https://cpes.org.uk/om/files/original/e0dd11c251a90749f8244cf48498c330.pdf 900ba503ff72b0eb31610503333e39f6 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Cambridge Economic Policy Review Volume 8 No 1 Date A point or period of time associated with an event in the lifecycle of the resource April 1982 Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Text Any textual data included in the document Chapter 4 The 'New Cambridge' hypothesis and fiscal planning by Michael Anyadike-Danes It is some years since the idea of an aggregate private expenditure function was mooted by members of the Cambridge Economic Policy group. The intention was to show that if there is a functional relationship between total private disposable income and total private expenditure strong conclusions can be drawn about the interaction between fiscal policy, the current balance of payments and the flow of national income. This 'New Cambridge' hypothesis never found favour with the economics profession partly because it was at first wrongly specified; but the main objection was that the aggregation of personal consumption with corporate investment is inadmissible in principle. Yet while the aggregate private expeilditure function has generally been regarded as unacceptable commentators have, to an increasing extent, found it acceptable to attach significance to the aggregate flow of public expenditure less the aggregate flow of taxation and other public income - namely the PSBR*. Indeed the size of the PSBR has come to occupy a central, even dominant, role in the discussion of fiscal policy. Yet anyone who tries to determine what is a reasonable target for the PSBR, unless they are simply inventing arbitary numbers, is necessarily making, by implication, a judgement about the balance between aggregate private income and expenditure. For if, as is argued against the 'New Cambridge' view, there is no reason to expect the private sector's aggregate financial balance to behave in any particular way, then the PSBR itself has no operational significance. As it happens it seems evident that many people (in particular members of the British government) do not have any coherent idea of the interrelationship between the PSBR and the targets of macro-economic policy. They only have a general notion, which is nevertheless strongly held, that the PSBR is a bad thing which must be got down as a prelude to sound money and the creation of *As well as the difference between current receipts and expenditure on goods and services and transfers, the PSBR, as actually defined for the UK, includes certain transactions in financial assets which are ignored in what follows. 'real' jobs. We still believe in the 'New Cambridge' hypo- thesis (though modified from its original form) since it provides a basic framework within which fiscal planning can take place.* The first section of this chapter sets out a simple model of flows of national income and expenditure to show how assumptions about the balance between private income and private expenditure affect demand management and fiscal planning. The second and third sections present an updated account of our views on private asset accumulation and expenditure behaviour and their main implications for macro-economic policy. The final section compares actual annual changes in private expenditure in the UK, 1965-81, with those predicted by our hypothesis. 4.1 Private asset accumulation in a model of national income determination We start with the flow of funds identity where G is government expenditure (both current and capital) T is tax revenue X is exports M is imports PY is private income after tax and PE is private expenditure (consump- tion, investment and stockbuilding) all measured as annual money flows at current pnces. This identity forms the foundation for a simple model of national income determination which *Discussion of these matters has not been assisted by the publication of an extremely misleading paper by Chrystal ('The New Cambridge Aggregate Expenditure Function', Journal of Monetary Economics, Vol. 7, December 1981); misleading because his results are entirely meaningless (he made a simple, but basic, error in processing the data used in his empirical work) and his claim that the aggregate private expenditure function is a 'close approximation' to an identity is fundamentally mistaken. 33 can be used to display the nature of the ~elationship between fiscal policy, national mcome and the balance of payments. First, note that Gt is decided by the government, as are tax rates. The PSBR is the outcome of these decisions and whatever happens to national income itself. Thus PSBRt = Gt- Tt and Tt=:8tYt (2) where 9t is the average tax rate and Yt is total national income. Next assume that the value of exports is beyond the direct control of either the government or the domestic private sector but that spending on imports is sensitive to the level of national income. The current balance of payments is Bt = Xt- Mt and Mt= llt Yt (3) where Jlt is the ratio of imports to national income (there is no presumption that this will be constant over time). Now the private sector surplus of disposable income over expenditure must by definition be equal to its net acquisition of financial assets, i.e. to its purchases of financial assets less any increase in its outstanding liabilities. Thus the change in the stock of private net assets is =.1NSFAt PYt - PEt (4) where pyt = y t (1 - 8t) and NSFA denotes the stock of private financial assets less liabilities. As is evident from the identity, any private disposable income accruing within a year which is not spent within the year must be added to the private sector's net financial assets. Since we are dealing with an aggregate priate sector it must be the case that any net financial assets being accumulated are net liabilities of the two other sectors whose income or expenditure flows have not been represented - the public and the overseas sector. Assets and liabilities within the private sector itself cancel out; for example, company sector financial liabilities are matched by the same financial claims regarded as personal sector financial assets. The system, although still incomplete, can be solved for national income to obtain Yt = Gt + Xt- .1NSFAt_ et + llt (5) A further equation determining changes in the stock of financial assets is needed to complete the 34 model. A conventional 'closure' would involve disaggregating private expenditure (PE) into consumption (C) and investment (I) and explaining these components by independent functions, say, C = c( ...) I= i(...) (the actual explanatory variables are here left unspecified). These functions could then be substituted into equation (4) yielding, .1NSFAt = PYt- [c(...) + i(...)] With such a closure, asset acquisition is left as a residual. There is no guarantee that the pattern of asset accumulation over time implied by this closure will be a plausible one. To ensure that the pattern of asset accumulation is plausible, some restriction must be placed upon the behaviour of total private expenditure relative to income. This is the starting point for the 'New Cambridge' approach. The significance of assumptions about asset accumulation for fiscal planning derives in the first place from the flow of funds identity which we may now rewrite as PSBRt = .1NSFAt- Bt (6) The choice of a PSBR target by the government ha~ implications for the balance of payments whtch can be known if and only if .1 NSFA is predictable. Assum~ng that .1 NSFAt is predictable, fiscal plannmg ought to be a relatively straightforward affair but the government is faced by one very important constraint: it cannot choose targets for the balance of payments and the path of money national income independently unless it ~an control exports or the ratio of imports to mcome. It has to choose targets for these two variables in the light of an assessment of the benefits and risks of incurring a balance of payments surplus or deficit relative to the advantages or disadvantages of a higher or lower objective for money national income. Given the target chosen for the balance of payments and a prediction of private asset accumulation, the PSBR can be set from equation (6) and a combination of public spending plans and tax rates should be decided which will achieve this PSBR at the desired level of national income chosen jointly with the balance of payments target. So long as exports, the import ratio and private asset accumulation behave as expected, fiscal policy will yield the chosen outcomes for the balance of payments and national income. If .1 NSFA is predictable, the government can also readily assess the implications of unanticipated changes in exports and the import ratio or a failure to implement its own tax and spending decisions. Thus assumptions about .1 NSFA are crucial for assessing the stability of responses to fiscal policy and for the design of stable and stabilising fiscal plans. 4.2 The 'New Cambridge' hypothesis The 'New Cambridge' equation for aggregate private expenditure was originally, but wrongly, formulated as a flow relationship between private spending and current and lagged private income measured in real terms (i.e. deflated by the price index for total private expenditure). This formulation had implausible implications for asset accumulation, particularly under conditions of rapid inflation. The error was corrected in a reformulation published in 1976* which also provided a derivation of the expenditure equation from a simple assumption about the net stock of financial assets. The essential hypothesis, which we still take to be an appropriate presumption, was that the private sector's net stock of financial assets grows broadly in proportion to the flow of money income. A general formulation would be that the net stock of financial assets at the end of any period is proportional to a weighted average of income flows in the current and previous periodsi.e. n NSFAt =a ~ i=O (7) fwhere Wi = 1 and a is the constant of propor- tionality. This formulation allows for a variety of adjust- ment patterns. The net stock of financial assets will be augmented more or less quickly in response to a step change in the flow of money income depen- ding on whether the weights, wi, are concentrated on current income or on lagged terms. But, as will be shown below, what matters for the dynamics of private expenditure is not so much the pattern of weights as the magnitude of the constant of proportionality, a. This is a more-orless directly observable magnitude since when the income flow is not changing much it is the same as the actual ratio of the net stock of financial assets to private income. There are, moreover, reasons for our presumption that the ratio a will be constant, or will change rather slowly, over quite long periods of time. There is a considerable degree of habit and institutionalised convention in the timing of income receipts (e.g. weekly or monthly payment of wages and salaries) and of expenditure payments (e.g. trips to shops and settlement of accounts) which imply particular patterns of cash balances and debts. The practices adopted are not mainly matters of individual choice- an individual employee would not expect a change from weekly to daily payment to be met, nor could any *Cripps, Fetherston and Godley, 'What is left of New Cambridge?', Cambridge Economic Policy Review, March 1976, Chapter 6. one person choose to shop on Sunday if shops were conventionally closed on that day. Practices are, in fact, conditioned by conventions which, in normal circumstances, remain relatively unchanged over quite long periods of time. Any particular set of habits and institutional arrangements gives a certain pattern to the accumulation of money assets and liabilities. Conventions also determine how debt limits as well as reserves and cash l:!alances are planned in relation to the magnitude of income and expenditure flows. These considerations suggest why private assets and liabilities normally rise broadly in proportion to money flows of income and spending. The assumption about the net stock of private assets, formally expressed by equation (7), has quite strong implications for the behaviour of total private expenditure: expenditure must adjust to changes in the flow of income with an average lag precisely equal to the constant ofproportionality, a, in the relationship between the stock of assets and income.* Formally, PEt= PYt- 6NSFAt n+1 = (1-a w0 )PYt + a ~ Ui PYt-i i=1 (8) where Ui = wi-1 - Wi and un+1 = Wn (i = 1,n) n+1 Note that ~ Ui = Wo i=1 and therefore that the sum of the coefficients on income will be equal to unity. It can also be shown by some algebraic manipulation that n+1 n a ~ ui.i = a ~ Wi i=1 i=O Since the weights wi by assumption sum to unity, the average lag is n+ 1 a 2; Ui.i = a i=1 This result can be illustrated intuitively by a simple example. Consider the case of somebody who keeps a constant cash balance and spends income at a steady rate. One could imagine units of income entering the stock of cash and being withdrawn subsequently in the same order as they had come in. Each unit would in effect move through the stock of cash until it emerged to be *A fuller discussion of this proposition, first suggested by Wynne Godley, will be found in his book Macroeconomics, to be published in late 1982. 35 spent. The time taken for income to pass through the stock, and therefore the lag before it was spent, would be the number of days, months or years of income flow represented by the size of the stock of cash - or in other words the ratio of the stock to income. In order to aid exposition the above account of asset accumulation and expenditure behaviour has been deliberately simplified in one important respect. The private sector has access to loan finance for certain types of expenditure and the amount of loans outstanding is neither closely related to disposable income nor does the use of this loan finance affect the amount of any other type of expenditure. This implies that private expenditure as a whole is not so tightly constrained by the level of disposable income as has so far appeared to be the case. The component of aggregate private expenditure which most obviously displays these properties is stockbuilding -a large proportion of which is generally financed by loans from the banking system. Borrowing for stocks is not closely constrained by income, nor need bank lending for stocks have much effect on the finance available for the other, largely income financed, components of private expenditure. Expenditure on stockbuilding could, it seems, vary independently of the flow of disposable income without being offset by changes in other forms of private expenditure. Stockbuilding is not the only component of private expenditure to be loan-financed. Expenditure on consumer durables is often financed by hire purchase and house purchases are usually financed by mortgages. In these cases, though, the amount of credit outstanding bears a closer relationship to the level of disposable income. Loan finance only perturbs the determination of aggregate private expenditure for more than brief periods when it is genuinely independent of income levels and in practice there does not seem to be much private expenditure financed on such a basis, at least in the UK, other than stockbuilding.* 4.3 Implications for fiscal policy As mentioned earlier the observed ratio of the private sector's net financial assets to private income tells us, at least approximately, the magnitude of the average lag between private expenditure and private income. In Britain the stock of net assets is of the order of 60% of annual income. From this we may infer that the average lag in the adjustment of private expenditure must be less than one year and is probably about 7 months shorter than many macro-economists may have supposed. According to our hypothesis this lag should be stable. It follows that the demand effects of fiscal policy should be predictable and fairly quick-acting. *A more complete account will be found in Wynne Godley's Macroeconomics (forthcoming). The principles of fiscal planning, if the hypothesis is correct, should be straightforward. For the moment let us ignore perturbations due to loan-financed stockbuilding. The rate of net financial asset accumulation by the private sector as a whole will be governed by the rate of increase in money income (i.e. the combination of inflation and real growth). Ignoring any quirks in the lag structure, we may write 6NSFA = agPY where g is the rate of increase in nominal income and (l is approximately 0.6. For a zero balance of payments the PSBR should be equal to 6 NSFA. Thus, for example, if g were to be 10% per year the PSBR should be equal to about 6% of private income or 4% of national income. For a higher or lower rate of inflation or real growth, the PSBR should be proportionately larger or smaller. We have already pointed out that the government cannot freely and independently choose targets ·for the growth of nominal income and the balance of payments. Thus the figure for growth of nominal income which enters into fiscal planning must itself be the result of an assessment of inflation and the possibility of growth of real income within constraints set by exports, import penetration and balance of payments objectives. If the government chooses too low a figure for the PSBR, the growth of nominal income will be held back while the balance of payments moves into surplus. Too high a figure will raise the growth rate of income, pushing the balance of payments into unsustainable deficit. The most difficult problem for fiscal planning is in fact to judge the trends of external trade performance and the stability of external financing. In the absence of controls over movements of financial capital it is hard for the government to know in advance how ambitious it can be in expanding the PSBR to stimulate growth of domestic spending and income. It is widely supposed that risks of accelerating inflation must also constrain fiscal expansion. There is nothing in the model set out in this chapter either to confirm or to deny such a possibility. Controversies about inflation are discussed in the next chapter of this Review. Our view, which has to be argued within a more complete model of the economy incorporating determinants of inflation and therefore beyond the scope of this chapter, is that fiscal planning should normally accommodate inflation because the costs of attempting to squeeze inflation by holding down growth of nominal private income (and thereby inducing a balance of payments surplus) are far too high in terms of the sacrifice of real income and employment which must follow. It remains to consider how fiscal policy should be modified to allow for loan-financed stockbuilding. In a very open economy such as the UK 36 something like one half of fluctuations in demand attributable to the stock cycle impinge directly on purchases of imports and therefore do not affect domestic income. Some fraction of stockbuilding, perhaps one-quarter or one-third, may be financed from income or income-related debt; this implies offsetting adjustments of consumption and fixed investment spending and is therefore not a source of instability in aggregate demand or income. It seems probable, therefore, that only a small fraction of the fluctuations in stockbuilding has any destabilising impact on domestic income and economic activity. From this viewpoint, therefore, the government might well ignore the stock cycle in its fiscal planning, allowing the effect to come through in the form of related fluctuations in the balance of trade. For the purposes of deciding a figure for the PSBR, the government's calculations should be based on estimates of the balance of payments and LlNSFA which have been corrected for the stock cycle. The case for ignoring the stock cycle is not so entirely convincing, though, when international interdependence is taken into account. Synchronised stock cycles can and do cause substantial fluctuations in exports which feed into domestic spending and income. It might be hoped that if many or most countries followed stable fiscal policies the stock cycle would damp down of its own accord. But international stock cycles have caused enough trouble in the past to suggest that, despite the risks of miscalculation, governments ought to make at least cautious fiscal adjustments to try to offset them. 4.4 Actual and predicted changes in aggregate private expenditure, 1965-81 To conclude this account of the present state of our work on private spending and net financial asset accumulation, this section compares actual changes in private expenditure since the mid 1960s with the changes predicted by the specific version of our hypothesis which is incorporated in the CEPG model of the UK economy. This only provides a crude test of the hypothesis in the form in which it impinges on flows of spending and income. We hope in future to provide a fuller empirical evaluation using data on private sector assets and liabilities whose pattern and changes through time should help to reveal the motivations and institutional constraints which govern net financial asset accumulation. The measure of private expenditure used for prediction purposes in the CEPG model is consumers' expenditure plus private fixed investment measured at current market prices*. The main explanatory variable is private disposable income to which is added the net inflow of direct investment from abroad (since this constitutes ali additional flow of finance for investment). The equation includes an adjustment for changes in hire purchase credit and for borrowing to finance a part of stock building (stock appreciation is included in the definition of both stockbuilding and income). Finally, a somewhat arbitrary allowance is made for the effect of changes in real interest rates on net asset accumulation. Formally, the equation may be written as Ct + IPt= [(1-ilw0 )PYt +a~ Uj PYt-iJe-(~1 ilTt+~2 Llrt-1) i=1 where C is consumers' expenditure IP is private fixed investment r is the real interest rate HP is the value of hire-purchase debt and S is the book value of stocks and work-in-progress, all measured in money terms at current prices. The value of the average lag, a, is assumed to be 0.75.* The weights w0 and is 0.7 and ~ui = 0.3 with a Ui are set such that w0 geometrically declining pattern. The coefficients on changes in the real interest rate,~~ and ~2. sum to 0.45 and the pro- portion of stock accumulation financed from income, y, is assumed to be 30%. . Chart 4.1 compares annual percentage increases in private consumption and fixed investment predicted by this equation with actual increases from 1965 to 1981. The prediction errors are generally about I% which is well within the limits of accuracy to which private income and expendi- ture flows are measured. It should be noted that the form and the parameters of the equation were chosen on a priori grounds and that it contains no constant or fitted trend term and this makes the prediction test quite stringent. We conclude that the evidence of annual time series for private income and expenditure is consistent with our hypothesis. *For presentational purposes the version of the equation used here differs formally from that in the model (for which see the CEPG Technical Manual, Seventh edition, DAE, July 1981). Private investment has been moved to the left hand side and the equation has been multiplied by the consumers' expenditure deflator. These transformations in no way affect the equation's predictive properties. *The value of a being used here is larger because the relevant stock of financial assets is larger when allowance is made for the banking system's contribution to the finance of stockbuilding. 37 Chart 4.1 Actual and predicted changes in private expenditure, 1965-81 % (percentage increases in the money value of consumers' expenditure plus private fixed investment over the previous year) ,.1 rI r</ 't I I I I I \ \ I I I I I I I I I I 1965 1970 ---actual --------predicted 1975 1980 1981< · 38 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Chapter 4 The 'New Cambridge' hypothesis and fiscal planning Creator An entity primarily responsible for making the resource Michael Anyadike-Danes Source A related resource from which the described resource is derived Cambridge Economic Policy Review Volume 8 No 1, pages 33 - 38 Date A point or period of time associated with an event in the lifecycle of the resource April 1982