Circuitism – its monetary theory and model of the
money circuit


The literature of Circuitism deals with two different subjects.[1] One subject is monetary theory, i.e. the functioning of the present-day money and banking system. The other subject focuses on the circulation of money. The Circuitist model examines how money originates in the banking sector by being credited to firms, circulates from the firms to wage earners, and flows back via the firms to the banking sector whereupon the money is extinguished.

Circuitism distances itself from neoclassical economics as well as from orthodox Keynesianism. Notwithstanding, Graziani makes positive reference to Keynes (but 'mostly the Treatise on Money, much less the General Theory'[2]) as well as to Kalecki and later representatives of Postkeynesianism (Moore, Minsky, Kregel, Davidson). He also acknowledges earlier contributions by the Swedish and German schools, e.g. Wicksell, Schumpeter and Hahn.

Circuitism is seen as a Postkeynesian theory, or an offspring of Postkeynesianism. This certainly applies to the monetary theory. The model of the money circuit, however, as will become apparent in the second part of this paper, looks rather like a new classical and Neomarxist construction of the money-mediated economic process dominated by banking capitalism. For the rest, the analytical perspective tends to be supply-side (from the firms' point of view) rather than Keynesian demand-side.

Under the angle of New Currency Theory as pursued on this website, there is agreement as well as partial disagreement with the monetary theory of Circuitism. With regard to the model of the money circuit, it appears as if that 'circuit' were just another case of over-simplified model-building.

> back to top

Monetary theory

Under descriptive aspects the monetary theory of Circuitism reflects the state of knowledge of Postkeynesianism in recent decades. It builds on the notion of endogenous credit money in the form of deposits. Bank credit creates deposits, while the reverse, deposits serving to fund bank credit, does not hold true, except for formally independent investment units of banks such as investment trusts; these, however, are nonbank financial institutions.[3] Modern fiat money has no coverage by other monetary items. It is pure purchasing power in general and regular use.[4]

Circuitism has developed an explicit understanding of what it means to have a monetarised and financialised economy, in contrast to a mere barter economy, as is characteristic of neoclassical theories of market equilibria, that operates on 'exogenous' money and finances itself on the basis of recycled savings only. The money, or capital respectively, that pre-finances production and trade, however, is basically created as primary bank credit. Its existence, according to Circuitism, alters the structure and inner workings of the economy. In general, as Graziani says,

'money is never neutral' and 'is, at the economic level, a source of profits and, at the social level, a source of power.' – 'Since access to money and credit is a key factor in a wage economy, producers of money and credit … enjoy a privileged position and are admitted as such to a share of total product.'[5]

According to Circuitist teaching, modern economies depend on how credit and money are managed. Likewise, the levels of employment and income are thought not to be determined by relative prices, but by the conditions of credit funding that are jointly decided upon by banks and firms, with the banks holding the whip hand. This stresses the Circuitists' view of 'the power of banks' and the overriding position of the banking industry as a pivotal player in the economy.[6]

Graziani states on various occasions that central banks, as core institutions of the monetary system, must not be merged with 'the government' into one and the same category, i.e. the public sector, as opposed to the private and foreign sectors. Lumping together government and central bank creates confusion about monetary and fiscal functions and is bound to result in unreal ideas about the creation and circulation of money.

According to Circuitists, government has no role in creating money today, as 'the monetary base is being created by the banking system, and not as being the consequence of a government deficit'.[7] Whether this still is the case indirectly, remains open to question. The relevant aspect here is Graziani arguing in favour of a separation of monetary and fiscal functions, not to say separation of state powers between the government and the central bank. This is contrary to Keynes-inspired models of public-private sector balances, and the idea that government debt equals the creation of sovereign currency. This idea is to be found in the approach to sector balances by Godley/Lavoie, and more pointedly so in Modern Money Theory (MMT), another strand of Postkeynesian origin whose monetary theory otherwise overlaps with Circuitism in a number of aspects.[8]

> back to top

Neglect of the dysfunctions of banks' primary credit creation

Despite the Circuitists' emphasis on the power of banks, and recognition of the bias towards privileged financial-capital revenue, there is, as is the case in Keynesianism in general, no criticism of the banking industry's systemic position as a monetary power. Circuitist theory seems to content itself with the description of the money and banking system, while being reluctant to evaluate the situation and give policy advice. Even if it could be read into some text passages that Circuitists may want to curb the power of banks, they have not, to my knowledge, contemplated reshaping the monetary and banking system.[9]

Still more remarkably from a New Currency point of view, Circuitists see no fault in fractional reserve banking and do not attribute financial instability and recurrent banking and financial crises to the regime of bank money such as it stands today. This is certainly typical for standard textbook economics and Keynesianism, but also for most scholars of Postkeynesianism, through to Circuitism and MMT. Quantities of money and GDP-disproportionate overshooting of pro-active primary credit and thus money creation are not an issue. This is a rare case of collective neglect of an issue – the quantity theory of money – that used to be a common topic for centuries. With special regard to Circuitism this also results from its version of the Banking-School's real bills doctrine as discussed below.

> back to top

Ambiguity regarding the interplay between banks and central bank

In various passages Graziani is unclear, as Keynes was, as to whether banks or central banks have the lead in creating money. In both Keynes and Circuitism there seems to be a prevailing belief in the effectiveness of central-bank reserve positions and base rates for controlling, or at least influencing, the rate and pace of banks' credit creation. Graziani even seems to refer to the multiplier model (rightly given up in much of Postkeynesianism) when he explains that the Circuitist theory 'stresses the fact that the credit potential of the banking system depends on the monetary base, the reserve ratio'.[10]

On the other hand, Graziani endorses the banks' ability for pro-active primary credit creation, explicitly referring to Keynes's statement in the Treatise on Money that 'there is no limit to the amount of bank-money which the banks can safely create, provided that they move forward in step.' And, it has to be added, as long as central banks always accommodate the banks' demand for reserves.

On balance, Circuitism remains unclear about the question of who has the pro-active lead in creating money; unlike the accommodationist view of Postkeynesianism, to which there is no doubt that the banking industry has the lead in creating credit and deposits, while the central bank of a currency area reactively accommodates the banks' demand for fractional re-financing in the form of reserves and residual cash.[11] In consequence, primary bank credit determines the entire money supply. The creation of bank money certainly depends to a degree on the demand for money from firms, government and households. This, however, does not alter the position of the banks as the pivotal actors in the money supply chain, in that they decide selectively on whether, how much, for what and for whom they create primary credit, thus bank money. 

> back to top

Token money, paper money, and credit

There is another ambiguity when Graziani says that a monetary economy

'must be using a token money, which is nowadays paper currency.' – 'Nowadays, money is paper money introduced into the market by means of bank credit'.[12]

This can be misunderstood. Bank credit creates deposits (bank money-on-account), not banknotes (paper money). As far as customers demand payout of deposits in cash, the banks need to obtain the coins and banknotes from the central bank, whereby the monopoly of coining rests with the national treasuries, while the central banks have the monopoly on banknotes since the 19th century. In contrast to bank money, the banks have to finance paper money and coins to 100%, not just fractionally.   

Moreover, Graziani's 'nowadays' is now a long time ago. Depending on the country, paper money came into use in Europe around 1700 and became the predominant form of token money from about 1800 to the middle of the 20th century. However, throughout that time it never existed in its own right (except for a few short periods as colonial bills, continental dollars and Greenbacks in the history of the US).[13] Paper money was not constitutive for the monetary system, but rested on traditional silver and gold currency, or a central national reserve of gold bullion under the gold standard from 1844 to 1971, since then just on primary central-bank credit (= central-bank money = reserves).

At the source, modern money is non-cash money-on-account in a bank or central-bank account. Physical cash, as long as it remains in use, is exchanged out of and back into the basically non-cash money supply. For the time being this also holds true for e-cash. Whether e-cash, while physical cash is vanishing, will become another privileged domain of the banking industry that is ever more detaching itself from central banks and legal tender, or whether e-cash as legal tender will serve to restore the sovereign monetary prerogatives of the currency, money and seigniorage, this is being decided today in the further course of contemporary history.[14]

> back to top

Wrong identification of token money and credit money

A fundamental reason for not seeing problems with fractional reserve banking seems to be that teachings in Keynesian filiation identify token money with credit money, and take this as the unquestioned natural state of affairs. For once, though, Graziani makes a distinction between money and credit (as any Currency scholar will do), in that he declares money to be more than credit:

'something different from a regular commodity and something more than a mere promise of payment; ... money has to be accepted as a means of final settlement of the transaction, otherwise it would be credit and not money.'[15]

The observation that credit creates a mutual obligation to pay, whereas the transfer of money discharges an obligation to pay, captures an important aspect of the matter.

Beyond that passage, however, and much like Postkeynesianism and MMT, Graziani falls back to an absolute identification of money with credit, that is, token money with credit money, as an alleged historical fact and necessity from the archaic beginnings of civilisation. This ignores about 2,500 years of coin currencies where the rulers of a realm – the pre-modern State, in a sense – enjoyed the genuine seigniorage from minting coins and spending these into circulation free of debt; which of course can be done in much the same way and more easily with modern money-on-account and mobile e-cash. Credit money certainly is token money, but token money is not necessarily credit money.

> back to top

The rhetoric about endogenous and exogenous money

Postkeynesianism has developed the notion of endogenous and exogenous money, and holds the view that money (credit) in the modern economy is endogenous.[16] This can be endorsed, and yet may be misleading in a specific sense.

The distinction can be traced back to a narrative created by Adam Smith and lateron also Carl Menger in the 1870s (neoclassical Austrian School). According to this narrative, money is imagined to have emerged as a spontaneous creation in archaic barter and early market processes, originally as commodity money (livestock, grain, salt, silver), then coins made of silver, copper and gold, and more recently also as credit-based paper money originated by individual market participants. At the time of Smith and Menger, this idea was directed against money creation by alleged 'outsiders' to the economy, in particular political or religious authorities.

What is really known about archaic and traditional societies, however, provides evidence to the contrary. In the archaic beginnings, money was developed as a unit of account for documenting and clearing claims and obligations (debt, tributes). This took place in the extended household-economies of the worldly and religious rulers of the time and the related chains of provision. When coins were introduced much later, 2,700 years ago, coining was under the control of those rulers from the beginning.[17] The sequence of commodity monies > precious-metal coins > credit-based paper money is probably correct, not however the postulate of 'spontaneous' barter and market economies including the 'spontaneous' creation of money. Rather, the economy developed around the courts and temples of the rulers of a realm, under their control, including control of the monetary and financing practices at subsequent stages of development.

Seen in the light of historical knowledge and contemporary facts, it is not any economic agent who can create 'endogenous' credit or money. Today, only banks and national central banks, that is, monetary institutions, create money. (Treasury coins now count for less than 1 per cent of the money supply). Private currencies, such as not-for-profit complementary currencies or speculative bitcoins, and other private means of payment beyond bank money do exist, but are not used as a general and regular means of payment. Unlike bank money they are not official money, i.e. de facto authorised money in addition to legal tender. Also payment in kind, or the transfer of bills of exchange, debentures or other financial assets in lieu of payment in official money are special cases representing exemptions to the rule. With the development of electronic payment systems run by the central banks, payment in official money is the rule more than ever before.

In neoclassical and Keynesian mainstream economics from about the 1920s, the narrative of 'money from outside vs inside the economy' was specified – or say, re-interpreted – so as to label legal tender from the national central bank or the treasury as 'exogenous', whereas bank money is deemed 'endogenous'. This is reflected in the two-tier model of the banking system, which, more precisely, is a double-circuit money system comprising the public circulation based on bank money, and the interbank circulation based on central-bank money (reserves).

The split between the two also expresses the prevailing situation of incomplete chartalism.[18] This means there are nation-state currencies, while the money supply consists of state money and private bank money in parallel, with the bank money over time having come to dominate the entire system. The situation has not been questioned since the interwar period. There may be some controversy on whether money creation is led by the supply side or the demand side. But only Postkeynesianism, Circuitism and the New Currency perspective of the monetary reform movement have posed the question of the extent to which 'the power of banks' dominates the system and determines the creation of money.

The Postkeynesian notion of endogenous vs exogenous money is nonetheless fallacious. Banks and central banks both create credit money in basically the same way. Both of them do it on demand. The banks, however, apply selective supply policies of their own, including proprietary trading beyond customer demand. The central banks today, by contrast, deliver as much money as the banks are demanding. Presently, central banks no longer intend to exert control over the quantity of money. If bank money is seen as endogenous in the economy, so too must central-bank money. If central-bank money is seen as exogenous to the economy, so too must bank money.

Considering the status of bank money as endogenous and that of central-bank money as exogenous is purely arbitrary. It represents ideological labelling which makes banks appear as 'insiders' of the economy, whereas the central bank appears to be an alien outside agency, similar to the way in which many economists see the role of government. This still goes back to the bourgeois ideology of the market economy as an 'extra-territorial' Robinson island beyond the state and society, based on private law with no role for public or state law. Amid all the justified criticism of the feudal state and mercantilism of the 17th to the 19th centuries, the fundamental and indispensable role of the state in modern societies, including the money system and creating a legal framework for the economy and finances, was not properly understood; which often enough is still the case today.

Speaking of 'exogenous' money would only make sense if an amount of money would be given, somewhere from an elusive 'economic outside', prior to the economic process without dynamically changing with the demand for and the supply of money. Exogenous money in this sense, however, does not exist in a modern economy. If something that comes close to an exogenous money supply did ever exist, it was the silver and gold of traditional coin currencies, andin concept, not in realitynational gold hoards under the old-industrial gold standard. Present-day fiat money, however, is always endogenous. In consequence, the distinction between exogenous and endogenous money is prejudiced and confusing.

A distinction analogous to endogenous vs exogenous and of the same meaning is the one between outside money (issued by the central bank and, maybe, the treasury) and inside money (issued by the banking industry).[19] The difference between the two wordings seems to be that 'outside money' is considered the more reliable, higher ranking asset in contrast to bank money, because 'outside money' comes from the central bank as the ultimate source of money and is also backed by the government, whereas banks in crisis are backed by no oneexcept their central bank and government. As a specification this is certainly correct. The terminology of inside vs outside nonetheless reproduces the ideological dictum according to which the banking industry is seen 'inside the markets' whereas the central banks are shunt off to an unreal, in fact non-existent position 'outside' the money and capital markets.

> back to top

Incomplete picture of credit creation                        

The Circuitist view of the creation of bank money is incomplete in two ways. Firstly, there is too narrow a focus on bank loans; secondly, bank loans in the Circuitist model just flow to firms. The latter aspect is dealt with in the next chapter.

With regard to the first aspect, the Circuitist view misses the fact that banks create primary credit not only by way of making loans and granting overdraft, but equally by purchasing securities or real estate, and even by paying for salaries, services, equipment and materials. A payment from a bank to nonbanks creates deposits, while payments from nonbanks to a bank delete deposits.      

In this respect, Graziani held a traditional view: 'A bank cannot buy commodities by means of its own credit (if it did so, it would require commodities from the market without giving anything in return).'[20] But of course, yes, banks can. They actually have to, because, since banks do not pay for salaries etc. in cash, they pay by crediting accounts, and when banks credit accounts they create primary bank credit and deposits. However, when banks receive payments from proprietary transactions in bank money, this creates an entry in the earnings account, while the deposits are deleted at the payer's bank and cease to exist.

More generally speaking, whenever a bank credits a giro account - held by nonbanks, nonbank financial intermediaries, or other banks apart from their central-bank account - new bank money is created, irrespective of the purpose of the transaction. Whenever payments are made to a bank from a giro account held by these actors, bank money is deleted.

How many excess reserves (payment reserves) a respective bank will ultimately obtain in the process, or will have to pay on balance, depends on the entirety of outgoing and incoming payments in both proprietary and customer transactions. In actual fact, the base of reserves involved in the process will be just a fraction of bank-money turnover at almost any point in time, even more so, the bigger a bank is (or a banking union with a central clearing unit).
Graziani obscures the matter further when saying that

'banks need to make use of interest payments made by firms in order to pay wages and salaries to their employees, buy commodities on the market, and possibly pay interest on deposits.'[21]

Interest payments to a bank, however, cannot 'be used' by that bank for making subsequent payments. Banks certainly need to keep up a certain balance of incoming and outgoing reserves (in order to avoid costly liquidity shortages). It does not make a difference, however, where the reserves come from, from proprietary or customer transactions, and whether a bank has enough reserves available when a payment has to be carried out, or whether the reserves have to be taken up, upon payment or afterwards, from the central bank or in the interbank market. In any case, banks do not need to have received a particular amount in interest payments in order to be able to pay salaries or purchase securities.

There is, however, an important difference between two kinds of transactions. One is bank purchases of securities, foreign exchange, derivatives, real estate, gold, commodities, and long-term equipment. The other is expenditure on salaries, services and materials. The difference between the two is that the former types of transactions (securities, etc.) can be booked as an asset on the balance sheet, whereas the latter transactions (salaries, bonuses, services, operational costs) have to be booked as an expenditure à fonds perdu in the profit-and-loss account, with no additional asset as counterpart, thus one-sidedly debiting the equity. The final profit or loss of a bank, a surplus or deficit in its equity account, depends on the balance of its earnings and expenditures as well as on gains or losses in the value of its assets.  

> back to top

The model of the money circuit 

The Circuitists' model of money circulation consists of a four-step sequence from money creation to its deletion:  banks credit firms > firms pay employees > employees buy what firms produce > firms pay back the credit to the banks.[22] According to Graziani, 'Circuit theory tries to consider the whole life cycle of money, starting with its creation by means of bank loans and ending with its destruction when these loans are repaid.'[23]

In actual fact, however, the Circuitist model does not satisfy the claim of representing 'the whole life cycle of money'. Beyond the bank-firm-relationship, the model does not make a distinction between real-economic and financial transactions, in particular non-GDP-related transactions; it does not systematically differentiate between banks and financial intermediaries; it reproduces the old-industrial concept of production and consumption; equally, it reproduces the old-industrial concept of capital and labour; it blinds out the interdependencies with foreign economies; equally, it blinds out the fundamental economic role of the state; it thereby also reproduces the lopsided idea of 'primary' allocation and distribution by the private economy, and 'secondary' redistribution by the government.     

> back to top

Old-industrial logic of capital vs labour

What firms sell to and buy from other firms is blanked out in the model on the grounds that this affects payments among firms, i.e. trade within the same 'sector', where expenditures and earnings are netting out. The model thus considers banks, firms and wage earners as the three sectors of the economy, in methodological analogy to other Keynesian sector-account mechanics (which, though, are based on different types of sectors, i.e. a private, public, and foreign sector). According to the three sectors of the Circuitist model, 'real output gets divided into real wages, industrial profits, and financial profits'.[24]

Circuitism defines the economy as a wage economy. This is not the only aspect under which the model is reminiscent of 19th century theories of division of labour (forerunners of today's life-cycle and chain analyses) and, in particular, theories of labour value according to which all value added can be retraced to labour employed, which is to say, wages paid. In Marxism this also includes wages foregone to the workers due to appropriation by the capitalist entrepreneurs and bankers.

The model thus reproduces the old-industrial logic of capital and labour. This certainly continues to be a defining component, and yet it is too simplistic in order to capture realities in a sufficiently differentiated way. For example, earned income must not be reduced to dependent wage labour, while it continues to be inappropriate to merge income earned by self-employed and small and medium-sized businesses with 'industrial capital'.

Equally, real-economic and financial investment is absolutely fundamental to modern economies, as Circuitism recognises by its definition of what a 'money economy' is. It thus is flawed to give capital revenues as such a negative connotation. Even if it does not apply to the lower classes, many dependent employees today have, to a degree, savings and other invested funds and thus benefit from capital revenue. Moreover, in today's individualised society, 'non-active' individuals represent about half of the population. They have to be supplied with money, but can no longer simply be treated as the family appendage of the wage earners and recipients of capital income. 

> back to top

Banks and nonbank financial institutions

In Circuitism – and not only there – the term 'bank' is not used consistently. In many passages 'bank' properly means a commercial monetary institutions that creates primary credit and deposits, refinances itself at the central bank if need be, and participates in the electronic payment system of the central bank. In other passages, however, the term 'bank' is also used for nonbank financial intermediaries that help to on-lend or to invest already existing bank money. Sometimes 'bank' is used as a generic term for both types of financial firms.

The difference between banks and intermediaries is that only banks are monetary institutions that create and delete bank money (deposits), whereby they do not on-lend or invest existing deposits. Financial intermediaries, by contrast, are not able to create deposits. They are no monetary institutions. They operate as managers and investors of already existing deposits.

The difference between banks as monetary institutions and nonbank financial intermediaries must not be blurred. When loans or bonds are redeemed to a nonbank creditor, the money involved is not deleted, but continues to circulate. At the same time, the banks continue to create additional credit (deposits) if this is in their individual business interest. Over time this can lead to problematic effects, as discussed at the end of this paper.

> back to top

The real-bills doctrine of Circuitism

The Circuitist model comes with its implicit version of the real bills doctrine. This was a central Banking-School position in the historical controversy with the Currency School. The doctrine maintains that banks will always create an optimum quantity of money, neither an inflationary overshoot nor a deflationary shortage of the money supply, and will thus always operate on the safe side as long as they create money for funding 'real bills', i.e. credit against bills of exchange or other securities issued by reputable enterprises, sound business, etc. Moreover, if banks do so, they should do it over the short rather than the long term. In the face of the realities of banking and entrepreneurship, this is just glossing things over, including a good measure of pretence of knowledge. In the end, as things naturally turn out, firms and banks often have not had enough knowledge, or things may change in an unforeseen way, so that initial expectations ascribed to projects and 'real bills' become unreal. What is more, many bankers, investors and adventurers, particularly in the upswing and climax of business and financial cycles, cannot stop themselves from leveraging up the stakes.     

Such market dynamics and related risk behaviour are no issue in Circuitism, as the existence of a huge non-GDP-related financial economy is non-existent in the Circuitist model. Instead, primary and secondary credit are assumed to fund real-economic expenditure of firms only. In this respect, the Circuitists' understanding of the financial economy is under-developed. In particular, it fails to see the difference between GDP-related and non GDP-related transactions – a flaw which again is typical not only for Circuitism. The growth of non GDP-related finance has been particularly important since around 1980, but it already existed in the beginnings of Circuitism.

In the Circuitist model, firms and banks are supposed to make appropriate predictions regarding business perspectives, the demand for what they intend to supply, future prices, the work force, etc. Firms are supposed to know how much money they will need, and they negotiate the credit conditions with the banks for funding the firms' planned activities. The firms' demand for money thus is supposed to have a 'real' and reliable foundation, and banks accordingly cannot fail in serving that demand – in fact another real bills doctrine, all the more, as banks in the last decades have created primary credit predominantly for financial investment, including real-estate bubbles, and for sovereign-bond bubbles by financing governments far beyond reasonable levels of indebtedness.

> back to top

Is there a necessary sequence in money circulation? 

The Circuitist model states a specific sequence in the circulation of money. Developing some such sequence, as an alternative to older classical and neoclassical positions, was among Keynes's research desiderata. Among those older models was the distinction between basic industries, or capital-goods industries, and consumer-goods industries (economic sections I and II in Marxist economics), as well as the Austrian-School five- to seven-step production model that is underlying its capital theory.[25]
The centrepiece of the Circuitist model is bank credit to firms.

Fontana: 'Money ... is mainly the flow of bank deposits demanded by firms to finance the production of goods and services'.[26]          

Graziani: 'Negotiations between banks and firms on the money market determine the amount of credit actually granted and the rate of interest charged to firms. ... The model is money market first, labour market second, the first determining the conditions for the latter. ... The total amount of money is a debt of the firms to the banking sector and a credit of wage-earners to the same sector'.[27]  

Circuitists certainly do not ignore credit to other actor groups, such as consumer credit, mortgages, or sovereign bonds. Rather, they consider these segments of the banking and financial business to be of secondary importance, implicitly maintaining that everything is determined by the gravitational circuit between capital (banks and firms) and labour (wage earners).

This is overly reductionist. The Circuitist model does not correspond to the empirical pattern of present-day credit and money creation. The 'circuit' represents just one component in a wider picture. Close to 60% of bank lending today is allotted to mortgages. The rest is spread over government debt, student loans (in the US), consumer credit (overdraft, car, credit-card and home-equity-line credit) as well as lending to firms.[28] The latter, of course, continues to be part of the banking business, but is in no way predominant and applies to small and medium-sized enterprises rather than big companies. Industrial corporations no longer depend on bank credit to a major extent. They tap the secondary credit market, i.e. on-lending of already existing bank money, for example, by way of issuing corporate bonds, shares, or taking up money from investment funds. Moreover, large multinationals now run banks of their own (which is a questionable development in terms of separation of monetary, fiscal, financial and real-economic functions).

In recent decades, furthermore, liquid bank money (i.e. money circulating by way of primary and secondary credit, as represented in European M1) has grown several times the nominal GDP that represents the economic product at current prices (including consumer price inflation). This is to say that the lion's share of credit went into financial non-GDP transactions, into self-referential and quite often purely speculative financial portfolio trading. The disproportionate growth of such non-GDP transactions creates bubbles so that even mortgages and sovereign bonds, normally considered as conservative investments, can turn into high-risk exposures.

According to classical and neoclassical views, credit should first flow into capital expenditure, especially into private investment in productive capacities, not immediately into consumption, and less so into government expenditure, both of which are supposed to result in inflation. In the beginning of industrialisation, with productive capacities at a low level of development, and potential consumptive demand still unsatisfied for the most part, the idea of 'investment in productive capacities first, consumption second' made some sense.

Programs of government expenditure since the 1930s, in particular Keynesian demand-side policies after WWII, established a complementary alternative to the classical attitude. In the context of a structurally entrenched lack of effective demand, additional government expenditure and increased wages were ascribed a positive role of their own, because this prevents deflationary depression and, as long as capacities are underused, does not entail an important risk of inflation. One reason is that productive capacities today are much higher and connected through global markets. This means that supply chains are flexibly adaptive. Increased demand for goods and services triggers a swift increase in supply rather than inflation, the more so as long as the cost level in new industrial and developing countries is much lower than in developed countries.

This is not to justify the abuse of Keynesian demand-side policies by permanent deficit spending as a bad political all-seasons habit detached from the real business cycle, resulting in all-too high levels of government expenditure and debt. In many cases, moreover, state interventionism actually contributes to reinforcing rather than overcoming problems of structural entrenchment.

Circuitism does not contribute to analysing such questions. Its narrow focus on 'banks financing firms' in fact leaves a somewhat dated impression. Classical, Marxist and neoclassical economists until the 1920s used to think in these terms, including R. Hilferding's and R. Luxemburg's notion of financial capitalism as banking capitalism. During the century since, the arena of actors and the complexity of the financial and real economy have considerably evolved. Focussing questions of money and finance too narrowly on the firm now represents an old-industrial bias of economics that may have had a point from around 1800 until about the 1960s. One consequence of this has been building the industrial welfare state too narrowly upon the relationship between employers and wage labour. Over time this has become another fiscal and financial-market problem without lastingly solving respective social problems.

A positive aspect in the model is that Circuitists reject a special focus on investment in real-economic capacities. They rather refer to 'the monetary cost of output in general', i.e. capital expenditure with a broad meaning, including wages.[29]

Notwithstanding, the firms in the Circuitist model still stand for the production of consumer supplies, the wage earners for the effective demand which absorbs the consumer goods and household-related services supplied. Again, this is just one part of the whole picture. It reproduces, however, the concept of a linear vertical production chain of an economy entirely aimed at 'final' private consumption. This is reflected in today's national accounts and the Circuitist model which describes the economy a reduced linear circle from the creation of bank money for financing the production of goods by firms, which includes paying wages, to wage-earning consumers buying consumer items and household-related services, so that firms are able to pay back the credit, plus interest, to the banks. This may be handy, and, as is known, Henry Ford thought this way a hundred years ago. Today, it does not sufficiently reflect the multifaceted realities of the economy, less so the financial economy and the total picture of the circulation of money.

The real-economic supply-side channel is of course an important one, but there is a wider picture – supply-side and demand-side impulses; by firms, private households and public units; life cycles of technologies, products/services and markets as well as vertical and horizontal chains of provision; including the arbitrariness of classifying steps therein as 'productive/investive' or 'consumptive'; the existence of a GDP-related and non-GDP money circulation. All this suggests the development of a more complex understanding of what money circulation actually encompasses.

It follows from this that there is no 'natural' sequence in the circulation of money. Instead, there can be various ways of channelling new money into circulation, and many ways in the further circulation of the money. There is no such thing as a strictly necessary sequence regarding the creation, circulation and deletion of money. Respective doctrines can safely be dropped. The important thing is that enough money for GDP-related purposes can be obtained by financial institutions, firms, private households and public bodies whenever they need it.

The above criticism of the circuitist model is certainly of a relative nature. Even if outmoded to a degree, the model maintains a core of truth which is certainly more adequate than, for example, the Postkeynesian model of sector balances which confuses its sectoral book-keeping model of the economy with an analysis of monetary and financial dynamics.

> back to top

Classical bias regarding the 'secondary' economic status of the state

The Circuitist model reproduces the classical view regarding allocation through enterprises, followed by primary distribution between capital and labour, and secondary redistribution by way of taxes through the state.[30] This too has been taken as the natural state of affairs up to the present day. It is a common feature in almost all approaches to economic modelling. Economic models often start from a version without the state (and no foreign economy), and then proceed to less-simplified versions including the state (and possibly a foreign economy). What nevertheless remains is the coding of the 'economic base' as primary/productive/ entrepreneurial/competitive, and the 'state superstructure' as secondary/unproductive/bureaucratic/power-dominated/monopolist, or in the same vein.

Once more the real picture is different, and a rather mixed one on both fronts. In modern corporate economies, private-sector companies tend to be heavily bureaucratised; most markets – not only labour, but also goods and services as well as certain segments in finance – are power-dominated and oligopolistic; factor allocation is often inefficient.

Regarding the relationship between the 'economic base' and the 'state superstructure' one cannot really say in the ongoing process which is the chicken and which is the egg. With the government sector representing 35–60% of GDP, and with much of the additions to the money supply created due to government demand for additional money, the government's economic status is anything but secondary. In actual and historical fact, the state represents the institutional, legal and infrastructural base of the economy which includes private as much as public firms, institutes, organisations and households. Seen like this, the state would appear as the 'primary basis' rather than a secondary redistributive 'superstructure' (which is not advocated here). Marx and Engels' utopia of communism was somewhat grotesque when they expected the dwindling-away of the state (an element of early 19th century social romanticism, at the time present in liberalism as much as lateron in both anarchism and socialism).

Government expenditure is certainly not always optimal, but so too is private capital expenditure. Labelling government expenditure per se as 'unproductive' and private capital expenditure as 'productive' is ideological humbug. Government expenditure feeds mass purchasing power more effectively than capital expenditure does by itself. Government provides necessary administrative and judicial functions, also and specifically for the economy. Government provides general infrastructures and a wide range of public services, including health and education, the biggest sectors of the economy. How much of this is to be classified as consumptive or investive, and how much contributes to 'reproducing' or 'producing' human capital, is a highly arbitrary and futile question, comparable to the flawed 19th century theories of productive and unproductive labour. In the transition from traditional to modern societies, all of these questions are backward-looking to pre- and old-industrial stages of development.

> back to top

Lost in the circuit

The model of the money circuit raises yet another old issue, that is, the 'loss' of a certain quantity of money in circulation, because money owners may retain some part of their money rather than spending it or lending it to others who spend it.

Graziani: 'If wage earners decide to keep part of their savings in the form of liquid balances … firms will get back from the markets less money than they have initially injected in it … there has been a loss in the circuit.[31]

Fontana: 'Circuitists have explained that for any given production process an increase in money holdings by wage earners is exactly the same as a loss of liquidity, and hence an equivalent increase in bank debts by firms.'[32]

In a way this is right, but reminiscent of the notion of hoarding of money in traditional economies. Graziani actually uses the term hoarding for that part of savings that are not on-lent or invested by wage earners on the secondary credit market.[33] The subject was treated in Keynes as the problem of liquidity preference.

Much in this context, however, rather than holding liquidity, is about avoiding liquidity and debt, or reducing them, respectively. Banks systematically avoid holding more liquidity than is necessary in order to make current payments. The actual question here is whether banksas primary credit creatorsare prepared to lend or invest, supposing there are potential debtors who appear to be creditworthy and are prepared to go into debt. Avoidance of currently not necessitated liquidity also applies to nonbank financial intermediaries, real-economic firms and public households in a similar way.

Private households, too, do not hoard large quantities of cash under the mattress, and normally there is not much idle money in current bank accounts either. Rather, money which is currently not needed is parked as a savings or time deposit in M2/M3, about 10–20% of income in advanced countries.

What is actually relevant with regard to the thesis of a 'loss in the circuit', is the fact that M2/M3-positions are non-liquid, even though they can be liquidated within a couple of weeks or months. (If there are M2-positions which can be liquidated any time, these are mis-classified positions which actually belong in M1). M2/M3-positions in fact represent a reduction in liquid money M1. M2/M3-positions, contrary to popular belief, do not serve to fund loans or investment. Credits in a savings or time account with a bank represent inactivated, non-circulating bank money.

It would nevertheless be inadequate to interpret M2/M3-items as 'hoarded currency'. Rather, M2/M3 is part of the equity of those households, beneficial to the entire economy, particularly in a life-time or other long-range perspective. With endogenous money, the banking industryor whoever else exercises the privileged prerogative of creating moneycan fill any supposed gap in M1.

Liquidity preference is certainly real. It causes a need for maturity transformation, or tolerated maturity mismatch, respectively. Equally, liquidity preference represents pro-cyclical behaviour. It waxes and wanes in the rhythm of economic and financial cycles. However, it does not cause cycles and crises by itself. Should there really be a 'Circuitist' problem with regard to absorbing what the firms have produced, this has not necessarily to do with liquidity preference and liquidity shortage. Nor is it, under present-day conditions, a fundamental problem when firms have to borrow additional money in order to finance current business. The central role ascribed to liquidity preference in Keynesian and Postkeynesian economics is overdone.  

As an alternative to M2/M3-savings, money can be invested in sovereign and corporate bonds, stocks, and funds of various types. This represents secondary credit which keeps the bank money circulating in M1, thus preventing it from being 'lost' in the circuit. The question remains, nonetheless, whether the money is lost to the real economy. The secondarily on-lent or invested part of the money supply can serve to fund real expenditure of firms, government and consumers (as is the case, for example, with IPOs of bonds and stocks). But the money may also go into non-GDP transactions (such as after-IPO trading of securities as well as mergers and acquisitions, hostile takeovers, or speculative investment in foreign exchange, real estate, commodities and derivatives).

In a way, the 'loss' in the money circuit and the supposed liquidity gap resulting from it, is an analogy to another question that has haunted theologists and social philosophers for centuries, i.e. the interest gap, the question of how, with a given money supply, to pay interest on top of repaying the principal. Circuitism is preoccupied with its version of a liquidity gap, while not looking into the question of an interest gap. Discussing the latter might anyway be going too far here. But I would like to remark that problems such as the Circuitist liquidity gap, or the Anarchosyndicalist interest gap, represent methodological artefacts resulting from the reductionist nature of the respective model considerations.

Is it not simply the case that in a monetarised and financialised economy, all kinds of expenditure, or income respectively, must initially and perpetually be funded to a considerable extent? And that with endogenous money there is basically no reason why this should not be ensured? Things would be different with exogenous money, but that is not part of our reality. Whether things may somewhat be different in a post-growth economy remains to be seen.

> back to top

Circuitist literature referred to

Graziani, Augusto 2003: The Monetary Theory of Production, Cambridge University Press.

Graziani, Augusto 1990: The Theory of the Monetary Circuit, Économies et Sociétés, No.7, 1990, 7–36.

Fontana, Giuseppe 2000: Post Keynesians and Circuitists on Money and Uncertainty, Journal of Post Keynesian Economics, Fall 2000, Vol.23, No.1, 27–48.

Parguez, Alain / Seccareccia, Mario 2000: The credit theory of money -The monetary circuit approach, in J. Smithin (ed), What is Money?, London/New York: Routledge, 101–123.  

> back to top


[1] Except Augusto Graziani, who is the main reference here, further important Circuitists include Bernard Schmitt, Alain Parguez, Alvaro Cencini, Frédéric Poulon and Giuseppe Fontana.

[2] Graziani 2003 23.

[3] Graziani 1990 8-10, 16, 29; Graziani 2003 11, 23-29, 82.  

[4] Fontana 2000 42.

[5] Graziani 1990 29, 2003 26.

[6] Graziani 2003 58-62; 1990 8, 11–29.

[7] Graziani 1990 29, 7, 11.

[8] Godley, Wynne / Lavoie, Marc 2007: Monetary Economics, London: palgrave/ macmillan. - Wray, Randall 2012: Modern Money Theory, Palgrave/Macmillan. - Tcherneva, Pavlina 2006: Chartalism and the tax-driven approach, in: Arestis, Philip / Sawyer, Malcolm (eds.), A Handbook of Alternative Monetary Economics, Cheltenham: Edward Elgar, 69–86.

[9] In his journal article from 1990, p. 18, Graziani once mentions 'the possibility for the central bank to create money in order to finance the government deficit. Central-bank money thus created is no longer debt of commercial banks, but debt of the government towards the central bank.'

[10] Graziani 1990 16.

[11] Moore, Basil 1988: Horizontalists and Verticalists: The Macroeconomics of Credit Money, Cambridge University Press. - Palley, Thomas 1993: Competing Views of the Money Supply, New School for Social Research, New York, Dept. of Economics, July 1993, docs/articles/macro_theory/monetary_supply.pdf.

[12] Graziani 1990 10, 2003 11.

[13] Cf. Zarlenga, Stephen 2002: The Lost Science of Money, Valatie, NY: American Monetary Institute, chapters 14–17; Hixson, William F. 1993: Triumph of the Bankers, Westport, CT: Praeger, chapters 7, 8, 11 – 19.

[14] Cf.

[15] Graziani 1990 11–12, 2003 61–62.

[16] Cf. Rochon, Louis-Philippe 1999: Credit, Money and Production. An Alternative Post-Keynesian Approach, Cheltenham: Edward Elgar, p.15, 17, 155, 163, pp.166. - Keene, Steve 2011: Debunking Economics, London/New York: Zed Books, pp.358. - Moore, Basil 1988: Horizontalists and Verticalists: The Macroeconomics of Credit Money, Cambridge University Press. 

[17] Cf. the contributions of Henry, Hudson, Gardiner and Ingham in Wray, L. Randall (ed.) 2004: Credit and State Theories of Money, Cheltenham: Edward Elgar. - Graeber, David 2011: Debt. The first 5,000 years, New York: Melville House Publishing.  

[18] Chartal theories of money see the money system and the money itself as a creature of the law. The lasting existence of money depends on being backed up by the state. Major references are the British Currency School of the 1830–40s, the State Theory of Money by Fr.  Knapp (1905), and, soon thereafter, the writings by A. Mitchell-Innes. The latter authors, however, advocated incomplete chartalism, where the money is not necessarily state money, but can also be private money, in particular credit money created by the banking sector.

[19] Lagos, Ricardo 2006: Inside and Outside Money, Federal Reserve Bank of Minneapolis, Research Department Staff Report 374, May 2006. - Roch, Cullen 2012: Understanding Inside Money and Outside Money, Pragmatic Capitalism,­standing-inside-money-and-outside-money.

[20] Graziani 1990 15.

[21] Graziani 1990 27.

[22] Graziani 2003 26–31; Parguez/Seccareccia 2000.

[23] Graziani 2003 21.

[24] Graziani 1990 27.

[25] Cf. Jesús Huerta de Soto 2009: Money, Bank Credit, and Economic Cycles, Ludwig von Mises Institute, Auburn, Alabama, Chapter 5, 265–396.

[26] Fontana 2000 42.

[27] Graziani 1990 12.

[28] Jordà, Òscar / Schularick, Moritz / Taylor, Alan M. 2014: The Great Mortgaging: Housing Finance, Crises, and Business Cycles, NBER Working Papers, No. 20501, Sep 2014, National Bureau of Economic Research. - Economist, March 28th, 2015, 16.

[29] Graziani 1990 14.

[30] Also cf. Parguez/Seccareccia 2000 pp.425.

[31] Graziani 1990 13.

[32] Fontana 2000 43.

[33] Graziani 2003 32.

> back to top