It is different for fiat money, which can be issued by the state potentially in arbitrary quantities through the monetisation of budget deficits or open market operations. Marx generally agrees with the QTM that, if an increasing quantity of fiat money is forced into circulation, its exchange value would decline permanently (inflation; see section 8.3). Although fiat money is a suitable means of circulation, it is unsuitable for hoarding in the same scale as gold, because its domestic (and external) exchange value is unstable. This instability derives from the absence of a direct relationship between the supply of fiat money and capital accumulation.²⁰

It is different again with convertible money issued by the banks or the state, and with credit money. The value of convertible money fluctuates around the value of the gold that it replaces, temporary discrepancies occurring naturally during the cycle.²¹ Inflation is possible in this system if the sphere of circulation is flooded with paper notes, but this process is limited because arbitrage makes it impossible for commodity prices to deviate permanently from their gold prices of production. However, this is likely to be neither a smooth nor a purely monetary process. Sudden disruptions of exchange, recessions, and fully-fledged monetary and economic crises, in which the money commodity plays an important role as means of payment and means of hoarding, are some of the ways in which the value of gold is made compatible with the exchange value of money.²²

Finally, contemporary monetary systems include primarily two forms of money, inconvertible paper currency issued by the central bank (legal tender that discharges all debts) and credit money produced by the commercial banks (liabilities of private financial institutions, including deposits and banknotes, offering a potential claim on another form of money). The quantity and the exchange value of credit money are indirectly regulated by the advance and repayment of credit, that is, by the processes of production and accumulation and, at a further remove, by the central bank’s influence on the operations of the financial system.²³

Temporary discrepancies between the supply and demand for credit money are inevitable, for two reasons. First, and more generally, the empirical determinacy of the quantity and velocity of money declines as the analysis becomes more concrete. They depend on social conventions, including the property relations, the financial rules and regulations, the structure of the financial system and its relationship with production, the international relations, the degree of concentration of capital, and other variables that make ‘supply’ and ‘demand’ difficult to determine even theoretically. Second, and more specifically, even though the supply of credit money necessarily corresponds to individual demand (credit money is always created in response to a loan request), the total credit supply may not reflect the needs of the economy as a whole. This is clearly the case when speculative loans help to inflate a real estate or stock market bubble, or when banks unwittingly finance the production of unprofitable or unsaleable goods. Excess supply is likely especially when a climate of optimism is fostered by rises in the prices of financial assets, which feed upon ongoing optimism and increase it even further. In other words, excess credit, fuelled by speculation or the surge of accumulation, may lead to price increases but, barring state intervention (see section 8.3), this process is limited by the unavoidable increase in financial instability and the possibility of crisis.²⁴

These mediations in the determination of the exchange value of money do not imply that it is wrong to posit, ex post, a monetary equivalent of labour, as in equation (8.1). However, focus upon the MEL tends to conflate levels of abstraction, and it may obscure the contradictory elements in its determination, in which case financial instability and the possibility of crisis have to be reintroduced at later stage arbitrarily and, potentially, in an impoverished manner (see section 2.2.2).

8.2 Money and prices of production

The transformation of values into prices of production (see chapter 7) has important implications not only for Marx’s analysis of the forms of social labour, but also for his monetary theory. The contemporary predominance of inconvertible (valueless) paper money poses a challenge to one aspect of Marx’s theory: it is not immediately clear how commodity values are measured, and expressed as price, if money has no intrinsic value.²⁵ This potential limitation has severe implications, because analysis of contemporary problems including the supply of credit money, inflation, and exchange rate determination, depend upon a satisfactory explanation of valueless paper money.

Some writers have refused to address this problem, either because they see inconvertible paper money as an illusion or a temporary aberration,²⁶ or because they postulate that money is never a commodity and is, instead, always created by the state.²⁷ However, the latter fails to explain how valueless money measures value. This section develops Marx’s analysis of the value forms, showing that it is fully compatible with inconvertible paper money. Limited to this aim, it does not review the historical process of displacement of precious metals from circulation, or the structure of contemporary monetary systems.²⁸

The explanation of valueless money departs from equivalent exchange. At the highest level of abstraction (see section 8.1), equivalent exchange includes commodities produced in equal RSNLTs, C-C’. The mediation of exchange by money, C-M-C’ modifies the meaning of equivalent exchange, because the commodity owners no longer necessarily hold the same value continuously, even in the case of gold money (because of abrasion). More generally, mediation of exchange by convertible paper money shows that what matters in exchange is not the intrinsic value of money but its exchange value (we ignore contingent price fluctuations).

The concept of equivalent exchange shifts again with the transformation of values into prices of production. After the transformation, equivalent exchange no longer involves commodities produced in equal RSNLT. Rather, it involves commodities whose production yields the same profit rate.²⁹ By the same token, equations (8.1) and (8.2) no longer generally hold, the MEL can be determined only after commodity prices (rather than before, as used to be the case), and it holds at the aggregate level, but not for every commodity. More generally, prices of production are not determined through a one-to-one ideal relationship between commodity values and the value of the money-commodity. Rather, they are determined simultaneously by the rate of valorisation of the advanced capitals (see sections 2.1.2 and 5.3):

p=(pA+wl)(1+r)(8.3)

This equation reflects Marx’s view that ‘price of production is … in the long term … the condition of supply, the condition for the reproduction of commodities, in each particular sphere of production.’³⁰

The shift in the price form due to the transformation implies that no single commodity can fulfil the function of measure of value. At this stage in the analysis, value measurement includes the assessment of the growth rate of the advanced capitals and the establishment of equivalent exchange in the relatively complex sense above, through a consistent relative price system. In other words, the money-commodity no longer measures values independently of the other commodities and production processes, as was the case previously (when absolute prices were logically determined before the relative prices).

After the transformation, the value of gold, its value composition, turnover time, and so on, are relevant only for the determination of the absolute price level (absolute prices are now logically determined after the relative prices). This is the case because, at this level of analysis, the measure of value is no longer the money-commodity but the general profit rate, which is the peg of the relative price system. Equation (8.3) shows that prices of production are determined by the current price of production of the inputs, marked up such that each capital draws the average rate of profit (this rate is limited by the total surplus value, see section 4.1).³¹ At this stage in the analysis, gold money may be abolished, with no bearing upon the stability of the economy or our ability to understand it.

Once gold is withdrawn from circulation absolute prices can be maintained at the previous level (which is usually the case when governments decree that the currency is no longer convertible into gold), or shift to any arbitrary level (if there is a monetary reform). A developed capitalist currency is, therefore, the complex unity of a measure of value (the general profit rate) and a medium of exchange (which may take the form of gold, copper, paper, electronic impulses or whatever), and that fulfils the functions identified in section 8.1. Even under a gold money system, the analysis above shows that, under capitalism, gold is never the sole measure of value nor an adequate means of circulation. In the development of theory, however, the role of commodity money is indispensible.

8.3 Credit, money and inflation

Marxian analyses of money, credit and crises can be developed in different directions, in order to illuminate a broad range of contemporary phenomena. This section illustrates the potential usefulness of these approaches through a critical review of three theories of inflation emphasising, respectively, distributive conflicts, monopoly power and state intervention on the dynamics of credit money.³² This is important for three main reasons. First, inflation poses an intriguing theoretical challenge.³³ Mainstream analyses, usually inspired by the QTM, have unacceptably weak foundations, including perfect competition, full employment and costless adjustment between static equilibria. In contrast, Marxian (and other political economy) contributions are promising, but remain relatively undeveloped. Second, advances in the understanding of inflation can easily be extended to the study of deflation, and both are currently important.³⁴ Third, inflation and conventional anti-inflation policies usually have high economic and social costs. They often lead to higher unemployment, lower real wages, higher rates of exploitation and shift the income distribution and the balance of social forces towards capital and, especially, financial interests. It would clearly be important to develop alternative analyses, in order to confront inflation and the consequences of conventional anti-inflation policies.

Two difficulties have frustrated attempts to develop Marxian analyses of inflation. First, inflation is a highly complex process that involves a wide range of determinants at different levels of abstraction, among them production, the supply of money, interest rates, the industrial and financial structure, external shocks, distributive conflicts, and many other variables. It is very difficult to order these influences systematically within a cogent theory. Second, it is especially difficult to explain inflation in inconvertible monetary systems, drawing on the anti-quantity theory tradition of Steuart, Tooke, Marx, Kalecki, and most post-Keynesians writers. Simply put, it is difficult to develop a theory of inflation whilst simultaneously preserving the claim that the needs of production and trade call money into circulation (endogeneity), and admitting that money may influence ‘real’ variables (non-neutrality). This exercise becomes even more complex when it involves different forms of money, issued by the state and by the commercial banks, each of them with a particular relationship with capital accumulation. In spite of these difficulties, this section shows that it is possible to outline the general conditions for inflation.

Conflict inflation

Non-mainstream economists of very different persuasions, including many Marxists and most post-Keynesians and neo-structuralists, argue that distributive conflicts are usually the most important cause of inflation (this approach is appealing to some Marxists because it apparently vindicates the notion of class struggle).³⁵

Conflict analyses are inspired by cost-push theories, which were popular between the 1950s and the 1970s. They usually depart from equilibrium, and assume that the money supply is fully endogenous, that fiscal and monetary policies are passive, and that key agents (especially the monopoly capitalists and unionised workers) have market power and can set the price of their goods or services largely independently of demand. Inflation arises because the central bank validates incompatible demands for shares of the national income through monetary accommodation or its support for the financial system, in order to ensure financial stability and the continuity of production.³⁶ The inflation rate is usually a positive function of the size of the overlapping claims, the frequency of price and wage changes and the degree of capacity utilisation, and a negative function of the rate of productivity growth (the basic model can be refined endlessly by incorporating target income levels, expectations, reaction functions, and limits on the wage claims because of unemployment, or on the mark up because of competition).

The most important shortcoming of the conflict approach is the absence of a clear internal structure. This approach is compatible with widely different theories of value, production, employment, demand, income and distribution, and with different rules of determination of the target income levels. Classes are sometimes seen as partners, in which case it is relatively easy to achieve economic stability through negotiated incomes policies. Alternatively, a theory of exploitation may be used; in this case, economic stability can be achieved only through the subordination of the workers by force. This flexibility makes conflict analyses potentially appealing to a wide audience; however, it is vulnerable to the charges of arbitrariness and lack of analytical rigour. In particular, inflation generally starts from a dislocation that shifts the economy away from a Pareto-optimal equilibrium. ‘Apportioning blame’ is, therefore, implicitly at issue, and alternative economic policies are usually assessed in terms of their ability to make the economy return to the initial equilibrium. It is not usually explained how that equilibrium was originally determined, or why it merits return. Moreover, it was shown in sections 2.2.2 and 4.1 that capitalists and workers do not confront each other directly over the shares of the national product, firstly because the wages are advanced, whereas profit is the residual and, secondly, because disputes generally involve income levels rather than shares.

Indeterminacies such as these can be eliminated only through an organic relationship between the conflict approach and a broader economic theory. Unfortunately, many such connections are possible, and none is necessary. In other words, conflict theories, as they are usually presented, are typically ‘middle range’.³⁷ They derive from a set of stylised empirical observations (e.g., agents exercise claims over the national product through the sale of their goods), and transform these observations into structures that are used to explain these stylised facts (e.g., distributive conflict leading to inflation). This approach conflates cause and effect, because it presumes that, since inflation has distributive implications, income disputes cause the process; and the analysis is unsound because it is not grounded by a broader structure that supports its elementary concepts and contextualises its conclusions. The lack of a theory of production implies that the state’s role and policies cannot be adequately grounded either, and they usually derive from a further set of stylised facts. Consequently, the rationale for, and the power of, economic policies are left unexplained and depend heavily on the analyst’s preferences.

In spite of these important limitations, the conflict approach is potentially relevant. Distributive conflicts must be part of any inflation theory, for inflation would not persist in the absence of dissatisfaction about the level and/or distribution of the national income, and the monetisation of these claims.

Monopoly capital, underconsumption and inflation

Many Marxists argue that inflation is associated with the increasing market power of large corporations and underconsumption, most clearly for the monopoly capital school.³⁸ This approach argues that monopolies are the most dynamic firms and the largest investors, employers, producers and exporters. In order to maximise economic growth the state supports the monopolies through purchases, cheap infrastructure, tax breaks, subsidies for research and development, and so on. More broadly, the state spends huge sums in civil servants’ wages, consumables and public investment, funds health, education and defence expenditures, and makes large transfers associated with social security. These expenditures support monopoly profits directly through purchases, and indirectly through transfers to their customers. Interventionist policies of the welfare state delivered unprecedented economic stability, high employment and rapid growth, especially between the late 1940s and the late 1960s. However, they also contributed to persistent budget deficits, rising public debt and creeping inflation. In sum, inflation is the result of interventionist economic policies trying to ensure full employment and social stability, in an economy constrained by monopoly power and pricing strategies.³⁹

These are important insights, but this approach is theoretically fragile. It does not include a theory of monopoly power or pricing, other than a collation of the ideas of Hilferding (for whom monopolies impose prices above the prices of production in order to reap extra profits) and Kalecki (for whom monopoly power is a stylised fact and monopolies reap extra profits because of their market power).⁴⁰ The influence of monopoly on the circuit of capital and income distribution is not explained, and the role of demand and other limits and counter-tendencies to the concentration and centralisation of capital are almost invariably ignored.

The theory of the state is also left unclear, and what is said is potentially contradictory. On the one hand, the state manages the economy relatively autonomously in order to ensure the reproduction of capital as a whole, which requires the accommodation of the interests of different fractions of capital and of the workers, and is best achieved in a democracy. On the other hand, the state is a tool of powerful (monopoly) interests, and its policies are limited by the need to obtain their consent, in which case fascism is a clear possibility. Finally, the linkages connecting monopoly power, state policies and inflation are left mostly unexplained. There is no clear theory of money, credit or finance, except for the presumption that money supply responds passively to monopoly demand or to state command, and that (largely unexplained) financial developments are contributory factors. How this leads to inflation is left unclear.⁴¹ More generally, the causes of inflation shift between monopoly pricing decisions and excess demand induced by the state (which is, paradoxically, the result of state attempts to avoid underconsumption).⁴² The distributive impact of inflation is not analysed, except to argue that monopolies benefit at the expense of the workers and other groups receiving nominally fixed revenues. It is unclear how this relates to the theory of wages or of exploitation.⁴³

Extra money inflation

In the mid-1970s an alternative analysis was outlined, in which inflation is the result of a permanent increase in the relationship between commodity prices and values, caused by a discrepancy between the supply and social demand for money.⁴⁴

The analysis departs from the circuit of capital. The productive circuit begins when capitalists draw on previously accumulated funds or borrow newly created credit money in order to finance production. Injection of these funds into the economy increases the ratio between circulating money and output value. If more output is produced and sold additional income is created, which cancels out the initial shift in the relationship between money and value. However, if the output cannot be sold at its price of production the firm suffers a loss that may be absorbed in two ways. If ‘market rules’ are respected, a well-defined set of agents bears the cost, usually the firm or its bank. This type of solution can be destabilising, because it may systematically lead to unemployment, capacity underutilisation, the deterioration of the working conditions, and financial fragility.

Alternatively, the loss may be socialised if the debt is refinanced or if the firm receives a state subsidy (in the extreme, it may be nationalised and ‘restructured’ with public funds). In either case, there is an injection of purchasing power that perpetuates the initial discrepancy between the circulating money and the output; in other words, the initial (presumably transitory) increase in the monetary equivalent of labour becomes permanent. The money injected into the economy through a violation of ‘market rules’ is extra money.⁴⁵ Extra money may also be created by central bank support to the financial institutions, by non-sterilised balance of payments surpluses, or by corporate or household dissaving or borrowing for speculative purposes.⁴⁶ Extra money typically increases the nominal income or the liquid wealth of the consolidated non-financial sector in spite of the constant value of the output, and regardless of the existence of equilibrium, currently or in the past. If the extra money induces a quantity response, the previous relationship between value and money may be restored; otherwise, the monetary expression of labour rises: this is extra money inflation.

Extra money inflation may be facilitated by the monetary policy stance, but the state cannot be generally ‘blamed’ for it because extra money is routinely and necessarily created by private decisions that are not subject to state control (including bank loans). Moreover, even if the extra money is created by the state it is impossible to know in advance whether it will have a quantity or price effect, or both (targeting is possible, but necessarily imprecise). In due course, discrepancies between the quantity of circulating money and demand will tend to be eliminated by changes in output, velocity or hoards. However, these adjustments take time, and they may create additional instability through their effects on prices, the exchange rate, the balance of payments or the interest rate. If these monetary discrepancies are continually renewed, they can lead to persistent inflation, severe balance of payments disequilibria and prolonged economic stagnation, which demonstrate the non-neutrality of money and its potential influence over the accumulation process.

Long-term inflation may derive from the attempt by the state to deliver continuous economic growth, or from the attempt to avoid deflation when growth falters. More generally, in the upswing, extra money is provided mainly by the private sector with the support of the central bank, in order to finance consumption and new investment. Therefore, growth necessarily breaches the established relationship between value and money, and it is always potentially inflationary (depending on the supply and import responses). As the economy grows, disproportions and bottlenecks inevitably develop, financial structures become more fragile and, unless cheap imports are readily available, prices (and, possibly, wages) tend to increase. At this stage, the crisis erupts either spontaneously, because of the balance of payments constraint or because contractionary policies have been adopted. If the crisis becomes acute and deflation looms, the state will usually intervene and deliberately inject (or facilitate the private creation of) extra money.⁴⁷

In spite of their apparent similarity, the theory of extra money inflation is incompatible with the QTM. The quantity theory’s assumptions that money supply is exogenous, that money is only a medium of exchange and that money is not hoarded are unacceptable from the perspective of the extra money approach. First, this approach argues that extra money is regularly and spontaneously created by the interaction between the central bank, commercial banks and firms, and that its quantity cannot be controlled, or even known precisely, by the state. In contrast, the quantity theory presumes that the banking system is always fully loaned up, and that the central bank can determine autonomously the supply of money directly (through the monetisation of government budget deficits or purchases of government securities) or indirectly (through changes in compulsory bank reserves, which should lead unproblematically to changes in the outstanding stock of loans). Other sources of changes in the supply of money are usually ignored, and the possibility that changes initiated by the central bank may be neutralised by hoarding, loan repayments or by a compensatory change in bank loans are generally neglected by the QTM.

Second, extra money is non-neutral in the short and in the long run; it may change irreversibly the level and composition of the national product and the structure of demand, depending on how it is created and how it circulates. In contrast, the QTM presumes that money is neutral in the long and, in extreme cases, even the short run. Third, the effects of extra money (whether quantity, price, or both) cannot be anticipated. All that one can say is that high capacity utilisation and activist state policies increase the probability of extra money inflation, but there is never a simple relationship between them. In contrast, for the QTM the relationship between money supply and inflation is usually straightforward. Because of the underlying assumptions of perfect competition, full employment, and money neutrality, a change in the supply of money (initiated by the central bank and automatically propagated by the commercial banks through the money multiplier) unproblematically leads to a predictable change in the price level.

The extra money approach can provide the basis for the development of a theory of inflation which incorporates the main claims of the labour theory of value, and the valuable insights of other Marxian analyses of inflation. However, this approach is still undeveloped at critical points, and it suffers from deficiencies and ambiguities that need to be addressed urgently. For example, the analysis of the supply of central bank and credit money is usually simplistic, and it would benefit from greater exposure to, and confrontation against, recent post-Keynesian developments,⁴⁸ circuitist contributions,⁴⁹ and the works of Kalecki.⁵⁰ At a more concrete level of analysis, the valuable contributions of Minsky on the financial instability of modern capitalism need to be evaluated in detail and incorporated into the analysis when this is warranted.⁵¹

In addition to this, much work remains to be done in order to make the structures and categories in the extra money approach fully compatible with those of Marx’s theory of value. For example, the relationship between the supply of money and the monetary expression of labour is usually left unclear, the extra money approach often shifts arbitrarily between levels of analysis and leaves ambiguous the stature of competition. Finally, further work is necessary to distinguish between price increases caused by extra money and those caused by other types of money supply growth. This would help to clarify the residual ambiguity between the extra money approach and the quantity theory of money, especially with respect to the role of excess demand as the trigger of inflation.

Addressing these issues systematically will make it possible to incorporate other important phenomena into the analysis, for example financial development and financial and capital account liberalisation. It will also make it possible to analyse concrete problems such as the potentially inflationary impact of the public debt overhang, whose increasing liquidity may be synonymous with the injection of extra money into the economy.⁵²

8.4 Conclusion

Marx’s theory of money has often been examined exegetically, as if it were fully developed and significant only because of the succession of value forms in chapter 1 of Capital 1. This viewpoint is infertile. This chapter has shown that Marx’s theory can be developed in important ways, including the explanation of inconvertible money and inflation. The former is important because it shows that Marx’s approach is internally cogent and it does not conflict with the facts of modern capitalism. The latter is an important current problem; tackling it creatively and consistently is relevant politically, and it demonstrates the vitality of Marx’s approach.

This chapter has also shown that the transformation of value into prices of production modifies money’s function as measure of value, and the homogenisation of labour. It does not affect the other functions of money, or the normalisation and synchronisation of labour. Finally, this chapter has shown that the theory of inflation should focus upon the same concepts. Inflation is a macroeconomic process that affects the expression of value as price, and it influences the relationship between total output and money.

The analysis of inflation needs to be developed much further, but some of its policy implications are already clear. First, inflation can be functional, but its dysfunctional aspects gradually tend to become predominant when inflation rises. In particular, economic calculus becomes increasingly complex and capital restructuring becomes more difficult because inefficient capitals and productive processes are preserved, rather than being annihilated by ‘market’ processes. Second, inflation leads to financial crisis by its cumulative character, through the formation of increasingly unstable debt structures. Crises may be postponed by increasing the supply of extra money, but this may lead to hyperinflation. Third, there can be inflation purely for monetary reasons, usually associated with speculative bubbles involving housing, the stock exchange and other assets, which can harm real accumulation by draining it of funds. Fourth, permanent inflation is not inevitable, whatever the power of the banks, monopolies or the workers. However, financial deepening, the concentration of capital, the reduction of trade flows, and worker militancy increase the vulnerability of the economy to inflation, and the difficulty to reverse the process once it is under way.