Cambridge Journal of Economics, 6(3), September 1982, 295-6.
During the 60-odd years since the death of Keynes there have existed two, fundamentally different, paradigms for macroeconomic research, each with its own fundamentally different interpretation of Keynes's work. [n.1] On the one hand there is the mainstream, or neo-classical, paradigm, which is based on the premise that economic activity is exclusively motivated by the aspirations of individual agents. At its heart this paradigm requires a neo-classical production function, which postulates that output is the result of combining labour with capital in such a way that, provided all markets clear, there will be no involuntary unemployment while the national income is distributed optimally and automatically between wages and profits. If markets do not clear because wages or prices are 'sticky', the same structure will generate determinate, if sub-optimal, disequilibrium outcomes and, for many economists, it is the possibility of such stickiness that defines Keynesian economics. The key assumption that individual welfare maximization is the universal mainspring is not consistent with the view that firms have an independent existence with distinct motivations, because optimum prices, output and employment are all decided for them by the location of aggregate demand and supply schedules. And as production is instantaneous, while supply is brought into equivalence with demand through the market-clearing process, there is no systematic need and therefore no essential place for loans, credit money or banks. The concept of 'money' is indispensable, yet money is an asset to which there is not, in general, a counterpart liability and which often has no accounting relationship to other variables. Mainstream macroeconomic theory is a deductive system which needs no recourse to facts (though it may be 'calibrated' with numbers) and lends itself to analytic solutions.
The alternative paradigm, which has come to be called 'post-Keynesian' or 'structuralist', derives originally from those economists who were more or less closely associated personally with Keynes such as Joan Robinson, Richard Kahn, Nicholas Kaldor, and James Meade, as well as Michal Kalecki who derived most of his ideas independently. So, far from being a deductive system, the post-Keynesian vision is underpinned by 'stylised facts' recognizing the manifest existence of institutions, together with regularities and magnitudes in the economic data which can be checked out empirically. Central to this system of ideas is that, in a modern industrial economy, firms have a separate existence with a distinct set of objectives, for example, to make enough profits to pay for growth-maximizing investment. Rejecting as chimerical the concept of the neo-classical production function,[n.2] post-Keynesians hold that, in an uncertain world, firms, operating under conditions of imperfect competition and increasing returns, must decide how much to produce and how many workers to employ, what prices to charge, how much to invest, and how to obtain finance. It will be the pricing decision which, in general, determines the distribution of the national income between wages and profits. And as production and investment take time while expectations are in general falsified, there is a systematic need for loans from outside the production sector which generates acceptable credit money endogenouslyㅡin other words (in accordance with common observation) there must exist a banking sector. According to post-Keynesian ideas, there is no natural tendency for economies to generate full employment, and for this and other reasons growth and stability require the active participation of governments in the form of fiscal, monetary and incomes policy. And it will probably impossible to derive analytic solutions which describe how economies as a whole evolve, particularly as institutions and behavioural patterns change drastically through historical time.
Luigi Pasinett (2005) laments the fact that post-Keynesians have progressively failed to establish 'a permanent winning paradigm'. And indeed, while pockets of stubborn resistance remain, the post-Keynesian tradition has now been virtually written out of the literature; it has lost out to the mainstream in terms of how the subject is taught, what the 'top' learned journals will accept, where research money is allocate, how appointments are made and how empirical models are built. Pasinetti attributes this collapse in large part to the personal characters of the formidable economists who directly succeeded Keynes, maintaining (correctly in our opinion), that they did not admit outsiders to their circle or sponsor their work. But Pasinetti also points to 'a lack of theoretical cohesion in the various pieces which emerged from the Keynesian School', which 'paid scant attention to the fundamentals on which an alternative, but coherent, paradigm could be built'. He suggests that 'a satisfactory blueprint that could house, beneath one single roof, the development of the existing ideas along the Keynesian lines ... is still lacking' ... and that there is a need for 'an account of what happensㅡas Keynes put itㅡin a "monetary production economy", which is more complex than a pure exchange stationary economy, because it is intrinsically dynamic, continually affected by history subject to changes both in scale and structure'. This is an admission that post-Keynesian economics up to the present time simply does not cover the ground.
Geoffrery Harcourt (2001: 277) in similar vein observes that post-Keynesians have been following the Marshallian/Keynesian method which 'consists at looking at parts of the economy in sequence, holding constant or abstracting from what is going on, or at least the ^effects^ of what is going on elsewhere, for the moment', in the hope that it would be possible eventually 'to bring all our results together to give a full, overall picture'. Harcourt thinks that 'this may be one of the reasons why ultimately both Marshall and Joan Robinson thought that they had failedㅡnot from realizing that by following the procedure they were attempting the impossible, but because the ^procedure^ itself was at fault'. While neo-classical economists have general equilibrium theory and computable general equilibrium models that helped capture the overall implications of their vision and the interdependence between markets and sectors, post-Keynesian economics could only offer the Sraffian model as a formal tool to tackle production interdependencies and relative prices, but which, ironically, did not and could not deal with the crucial Keynesian issues of output, unemployment, inflation, financial flows and pieces, with no account of how the system as a whole worked. There is no statement which characterizes how post-Keynesian theory can underlie the way in which an industrial capitalist economy works ^as an organic while^. Despite valiant efforts, such as the book by Eichner (1987), there is no post-Keynesian textbook which covers all of the monetary macro ground as a coherent whole.[n.3]
In writing this book it has been our aspiration to lay the foundations for a methodology which will make it possible to start exploring rigorously how real economic systems, replete with realistic institutions, function as a whole. Our starting point, though a little intricate for an introduction, is yet so simple that we propose to plunge straightaway ^in medias res^[Latin for 'into the middle of things].
The standard textbook introduces macroeconomic concepts via the national income identity. Thus total production, or gross domestic product (GDP), is defined as the sum of all expenditures on goods and services or, alternatively, as the sum of all incomes paid for production of goods and services. More precisely, the GDP (assuming the economy to be closed) is made up of personal consumption, investment and government expenditure on goods and services; looked at from the income side, it is made up of income from employment and profits. All these concepts are introduced as 'real' variables, the GDP being an economy's total volume of production. Writing these identities formally we have:
C ＋ I ＋ G ＝ Y ＝ WB ＋ F (1.1)
where C is consumption, I is investment, G is government expenditure, Y is GDP, WB is the wage bill and F is profits.
And that is about it, so far as accounting goes, though when it comes to studying the consumption function the student will quickly have to learn that personal disposable income is give by:
YD ＝ Y － T (1.2)
where YD is personal disposable income and T describes all taxes and transfers received or paid by the government. Equation (1.2) builds in the implicit (but counterfactual) assumption that all profits are instantaneously distributed to households.
Decomposing the wage bill into a quantity of employment times a wage rate and postulating the existence of a rate of interest, a stock of money, a price of real output and a stock of fixed capital equipment, we have enough concepts to erect the 'core' model of the so-called neo-classical synthesis, which constituted mainstream macroeconomics at least until the 1980s and from which more recent schools of thought (e.g. Rational Expectations, Real Business Cycles, New Keynesian) are directly descended. By this model, in its basic manifestation, the demand for output is determined by consumption and investment functions, the profit-maximizing supply of output is determined by the marginal product of labour and the real wage. The demand for real money balances is determined by income and the rate of interest, while the supply of money is exogenous and given. The entire system is in market-clearing equilibrium when all three demands are in equivalence with all three supplies, yielding determinate values for all the components of the national income as well as for employment and for each 'price'.
Although every author will have his or her own gloss on how exactly this model works, and what happens if in various ways it doesn't work, we have no doubt whatever that this account does fairly summarize the core model which dominated the scene for so long. [n.5] The purpose of reproducing it here is not to criticize it, but rather to set up a clear reference point in terms of which we can clearly deploy a radically different way of viewing the world and setting up a research agenda to explore it.
[n.5] The classic expositions are to be found in Patinkin (1965) and Modigliani (1944, 1963). The basic model is not changed when markets fail to clear. It is, for instance, commonly argued that 'Keynesian economics'ㅡusing this modelㅡis encapsulated by assuming that the nominal wage is exogenously determined, in which case the supply of labour can exceed the demand, causing, and suggesting the cure for, unemploymentㅡin advance of any empirical investigation whatever.
The difference between the world to be deployed in the following chapters and that introduced in most textbooks is well introduced by first fitting the variables described in equation (1.1) into a matrix such as that shown in Table 1.1, which brings out the fact that each variable is a transaction between two sectors which takes place in some given period of time.
.JPG)
The second column of Table 1.1 does nothing more than reproduce equation (1.1) in a vertical arrangement. The other columns show the transactions implied by the component parts of equations (1.1) and (1.2). Thus, for instance, consumption is a receipt by the business sector and a payment by the household sector. The only thing which might be unfamiliar to a student is the third column, which describes the capital account of the business sector. But there should be no difficulty about the meaning and significance of this; sales of investment goods give rise to receipts by the business sector like any other sales. But these receipts will have to come (at this level of abstraction) from payments by the business sector itself, which is assumed to do all the investing.
But now it is easy to see that this system of concepts is seriously incomplete. Consideration of the matrix immediately poses the following questions. What form does personal saving take? Where does any excess of sectoral income over expenditure actually go toㅡfor it must go somewhere? Which sector provides the counterparty to every transaction in assets? Where does the finance for investment come from? And how are budget deficits financed?
There is an obvious answer to these questions, which follows from an elementary knowledge of the way the real world works and which can be quickly verified by inspecting the Flow-of-Funds tables published by the Federal Reserve in the United States, which provide data relating to every quarter since 1952.
Table 1.2 completes and rectifies the story adumbrated in Table 1.1, showing a relatively simple comprehensive system of accounts which describes all the intersectoral transactions implied by the Table 1.1 concepts but not shown there.
The upper, national income, part of the table reproduces Table 1.1, with the important difference that the usual assumption that all profits are distributed has been dropped. Instead some proportion of profits is transferred to firms' capital account, where it may be used to finance investment. [n.6] The lower, flow of funds, part of the table could have been completed in various different ways depending on the degree of detail and the simplifications deemed appropriate. However, it will be a cardinal principle applying here and to every array of concepts we shall deploy in the future that all rows and all columns sum to zero, [n.7] thus ensuing, in the catch-phrase, that 'everything comes from somewhere and everything goes somewhere'.
[n.6] Table 1.2, although an improvement over Table 1.1, still omits several relevant features, such as interest payments, and it assumes that the central bank is amalgamated with the government. A more complete matrix will be introduced in Chapter 2.
[n.7] For this reason the closed economy described above could not be 'opened' by adding a column describing exports and imports since this will not normally sum to zero. The solution will be to include all trading partners in the matrix, as will be shown in Chapter 6 and 12, making a larger closed system in which there is no place for a balance of payments column.
However, no sooner does one contemplate filling in the assets which are acquired by households than a second important inadequacy of Table 1.1 immediately becomes manifest. Households may (for instance) acquire credit money as an asset, but where is the counterpart acquisition of liabilities to be found? And firms may require loans to finance investment in excess of retained profits, but from where are these to come? The answers are obvious as soon as the questions are asked. The matrix cannot be completed unless a whole new sectorㅡa banking sectorㅡis introduced into the elementary system of concepts.
- In column 1 the saving of the personal sector is assumed to go entirely into cash[－ΔHh], credit money[－ΔM], government securities[－ΔBh], and newly issued equities[－Δe·Pe].
- There are no entries in column 2 because profits are defined as the residual between current inflows and outflows.
- In line 5 profit are in part distributed[－FDf] and in partㅡin practice by far the greater partㅡundistributed[－FUf].
- In column 3 the funds in excess of retained profits required for investment are assumed to come in part from the issue of equities[＋Δe·Pe], with the balance coming from loans[＋ΔLf].
- In column 5 the government is assumed to finance any deficit by the issue of securities[＋ΔB] and cash[－ΔH].
- Finally in column 4 we have the banks' transactions in assets which comprise the genesis of loans and credit money and which bring these concepts firmly into the most basic accounting structure, and they also say, non-trivially, that any gap between these two supplies[? loans and credit money] must always be matched exactly by net accumulation by banks of cash[? －ΔHb] and government bills[? －ΔBb]ㅡfor the balance of banks' transactions in assets must sum to zero if only because every other row and column in the table sums to zero.
All entries in the flow-of-funds sections of Table 1.2 describe changes in stock variables between the beginning and end of the period being described. [n.8] Thus the evolution of historic time is introduced into the basic system of concepts. The transactions in asset stocks in Table 1.2 imply the existence of an interlocking system of balance sheets, described in Table 1.3. These balance sheets measure the levels of all stock variables at some given point of time. And it is the configuration of stock variables which is providing the link between each period of time and that which follows it.
[n.8] The variables are defined in the matrix. The term e in the final line describes the number of equity titles and Pe describes their price.
The evolution of the entire system may be characterized (at the level of accounting) by saying that at the beginning of each period, the configuration of stock variables (i.e. all physical stocks together with the interlocking system of financial assets and liabilities) is a summary description of (relevant[n.9]) past history. Then the transactions described in Table 1.2 heave the stock variables from their state at the beginning of each period to their state at the end,[n.10] to which capital gains will have to be added.
For this system of accounting identities to hold, all variables must be measured at current prices, since they describe the sums of money that actually change hands each periodㅡotherwise, unless there is no change in any price, the columns would not add up to zero. [n.11] Yet a number of key decisions regarding, in particular, production, consumption, investment and many kinds of government expenditure are taken in terms of real, physical quantities. So we shall at some stage have to describe, at the level of accounting (i.e. before considering behaviour)[,] how prices translate nominal into real variables, thereby determining the distribution of the national income.
1.3 Endeavour
We can now disclose in a nutshell the nature of the task we have set for ourselves. We are going to define a series of evolutionary models, each of which describes an economy moving forward non-ergodically in historical time, as Paul Davidson (1988) would put it. We start with truly primitive models containing a mere handful of equations and end up with relatively elaborate models containing one hundred or more equations. Each model must account for every single one of the variables contained in the relevant transactions and balance sheet matrices. So in sharp contrast with the Marshallian method, we shall always be exploring the properties of complete systems, never assuming that we can consider one topic at a time in the hope that the rest of the world stays in place while we do so.
The method will be to write down systems of equations and accounting identities, attribute initial values to all stocks and all flows as well as to behavioural parameters, using stylized facts so well as we can to get appropriate ratios (e.g. for the proportion of the national income taken by government expenditure). We then use numerical simulation to check the accounting and obtain a steady state for the economy in question. Finally we shock the system with a variety of alternative assumptions about exogenous variables and parameters and explore the consequences. It will be our contention that via the experience of simulating increasingly complex models it becomes possible to build knowledge, or 'informed intuition',[n.12] as to the way monetary economics must and do function.
The use of logically complete accounts (with every row and every column in the transactions matrix summing to zero) has strong implications for the dynamics of the system as a whole. This completeness carries the implications that once n-1 equations are satisfied then the nth equation will be found to be satisfied as well and for this reason must always be dropped from the computer model to avoid overdetermination. If the accounting is less than complete in the sense we use, the system dynamics will be subvertedㅡrather as though we were trying to operate a hydraulic machine which had leaky pipes.
( ... ... )
( ... ... )
( ... ... )
( ... ... )
( ... ... )
1.4 Provenance
Over the past few years, centered along the axis of the New School University and the Levy Economics Institute, both located in New York State, there has been a revival of interest in the stock-flow consistent approach to macroeconomic modelling, or what we could call a sectoral monetary stock-flow consistent approach.[n.13] The purpose of the present book is to feed this revival, in the hope that an accessible introduction to stock-flow consistent macro-economic modelling will induce more students and more colleagues to adopt and develop such an approach. Our belief is that, if such an adoption occurs, macroeconomics in general and heterodox economics in particular should become sounder and more transparent.
( ... ) In broad terms, one can identify two schools of thought which actively developed a series of models based on the stock-flow consistent approach to macroeconomic modelling, one located at Yale University and led by the Nobel Prize winner James Tobin, and the other located at the Department of Applied Economics at Cambridge University and led by one of the present authors (Wynne Godley). To a large extent, both groups worked independently, at least until a conference on Keynes that was organized in Cambridge (UK) in 1983. The Yale group, also known as the 'pitfalls approach' or the New Haven school, focused its attention on portfolio and asset choice; its inspiration was essentially neo-classical and based on a practical variant of general equilibrium theory. The Cambridge UK group, which was known as the Cambridge Economic Policy Group (CEPG) or the New Cambridge School, used the stock-flow consistent framework mainly for forecasting whether an expansion was ^sustainable^, as Godley (1999c) still does today, and to discuss the balance of payments problems that were then plaguing the United Kingdom.
cf. Godley (1999c) Seven Unsustainable Processes: Medium-Term Prospects and Policies for the United States and the World, Strategic Analysis, The Levy Economics Institute of Bard College.
Both research groups faded in the middle of the 1980s, as their funding was cut off, ( ... )
The more recent work of Godley (1996, 1997, 1999a), which has led to the creation of the present book, owes a substantial debt to Tobin, most particularly the work of Tobin as it appears in Backus, Brainard, Smith and Tobin (1980), which presented the most explicit and most empirically-oriented version of the research programme that was being pursued at Yale University on the stock-flow consistent approach to macroeconomic modelling.[n.15] ( ... )
1.5 Some links with the 'old' Yale school [n.17]
In his Nobel lecture, Tobin (1982a: 172-3) identified the main features that distinguished his work. Four features stood out, and they certainly apply to the present book.
1. Tracking of stocks and precision regarding time;
2. Several assets and rates of return;
3. Modelling of financial and monetary policy operations;
4. The budget constraint and the adding-up constraint.
( ... ... )
Feature (2) says that a comprehensive model should have several assets and several rates of return. Tobin objected to the standard representation of the IS/LM model, which has only one explicit rate of return, the bill rate, and one explicit asset, money. Since financial relations are so important in a modern economy, a sophisticated financial framework must be developed to understand the various interaction between borrowers and lenders, as well as the role of the banking system. In our book we shall experiment with various numbers of assets, and various kinds of assets. ( ... )
Feature (3), the modelling of financial and monetary policy, will be a key part of our book. How the stocks of the various assets are supplied, in particular by the monetary authorities and the government, will be described in detail. ( ... ) Indeed, how the banking and financial systems are precisely being modelled constitutes one of the major differences between the Yale approach on the one hand and the New Cambridge approach which is being advocated here.
Finally, there is feature (4), which says that agents must respect their budget constraint, both in regard to their expectations and when they assess realized results. In the case of expected results, this is sometimes referred to as Walras' Law, as does Tobin in his Nobel lecture, but we would rather refer to a budget constraint or to a system-wide consistency requirement. In a water-tight accounting framework, the transaction flows of the ultimate sector are entirely determined by the transaction flows of the other sectors. Indeed, we shall see that this consistency requirement always implies a ^redundant^ equality. Feature (4) means that there cannot be any ^black hole^. In the words of Godley and Cripps (1983: 18), 'the fact that money stocks and flows must satisfy accounting identities in individual budgets and in an economy as a whole provides a fundamental law of macroeconomics analogous to the principle of conservation of energy in physics'. While consistency is required at the accounting level, it is also required at the behavioural level. This consistency requirement is particularly important and useful in the case of portfolio choice with several assets, where any change in the demand for an asset, for a given amount of expected or end-of-period wealth, must be reflected in an overall change in the value of the remaining assets which is of equal size but opposite sign.
The above four features distinguish the work of Tobin and that of the New Haven school, along with the work of individuals such as Turnovsky (1977), compared with that of standard mainstream macroeconomics.[n.18] The same features apply to the work and the approach being presented in this book. Thus, on the methodㅡconsistent accounting, consistent stock-flow analysis and consistent adding-up constraints on behavioural relationshipsㅡthe New Haven school and the New Cambridge school are in agreement.[n.19] In addition, as already pointed out, the modelling of portfolio behaviour by households in the present book is essentially being inspired by the method propounded by Brainard and Tobin (1968).
[n.18] For instance, in Hicks's (1937) famous IS/LM model of Keynes's ^General Theory^ (1936), investment is carried on, and savings occurs, while the supply of money is assumed to be exogenous to the model. What happens to wealth or debt at the end of the period is never discussed. Whereas the money stock ought to be an endogenous variable, determined by the system, it is assumed to be exogenous and controlled by the monetary authorities. As pointed out by Tobin (1982a: 187), 'the convenient strategy is to model the determination of asset prices and interest rates as a temporary stock equilibrium independent of flows of new saving'. The stock-flow consistent approach to macroeconomic modelling, advocated here and advocated by Tobin, precisely goes beyond this temporary equilibrium, where time seems to be frozen and the flows of investment and household saving have no impact on fixed capital, debt, wealth, and money stocks. The IS/LM model is only one slice of time (Tobin 1982a: 172), and a bad one at that.
However, agreement on the method does not preclude disagreement on the model. ( ... )
1.6 Links with the post-Keynesian school
( ... ... )
( ... ... )
( ... ... )
( ... ... )
This focus on the monetary side of production, debt and portfolio behaviour requires a serious examination of the banking system and of the financial system more generally. Banks and their balance sheets have to be fully integrated to the production process, and interest flows have to be taken into account explicitly. Our accounting framework will allow us to do just that. In addition, this framework will allow us to describe and understand the monetary circuit, that is, the monetary creation, circulation and destruction that accompanies production and wealth creation. The role of government expenditures, and their link with monetary creation and interest rates prevailing on government securities will also be understood through the use of the same rigorous accounting framework. In particular, that the money stock is ^endogenous^, as post-Keynesians such as Kaldor (1970a, 1982) and Robinson (1956, ch. 23) have long asserted, will be a crucial element of our models.
Another feature of post-Keynesian models, which can be associated with the ^principle of effective demand^, is that market clearing through prices does not usually occur except in financial markets. The real markets, those for products and labour, are assumed to be demand-led. Full employment of labour is not assumed, nor is full employment of capacity, although, in the later chapters, where the possibility of inflation is introduced, high levels of employment or capacity will be assumed to generate inflationary pressures. In that sense, one can say that our later models will be demand-led but eventually supply constrained. Post-Keynesians believe that if market forces based on price clearing were to act on the labour market, they would generate instability. As to the product markets, when dealing with the simplest models it will be assumed that supply adjusts to demandㅡthe reverse of Say's lawㅡwhile when dealing with more realistic models there will be another sort of quantity adjustment, a partial one, through inventories. It follows that the models to be described are typically ^Keynesian^: product market clear through quantity adjustments, and the models are ^demand-led^. The lack of production capacity, brought about by insufficient past investments will not be discussed here although it may provide a possible explanation of current unemployment.
( ... ... )
1.7 A sketch of the book
1.7.1 How the book was written
( ... ... )
( ... ... )
1.7.2 And how it should be used
As most of our models do not lend themselves to analytic solution, we strongly recommend readers to carry out simulations for themselves (Table 1.4). It will be via the experience of trying out alternative values for exogenous variables and parameters – and, indeed, by changing the models themselves─that major intuitions will be achieved. It will be found that key results will be far less arbitrary (less open to the ‘garbage in garbage out’ gibe) than one might suppose. The reader will be able quite easily to verify our results and conduct his or her own experiments because our colleague Gennaro Zezza has set up every one of our models (complete with data and solution routine) in a form that can be readily accessed.[n.24]
[n.24] at www.gennaro.zezza.it/software/models.
( ... ... )
Appendix 1.1: Compelling empirical failings of the neo-classical production function ( ... ...)
Appendix 1.2: Stock-flow relations and the post-Keynesians ( ... ... )
.JPG)
a.JPG)
.JPG)
글
