Macroeconomics 9780195635348, 0195635345

Citation preview

MACROECONOMICS

Edited by

PRABHAT PATNAIK

DELHI

OXFORD BOMBAY

UNIVERSITY CALCUTTA

1995 \ .

\

PRESS

MADRAS

O xford U niversity Press, W alton Street, O xford 0X2 Oxford New York Athens Auckland Bangkok Bombay Calcutta Cape Town P ar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madras Madrid Melbourne Mexico City Nairobi Paris Singapore Taipei Tokyo Toronto and associates in Berlin Ibadan

©

O xford University Press 1995 ISBN 0 19 563534 5

Typeset by Rastrixi, New Delhi 110070 Printed in India at Wadhwa International, New Delhi 110020 and published by Neil O'Brien, Oxford University Press XMCA Library Building, Jai Singh Road, New Delhi - 110001

|

Note from the General Editors As economics advances rapidly and becomes both more mathematical and statistically founded, the need arises to interpret its general prin­ ciples in the context of specific economies. In India, students are taught the latest models but it is usually left to them or the rare teacher to relate the models to the Indian context. The present series is an attempt to rectify this lacuna. Each book in this series presents the latest developments in a field and enunciates these in the context of the Indian economy. The series was conceived with senior undergraduate and post­ graduate students in mind. The aim is to provide accurate and interest­ ing books with contributions from leading economists. While each book has a volume editor, we—the general editors— work with the volume editors in order to try and maintain some common norms and standards for the series as a whole. Initially there was a third general editor, the late Sukhamoy Chakravarty. He was actively involved in the planning of the first few books in this series; and was a great source of inspiration to us till the last days of his life. Kaushik Basu Prabhat Patnaik

»

Contributors P rabhat P atnaik

Centre fo r Economic Studies and Planning, Jawaharlal Nehru University, New Delhi A mitava K rishna D utt

Department o f Economics, University o f Notre Dame, Indiana A miya K umar B agchi

Centre fo r Studies in Social Sciences, Calcutta C . P . C handrasekhar

Centre fo r Economic Studies and Planning, Jawaharlal Nehru University, New Delhi J ayati G hosh

Centre fo r Economic Studies and Planningf Jawaharlal Nehru University, New Delhi V . P andit

Delhi School o f Economics, Delhi

Contents Introduction: Some Indian Themes in Macroeconomics P r a b h a t P a tn a ik

Open-economy Macroeconomic Themes for India A m itava K rishna D u it

Closed-economy Structuralist Models for a Less Developed Economy A m iy a K umar B agchi

The Macroeconomics of Imbalance and Adjustment C . P. C handrasekhar State Intervention in the Macroeconomy J ayati G hosh

Macroeconomic Character of the Indian Economy: Theories, Facts and Fancies V. P a n d it Classified Bibliography Name Index Subject Index

Introduction: Some Indian Themes in Macroeconomics P ra b h a t P atnaik

The term ‘Indian themes’ used in the title of this introduction may appear intriguing to many. Are there any specifically Indian themes in macroeconomics, except in the purely descriptive sense of whatever research work happens to have got done on the Indian economy? It is my view that there are. In the process of applying the tools of analysis of conventional macroeconomics to the Indian economy, a literature has developed over the years which constitutes a distinct theoretical corpus. This literature may have a family resemblance with the parallel literature developed in the context of Latin America and other underdeveloped economies, but it differs in its concerns and perceptions from those o f conventional macroeconomics. Different segments of this literature are reviewed in the articles which follow in the present volume. The purpose of this introduction is to highlight, to critically examine, and to trace the history of the development of som e of these overall concerns and perceptions.

1 T he differentia specifica of Indian discussions on macroeconomics has been to my mind the explicit introduction of an agricultural sector and the manner in which its responses shape the overall behaviour of the economy. V.K.R.V. Rao’s celebrated article (1952) on the multi­ plier was a pioneering one in this respect. The vast literature that has followed has never lost sight of the need to incorporate this sector which, though important in our context, is usually missing in western macroeconomics discussions.

2

P r a b h a t P a t n a ik

The basic difference between agriculture and manufacturing in the short period has been commonly seen to lie in the fact, emphasized by Kalecki1, that prices in the former are demand-determined and in the latter cost-determined. The starkest scenario is of a perfectly inelastic supply curve in the former involving only price adjustment and of a perfectly elastic supply curve in the latter involving only output adjustment. In such a world if the agricultural good required per unit of manufacturing output is fixed (because wages, spent entirely on this good, are at a subsistence level and the raw material needs are technologically determined), then the maximum manufac­ turing output is determined by the surplus of the agricultural sector; and demand management by the state can only realize this maximum, but cannot eliminate excess capacity if it exists at this level. This last conclusion does not of course necessarily depend on the assumption of a horizontal supply curve in manufacturing (and the associated assumption of mark-up pricing owing to the prevalence of monopolistic elements). As Rakshit (1989) has shown, in any economy, including a barter economy, if there exists one commodity, whose supply is perfectly inelastic, in terms of which all other prices have a floor, and for which there is excess demand at these floor prices in conditions of full capacity production of other goods, then such full capacity production can never be achieved even under competitive conditions. This indeed is the essence of Keynes’ theory, as he himself perceived it in Chapter 17 of The General Theory: the low elasticity of production and substitution of money, together with the fact that wages are fixed in money terms, was the basic reason why full employment could not be achieved. In India-type economies, agricultural output in the short run can play the same role that money did in Keynes’theory. Whatever might be the specific assumptions made about the supply curve and the pricing behaviour in the manufacturing sector, if the output of this sector is constrained by the availability of agricultural supplies, the question arises: how does this constraint actually become operative in practice? How in short does the excess demand for the agricultural good get eliminated to give a single-period equilibrium if real wages are bounded below? One route explored in Patnaik (1972) is to postulate a degree of inflation-sensitivity on the part of the state for political reasons. The state wishes to maximize its 1 Kalecki (1971), p. 43.

Introduction 3 investment subject to the constraint that real wages in the manufac­ turing sector should not fall below a certain floor level. It adjusts its real spending in order to achieve a single-period equilibrium at this level of real wages; with a given money-wage rate for manufacturing workers, and mark-up pricing in this sector (which is more realistic but not logically necessary), the prices in both sectors are determined. Of course a more plausible story, involving a sequence of periods, would run as follows: for a sequence of exogenously-given agricul­ tural outputs (or supplies), it is not that the real wage in manufacturing is always at the floor (or ‘subsistence’) level, but rather it fluctuates around it. State spending is jacked up when wages are above this level and pushed down when they are below, but never quite succeeds in hitting the target and, therefore, itself shows fluctuations. With private spending determined through some version of a Kalecki-type story the economy can exhibit cyclical growth around some trend linked to an exogenously given growth rate of agricultural output (Patnaik, 1972). An additional route by which single-period equilibrium can be attained has been discussed in an elegant model by Bose (1989).2 If all manufacturing wages are not spent on food alone but a part expended on an industrial mass consumption good, and if this part consists simply of what is left over from the wage bill after the purchase of a fixed minimum level of food per head of employed workers (an extreme case of Bose’s more general discussion), then the maximum employment in the manufacturing sector is determined once again by the magnitude of the food surplus. Larger or smaller autonomous real expenditures by the state and the capitalists (on investment and consumption) simply get reflected in smaller or larger outputs of the industrial mass consumption sector. If on the first route single-period equilibrium gets established when there is a lower bound to the real wage, through adjustments in autonomous expenditure by the state, on the second route this happens through variations in the output of mass consumption goods. The first is an adaptation of Joan Robinson’s 'inflation barrier1, while the second is based on the finding by Krishnan (1964) of a negative relationship between the price of food and the demand for textiles in the Indian context (see also Patnaik, 1972). 2 This particular model of Bose is discussed in Bagchi’s paper in this volume.

4

P r a b h a t P a t n a ik

To be sure, in neither route is there a floor real wage as such. With a fixed money wage, explaining short-period equilibrium is not a problem. But an agricultural constraint upon manufacturing growth exists precisely because manufacturing real wages cannot go on falling for ever. So, this constraint must be mediated in some way. And the above discussion draws attention to two possible mediations, o f which clearly the first has to be the decisive one in the context of a sequence of periods. In what sense, however, can one talk of an agricultural constraint over a sequence of periods, i.e. in the medium or long run? Even if the short-run distinction between agriculture and industry is a valid one, can’t public investment which has a ‘crowding in' effect upon private investment in both sectors and especially in agriculture (Rao, 1972; Sen, 1 9 8 1 ), and which in the Indian context o f a planned economy of sorts can be moved from one sector to another for removing bottlenecks, be always so earmarked as to make an agricul­ tural constraint impossible, except as a transitional phenomenon? Kalecki (1972) was of the view that in the absence of radical land reforms, if the peasants are subject to disincentives arising from tenancy and usury, there is a ceiling to the rate o f growth of agricultural wage goods; once the economy hits this, the agricultural constraint becomes dynamically-binding unless real wages can be pushed down ad infinitum. A second way, apart from such an exogenously given ceiling growth rate in agriculture, that one can talk about an agricultural constraint is based on the view that while the agricultural growth rate can be raised to any level, a price has to be paid in terms of the capital-output ratio. Not only would a higher growth rate entail a higher capital-output ratio, but what is more, the maintenance of any such growth rate through time would entail a rising capital-output ratio, at any rate within the extant agrarian relations. The rise in energy-intensity o f agriculture accompanying the Green Revolution lends credence to this view. Balanced growth not tethered to some exogenous agricultural growth rate in other words can be achieved, but for any given investment ratio in the economy the rate of balanced growth would keep slowing down. This view is not synonymous with Ricardianism, since its validity is confined to the extant agrarian relations. Nonetheless its immediate practical relevance in explaining the actual macroeconomic behaviour o f the Indian economy over any stretch of time is open to doubt. At any rate in the post-Green Revolution period the nature o f the

Introduction 5 immediate binding constraint upon the growth rate of the economy, especially of the manufacturing sector, was seen to lie elsewhere. The Green Revolution, while not accelerating the growth rate in the agrarian sector, more or less sustained the earlier trend, which in its absence would not have been possible. It was also marked by increased commoditization of output. And yet for a long period of over a decade after the mid-sixties industry continued to experience much lower growth rates than it had previously. It is in this context that the so-called agricultural constraint ceased to carry much credibility. Something was clearly happening in the economy which could not be adequately captured by it. While the emphasis on a dynamic agricultural constraint was not altogether abandoned (Sen, 1981), macroeconomic discussions broke new ground. M itra (1977) and Chakravarty (1977) shifted emphasis to the question of the terms of trade between agriculture and industry. The Kaleckian distinction between cost- and demand-determined prices was replaced by the view that in the Indian economy where the state intervened in price fixation in both the sectors, especially after the mid-sixties, the terms of trade were determined by the political pressure exerted by different classes upon the state. Indeed Mitra used the concept of the terms o f trade in a far more-inclusive sense than is usual: each class strives to tilt the distribution o f income in its own favour and for this purpose uses both market as well as extra-market instruments (i.e. influence upon the state) to gamer for itself higher prices, a plethora of subsidies and tax concessions; terms of trade refer to this whole complex of benefits, constituting a relative distribution o f social income, on the basis of which a particular class, controlling a particular commodity, makes it available. His argument was that the rural rich had succeeded in tilting the terms of trade in their favour starting from the mid-fifties and especially after the mid-sixties and that this was responsible for the atrophy of the growth process. This was also echoed in Chakravarty’s remark that the increase in the political power of the rural rich after the Green Revolution was partly responsible for the favourable tilt in the terms of trade for agriculture after the mid-sixties. O n the link between the terms of trade and industrial growth two distinct strands of argument were advanced. On the demand side, M itra argued that a terms of trade shift in favour of the rural ricb (clearly sectoral terms o f trade in his conception were misleading, the relevant concept being the class terms of trade they gave rise to),

6

P r a b h a t P a t n a ik

reduced the real incomes o f both the rural poor as well as the urban workers; since at the margin these classes had a higher propensity to consume industrial goods, this resulted in a relative shrinking of the market for industry causing a slowdown in its growth. At the same time profit margins in industry also shrank, partly because of higher raw material prices and partly because the decline in industrial real wages was offset in part by lower profit margins. This too had an adverse impact upon investment in industry. Chakravarty’s emphasis was more on the latter strand. Terms of trade movements in favour of agriculture did not so much lower real wages as raise product wages in industry. They also tended to lower the overall savings rate of the economy. The net result was not so much a problem of aggregate demand, but a lower investment ratio in the economy resulting in a lower growth rate. The demand argument was clearly an incomplete one. If terms of trade movements led to a reduction in industrial demand and thereby in aggregate demand, why couldn't the state run an appropriate budget deficit, financed by borrowing from the rural rich, to make up for this reduction? After all India was not a laissez-faire economy but one that was supposed to be a planned one. Was there not a mode of intervention by the state that could have overcome deficiency of aggregate demand? This question at least had to be posed and answered in order to make the demand argument complete, a comment that is equally relevant in the context of other explanations, e.g. Nayyar (1978), which did not explicitly talk o f terms of trade but which linked industrial stagnation to growing inequality in income distribution.3 To say this is not to suggest that these explanations lacked weight, but simply that they needed fuller specification. The need for such a fuller specification in the context of the terms of trade argument was underscored in a comprehensive paper by Dasgupta (1989). Since the most significant instrument for tilting the terms o f trade in favour of agriculture is an increase in procurement price, Dasgupta's examination of equilibrium and comparative statics in a dual foodgrain regime is significant. He examines what for simplicity can be called ‘static expectations equilibria': since the free market price expected by the producers affects ceteris paribus what the government can procure, for any given expected price we can 5 An extensive literature exists on the constraining role of demand on India's economic performance. Apart from Nayyar, see for instance Bagchi (1970) and Raj (1976).

Introduction 1 think of a short-period equilibrium, and among these would be one equilibrium where the expected and the actual free market prices coincide; he is concerned with that equilibrium. Among other things he shows that a rise in procurement price, if it is accompanied by a certain level of increase in food rations at a fixed issue price, would have an unambiguous effect of increasing industrial demand, no matter what the value of the marginal propensity to consume the industrial good among the rich (assumed to be positive and identical for the urban and the rural rich), even though the equilibrium free market price also increases. The reason for this result is obvious: once we bring in procurement operations, a change in the procurement price is associated with a change in the budget deficit as well, so that mere comparisons of marginal propensities to consume the industrial good among the gainers and the losers from a price change cease to suffice (see also Das, 1989). A n attempt was made in Patnaik, Rao and Sanyal (1976) to complete the demand strand of the terms o f trade argument, eschewing dual market considerations, in the following manner: if a unit of income accruing to the rural rich gives rise to expenditures that are directly and indirectly more ‘food-intensive* than a unit of income accruing to the industrial workers, then a shift in terms of trade that shifts income distribution from the latter to the former would cause a shrinking of industrial output, no matter what the state did. To be sure the single-period equilibrium in such a case would be an unstable one f o r any given level o f autonmous real expenditure by the state a n d the capitalists, but this instability would not matter since the state can always adjust its expenditure to prevent any cumulative move­ ment* upward or downward in food prices relative to money wages (see also Bose, 1985). The problem with the argument, however, lay in th e fact that the empirical validity of the assumption about ‘food intensity* was open to question. T h e second strand of the terms of trade argument which relied not on th e demand-side effects but on the investment ratio was a more ro b u st one. According to this strand it is not the effect of a movement o f th e terms of trade in favour of the rural rich upon aggregate demand that is important but the effect on the composition of aggregate dem and, i.e. its division between investment and consumption. And if th is division moves in favour o f consumption owing to the terms o f trade shift, growth is bound to atrophy. But in raising this question the term s of trade argument already began to point to something more

8

P r a b h a t P a t n a ik

general. Any tendency for the composition o f aggregate demand to shift in favour of consumption would have a growth-reducing effect; terms of trade movements could constitute one factor causing such a tendency, but there could be other factors as well. In other words, the terms of trade argument which broke away from the agricultural constraint view, paradoxically opened the way for the discussion to proceed beyond terms of trade per se.

2 In explaining overall investment behaviour in the economy, a crucial role was generally assigned to public investment. At an empirical level Srinivasan and Narayana (1977) found the slowing down of the growth of public investment to be a significant factor behind the deceleration of industrial growth after the mid-sixties. This position was endorsed by Shetty (1978) and Patnaik and Rao (1977) and the macrotheoretic discussion veered around to the question of why such a slowing down should occur. The budgetary position of the govern­ ment became the obvious focus of attention. The argument linking the government’s budgetary position with the overall growth rate of the economy can be reconstructed as follows (see Patnaik, 1984): consider a one-good economy, eschewing any distinction between agriculture and industry, where productivity growth is ignored for simplicity, where government consumption is a fixed fraction of output, and where government revenue (raised entirely through taxes on profits) is a fixed fraction of profits. The government seeks to maximize investment subject to the maintenance of a certain floor real wage rate for political reasons (the inflationary barrier). All wages and a fixed proportion of post-tax profits are consumed and private investment (we ignore all depreciation for convenience) is simply a multiple (< 1) of private savings, the multiple itself being dependent upon the degree of capacity utilization. In such a world, budgetary policy would always ensure that output never falls below full capacity output and the wage rate never rises above the floor level. Surplus, i.e. output minus the wage bill, would bear a fixed ratio to output and so would consumption (by the government and the capitalists). The investment ratio would, therefore, be deter­ mined and the growth rate would simply be this ratio times the output-capital ratio. Any ceteris paribus decline in the tax ratio, or

Introduction 9 increase in the govemment-consumption ratio would be associated with a lower growth rate. On the other hand, with a parametric change in either o f these ratios, if the growth rate does not decline then the floor real wages must fall. In short, a rise in the ratio of the current-account budget deficit to full capacity output must result either in a lower growth rate (if floor wages are to be maintained) or in a lower level of floor wages (if the growth rate is to be maintained), or in some combination of the two. This proposition can be reformulated in terms of inflation rates. If the money wage rate in any period is simply some desired real wage rate times the expected price and if workers have static expectations (i.e. the price in any period is expected to persist into the next), then a lower level of the actual real wage rate must be associated with a higher rate of inflation. The government's constraint in terms of maintaining some floor level of real wages can then be equivalently stated as maintaining some maximum rate of inflation. Thus, for any given ratio of current-account budget deficit to full capacity output, one can think of an upward sloping ‘growth-inflation frontier' on which the government can choose any point. Since it wishes to maximize the growth rate subject to the maintenance o f some maximum inflation rate, the point it chooses gets determined. An increase in the ratio of the current-account budget deficit to full capacity output entails a shift in the position of this frontier, and unless the constraint in terms of the maximum inflation rate is relaxed, the growth rate must slow down. Now this simple one-good model can be made more elaborate to explore the implications of terms-of-trade movements. It is intuitively clear however that a terms-of-trade shift in favour of the rural rich, since it would entail larger food subsidies for the maintenance of a given floor level of real wages would be exactly analogous in its effect to a cut in the tax rate on profits, and hence would ceteris paribus entail a lower growth rate. But, then, so would an actual cut in the effective tax rate upon the rich through, for instance, growing tax-evasion or an increase in concealed transfers of various kinds to them. The focus then shifts to taxes paid and transfers obtained by the rich, i.e. both agricultural and industrial capitalists taken together, rather than terms of trade per se. What is more, once we abandon the assumption of a homogeneous category of workers and introduce the possibility of divergent moneywage movements between urban and rural workers, or divergent

10

P r a b h a t P a t n a ik

movements between money incomes of rural workers and sections of peasantry, on the one hand, and those of urban workers, on the other, it is even possible that larger transfers to the rich as a whole, e.g. larger transfers to the urban rich without a reduction to the rural rich, can be accommodated without lowering either the growth rate or the floor real wages of the urban workers, by turning the terms o f trade against agriculture: the adverse consequences of this would be felt by the agricultural workers and sections of the peasantry rather than the rural rich if the latter can depress, relatively, the money incomes of the former to a sufficient extent. Something of this sort may have happened in the period of a decade or more after the mid-seventies, when the terms of trade moved against agriculture, the urban real wages as well as the growth rate improved and yet the government’s current-account budgetary situation deteriorated (see Patnaik, 1992). What the above points to incidentally is a certain fallacy inherent in looking at rates of inflation per se, a fallacy of what one might call ‘price-fetishism*. The crucial concept of inflation which is relevant in our context is inflation in terms of the wage-unit when relative wages (or incomes per unit work) of the different sections of the working population remain constant, or inflation in terms of the wage unit of particular sections of workers when the relative wages change. ‘Income deflation’, i.e. a lowering of the profile of money incomes relative to prices greater than would have occurred otherwise, of the working population as a whole or of particular sections of it, can have the effect of bringing down the nominal rate of inflation, even though it has the same real consequences as a higher nominal inflation without the altered money-income profile. The government has often claimed credit for ‘inflation-control* even though all that such 'control' has meant is having the same consequences as inflation (perhaps in a concentrated manner for some particular group) but at a lower nominal rate of price increase. If the rise in the ex-ante share of private post-tax surplus in full capacity output (ex-ante in the sense of what would prevail along a particular growth path and with a particular floor wage)4 leads to a lowering of the growth rate, then the result of course is economic 4 A rise in the ratio of current budget deficit to output necessarily implies, for a given level of government investment and current account balance of payments deficit relative to output, a rise in the share of post-tax private profits. Since government profligacy directly leads to private enrichment, whether we locate the genesis of the problem in one or the other becomes somewhat immaterial.

Introduction

11

atrophy; and this is all the greater if the ex-ante share keeps rising. If, on the other hand, the growth rate is kept unchanged, the result is a squeeze on the living standards of the working population or some sections of it. The squeeze need not be absolute if there is productivity growth (in which case, of course, the argument has to be suitably reformulated); but with productivity growth there emerges the addi­ tional problem that extant labour reserves may get perpetuated or may even increase in relative terms. Moreover, even if productivity does increase, some segments, even large segments, of the working population could still experience a squeeze that leaves their real living standards more or less unchanged, all of which contributes to social instability. Is there, however, such a tendency? A political economy literature has developed, which despite the divergent methodological perspec­ tives o f the authors, would answer this question in the affirmative. To be sure not all strands of the political economy literature deal with this question. Dandekar (1981) for instance is more concerned with defining the nature of exploitation in Indian society, and for that purpose formulating criteria for ‘unequal exchange* (see also Sau, 1978), than with the dynamics of the economy. But for several other strands, the question of dynamics has been at centrestage. Mitra’s book (1977) can be read at two distinct levels. At one level which is its immediate concern it is about the terms of trade extracted by the landlords. At another level, however, it is about the struggle over distributive shares, in which the workers, taking urban and rural workers together, lose out in relative terms, which, as long as the polity does not turn authoritarian, must necessarily impinge on the growth rate. Bardhan (1984) sees the state as resting upon a coalition of three classes, the capitalists, the landlords and the *upper bureaucracy*. Their increasing claims produce a fiscal crisis which progressively snuffs out the economy’s growth performance. Patnaik (1988) sees the landlords and the capitalists as constituting the ruling class-alliance, and the increase in their ex-ante share as part of a process of primitive accumulation, though with a difference: unlike in the classical case, this primitive accumulation is not the harbinger of a vibrant capitalist order. A noteworthy feature of this strand of the macroeconomics dis­ cussion has been its emphasis on what one may call ‘overconsumption*, as opposed to the usual ‘underconsumption* argument. A rise in the ratio of the current budget deficit along any

12

P r a b h a t P a t n a ik

growth path leads to a lowering of the growth rate for any given floor wage rate because the ratio of consumption to output increases. Suppose, however, all post-tax private surplus was saved (though not necessarily invested) so that this ratio did not increase. Would there still be a slowing down of the growth rate? In other words, is there any difference between tax-financed public investment and borrow­ ing-financed public investment (leaving aside considerations of pos­ sible interest payments on public loans) other than the fact that the latter may be accompanied by higher consumption? Mrs Robinson, in response to Paul Baran’s argument that there was a difference (other than with respect to consumption) between the two cases had accused Baran o f dragging in the Quantity Theory of Money;3 elsewhere she had called this argument ‘the humbug of finance’. Whether or not one believes this to be humbug, the need for examining the argument cannot be gainsaid. But unfortunately not much attention has been paid to it, and this in turn is a reflection o f what I think is a basic shortcoming of the Indian macroeconomics discussion, namely its inadequate exploration of how exactly money 'matters’ in the Indian context.

3 By and large, the Indian macroeconomics discussion continues to be trapped within the neo-Keynesian-monetarist dichotomy. Authors can be neatly pigeonholed into two different groups: those who believe in endogenous money and carry on their analysis in terms of flow decisions on the assumption of given money wages during the single period, and for whom therefore money does not really matter (the Quantity equation being read from the 'right to the left’); and those for whom there is a given money supply, determined by the magnitude of reserve money and the money multiplier, and the level of money income is determined by a constant stock-flow ratio in the single period which may change slowly over time. Our monetary institutions are of course different from those prevailing in the advanced capitalist countries. We do not, for instance, have a bill-market worth the name, and until recently the capital market was dominated by financial institutions which doled * See Joan Robinson (1966), p. 90.

Introduction

13

out funds at fixed rates of interest. In short, there can be little quarrel about the fact that the interest rates for much of the period we are concerned with were administered, though these changed from time to time, and hence did not clear any market. A monetarist under these conditions would have to fall back on the Keynesian concept (of the Treatise) of a 'marginal fringe of unsatisfied borrowers’ to explain why the money multiplier would operate. A proponent of endogenous money would have to believe that in the credit market, the short side is represented by demand. But if they so believe, then the difference between their arguments, especially on the effects of a fiscal deficit (abstracting from the external sector), can be stated as follows: the neo-Keynesian would argue that a fiscal deficit, apart from its primary impact, gives rise to subsidiary expenditure flows owing to its income and wealth effects, and that these are all that matter; a monetarist on the other hand would argue that all these are subsumed under its overall effect arising from the fact that it boosts reserve money, and that the latter therefore captures its final impact. Neither of these positions, however, carries us very far. The endogeneity argument is vitiated by the fact that credit rationing does occur, and it has at least to be reckoned with. The monetarist argument is vitiated by the fact that banks are not continuously fully-stretched, that the value of the money multiplier does move around a great deal, as does the credit-deposit ratio, indicating that over certain periods banks are more liquid and over other periods they are less so. Both positions, moreover, suffer from at least two obvious limitations. The first relates to the fact that tightening of credit not only appears to have an impact on the rate of inflation, but to do so often with extraordinary rapidity. While the fact of its having this impact militates against the neo-Keynesian argument, the fact of its doing so with such rapidity militates against monetarism. What is involved in many of the inflationary episodes in other words is the build-up of speculative pressures based on inflationary expectations. Now, to be sure, inflationary expectations figure prominently in contemporary monetarism, though as Kaldor once remarked they are not a monopoly of monetarism;6 but the inflationary expectations underlying specula­ tive behaviour are an altogether different animal from those which figure in monetarist theory. A given increase in money supply caused by an increase in credit for speculative stock holding would have very 6 Kaldor (1986).

14

P r a b h a t P a t n a ik

different price implications compared to the same increase in money supply caused by an expansion of credit for purposes of ‘normal’ expenditure. Theoretical considerations of commodity speculation and the effect of the state of credit upon it have not been adequately incorporated into the Indian macroeconomics discussion. This leads on to the second point. The impact of monetary expansion upon asset markets has not been clearly distinguished from that on goods markets. Monetarism, postulating in the short run, a constant income velocity of circulation, abstracts altogether from considerations of what happens in asset markets. But even Keynesian­ ism, no matter what version of it one considers, does not pay sufficient attention to it, notwithstanding Keynes’ elaborate discussion of the speculative demand for money. The problem lies in the fact that unlike Irving Fisher whose transactions demand for money included what is demanded both for goods and for asset transactions, Keynes confined transactions demand exclusively to goods transactions. In the process, however, an element which is not only of practical importance (after all, the monetary expansion which fed the stock-market scam in India was required for asset-transaction purposes), but also an equilibrating mechanism, disappeared from analysis. Let me clarify this last point. Consider a world (ignoring all government transactions for the purpose of the present argument) in which money wages and money supply are given. Investment is interest-inelastic because of the prevalence of oligopolies and can also be taken as given in real terms in the short period. Pricing is prime-cost-plus, and all wages and a fixed fraction of profits are consumed. With aconstant income velocity of circulation, the demand for money for income-transaction purposes gets determined. Now suppose there are two non-reproducible assets in this economy, for each of which economic agents have unit-elastic, certain, but diver­ gent price expectations. The own rate of interest of each of these assets for each economic agent is completely unrelated to its money price but declines as the agent holds more of the asset. But, for all agents and for all assets, the sum of the own rate of interest and the expected rate of price appreciation (i.e. what Kaldor called the ‘own rate of money interest’)7 is always positive. These are by no means far-fetched assumptions: one can easily think of land and gold as two possible examples of such assets. In such a world,‘if the supply of 7 Kaldor (1964).

Introduction

15

money exceeds the demand for it for income-transaction purposes, there can be no monetary equilibrium according to the logic o f stand­ ard theory, since nobody will wish to hold money as an asset. But there would be an equilibrium in such a world if there is an assettransaction demand for money: since asset transactions must occur because of the divergence of views, asset prices cannot rise in an unlimited fashion with a given money supply. Putting the matter differently, speculative demand for money, which does involve asset choice, refers to that element of the demand for money which arises for its non-use. It does not incorporate that element of the demand for money, which also deals with the realm of asset choice but which arises because money is used. And the latter is what I mean by the asset-transaction demand for money. Once we take account of it, no unique relationship between money supply and money income can be postulated. To be sure this is a conclusion which Keynes also derived by making the income velocity of circulation a function of the rate of interest, but this same conclusion holds even in a world with fixed interest rates and credit rationing, since the income velocity of circulation of money becomes a function of asset-prices. Indeed in the above example, if the minimum of the own rates of money interest of the different asset for the different agents lies above the administered interest rate which banks charge on credit (which for simplicity we assume to be single rate), then the banks will always be operating at the credit ceiling (i.e, the money multiplier analysis will hold), and yet there will be a monetary equilibrium quite different from what the Quantity Theory postulates. At the very least the above two arguments suggest that even any standard analysis of monetary equilibrium has to be done for an India-type economy in a manner more complex than has been attempted hitherto, on the basis of a wider set of specifications and incorporating many more variables. But they perhaps suggest some­ thing more, namely, in a world with credit rationing, who gets the credit and who is rationed out may itself affect the equilibrium configuration, in which case having a more inclusive macro-model itself would not be adequate; one has to specify in addition the precise rationing rule (see Patnaik, 1992b). While the credit structure, and more generally the financial institutional structure, has been studied in a number of reports, e.g. the Banking Inquiry Committee, the Dahejia Committee, and the Chakravarty Committee, and much

16

P r a b h a t P a t n a ik

s

econometric work of interest (referred to in V. Pandit's paper in this volume) as well as a certain amount of theoretical work, e.g. by Bose (1987) and Rakshit (1982,1986), has been done, this is an area, which requires a considerable amount of further research. This lacuna in our macroeconomic research, however, does not in my view invalidate in any way the arguments summarized in the earlier sections; what it does is to rob them of roundedness, and also make debates on short-term policy issues ill-founded. Let me, therefore, pick up the threads once again from the earlier sections.

4 A major hallmark of the models mentioned above was their ignoring of the external sector. The implicit assumption underlying them was that the exchange control regime in operation more or less ensured a balance between imports and the autonomously-determined level of foreign exchange availability (consisting of export earnings and foreign aid), with frustrated import demand spilling over into an equivalent demand for domestic goods. In such a world it clearly makes sense to concentrate attention on domestic developments for the purpose of examining the dynamics of the economy. In making this implicit assumption, however, a dimension of analysis was lost sight of, which in retrospect has proved to be the crucial factor underlying the regime change currently occurring in India. To see this, let us extend the analysis of section 2 by incorporating the external sector. Suppose, to start with, the economy happens to be on a steady growth path, determined by the full capacity-investment ratio and the capital-output ratio, with the exchange-control regime ensuring foreign exchange balance along the path. This would mean not only that the ex-ante import propensity of the economy (in the absence of controls) is greater than the ex-post import propensity, but also that the profile of ex-post import propensities through time is such as to ensure foreign exchange balance in every period for any given autonomous rate of growth of foreign exchange availability. Now suppose the current-account budget deficit increases, but the government pursues a policy which keeps the economy on the same, growth path with the same foreign exchange availability profile. There would in such a case be two distinct kinds of pressure on the government. The first, as we saw earlier, is a decline in real wage

Introduction 17 (we ignore productivity growth for simplicity) which is ensured through a higher rate o f inflation along the growth path; and this has a political fall-out. The second consists of the following: this decline in real wage would be accompanied by a rise in the share o f post-tax profits in output, and of consumption out of profits. Now, if at the margin the ex-ante import propensity out o f profits is greater than that out o f wage incomes, the overall ex-ante import propensity of the economy would increase. To remain on the same growth path and therefore to maintain the same profile o f ex-post import propensities, the exchange-control regime would have to be tighter than would have been the case otherwise (tightness being measured in terms o f the difference between ex-ante and ex-post import propensities). Thus the discontent among both the poor as well as the rich would be greater among the poor because their real income would decline, and among the rich because the actual volume of imported goods available to them relative to their total import demand in any period would be smaller. In the face of this twin-pressure, the tendency o f the government would be to obtain larger foreign borrowings which would keep both domestic inflation as well as the pressure on the exchange regime in check while maintaining the growth rate. And if the current-account budget deficit keeps rising, the recourse to foreign borrowing, over and above exports and foreign-aid availability, would also keep rising relative to output. This, however, is not a sustainable state of affairs. Sooner or later the country would face an external financial crunch, when it would have to cut back the ratio o f its external net borrowing to output Now it may appear at first sight that the trajectory o f this adjustment is simply identical, though in the opposite direction, to that which saw an increase in the ratio of net external borrowing in the first place, i.e. it simply has to lower the ratio of the current budget deficit to output and that this would automatically lower the ratio o f net external borrowing to output without departing from full capacity use and lowering the growth rate. This, however, is an erroneous supposition even logically. There is, in other words, a logical asymmetry between the tw o situations which arises for the following reason. An expansion of demand at full capacity use cannot possibly raise output which is at its maximum, but a contraction in demand can lower output. Unless therefore an effort is made to ensure that the effect of this contraction is felt exclusively on imports, the economy would go into a recession in the sense of slipping below full capacity

18

P r a b h a t P a t n a ik

use. To put it differently, unless the ex-post import propensity is lowered sufficiently to permit full-capacity use with lower external borrowing, a mere reduction in the current budget deficit would reduce external borrowing requirements through the instrumentality o f a recession. Typically, however, a country which has run into a balance of payments crisis owing to a large debt pile-up is forced into IMF conditionalities which insist upon a ‘liberalization* of imports. Such ‘liberalization* has an effect which is the very opposite of what is required to ward off a recession: instead of lowering the ex-post import intensity compared to the pre-adjustment situation, it actually increases this intensity to its ex-ante level, which enforces a further reduction in output in order to balance the payments. Putting the matter differently, adjustment to a lower level of external borrowing is associated with recession for two quite distinct and mutually complementing reasons: with an unchanged ex-post import propen­ sity, there would be a recession anyway as the budget deficit is cut to lower external borrowing requirement; additionally, with import ‘liberalization* the actual import propensity increases. This neces­ sitates a further contraction o f the economy to achieve this lower borrowing target.8 This contraction, however, refers to the adjustment process. What happens over a longer period? Unless the rate o f growth of autonomous net foreign exchange inflow is stepped up compared to the pre-adjustment path, the rate of growth of output in the economy would be less than or equal to what prevailed earlier. It is obvious that in any economy which balances its payments, if the average import propensity is constant, the rate of growth of output must equal the rate of growth of autonomous net foreign exchange inflow. A post ‘liberalization* economy would have a higher level of import propen­ sity and hence, as we have seen, a lower level o f output as a consequence of adjustment; but if the rate o f growth o{ net auto­ nomous foreign-exchange inflow continues to be the same as before, its output growth-rate cannot possibly be any higher than before. In fact, to the extent that the earlier output growth rate exceeded the growth rate of this inflow and was sustained on the basis of a declining ex-post import intensity, the new output growth rate can only be lower than before. 8 This point is developed in the paper by Chandrasekhar in the present volume.

Introduction

19

It may be argued at this point that exchange rate adjustments could possibly break the dependence of the economy's growth rate upon some allegedly given levels of import propensity and autonomous growth rate of net foreign exchange inflow. Surely these two entities become variables influenced by the exchange rate. How then can we treat them as autonomously determined parameters? To discuss the influence of the exchange rate, however, we have to go beyond the simple one-good model that has been used hitherto, but all that is said below does not invalidate the conclusions derived till now from the one-good model. In any economy which uses an essential input that is imported (i.e. is a 'basic good' in Sraffa's terminology), such as, say, oil in the Indian context, and where pricing is 'mark-up' (as would be the case when the economy has slipped below full capacity), a nominal exchange rate depreciation would give rise to a real exchange rate depreciation, only if either the real wage rate or the profit-margin declines.9 If, as is likely, the profit-margin, determined by the mark-up, remains unchanged, the necessary conditions for a real depreciation is a fall in real wages. If the commodities released by such a fall are not easily exportable directly or cannot easily substitute existing imports, the effect once again is recessionary. To be sure, to the extent that such a depreciation-engendered recession has a foreign exchange conserving effect (which it may not have if expectations of further depreciation lead to capital outflows), the required cut in the budget deficit for balancing payments during the adjustment phase is less. In other words, unless the government is bound by independent deficit-cut targets as part of the 'conditionalities' (in which case the two recessionary effects would be additive), depreciation obviates to an extent the need for a deficit cut. But we are concerned here with a separate issue. Clearly, for a depreciation to work at all, the share of wages must fall; and even if it does fall, the effects of depreciation (assuming all along that competitive depreciations are not occurring elsewhere) can make themselves felt only over a certain more or less prolonged period of time, and depend upon how sensitive the demand for exports and for import substitutes is to relative prices. More importantly, however, the growth of exports is not identical with the growth of net foreign exchange inflows. If an exchange rate depreciation is accompanied by expectations of a further depreciation, 9 This is an adaptation of Kador’s proposition in 'What is Wrong with Economic Theory?*, reprinted in Kaldor (1978).

20

P r a b h a t P a t n a ik

the growth of net foreign exchange inflows would slow down even if the growth of exports increases, in which case, notwithstanding some effect o f depreciation in reducing the import propensity of the domestic economy, the overall rate of growth of the economy consistent with payments balance may still be less than in the pre-adjustment phase. It may be thought that this only means that the exchange rate should be depreciated still further, but frequent depreciations o f the real exchange rate only strengthen expectations o f further falls in it and can turn out to be further counterproductive. This argument about expectations has an important implication, namely in the absence of inelastic price expectations in the foreign exchange market, a single-period equilibrium may not exist at all under a flexible exchange-rate regime at given levels of government consumption and investment, capitalists* consumption and invest­ ment, and the tax ratio. In appreciation of this fact, it is often suggested for underdeveloped economies that they should not allow capital account convertibility of the currency even while permitting current account convertibility. But the distinction is a spurious one, since any non-repatriation of export earnings amounts to a de facto capital outflow. In short, the foreign exchange market is not like the market for matchboxes. The argument that if this market is 'freed1, the price would always move to ensure an equilibrium is a naive one. Not only do movements in this price have serious implications for income distribution, but what is more an equilibrium itself may not exist within some reasonable range of price movement, or even at all. In practice, therefore, a government operating within a democratic polity and, hence, perforce sensitive to inflationary squeezes on real wages, would be constrained to halt at some point the downward slide of the real exchange rate. And this brings us back to the original issue, namely any adjustment of an economy to a lower level of external borrowing requirement that is sought to be achieved while abandoning all measures of policy control over its import propensity, would entail not only a recession during the adjustment phase itself, but also lower output growth thereafter, unless the autonomous growth of net foreign exchange inflow (at this floor exchange rate) is higher than before. The constraints upon export growth arise no doubt on the demand side, especially in a period such as now when the advanced capitalist economies are passing through a recession whose end is nowhere in sight. Additionally, however, the very mode of adjustment creates

Introduction 21 serious supply-side constraints. In discussing adjustment above, we assumed that the government cuts the current budget deficit. As a matter of fact, however, deflating the economy takes the form o f a cut in government investment for a variety of reasons: first, it is easier; and secondly, IMF conditionalities fix targets in terms o f the fiscal deficit rather than the current budget deficit, and, since these targets are time-bound, the rapidity with which investments can be cut makes them a particularly popular candidate for using the axe on. This is precisely what has happened in the Indian economy. Investment cuts, however, create supply constraints later, especially in infrastructure, which also hamper export growth. In the context of this scenario, it is often argued that if the government privatized some of its producing units, the resources it would obtain therefrom would enable it to carry out its investment programme without any cuts. Indeed, the proceeds from the sale of public sector units are sometimes considered equivalent to revenue and hence such sales are seen as a means of closing the fiscal deficit. This, however, is logically untenable. Consider the case where the public sector units are sold domestically. Unless the purchase is financed through an equivalent voluntary reduction in the level of flow expenditure that would have been otherwise incurred, either by the purchasers themselves or by somebody else from whom they borrow indirectly, there is no reduction in ex-ante absorption in the economy. If, for instance, the purchase of public sector units is financed by borrowing from banks which happen to have excess reserves, then the fact that the proceeds in accounting terms can be shown as closing the fiscal deficit makes not an iota o f economic difference: from an economic point of view, it is as if the deficit has not been closed. Matters are, o f course, different when the purchasers are foreign investors who pay for it through foreign-exchange inflow. In this case, the balance of payments position during the period in question does improve even without any investment cuts on the part o f the govern­ ment, so that it is possible to argue that sale of public sector units is a means of sustaining the government’s investment programme. But involving as it does the transfer o f crucial national assets to foreign owners, it is a singularly high social price to pay for the dubious luxury of maintaining a ‘liberal’ trade regime. Indeed, much confusion prevails on the question of foreign invest­ ment. There is a general feeling that any inflow o f direct foreign

22

P r a b h a t P a t n a ik

investment is beneficial for the economy since it adds to domestic production and employment. This, however, is erroneous. A distinc­ tion must be drawn between foreign investment that comes in to meet the domestic market and foreign investment that comes to locate production in the domestic economy for the purpose of exporting. While the latter does benefit the economy, the same cannot always be said of the former. Since the import content both in its capital cost as well as in its current cost tends to be higher than is the case, or can be made to be the case, with domestic producers, if foreign capital for meeting the domestic market displaces domestic capital, it can become a potent means for economic retrogression. An example should illustrate the point. Suppose direct foreign investment worth Rs 100 comes into the economy and finances the capital cost consisting solely of imported equipment. A domestic producer un­ dertaking the project would have used, or could have been asked by the government (especially if it was a public sector unit), to use local equipment. Owing to the foreign investment, therefore, Rs 100 worth of local equipment ceases to be produced, which is an example of de-industrialization. Additionally, the import content of the current cost is higher for the foreign firm than would or could have been the case with the domestic firm; and then there is outflow in the form of dividends, royalties, technical fees, etc. As a result, the import intensity of the economy goes up so that for any given growth rate of net foreign exchange inflow, to which this particular act of foreign investment contributes nothing (since it does not export), the growth rate of the economy slows down. No doubt we have talked of a single instance of foreign investment, but a more general argument involving a sequence of periods can be easily constructed to which this gives an insight. Aware that the rate of growth in an open ‘liberal* regime depends crucially on the rate of growth of net foreign exchange inflow, of which export growth must be the key component, many writers have seen in agriculture a promising source of export growth. Leaving aside the question of how feasible this is, there is an implication which has been missed. Agricultural growth, it was argued in section 1, depends cru­ cially upon the provision of infrastructure and hence upon public investment in irrigation, power extension, etc. If these are being cur­ tailed on account of the need to keep down the fiscal deficit, so that the trend level of agricultural growth is not increased, a higher rate o f growth of agricultural exports must mean a lower rate of growth o f

Introduction 23 domestic availability o f agricultural goods, and hence a lower rate of growth of per capita availability, including of food, than hitherto. Since the increase in per-capita food availability in the post-independence period has itself been marginal, the real achievement, if it can be so called, being the arrest of the decline which had been occurring during the last half-century of colonial rule, the new scenario, if it comes to pass, would witness a worsening of the already precarious food situa­ tion of the masses. We could well have a re-enactment of the pre­ independence experience when export growth was accompanied by declining per capita food availability and growing poverty of the people. At any rate, the question of agricultural exports brings us back to the agriculture-industry relationship with which we started this introduction, a relationship that continues to be important for Indian macroeconomics.

5 One may believe, though I do not, that with the current change in regime in the Indian economy, a good deal of direct foreign invest­ ment would flow in for meeting the international market; one may believe, though I do not, that even otherwise the Indian economy would get launched onto a higher growth path as a result of the new policies producing a far better export performance than hitherto. In short, one may support the current regime change which I do not. But no matter what position one takes on policy debates it has to be founded upon a consistent macroeconomics. Policy debates in other words necessarily spill over into debates on macroeconomics. Unfor­ tunately there has as yet been too little of the latter in the current Indian context, but in the coming months debates on macroeconomic perspectives are likely to assume significance. For this it is useful if such macroeconomics discussion as has occurred in the past is made available in the form of surveys of particular areas, to a wider academic audience. And that is the purpose of the present book. The papers which follow are not meant to be original articles on particular subjects. They are meant to be surveys of the literature, with different degrees o f comprehensiveness which are a matter of choice of the authors; the authors* own positions and perspectives come out through these surveys. Readers may or may not sympathize with these positions, but since the authors always present views

24

P ra bh a t P a tn a ik

opposed to their own at some length, they are presented with the opportunity to follow up whatever strands of argument take their fw cy. The scheme o f the book perhaps requires a word of explanation. The first essay by Amhava Krishna Dutt is a broad and fairly comprehensive survey o f the entire field of Indian macroeconomics literature. This is followed by three surveys of particular areas. Amiya Kumar Bagchi looks critically at the literature on the agricultural con­ straint which was referred to in section 1 above. C.P. Chandrashekhar looks at the macroeconomic models of adjustment. Since on this, as mentioned above, the specific Indian discussion has as yet been insufficient, and those arguing for the current set o f policies have tended to fall back upon macroeconomic models worked out at the IMF, the thrust o f his essay is directed at a critical lode at the IMF models, the Folak model to start with and the Khan-Knight model which came later (the two incidentally are quite different as he argues in the paper). Jayati Ghosh surveys the literature on the state. This may appear somewhat unusual, but fundamentally underlying macroeconomic debates are alternative theories o f the state. Keynesian macroeconomics for instance, taken in its totality, viewed the state as a supra-social entity, capable of rising above particular interests and ordering the economy to function in a rational manner to achieve humane ends. Subsequent critiques of Keynesianism have taken three different forms: first, the market does not function as badly as Keynes made out; secondly, the functioning o f markets in a regime o f pervasive state intervention is quite different from its functioning under laissez-faire so that state intervention brings in its train new kinds o f contradictions; and thirdly, the functioning o f the state is quite different from what Keynes visualized. This last is ip the fact an overarching argument, in the sense that both the second and the third strands o f criticism are based explicitly or implicitly on alto­ gether different theories o f the state, not as an agency to end all economic problems, but one with its own peculiar constraints and modes o f behaviour, which are socially conditioned. Bringing out ex­ plicitly the different perspectives on the state and the macroeconomic theories they give rise to is therefore a worthwhile exercise. Finally, V. Pandit surveys the large and rapidly-growing Held o f macroecono­ metric modelling which purports to throw light on the relevance of alternative macroeconomic positions. What it has achieved so far may be matter o f debate; what it can achieve raises deep methodological

Introduction 25 issues which have been debated since Keynes* own time (including by Keynes himself apropos Tinbergen’s work). But there is no gain­ saying its growing appeal and vigour. Macroeconometrics has not only come to stay, but has become an increasingly influential member of the household, and accordingly finds its due representation here. There are obvious gaps and lacunae in this volume. Some of these gaps, e.g. the absence of any survey of the literature on pricing behaviour and on the labour market and the inadequate review of the work on money, would, hopefully, be filled in by subsequent volumes in this series. Meanwhile, I hope that the present volume which has grown large enough as it is would be found useful by the readers.

REFERENCES 'Long-Term Constraints on India's Economic Growth*, in Robinson E .A .G . and M. Kidron (ed.), Economic Development in South Asia, Macmillan, London. B ardhan , P.K. (1 9 8 4 ), The Political Economy o f Development in India, Oxford University Press, Oxford. B ose A mitava (1 9 8 5 ), ‘On th e Macroeconomics of Development Models* (mimeo.), Indian Institute of Management, Calcutta. ------(1987), ‘Money and Commodities* (mimeo.), Indian Institute of Management, Calcutta. ------(1989), ‘Short-period Equilibrium in a Less Developed Economy*, in Rakshit, Macroeconomics o f Developing Countries. C hakravarty , S. (1977), ‘Reflections on the Growth-process in the Indian Economy*, in C . Wadhwa (ed.). Some Problems o f Indian Economic Policy, Tata McGraw-Hill, New Delhi. D an dekar , V.M. (1981), The Worker-Peasant Alliance, Orient Longmans, Calcutta. D as C handana (1989), ‘Food Policy in a Dual Economy*, in Rakshit, Macroeconomics o f Developing Countries. Dasgupta, D. (1989), ‘Procurement Price, Market Price, Employment and Effective Demand*, in Rakshit, Macroeconomics o f Developing Countries. K a ldo r , N . (1 9 6 4 ), ‘Own Rates of Interest* in his Essays on Economic Stability and Growth, Duckworth, London. ------(1976), ‘Inflation and Recession in the World Economy*, Economic Journal, reprinted in Kaldor, Further Essays. B agchi, A .K . (1 9 7 0 ),

26

P r a b h a t P a tn a ik

------(1978a), ‘What is Wrong with Economic Theory?’ reprinted in Kaldor, Further Essays. ------(191%b),Further Essays on Economic Theory, Duckworth, London. K alecki , M . (1 9 7 1 ), Selected Essays on the Dynamics o f the Capitalist Economy, CUP, Cambridge. ------(1972), ‘Problems of Financing Economic Development in the Mixed Economy', in Selected Essays on the Growth o f Socialist and Mixed Economies, CUP, Cambridge. K rishnan , T.N. (1964), ‘The Demand for Mill-Cloth in India: A Study of the Interrelationship Between Industry and Agriculture*, Artha Vijnana, Dec., 4. M itra A shok , (1977), Terms o f Trade and Class Relations, Cass, London. N ayyar D eepak , (1978), ‘Industrial Development in India: Some Reflec­ tions on Growth and Stagnation*, Economic and Political Weekly, Special Number. P atnaik , P . (1972), ‘Disproportionality Crisis and Cyclical Growth*, Economic and Political Weekly, Annual Number. ------ (1984), ‘The Market Question and Capitalist Development in India*, Economic and Political Weekly, Special Number. ------(1988), Time, Inflation and Growth, Orient Longmans, Calcutta. ------(1992a), ‘Terms of Trade and the Rate of Inflation*, in Kaushik Basu and Pulin Nayak (ed.). Economic Theory and Development Policy, OUP, New Delhi. ------(1992b), ‘Credit, Inflation and Deflation*, Economic and Political Weekly. P atnaik , P. and S.K. R ao (1977), '1975-76: The Beginning of the End of Stagnation?*, Social Scientist, Jan-Feb. P atnaik , P., S .K . R ao and A. S anyal (1976), ‘The Inflationary Process: Some Theoretical Comments*, Economic and Political Weekly, 23 October. Raj, K.N. (1976), ‘Growth and Stagnation in Indian Industrial Development*, Economic and Political Weekly, Annual Number. R akshit , M. (1982), The Labour Surplus Economy, Macmillan, Delhi. ------(1986), ‘Monetary Policy in a Developing Country* (mimeo.), WIDER, Helsinki. ------(1989a), ‘Effective Demand in a Developing Country*, in Rakshit, Macroeconomics o f Developing Countries. ------ (1989b), (ed.). Studies in the Macroeconomics o f Developing Countries, OUP, Delhi.

Introduction 27 R ao , S.K. (1972),

4Inter-regional Variations in the Growth of Agriculture and Population in India’, unpublished Ph.D. thesis, Cambridge. Rao, V.K.R.V. (1952), ‘Investment, Income and the Multiplier in an Underdeveloped Economy’, Indian Economic Review. R obinson , J. (1966), Economic Philosophy, Penguin, Harmondsworth. Sau, R. (1978), Unequal Exchange, Imperialism and Underdevelopment, OUP, Calcutta. S en A bhuit , (1981), ‘The Agricultural Constraint to Economic Develop­ ment: The Case of India*, unpublished Ph.D. thesis, Cambridge. S h etty , S.L. (1978), ‘Structural Retrogression of the Indian Economy since the Mid-Sixties*, Economic and Political Weekly, Annual Number. S rinivasan , T .N . and N .S .S . N arayana (1977), ‘Economic Performance since the Third Plan and its Implications for Policy*, Economic and Political Weekly, Annual Number.

Open-economy Macroeconomic Themes for India A m ita v a K rish n a D utt

1. INTRODUCTION The Indian economy has a low level of interaction with the rest of the world: exports, imports, and foreign debt, as percentages of total output, are extremely low by international standards.1 Perhaps this explains why the bulk of the analytical discussion on the Indian economy focusses on closed-economy issues such as the domestic savings-investment process and its relation to income distribution and government activity, or the dynamics of the agricultural sector and its interaction with the rest of the economy. However, open-economy issues have attracted a fair amount of attention at certain times when they involve important policy debates, as is the case at present.2 For other times, however, open-economy 1 According to The World Bank's World Development Report 1988, in 1965 and 1986, respectively, exports of goods and non-factor services were 4 and 6 per cent of GDP, the corresponding percentages being 19 and 14 for low-income economies other than China and India, 15 and 21 for middle-income countries, and 12 and 17 for industrial market economies. Imports as a percentage of GDP for India for the same years were 6 and 8, compared to 22 and 22 for the other low-income economies, 17 and 21 for middle-income countries, and 12 and 17 for industrial market economies. Total long-term external private and public debt as a percentage of GNP for India stood at 15.1 in 1986, the lowest figure for all low- and middle-income debtor nations with the exception of China. 2 In the early phase of India's industrialization the reliance on foreign aid attracted a fair amount of attention. Bhagwati and Chakravarty's (1971) survey, which was first published in 1969, devoted almost 12 per cent of the space (excluding the introductory and concluding sections) to open-economy issues. However, these received far less space than issues relating to planning and agriculture. More recently, with the increasing

Open-economy Macroeconomic Themes for India 29 issues should not be glossed over because closed-economy macroeconomic models usually have implications which are different from open-economy models, even when trade and other international flows are small; thus, open-economy approaches are necessary, even for showing the primacy of closed-economy factors? Moreover, as some have argued, it is possible that a low level of international interaction may be the cause of problems for a given economy. The purpose of this chapter is to survey the analytical openeconomy macroeconomics literature on the Indian economy. Sec­ tion 2 will examine some theoretical approaches and models which are relevant for analysing open-economy macroeconomic issues for India. Neoclassical, monetarist, two-gap, Keynes-Kalecki, structural­ ist, and North-South approaches will be discussed briefly. Section 3 will focus on key issues in Indian open-economy macroeconomics in the light of these analytical approaches. These issues relate to macroeconomic constraints on growth, trade policy (with reference to the debate between import substitution and export promotion), exchange rate management and other stabilization policies, and the effects of foreign capital flows. At the outset, four clarifying comments need to be made regarding the scope of this paper. First, our concern is with macroeconomic themes (such as the determination of aggregate output, growth rates, distribution, and the balance of payments) rather than with issues analysed using the tools of traditional microeconomics; thus, tools of traditional microeconomics will be used only where necessary. Second, our interest is in analytical issues rather than descriptions of events and policies except to the extent required for placing the discussions in their proper contexts. Third, our attention is focussed on the more recent literature, referring to earlier contributions only very selectively. Fourth, even within these boundaries, space limita­ tions force the exclusion of more specialized topics such as the macroeconomic implications of international technology transfers, the effects of multinational investment, and international labour mobility and the brain drain. liberalization of the economy, issues relating to trade and foreign investment are receiving greater attention. 3 It is worth noting that despite the low share of exports in GDP for the US (5 and 7 per cent in 1965 and 1985, respectively), the US macroeconomics literature is preoccupied with open-economy issues relating to import competition, the trade deficit, and foreign capital flows.

30

A m it a v a K r is h n a D u t t

2. OPEN-ECONOMY MODELS In this section we examine some models which have been used—or are relevant—for analysing open-economy issues for India. Neoclas­ sical, monetarist, two-gap, demand-constrained, and structuralist models are discussed in which the economy interacts with the rest of the world; North-South models are discussed for the bearing they have on India's role in the world economy. Our purpose is to provide a brief review of the central features of the different approaches (to make the discussion reasonably self-contained), to comment briefly on their relevance for the Indian economy, to survey specific applica­ tions of the approaches to the Indian economy and, finally, to provide a theoretical background to the issues discussed in the next section. The models we will examine will be simple ones illustrating the different approaches, and also reflect Indian conditions. Fixedexchange rates will be employed since the Indian exchange rate has been a fixed one (or one which is pegged to a basket of currencies).4 The relative insignificance of bond and stock markets, and govern­ ment controls on transactions in foreign exchange imply that asset market complications, aside from those concerning money (especially in the monetarist approach), can generally be swept under the rug.5 Finally, we will focus primarily on trade flows; government restric­ tions on international capital flows allow us to take them—at least as a first approximation—to be exogenous. Although we will comment on the relevance of the alternative models for the Indian case, our comments will be brief and illustrative in nature, rather than based on any detailed analysis. Rather than attempting to overcome the well-known problems of empirically 4 An exception is the neoclassical model, which does not require any assumption regarding the exchange rate. As we shall see below, the exchange rate arrangement currently in place may be called a managed floating system, and thus may ideally require some modification of the fixed exchange rate models discussed here to allow for exceptional issues. 5 In the presence of government foreign exchange controls an active black market for foreign exchange has evolved. Biswas and Nandi (1986) use the monetary approach to show that a disequilibrium in the domestic money growth, due to domestic credit creation, will have a spill over effect on the black market for foreign exchange, given fixed exchange rates. The macroeconomic effects of such black markets for the Indian case await careful exploration, although with the increased flexibility of the exchange rate the importance of this issue is probably on the decline.

Open-economy Macroeconomic Themes for India 3 1 choosing between alternative approaches, our purpose here is to understand the alternatives and to discuss their implications.

2.1 The neoclassical approach The neoclassical approach to the open economy is essentially based on the Heckscher-Ohlin-Samuelson (HOS) model, which assumes the clearing of competitive markets for all goods and factors through price variations. In its simplest version, the model considers an eco­ nomy producing two goods with two factors of production, capital and labour, which are given in fixed quantities, with constant-retums production functions allowing smooth factor substitution. Given the production functions and factor supplies, and the assumption that the two goods have different factor intensities, the production possibilities o f the economy can be represented by a convex set bounded by a production possibilities frontier. For any given price ratio between the two goods, profit-maximizing firms operating in perfectly com­ petitive markets will produce on the frontier where the price ratio line is tangential to it, maximizing the value of production at the given prices. In the absence of any distortions, this implies that a small country (which can trade any amount it wants to given the relative price in international markets) will maximize the value of its production at international prices with free trade. On the other hand, if the economy imposes restrictions on trade, it will divert production from the optimal point. If the economy's preferences can be represented by indifference curves, the economy will also achieve its highest welfare level with free trade. If the economy is labour abundant compared to the rest of the world, free trade (by the Stolper-Samuelson theorem) will also imply a higher labour share in income than with restricted trade, which may presumably be viewed as superior from a distribu­ tional point of view. Finally, under free trade a net transfer from abroad, or a net inflow of a factor from abroad will place the economy on a higher indifference curve.6 This simple model has been extended in several ways—to increase the number of commodities and factors, incorporate produced inputs, and introduce intertemporal considerations—to show that the benefits o f free trade and factor inflows occur in more general settings. 6 If the inflow of the factor results in a payment equal to the marginal product, there would be no effect on welfare.

32

A m it a v a K r is h n a D u t t

However, a large literature now exists to show that the propositions do not necessarily hold if certain other modifications are introduced, which modifications may be classified into three categories.7 First, if the economy is a large one (so that international prices depend on the amounts it trades), free trade is not the best policy and factor inflows can be welfare-reducing. Second, if the economy is ‘distorted* by the presence of monopolies, factor-price rigidities, economies of scale, and production externalities it can lose, for example, if it moves from a position of autarky or restricted trade to free trade, or if it experiences an inflow of factors from abroad. The economy may also lose if it experiences an inflow of foreign factors in the presence of trade restrictions. Third, certain dynamic factors, of which we mention two, may also alter the propositions. One, if dynamic learning effects are important in importing sectors, free trade may not be the best policy; the case of the infant industry has, under certain stringent conditions, been viewed as a legitimate exception to the free trade doctrine.8 Two, if free trade—as mentioned above—raises wage income and reduces rental income, and if rental receivers have a higher propensity to save than wage earners, aggregate saving in the economy could be reduced as a result of trade liberalization; if saving determines investment (as in neoclassical models), the growth rate of the economy could be adversely affected. A close variant of the HOS model is the non-traded goods model which departs from the HOS framework by distinguishing between traded and non-traded goods. The prices of the traded goods are determined in world markets, while those of non-traded goods are determined by domestic demand and supply. While the simplest models consider two composite goods (one traded and the othfer non-traded), recent models examine the effects of export expansion in particular sectors by introducing an additional traded good.9 Assuming that labour is the only mobile factor within the three sectors, an expansion of one of the traded sectors (due to technological improvement) causes a movement of labour away from the other 7 These propositions arc too well known in the trade theory literature to require any more than a brief mention here. For a fairly recent and comprehensive textbook survey see Bhagwati and Srinivasan (1983). 8 -See Krueger (1984) for a brief discussion of the stringent conditions. 9 See Corden (1984) for a survey. The models were originally developed to analyse the implications of a booming mineral sector, a phenomenon which has been dubbed the Dutch disease. The model, however, can be applied more generally.

Open-economy Macroeconomic Themes for India 33 traded goods sector, because of the expansion of the traded sector, and because of the expansion of the non-traded sector caused by the expansion of income in the expanding traded goods sector. If the contracting traded goods sector is interpreted as the general industrial sector, an expansion in non-industrial exports implies deindustrial­ ization. The neoclassical approach, in its HOS form, has been used for analysing the implications of trade policies in the Indian case by Bhagwati and Desai (1970), Bhagwati and Srinivasan (1975), and Bhagwati (1978); unsurprisingly, they find that import restrictions have led to losses from allocative inefficiencies. But this overlooks the fact that distortions may be present in the economy (since in the presence of distortions no general claims about welfare losses from trade restrictions can be made). Further, to find the direction and magnitude o f the effects of trade policy changes, it is necessary to go beyond the general theoretical models and use computable general equilibrium models using appropriate parameter values, introducing distortions and non-traded goods when necessary (see Dervis, et at., 1982). Parikh (1986) has developed a neoclassical ten-sector model specifically for India, where all goods are internationally traded. Income generated in production goes to several classes, each having a savings function and a linear expenditure consumption system. Taxes, subsidies, and transfers (including a food-rationing system) interfere with the market system, and aggregate fixed capital forma­ tion as a percentage of GDP is fixed exogenously. Given expected prices—and hence outputs— in any given period, prices vary to clear all markets, and the model is solved sequentially.10The model is used for examining the effects of a variety of parametric changes, including one which removes tariffs and quotas (on which we will report later). It may be argued that the assumptions of the basic neoclassical model—such as perfect competition and the full-employment of all resources including labour, and smooth substitution in production— do serious injustice to an economy such as the Indian one, where unemployment is high and where a variety of inflexibilities (such as those necessitating the importation of key raw materials and machines) exist. Dynamic issues such as savings, investment, and technological change may be of greater importance to LDCs than the economics of static resource allocation emphasized in neoclassical 10 See also Srinivasan (1986) for a description of the model.

34

A m it a v a K r is h n a D u t t

theory. While it is true that distortions, rigidities and dynamic factors can be and have been introduced in neoclassical models, it is legitimate to ask why it is necessary to start with the HOS framework and modify it to introduce these complicating factors. The danger of following this route is that the qualifications to the gains from trade and capital flow propositions can easily be forgotten; likewise the role of 'distortions* can be underplayed by treating them as curiosa which can be corrected with optimal policies, rather than viewing them as intrinsic aspects of the development process.

2.2 Monetarist models The neoclassical approach discussed so far does not explicitly intro­ duce money into the analysis. Monetarist models make neoclassical assumptions regarding the structure of the economy (although the simplest versions take an aggregative view and assume that the economy produces only one good) and examine the role of money in them. Developed by economists at the IMF and the University of Chicago, the primary concern of these models is to explain the balance of payments, which is viewed as a monetary phenomenon. The crux o f the approach can be grasped from the model developed by Polak (1957), which provided the early theoretical underpinnings of the IMF approach. Polak considers the quantity equation Y=VH

( 1)

where the velocity, V, is assumed to be a constant, Y denotes money income, and H the stock of money. Assuming, for simplicity, a money multiplier of unity, we have / / = /? + £>

(2)

where R denotes reserves and D t domestic credit, an exogenous policy variable. Assuming, also for simplicity, that all reserves are foreign exchange reserves and that capital flows are given exogenously, the change in reserves is given by the trade surplus, so that

(3) where the overdot denotes the derivative with respect to time, E denotes exports, Af, imports and F> capital inflows. Exports, are con­ sidered to be exogenously given and imports assumed to be an

Open-economy Macroeconomic Themes for India 35 increasing function of income. Substituting (1) into (2), differentiating with respect to time and substitution of (3) implies M-E-F=D

(4)

(when Y is in equilibrium and hence stationary), establishing the basic monetarist proposition, that the balance of payments deficit is caused by domestic credit creation and that a reduction in domestic credit creation will reduce the deficit. This simple model abstracts from many features that are present in other monetarist models. It does not distinguish between real and monetary variables, while other models do so by assuming that real income is at full employment. It makes the demand for money, (K/V), depend only on the level of income, while others make it also depend on the rate of interest and the expected rate of inflation. It assumes that the supply of money is always equal to the demand for money, as implied by (1), while others allow for excess supply of money and the consequent dishoarding of money which affects the goods markets. A simple model which incorporates this, and which assumes that the domestic price level is determined (through arbitrage) by the interna­ tional price level and the exchange rate, has been developed by Dombusch (1980). This model provides a simple analysis of the effects of devaluation in the monetarist approach: given the foreign price level, a devaluation increases the domestic price level, which reduces the real supply of money, causes an excess demand of money and therefore an excess supply of goods, and this improves the balance of payments. Finally, it makes particularly simple assumptions regarding exports and imports, while other models introduce additional determinants. Simple monetarist models have been applied to India by Talele (1984) and Sohrabuddin (1985), assuming monetary equilibrium as in the Polak model. Talele regresses the current account balance and also the overall balance of payments on changes in several monetary variables; a more complicated specification regresses the balances on gross domestic product, the domestic wholesale price index and an index of foreign prices (taking into account exchange rate changes). Sohrabuddin assumes that the real demand for money is a log-linear function of real income and the interest rate. Least square estimates in the two studies do not show that reserve changes can be explained adequately in terms of the models; but given the simple versions of the monetarist model considered (without taking lags into account,

36

A m it a v a K r is h n a D u tt

for example) this cannot be interpreted as firm evidence against the monetarist approach to the balance of payments for India. A more complex monetarist macroeconometric model has been applied to the Indian case by Sundararajan (1986), which departs from the Polak model in several ways. First, it makes the demand for money depend on real marketed income ($ince marketed income rather than total income— a large part of which is not marketed—is expected to affect the demand for money) and on the expected inflation rate (which is assumed to depend on current and past inflation rates). Second, it allows the real demand for money to deviate from the real supply and analyses hoarding behaviour as a consequence of these deviations. Third, it offers a more complicated treatment of exports and imports. The supply of exports depends on the price received by selling abroad relative to the domestic price level, on the level of output, and on the excess supply of real balances (since consumers buy more goods domestically, reducing export supply); the demand for exports depends on the foreign currency price of Indian exports relative to the price charged by competitors, and the real income of India’s trading partners. The foreign currency price of Indian exports is determined endogenously by equating the supply of, and demand for exports. The specification of imports takes into account govern­ ment restrictions on imports in the form of import licenses. The desired quantity of imports depends positively on the price of imported goods relative to the domestic price, the level of output (mainly for intermediate use), the exogenously given level of essential imports, and the excess supply of real money (which increases the demand for goods in general). The government permits imports taking into account both desired imports and the economy’s foreign ex­ change reserve position; this implies that actual imports, in addition to the arguments of the desired import function, depend on foreign exchange receipts from exports and exogenously given capital in­ flows, and the stock of reserves in the previous period.11 Simulations o f the model show decent fits to actual data, but given the fact that the model has a large number of exogenous variables whose values are taken from actual data, and that the simulations are not compared to those from an alternative model, it is difficult to accept this as overwhelming evidence in favour of the modified monetarist model for India. 11 For a similar, but more detailed, econometric model of Indian imports see Ghose, Lahiri, and Wadhwa (1986).

Open-economy Macroeconomic Themes for India 37 Monetarist models such as those described above arguably suffer from several problems, especially with regard to their application to India. First, they treat output levels as exogenous, so that aggregate demand (and policy affecting it) plays no part in affecting output levels even in the short run. To the extent that the exogeneity of output is due to the neoclassical assumption of full employment growth, this approach has all the possible defects of the neoclassical model. Second, since the models do not distinguish between the behaviour o f different classes in the economy, and in general do not allow the distribution of income to have any role in the model, they can say nothing about the relation between open-economy issues and the distribution of income. Third, their aggregative structure does not take sectoral complications into account. Finally, and more generally, the approach assumes domestic credit creation to be exogenously deter­ mined, not allowing it to respond endogenously to the demand for credit (see Moore, 1988). 2.3 Tw o-gap models The two-gap model focusses on the effect? of foreign capital inflows on savings and foreign exchange constraints.12 To fix ideas, consider a simple model of an economy which produces output with only capital and labour. Assuming that saving is a fraction s of real income Yt and ignoring fiscal policy, l =sY+T where / is real investment and 7, the trade deficit. Dividing through by the stock of capital in the economy, AT, we get g = (s/ak) + 1 where g is the rate of capital accumulation in the economy (assuming away depreciation), a* the technologically fixed capital-output ratio, and t ~ T/K\ we are assuming here that there is enough demand for goods and labour available to allow full capacity production. If F is the capital inflow into the economy net of factor payments (and foreign reserve reductions), this can be written as gs = (s/ak) + f 12 See, for example, Chenery (1961) and McKinnon (1964).

(5)

38

A m it a v a K r is h n a D u t t

where / = F/K, which shows the growth rate which can be achieved by an economy if it is savings constrained for any given level of foreign capital inflow (as a ratio of capital stock); a higher gs results from a higher s and f o r a lower ak. Assume also that a fixed fraction (1 - 6) of investment goods has to be imported and is not produced domestically.13 Denoting exports less imports competing with domestic production by X t the trade balance of the economy is (\ - 0 ) 1 - X = T Dividing by Kf setting x = X/K, and noting that t - f we then get

*/ = [ l / ( l - e ) ] [ / + * ]

(6)

which shows the rate of growth of the economy if it is foreign exchange constrained. If world trading prospects and the structure of the domestic economy fix x and 0, the equation shows that when net inflows increase (due to a higher level of foreign aid, say) the rate of growth o f the economy, as measured by gy, will increase. Note that the equation implies that for a given/, a higher g jcan be achieved by increasing 0 or x (which implies the expansion of exports or the reduction of competitive and investment good imports), and that a devaluation may increase the growth rate by increasing x. Finally, the actual rate of accumulation is determined by g = min [g, g} ]

(7)

If the economy is savings constrained an increase in /w ill increase the growth rate by the same amount, while in the foreign exchange constrained case, since 0 < 1, there will be a multiplied impact due to the fact that domestic saving is available to the economy to supplement the additional foreign savings.14 The model can be used 13 Some models also take into account other imports, such as imports of non-competitive intermediate goods which are usually proportional to output For simplicity we ignore these here, but see below. 14 When one of the constraints is not binding, some mechanisms must be assumed which satisfy the relevant equation ex-post. In this simple model it may be assumed that, when the saving constraint is binding, the economy uses up its extra foreign exchange by increasing competitive exports or reducing export efforts, so that x changes to take up the slack, and when the foreign exchange constraint is binding, s adjusts to take up the slack in the savings equation. Other adjusting mechanisms can also be considered; see Taylor (1983) and Bacha (1984).

Open-economy Macroeconomic Themes for India 39 to analyse the implications of parametric changes on the growth rate, taking care to distinguish between whether the economy is savings or foreign exchange constrained. The two-gap model has been applied to the Indian economy in several ways. First, early applications, such as that of Bergsman and Manne (1965), used it for analysing the foreign aid implications of different strategies of growth involving assumptions of domestic expenditure and import substitution. Second, the model can be used for estimating the growth implications of foreign capital inflows; a recent analysis by Mammen (1984) estimates the implications of an IMF loan, but it assumes that the economy is foreign exchange constrained, never allowing the savings constraint to apply. Third, the approach can be used for ascertaining whether an economy is foreign exchange constrained or savings constrained. Diwan (1968), finding that the regression coefficients o f both capital stock and ‘imports related to output* on national income are significant and the value of the multiple correlation coefficient are high, argues that India is constrained by supplies of both domestic capital and foreign capital; however, this analysis does not establish the direction of causation. A more appropriate procedure would be to explicitly write down appropriate behavioural equations and have a test to determine which of the two constraints is binding (Weisskopf, 1971), or better still, to allow different constraints to be binding in different periods using the method of switching regressions (see Gersovitz, 1982). Finally, the model can be used as the basis of a dynamic simulation model which allows the economy to switch from one type o f constraint to another over time. Bacha (1986) describes such a model where imports consist of not only capital goods but also intermediate goods (although with variable average import coefficients). Taking the economy initially to be foreign exchange constrained (so that the current account deficit is given and the level of exports is given by foreign demand), the model solves for the values of all the variables. Over time foreign capital income grows according to the foreign debt and interest payment, the debt grows according to the current account deficit, and capital stock (and productive capacity) according to investment. Restrictions are imposed on the solution so4hat investment must be positive, actual output cannot exceed capacity, and actual exports cannot exceed potential exports defined as a ratio of capacity. If foreign demand for exports exceeds potential exports, exports are no longer constrained by foreign demand and the economy is constrained

40

A m it a v a K r is h n a D u t t

by the domestic supply of exports. If the capacity constraint is reached the economy is no longer constrained by foreign exchange, but by capacity: exports are then determined by the gap between domestic absorption and capacity output. If investment becomes negative, it is set equal to zero and the savings rate is made to adjust to bring saving and investment to equality. Thus, the behaviour o f the economy can be analysed over time allowing it to switch constraints. Ahluwalia (1986) has applied this model to the Indian case, using parameters consistent with recent observed behaviour, and simulated it to ex­ amine the nature of external constraints on India's economic per­ formance. There are several features of the two-gap approach which, it can be argued, make it relevant for economies such as India. First, it assumes that labour is in unlimited supply, which is appropriate for labour surplus economies. Second, it emphasizes the inflexibilities and rigidities in less developed economies which have received much attention in the informal literature, such as those preventing substitu­ tion between capital and labour and between imported and domestic investment goods, and those preventing increases in net exports. On the other hand, the excessive rigidity of these models has been criticized by Little (1982) from the neoclassical perspective (because it neglects factor substitution in production) and Bagchi (1970b) and Chandra (1973) from the left (because it takes a given pattern of income distribution and spending, which fixes aggregate ratios despite differences in sectoral ratios and the possibility o f technological change). Moreover, it has been argued (see Krueger, 1984) that these inflexibilities were policy induced—caused by the tariff structures and overvalued exchange rate—so that these models shifted attention from these more basic policy problems to the effects of inflexibilities they generated. Findlay (1971) has shown, however, that if imports are allowed for consumption purposes, increases in domestic saving rates will raise the growth rate, but there will be an upper asymptote to how far it can be pushed, and that this condition is unchanged if imported and domestic inputs are substitutable in production and investment. Two-gap models may also be criticized on other grounds. They assume away all other constraints on growth. For example, although it has often been argued that stagnation in India is due to the lack of demand (see below), the models assume that there is enough demand for the economy to achieve full capacity. However, it should be noted

Open-economy Macroeconomic Themes fo r India 41 that the approach leaves open the possibility of introducing additional constraints; a third 'fiscal* constraint has been introduced by Bacha (1989) in which the level of foreign capital inflows limits the size of government investment which can constrain the growth rate of the economy (because of strict complementarity between public and private investment). Bagchi (1970a, 1970b) also criticizes Indian two-gap planning models for overlooking negative effects of foreign capital flows on domestic savings and the rate of growth of agricul­ tural output, and for treating foreign capital as shiftable when aid tying actually makes them non-shiftable. As we shall see below, however, these difficulties can be overcome in two-gap models by allowing the domestic saving rate to depend on foreign capital flows, by assuming a more disaggregated structure, or by calculating the effective amount of aid by subtracting losses due to tying.

2.4 Demand-constrained models While the models discussed so far ignore the problem of demand, the Keynesian approach takes output to be demand-determined. The simple textbook one-sector Keynesian model, modified to take rela­ tive price effects on trade flows into account, assumes that real consumption is a function of real income, that real investment is fixed (determined by the interest rate which is fixed by monetary policy),15 exports and imports depend on the relative prices of domestic and foreign goods, imports depend also on real income, and the domestic price level is fixed.16 Equilibrium is determined when output is equal to the demand for goods, so that Y= C(Y) + / + E (eP*/P) - (eP*/P) M (K, eP*/P)

(8)

where C is the consumption function, E the level of exports, M the level of imports, P* the price of the foreign good in terms of foreign currency, and e the exchange rate, and where other arguments in the functions have been suppressed. The equilibrium trade deficit in terms of the domestic good is given by 15 Note that this implies that domestic credit is endogenous. An alternative, using the IS-LM framework, is to have the money market cleared by interest rate variation with a given money supply. 16 The assumption that the price level is fixed makes this model a textbook Keynesian one rather than a truly Keynesian one.

42

A m it a v a K r is h n a D u tt

T = (ieP*/P) M (K, eP*/P) - E (eP*/P)

(9)

where Y is determined by (8). It is assumed that any level of the trade deficit is permissible, accommodated by foreign capital inflows or reserve changes. The model implies that a devaluation represented by a rise in ef given the Marshall-Lemer condition, will improve the trade balance and increase the level of output of the economy, import restrictions will reduce imports and increase output, and a restrictive credit policy will reduce investment spending and output and, by reducing imports, improve the balance of trade. The relevance of the Keynesian approach to less developed economies such as the Indian one has been questioned. Rao (1952) and Dasgupta (1954) have argued that unemployment in less developed economies such as the Indian one is not so much the result of insufficient effective demand, but of bottlenecks due to capital and land constraints. Without entering into the details of this issue, it may be noted that the Keynesian approach has found more acceptance in Indian analysis (Rakshit, 1982). Demand factors are also central to the Kaleckian approach adopted by Dutt (1984), Patnaik (1988) and Taylor (1983). Drawing on the approach of Kalecki (1971), this approach stresses the different consumption patterns of capitalists and workers (thereby introducing income distribution into the analysis), and has an explicit theory of prices in which firms set the product price by marking up prime costs in an oligopolistic environment which appears to be relevant for the Indian manufacturing sector.17 The open-economy Kaleckian models also depart from the Keynesian one described above by allowing for imports of intermediate and capital goods, thereby incorporating some characteristics of the two-gap approach. To examine the main features o f this approach we examine a simple model which borrows from the Indian models of Dutt (1984), Taylor (1983) and Taylor et al., (1984), as well as from the theoretical discussion in Taylor (1989). The economy produces one good with labour, an imported inter­ mediate good and capital, using a fixed-coefficients production function. Firms operate with an excess capacity of capital in an oligo­ polistic environment, and set the product price as a mark-up on prime costs so that p = (wat + eP *a) ( 1+2)

(10)

17 SeeChatterji (1989), for instance, although (his study assumes a closed economy framework.

Open-economy Macroeconomic Themes for India 43 where w is the money wage, a/ the fixed labour-output ratio, Pi* the price of an intermediate good in terms of foreign currency, a%the intermediate good-output ratio, and z the fixed mark-up rate deter­ mined by the degree of monopoly power. We assume, for simplicity, that the money wage is fixed.18 Workers (who earn wage income) spend their entire income on the domestic good, while capitalists (who earn mark-up income) save a fraction sc of their income, and spend a fraction, m, of their consumption expenditure on a luxury imported good, and the rest on the domestic good.19 The consumption demand for the domestic good is PC = wa,Y + (1 - sc) (1 - m) z (wat + eP*a$ Y

(11)

where C is real consumption and Y is real income and output. A fraction 6 of investment goods is domestically produced, the rest being imported.20 The rate of investment by firms depends positively on the rate of capacity utilization, since a higher degree of excess capacity is a deterrent on investment.21 Adopting a linear form we have //K =a + p«

(12)

where / is the level of real investment, K the stock of capital, and a and p are positive constants. Finally, exports depend positively on relative prices and the domestic capital stock, which we write in the form E = E(eP*/P) K

(13)

19 Changes in w as well as z could explain inflation; see Taylor (1983, 1989). Given the relative unimportance of inflation in India—compared to many Latin American economies—we have chosen to abstract from inflation in our discussion of this model. 19 While less developed economies such as India have reduced their imports of luxury imports to negligible levels, direct spending on luxury imports captures in a simple way the higher import intensity of domestic goods consumed by upper-income groups. See Hazari (1967). 20 The fixed ratios, o,- and 0, are consistent with econometric studies. See for example, Ghose, et al. (1986). 21 This assumption is discussed in more detail in Dutt (1984). The rate of invest­ ment can also be assumed to depend on the rate of profit and the rate of interest. The former can be shown to be positively related to the rate of capacity utilization (see Dutt, 1984), and the latter has been assumed to be constant, determined by Central Bank policy.

44

A m it a v a K r is h n a D utt

where P* is the price of foreign goods, E the level of imports in real terms, and E > 0. We abstract from government spending and taxation. Assuming that producers increase output in response to excess demand, equilibrium in the economy implies Y=C+QI + E Using (10) through (13), this implies that the equilibrium level of capacity utilization, u = Y/K is given by 0 a + E(eP*/P)

U= {[ i c +

m (l

- *c)] [ z /(l+ z )] + b, - 0p

J

(14)

where bi = ePi*a/P, the value share of intermediate inputs in unit price. The balance of trade deficit as a ratio of capital stock, is given by t =[ z / ( 1 +z)]m(l -j)w + V + 0 -Q )(a+ $u)-E (eP */P )

(15)

the equilibrium value of which can be solved by substituting the equilibrium value of u from (14). We assume that if f > 0 the trade deficit is financed by foreign capital inflows. This model can be used to examine the implications of several parametric changes. A rise in e (a devaluation) will increase the pnv duce price and, given the money wage, reduce the real wage, causing a shift in income distribution away from workers with a high con­ sumption propensity. Unless the devaluation increases exports suffi­ ciently, aggregate demand, and hence the level of capacity utilization, will fall, implying that devaluation is contractionary, the rate of accumulation will fall as well. The contraction in capacity utilization will reduce imports and this, with the increase in exports, will improve the balance of trade. A credit squeeze, represented by a rise in the interest rate which can be taken to reduce the investment parameter a will reduce u and improve the balance of trade as well. If the credit squeeze also causes a rise in the mark-up as firms try to cover higher 22 The precise condition for du/de > 0 is E ' > P *a/P . A variety of models reproduce this structuralist result without introducing all of the rigid features of this model. See, for example, Islam (1984) who emphasizes the existence of an inter­ mediate import, and Buffie (1986) who stresses the role of a capital good which requires domestic and imported components in fixed proportions, in an otherwise neoclassical model.

Open-economy Macroeconomic Themes for India 45 borrowing costs with a higher price, there will be a redistribution away from workers as the real wage falls as discussed in the case of the devaluation, worsening the income distribution, and if bi is not too high this will aggravate the contraction. Although this increase in the mark-up may reduce exports by reducing the competitiveness o f domestic goods, unless the export response in very strong it is likely that the credit squeeze will still improve the balance of trade. The implication is that a devaluation and a credit squeeze may improve the economy's trade and payments balance, but it may do so often by reducing its rate of growth and by worsening its distribution of income. While the Kaleckian approach takes into account some important characteristics of less developed economies, these models ignore many important features of the Indian economy. First, by assuming mark-up prices or fixed prices and demand-determined output, they appear to ignore the large agricultural sector of the economy where production is arguably resource constrained. Second, they assume away constraints to growth other than demand, namely capacity, infrastructural, and foreign exchange constraints, all of which may, at least at certain times, constrain growth. 2.5 Structuralist models The Kaleckian approach takes into account some important structural characteristics of less developed economies, including classes, oligo­ polistic markets, and imported intermediate and capital goods. More complete structuralist models can be seen as modifying and extending this approach by considering alternative macroeconomic factors which may constrain the economy, by introducing additional sectors, and by incorporating inflation dynamics. With inflation being rela­ tively less important for India,23 we examine a few examples illustrat­ ing the first two types of modifications. While the Kaleckian model considered above is demand con­ strained, alternative structuralist models examine other constraints, 23 The interested reader is referred to Taylor (1983,1989). In the one-sector models with mark-up pricing, inflation can be taken to be the result of changes in the money wage and the mark-up rate due to conflicting claims over income. Changes in exchange rates and imported intermediate good prices can also contribute. Perhaps more relevant for India, excess demand for agricultural goods which raise agricultural prices, can also contribute.

46

A m it a v a K r is h n a D u t t

such as capacity constraints and foreign exchange constraints. If demand (at the fixed mark-up) is high enough to take the economy to full capacity utilization, so that u = \/ak, where ak is the technologically-given capital-output ratio, the economy becomes capacity con­ strained. In this case, since capacity utilization becomes rigid up­ wards, we need some other adjusting variable to achieve equilibrium in the goods market. If the price (hence the mark-up rate) is the adjusting variable, a rise in domestic demand will have the effect of increasing the product price and reducing demand by squeezing labour income (with the given money wage) and exports (which fall as the competitiveness of Indian goods fall in world markets). If the rate of accumulation is the adjusting variable (perhaps due to crowding out through changes in the interest rate), investment is determined by saving. With the fixed mark-up, the rate of accumulation will be given by g = (1 / 9 ) { Is + m (1 - s)} [z/( 1 + z)] u + (b / a k) - E (eP*/P) } (16)

It may be noted that an exogenous change in the export function which increases exports will have to squeeze domestic demand by reducing the real income of workers in the first case, and reducing the rate of accumulation in the second case. Foreign exchange constraints can be introduced by dropping our assumption of the Kaleckian model that any trade deficit could be covered by capital flows and, assuming instead, as in the foreign exchange constrained regime of the two-gap model, that the capital flow as a ratio of domestic capital is given. In this case the model requires a new adjustment variable; possible candidates are the rate of investment (under the assumption that government expenditure is cut when the economy experiences foreign exchange difficulties, see Taylor, 1983) or the level of capacity utilization (see Bacha, 1984). Bagchi (1977) developed an early model of this type for India which assumed that the trade surplus is fixed by the government to build up foreign exchange reserves, so that it is a policy-induced foreign exchange constrained model. However, the model does- not have consumption, investment, and commodity balance equations, and does not specify what adjusting variable clears the goods market. Exports are assumed to be a positive function of property income (no explanation is given), imports a positive function of wage income and property income (with a higher propensity to import out of the latter), property income is assumed to be a positive function of income (which

Open-economy Macroeconomic Themes for India 47 is an income distribution equation), and an identity stating that wage and property income sum to national income closes the model. Bagchi shows that under appropriate parametric restrictions, a rise in the property share will increase exports and property income, but reduce income and wage income. One feature of this model, that the level of imports depends on the level of exports (implied in the fixed trade deficit assumption) is consistent with empirical work (Ghose et al., 1986), and should find its way into more macroeconomic models; the propensity to import out of export revenues can be taken to be less than one. Structuralist models also introduce additional sectors into the model. Perhaps most relevant for the Indian case, Taylor (1983) discusses a two-sector model with a fix-price industrial sector similar to the Kaleckian one described above, and a flex-price agricultural sector with fixed output in the short run and a market-clearing price. The implications of an expansion in exports in this model depend on which sector increases its exports: while an increase in exports for the industrial sector increases demand and capacity utilization in that sector, an expansion in agricultural exports reduces domestic supply and raises the price, and in the presence of strong Engel effects can squeeze industrial demand sufficiently to reduce non-agricultural output. On the basis of this fix-price flex-price structure Taylor, et al. (1984) construct a model for India with two flex-price agricultural sectors and three fix-price non-agricultural sectors (industry, services, infrastructure) with excess capacity. A monetary side is also appended to this real model to examine the interaction between real and financial variables. The interest rate affects marked-up prices by entering into prime costs, and negatively affects investment spending. An LM equation completes the model, where the change in money supply is determined by the fiscal cash deficit, the balance of payments surplus, and net domestic credit to the private sector. An alternative sectoral division stresses the traded-nontraded distinction. Sen's (1977) model distinguishes between a home goods sector producing non-traded wage goods, and an export sector which exports the major part of its output. Investment goods in the export sector have to be imported, while the home goods sector uses its own output for investment. Income in the two sectors is distributed between workers (who do not save and consume the home good) and capitalists (who save a fraction of their income and consume the export good). The level of exports depends on the relative price of

48

A m it a v a K r is h n a D u t t

the exported good. The price ratio between the two goods produced in the economy is assumed to be fixed by the government, and the latter is also able to transfer resources from the home to the export sector. The model implies that a rise in exports as a result of a state-sponsored expansion of exports is not always favourable: in­ creased exports allow a higher rate of investment in the exports sector since higher exports allow higher imports, with balanced trade (or a balance of payments constraint). However, increased export produc­ tion increases the wage bill in that sector, and the demand for home goods. With full capacity production in the home goods sector the investment in that sector falls as the government transfers the good to the export sector for the higher consumption need. The effect on aggregate investment is ambiguous, but depends on the elasticity of export demand and the surplus production per unit of export produc­ tion. While the results are interesting, and the model quite relevant for India, the model is not clearly developed and its assumptions not explicitly stated. Taylor (1983) discusses yet another model with two export goods and a non-traded good, similar to the neoclassical three-sector model. In the model, one good, a mineral, is entirely exported and produced at full capacity; for the Indian case this may be interpreted as exports with low domestic consumption and produced under supply con­ straints. A flex-price non-traded intermediate-cum-consumption good (infrastructure) is assumed to be in fixed supply (its price varies to clear the market), and a manufactured good, which is consumed domestically and also exported, is assumed to be produced under conditions of excess capacity with a fixed mark-up price. The model shows that an increase in mineral exports may reduce the level of production of the manufactured good by drawing away infrastructure from it; the possibility of this increases if the non-traded good is not important as a consumer good. The increase in the price of infrastruc­ ture also raises the price of the manufactured good, reducing exports which are price sensitive. Thus, greater 'mineral* export may well imply a reduced diversification in both production structure and exports.

2.6 North-South models It is sometimes argued that the structure and performance of an economy can only be understood if it is seen as a part of the world

Open-economy Macroeconomic Themes fo r India 49 capitalist system. Even without taking such an extreme view, it is possible to obtain some insight into the growth constraints and pos­ sibilities of an economy using a global perspective. It is in this context that models of North-South trade may be relevant for Indian openeconomy macroeconomics.24 Two types o f dynamic North-South models are relevant for us: those emphasizing structural asymmetries between the North and South assuming given technology and patterns o f specialization, and those showing cumulative causation in which technological change plays a central role. In the first type the pioneering contributions are those of Findlay (1980) and Taylor (1981, 1983). A common framework with which to examine these and other models is provided by Dutt (1990) where it is assumed that there are two regions in the world economy, the North and the South, each producing and exporting a single good. The Southern good is a pure consumption good and the Northern good is a consumption and investment good. Each good is produced with a fixed-coefficient production function using only capital and labour as inputs. Income goes to workers and capitalists in each region; workers everywhere consume all their income, and capitalists save a fraction of their income; in the North, consumption expenditure is distributed between Northern and Southern goods in a given proportion; in the South, workers only consume the Southern good and capitalists distribute their consumption between the two goods in a certain proportion. Factor mobility is assumed away in the basic model and balanced trade is assumed. This framework provides the basis for analysing the interaction between growth rates in the two regions, their relative stocks of capital, and the terms of trade, but is not sufficiently specified to solve for the values of the relevant variables. Findlay's (1980) model can be thought of as one in which the North grows under conditions of full employment and full capacity growth as in neoclassical growth models with an exogenously given rate of growth of labour supply, while the South has a fixed real wage and full capacity utilization. Taylor's (1981, 1983) model can be interpreted as having a similar South, but with a North which grows along Kaleckian lines with excess capacity, mark-up pricing and a desired accumulation function.25 A variety of other models can also 24 See Dutt (1988b) for a survey of the literature on North-South models. 25 The Findlay and Taylor models are not exactly as described here. Findlay, for example, allows factor substitution in production, and Taylor introduces a third region into the model.

50

A m it a v a K r is h n a D u tt

be developed, including those which allow imperfect competition in the South and hence the presence of excess capacity (see Dutt, 1990). These models may be used to understand the functioning of the world economy, and the constraints facing Southern economies, in various ways. First, the properties of the models at long-run equi­ librium in which both regions grow at the same rate have been analysed to show that the Southern growth rate is determined by the Northern growth rate which is determined internally in the North either by the growth of labour supply (as in Findlay’s approach) or by investment and saving parameters (as in Taylor’s approach). This is because the South is dependent on the North for its investment goods and can only import to the extent that the North will import the Southern good. Second, the parameters of the models can be endogenized to examine the effects of various dynamic mechanisms which could result in uneven development on a world scale by which the North forges ahead relative to the South. For example, Dutt (1988) analyses the role o f the inelasticity of Northern demand for Southern goods and international demonstration effects:26 as the North grows relative to the South it spends a large share of its consumption expenditure on Northern goods (since Southern goods are inferior), which reduces the export potential of the South; international demonstration effects then make the South consume more of the Northern goods, reducing both its saving and spending on the Southern goods. If these effects are stronger than the effect by which the expansion of the North expands the demand for Southern goods (in the absence of these effects), the South will fall relatively further behind. A crucial feature of these models is the assumption that all investment goods are produced in the North; actually the assumption can be modified to allow the South to produce some of its investment goods, assuming that an inelastically given core must still be imported. These results reveal the obvious importance of developing a capitalgoods base in Southern economies. The second class of models examines the implications of tech­ nological change. This approach follows Kaldor (1970) in arguing that a region which expands its exports and output more rapidly 26 Bagchi (1970b) and, more recently, Chakravarty (1984) stress the role of the international demonstration effect in the Indian context. Despite the banning of con­ sumer goods imports, international demonstration effects can work by increasing the imports of intermediate goods entering into domestically produced consumer goods.

Open-economy Macroeconomic Themes for India 51 increases its productivity faster (through the so-called Verdoom effects) and by becoming more competitive, can expand its exports even more rapidly. This implies that international development will be uneven, with richer regions growing faster than poorer regions. This approach raises two important issues. First, to what extent is growth export-led? While it is certainly possible for the spurt in growth to be the result of export growth, it is also possible that growth initially may be the result of an expansion in domestic markets which allows producers to reap significant scale economies to be able to increase exports. Second, to what extent are the productivity gains sector-specific? It has been argued that dynamic scale economies and learning effects are most important in manufacturing, so that economies which increase their production and export of manufac­ tured goods rather than agricultural goods will be able to forge ahead. Krugman (1981) has developed a model in which the North, by exporting manufactured goods which experience increasing returns to scale, is able to increase its profits and hence accumulation faster, leaving the South, exporting agricultural goods, behind. Dutt (1986) formalizes the argument that high-processing sectors such as manufacturing experience greater learning effects which increase productivity not only within the sectors, but lead to greater spin-off effects raising productivity all over the economy. The model shows that free trade may cause uneven development with the North accumulating capital more rapidly as the South stagnates and that, by protecting its manufacturing sector, the South can stem these forces o f uneven development. Uneven development in these models arise despite the assumption of the full employment of all resources, purely due to the dynamics of technological change. Since North-South models usually consider only two regions, it is not immediately clear to what extent they can illuminate the problems o f individual economies such as the Indian one. A higher level of disaggregation is necessary to make these models more useful for this purpose. An obvious modification o f the first type of models is to allow for a third region, the NICs. This could help to examine the effects o f the growth of the NICs for economies such as India which have not been able to develop their exports at such high rates. A similar modification of the second type of model could explain why the greater technological dynamism of the NICs has helped them to increase exports, and why this has made it more difficult for other less developed economies to increase their exports.

52

A m it a v a K r is h n a D u tt

3. FUNDAMENTAL ISSUES IN INDIAN OPENECONOMY MACROECONOMICS This section examines open-economy issues in macroeconomic con­ straints on growth, import substitution and export promotion as trade policies, stabilization policies for balance of payments adjustment, and the effects of inflows of foreign capital. The topics considered have different dimensions; our focus will be on their macroeconomic aspects (in our sense) which are of analytical interest.

3.1 Open-economy issues and constraints on growth The literature on the constraints to Indian economic growth has stressed open-economy issues only sporadically, most of the discus­ sion evolving around the internal constraints. We first discuss some open-economy considerations related to the internal constraints, and then discuss the foreign exchange constraint. (i) Open-economy issues and internal constraints In the early years of planning the constraints emphasized most were savings and the size of the capital goods sector. The savings rate was sought to be raised by government efforts at resource mobilization, and this was supplemented by foreign saving in the form of aid. Two-gap models were used to find the aid requirements as discussed above (generally assuming binding saving gaps). It has been argued that such capital inflows were not an unmixed blessing, a theme we postpone to section 4.4. The problem of a small capital goods sector was sought to be solved by the Mahalanobis (1953) strategy o f allocating a high share o f investment to the capital goods sector, based on a two-sector, non-shiftable capital model similar to the Feldman model used in Soviet planning popularized by Domar. This approach, however, has been criticized for ignoring open-economy issues (Bhagwati and Desai, 1970). First, it has been argued that in an open-economy there is no need for India to develop a capital goods sector since capital goods can be imported; indeed it would be uneconomic to do so since comparative advantage would dictate the production and export of other labourintensive goods in return for which capital goods could be imported. Harris (1972) has argued, however, using a three-sector extension of

Open-economy Macroeconomic Themes for India 53 the Feldman-Mahalanobis-Domar model, that if export growth is limited by slowly-growing foreign markets, the share of investment allocated to the capital goods sector would still determine the long-run rate of growth of the economy. As long as exports did not grow rapidly (a realistic assumption in the times, whatever the cause), the Mahalanobis strategy was still valid. Moreover, as Chakravarty (1987) has argued, the static comparative advantage argument over­ looks that fact that the creation of a capital goods base speeds up the rate of technological change in the economy through learning and spill over effects (which were referred to in connection with the North-South models above). Second, it has been argued that capital goods production was highly import intensive, and that this could create balance of payments crises for the economy. However, this overlooks the fact that the strategy reduces capital goods imports in the future, which has been seen above to be of great importance for less developed economies. The stagnation of the Indian economy after the mid-sixties produced a large literature on the constraints on Indian growth, stressing the possible role of several other constraints. The arguments were as follows. Agriculture constrained Indian industry from the supply side by restricting the supply of wage and intermediate goods and from the demand side due to low agricultural income (Chakravar­ ty, 1974, 1979; Raj, 1976; Mitra, 1977). A high level of inequality o f income created a demand problem by restricting the size of the industrial market (Bagchi, 1970; Chakravarty, 1979; Nayyar, 1978). Claims by different pressure groups on government resources (Bardhan, 1984) created a fiscal crisis of the state, reducing govern­ ment spending and constraining the economy from the demand side as well as the supply side because of the inadequate growth of infrastructure (Ahluwalia, 1985; Patnaik, 1987). Finally, industrial and trade policies pursued by the government resulted in widespread inefficiencies in the economy which restricted growth (Bhagwati and Desai, 1970; Bhagwati and Srinivasan, 1975). While these are all internal constraints, some of them make implicit open-economy assumptions. Ahluwalia (1985) points out that the agricultural-supply constraint argument presupposes a foreign ex­ change constraint which prevents the import of food. Further, all the demand-based arguments implicitly assume that the small country assumption in trade is not satisfied, since otherwise there would be unlimited foreign markets for Indian products. Estimates by Lucas

54

A m it a v a K r is h n a D u t t

(1988) for Indian manufactured exports, however, support this as­ sumption . While we shall return to the fiscal and policy constraints in connection with trade policy, we may here discuss at greater length the inequality constraint which is affected in several ways by openeconomy complications. A theoretical interpretation of the constraint is providedin Dutt (1984) where it is shown that in a closed-economy Kaleckian model a rise in the mark-up rate redistributes income from workers to capitalists (who have a higher propensity to save), reduces aggregate demand, investment incentives, and growth. An extension o f the model to the case of the open economy (Dutt, 1984) reveals several complications. First, if the rise in the mark-up rate raises the price of Indian exports, exports will fall given some price elasticity, exacerbating the demand contraction. Second, if high-income groups have a higher import component in consumption (as is suggested by Hazari, 1967), this redistribution will further reduce domestic demand by switching demand for foreign goods.27 Third, the presence of intermediate imports may weaken or even negate the inverse relation between the mark-up rate and the growth rate. A sufficiently high share o f imported intermediates in total costs, in relation to the price elasticity o f exports, implies that a higher mark-up will reduce foreign saving as imports amount to a lower share in income flows, which raises demand and hence the growth rate; plausible numbers, however, still imply an inverse relationship. Finally, it should be noted that the model abstracts from foreign exchange constraints; since an improved income distribution is expansionary, it is likely that this will imply greater import needs which would create foreign exchange problems for the economy in the absence of borrowing and the growth of exports. (ii) The foreign exchange constraint If the government is worried about the economy’s external indebted­ ness it will use its policy tools to keep the level of external borrowing within bounds. This implies that the economy will appear to be constrained in some other way (by the lack of demand, for example) 27 Estimates by Mohammad (1981) based on a closed Leontief input-output model with several income classes does not find a strong import-reducing effect of an income distributional improvement, but because it ignores capacity constraints, it exaggerates the expansionary effects of the income distributional change, and hence the expan­ sionary effects on imports.

Open-economy Macroeconomic Themes for India 55 even if it is actually (through its effects on policy) constrained by foreign exchange; it is only in times of extreme pressures on the external payments—usually unanticipated—that one hears much of foreign exchange constraints. Recent formal contributions on con­ straints on growth which incorporate the foreign exchange constraint explicitly include Sen (1981), Ahluwalia (1986) and Taylor (1988). Sen (1981) examines the role of a foreign exchange constraint in the Indian economy with a multi-sector structuralist model to deter­ mine terminal year equilibrium. The model is based on an inputoutput system with structuralist fixprice-flexprice assumptions, and money wages in each sector are assumed to increase with the cost of living. Each income class has its spending determined by a linear expenditure system, and imports are divided into competitive and non-competitive sections, the latter bearing a constant relation to output in each sector; exports, investment, and competitive exports are determined exogenously. The model allows for savings, foreign exchange, and agricultural constraints on growth, and calculates the rates o f growth determined by each of these constraints. For example, the foreign exchange constrained equilibrium is obtained \>y setting the non-competitive imports to zero and determining the rate of growth given exogenously given capital inflows. The model, using Indian parameter values. Shows that the savings constraint is dominated by the foreign exchange constraint, which is dominated in turn by the agricultural constraint; the actual growth rate in the economy is even lower than the agricultural one. implying a consid­ erable resource slack in the economy. This approach only determines the constraint for the terminal year and is not able to allow different constraints to bind at different times. A model which takes dynamics explicitly into consideration, but which abstracts from the inter-sectoral complications discussed by Sen, and especially the agricultural constraint, has been used by Ahluwalia (1986). This model, as mentioned earlier, assumes that the foreign exchange constraint is binding initially (so that the current account deficit is fixed and the level of exports is determined by external demand), but later allows the economy to determine exports by the pressure of domestic demand (if it produces at full capacity) or by the supply of exports (as a ratio of potential output). The simulation (done from 1983-4 to 1990-1 using parameters consistent with recent his­ tory) assumes that real external demand for exports grows at 5 per cent per year, and that the nominal current account deficit also expands at

56

A m it a v a K r is h n a D u t t

10 per cent nominally. It also makes some assumptions about conces­ sional aid flows and foreign interest rate payments. It implies that the economy experiences a severe foreign exchange constraint, reflected by a falling degree of capacity utilization; there is also a marked deterioration in the debt-service ratio. A higher level of availability of foreign exchange would have allowed higher levels of output and capital accumulation; thus a more rapid growth in exports is beneficial for the foreign exchange constrained economy. Taylor’s (1988) analysis of the macroeconomic constraints to growth in India incorporates the role of the foreign deficit and the external debt trap, in addition to constraints due to saving and productivity, agriculture, inflation and fiscal problems. Richer than both the Sen and Ahluwalia exercises, it is less formal because it uses accounting equations with plausible parameter values and not a fully specified model. Arguing that the economy is demand-constrained and not supply-constrained, Taylor show that the foreign exchange gap equation implies that a 1 per cent increase in capacity growth— because of higher capital and intermediate good imports—would raise the current account deficit by about 0.5 per cent, or raise the export growth rate to almost 14 percent. While these changes are non-trivial, they are within the range of the possible, and can be helped by sensible policies of import substitution. However, they can be hurt by indis­ criminate liberalization. Since foreign exchange problems do appear sometimes, but not always, it would be useful m future work to analyse them in a way which allows the economy to switch between the important con­ straints. One method is to use simulation techniques like that used by Ahluwalia (1986), incorporating constraints like agriculture and demand. An alternative method is to use the technique of switching regressions extending the method used by Dutta (1988), who only allows for a wage goods constraint imposed by agriculture, and a demand constraint.

3.2 Trade policies The debates on trade policy in India have mirrored the general debate in development economics on import substitution versus export promotion. Although much of the discussion has taken the form of the relative merits of the two strategies, we examine the macroeconomic aspects of the two separately.

Open-economy Macroeconomic Themes for India 57 (i) Import substitution The debates on the import substitution strategy in India have been influenced by actual changes in import policy. The debates during the protectionist phase—up to the mid-sixties—and those during the period o f liberalization— which has gathered strength in recent years have brought up similar analytical issues, but it is useful to discuss the two phases of the debate separately. During the earlier protective phase, the government sought to develop the industrial sector; imports of a variety of items was banned when they were domestically produced (regardless of cost), and those o f other items was severely restricted with quotas, high tariffs, and foreign exchange controls. These were supplemented by an elaborate system of industrial controls. Several arguments have been made to show how this system created various types of inefficiencies and slowed down the rate of growth of the economy (Bhagwati and Desai, 1970; Bhagwati and Srinivasan, 1975; Ahluwalia, 1985; see also Bhagwati, 1978). First, protection caused inter-sectoral inefficiencies due to the diversion of resources away from industries in which the economy has a comparative ad­ vantage. This can be seen as an application of the neoclassical approach in its simplest form, ignoring distortions and dynamic factors; since more complicated neoclassical theories do not imply that free trade is optimal, this argument is a weak one. Second, a more complicated version of this argument stated that protection was excessive, hap­ hazard, and devoid of economic rationale. Calculations of effective rates of protection and domestic resource costs showed the height, and wide range of levels of protection. The merit of this argument is very difficult to judge because of our limited knowledge of actual distortions and dynamic factors in the economy. The empirical calculations (Bhag­ wati and Desai, 1970) have also been challenged on methodological grounds (Bagchi, 1970b; Chandra, 1986). Other measures, comparing actual import and export prices (Nambiar, 1983, for example), show that the extent of protection is low and falling. Moreover, this is at best a critique of the type of protectionism pursued rather than of the strategy of import-substituting industrialization as well, unless an ap­ peal is made to the view that government, intervention is necessarily haphazard. Third, the system* promoted rent-seeking behaviour which resulted in the unproductive use of resources and adversely affected entrepreneurship. While the first effect creates problems in a resource-

58

A m it a v a K r is h n a D u tt

constrained neoclassical economy, the second can be problematic in the context of alternative visions of growth, including a Schumpeterian one. Fourth, heavy protection and other controls resulted in X-inefficiencies and other types of inefficiencies and waste within firms, and also sapped the incentives for technological change. While there is much evidence on this, Chandra (1986) cites several World Bank reports to show that the inefficiency of Indian firms, especially those with sufficient manufacturing experience, has been considerably ex­ aggerated. Fifth, protection resulted in a higher degree of monopoly power in the economy; this not only created distortions due to product market imperfections consistent with the neoclassical approach, but by raising price cost margins (see Katrak, 1980) had adverse aggregate demand effects according to the Kaleckian approach (Dutt, 1984). Sixth, protectionism, despite the special concessions allowed to ex­ porters, created export disincentives. According to the standard neoclassical approach protection to import-competing industries diverts resources away from exporting industries. Even in other ap­ proaches, for example in the Kaleckian model, tariffs raise imported intermediates prices and hence export prices. The non-availability of and delays in obtaining materials and parts has also created supply bottlenecks in export sectors (Wolf, 1982). The stagnation in exports can slow down growth in a variety of situations, including demandconstrained and foreign exchange constrained regimes. Seventh, protection increased the domestic profitability of a variety of goods with high import requirements; while this may have been true, it is likely that, in the long run, the cumulative direct savings of foreign exchange are likely to exceed the cumulative indirect expenses of foreign exchange. Finally, import substitution developed capital-intensive industries, thereby reducing labour absorption. This has usually been argued using closed Leontief demand-constrained models where there are no resource constraints (Mohammad, 1981) which do not make clear why the development of capital-intensive sectors preclude the development of other sectors. Savings-constrained economies with unemployed labour would have lower rates of labour absorption if they have a higher capital intensity; neoclassical models would imply a fall in the real wage with the protection of capital-intensive goods, but would have no problems of labour absorption. Several spirited defences of the strategy have also been made. First, it has been argued that import substitution has made a considerable contribution to the expansion o f output in India (Ahmad, 1968),

Open-economy Macroeconomic Themes for India 59 especially that of capital goods.28 This is especially important for demand-constrained economies since such expansion does not simply imply the reallocation of resources from other sectors as it would in the neoclassical approach. Second, Singh and Ghosh (1988) argue that, as a result of the policy, India is an exporter of many types of capital goods; this has relaxed the foreign exchange constraint for the economy. Third, it has resulted in a tremendous reduction in import coefficients in most manufacturing industries, especially of capital goods; this has eased the balance of payments constraints as well. Fourth, the policy has caused the economy to get technologically sophisticated (Lall, 1984), and India has emerged as a leading exporter o f technology among the NICs (see Singh and Ghosh, 1988); the learning and spin-off effects from capital goods production are obvious. Finally, it is argued (Singh and Ghosh, 1988) (hat India's reliance on domestic capital goods has insulated her to a large degree from foreign shocks (such the first oil shock). Here, however, the implications are not so clear-cut, as countries with already low levels o f imports have less flexibility regarding their imports, and hence less ability to adjust to the shocks (Srinivasan, 1988). As a result of the more recent liberalization measures, a large number of capital goods are allowed to be imported more freely, and import tariff rates have been considerably lowered, but the imports o f consumer goods (except for some essential items) are mostly banned. The critics of the earlier policy welcome these developments, though some argue that the change is too slow (Wadhwa, 1988b). A simulation exercise using the computable neoclassical model dis­ cussed above (Parikh, 1986) found that the removal of all trade barriers has a negligible impact on GDP (only a 0.3 per cent increase), not a very comforting finding for the critics. Srinivasan (1986), however, argues that the gains found by this model are negligible because they ignore issues such as rent seeking, monopoly creation and scale economies; it is not clear, however, that incorporating scale economies would increase the gains from liberalization. The critics of liberalization point out, first, that it has resulted in the growth of import coefficients (both of capital good imports as a ratio of total investment, and other goods as a ratio of GDP) and this 28 Indeed, it has sometimes been argued that industrial stagnation in the mid-sixties was related to the slowdown of import substitution where import substitution is measured by the decline in import-availability ratios. See Ahluwalia (1985) for a discussion.

60

A m it a v a K r is h n a D u t t

has had adverse balance of payments implications (Ahluwalia, 1986; Patnaik, 1987; Singh and Ghosh, 1988). It has been argued (Ray, 1987) that these effects are temporary and would be reversed as soon as liberalization improves efficiency and hence growth in exports; even granting this, there is the question of how long this process would take and how, if the economy faced foreign exchange con­ straints in the meanwhile, the payments deficit would be covered without the debt situation becoming explosive (Singh and Ghosh, 1988). Second, the domestic production of capital goods has been adversely affected by foreign competition (Singh and Ghosh, 1988). While this may be argued to be, simply, the reallocation of factors to more efficient uses according to the neoclassical model, even accord­ ing to the neoclassical approach with fully-employed resources, if capital goods exhibit greater economies of scale than other goods (as one would expect), losses from this type of reallocation are likely (Ramana and Bhattachaijea, 1985). Moreover, if the economy is not neoclassical but demand-constrained, there would be a net reduction in output, not only as a result of a rise in imports, but also because of a reduction in aggregate demand due to a fall in quota rents (Ocampo, 1987; Taylor, 1989). Third, if technological change is primarily caused by learning by doing, the fall in output will cause technological stagnation (Singh and Ghosh, 1988), contradicting the long-run productivity-raising neoclassical argument. A demand-constrained view is not necessary for this argument; even with a neoclas­ sical view, if capital goods are the harbingers of technological change, the fall in output of these sectors will cause problems on this score. Four'll, liberalization, by causing contraction in the capital goods sector reduces capacity utilization and profitability in these sectors; this reduces profits from public sector enterprises and has an adverse effect on government revenues (Ramana and Bhattacharjea, 1985), not only slowing down repairs and worsening efficiency in these enterprises, but also aggravating the fiscal constraint on growth. Final­ ly, Nambiar and Mehta (1987) find that the relative foreign price of a variety of industrial imports is inversely related to the level of import tariffs. They attribute this to imperfections in the world markets for these goods and argue that trade liberalization will imply a terms o f trade loss for the Indian economy and not reduce domestic prices substantially.

Open-economy Macroeconomic Themes for India 61 (ii) Export promotion Here we examine three issues: the relationship between import liberalization and export promotion, the macroeconomic effects of export growth, and the prospects for raising Indian exports. A large literature has developed attempting to show that import substitution and export promotion are competing strategies, and that import liberalizers and export promoters have been the successful performers, citing evidence from different countries (Bhagwati, 1978; Krueger, 1978). It has been argued that the main reason for this is that export promoters have moved from the system of incentives favouring import-competing industries to one which is closer to the free trade optimum. This reasoning is explicitly based on the neoclas­ sical HOS approach (either in its crude form or more sophisticated forms), or uses one of the arguments against import substitution given in the previous section. We have seen that these arguments, though having some merit because of the inefficiencies of the bureaucratic structure of controls, cannot be taken at face value especially in the crude form, and ignore many of the effects of liberalization stressed in approaches alternative to the neoclassical one. Moreover, the view that protectionism and export promotion are competing strategies relies heavily on the neoclassical HOS framework which talks of resource movements from exporting to import-substituting industries. In terms o f alternative frameworks, which allow employed resources to exist, while protectionism can raise input prices and make it physically difficult to obtain inputs for exporters and thereby dis­ courage exports, it is not difficult to provide special incentives to exporters to overcome these problems, as has been done in India.29 This would involve practising both import substitution and export promotion, which seems to have been done in Japan and South Korea. Moreover, efficiency can be improved by promoting domestic com­ petition and liberalizing the industrial licensing policy, without inviting foreign competition. While the efficacy of liberalization in raising growth remains a highly controversial issue, there appears to be more agreement that actual export expansion has favourable macroeconomic effects on the 29 One other argument according to which liberalization helps growth hinges on the link between liberalization and the greater willingness of international lenders to lend. This, however, involves an argument quite different from the efficiency argu­ ment, and points towards more political factors influencing the decisions of lenders.

62

A mhtava K r is h n a D u tt

economy. A large empirical literature using cross-section data at­ tempts to show that a faster growth of exports leads to a superitjr growth performance (see Ocampo, 1986 for a survey). These exer­ cises may reveal correlations, but not causality. Jung and Marshall's (1985) Granger-causality tests for 37 less developed economies do not provide* much support for the view that export growth causes income growth; for the Indian case, the data for 1960-79 fails to show causality either way between export growth and real output growth. In view of the inconclusiveness of these empirical controversies, it is preferable to focus theoretically on possible causal mechanisms. Aside from the efficiency effects of liberalization per se, which do not require the actual expansion of exports to affect the economy, the neoclassical approach could rely on higher savings and technology effects emanating from exporting sectors. However, Bhagwati and Srinivasan (1975), despite their strong pro-export promotion bias, do not find convincing evidence for these effects for India. One would in fact expect stronger arguments based on alternative approaches, especially the foreign exchange and demand-constraints, where greater exports would allow higher growth by increasing the ability of the economy to import intermediate and investment goods, and increase the demand for domestic goods in external markets. Bhag­ wati and Srinivasan (1975) show that increasing exports in the Eckaus-Parikh (1968) model with foreign exchange and savings constraints leads to faster growth, although this happens only after a reduction in absolute consumption in the economy for almost a decade, because of constraints on capacity. Given the higher labour content of exporting industries (Banerji and Riedel, 1980; Moham­ mad, 1981), export expansion would not only increase demand directly by expanding foreign markets, but would have the additional effect of increasing domestic markets by changing the structure of domestic production. If the higher growth in employment increases the labour share in the economy (by tightening labour markets) and makes income more equal, this would also tend to increase demand due to the reasons discussed in the Kaleckian model (Dutt, 1984) without creating an excessive burden on the foreign exchange situa­ tion (given the lower import-intensity of consumption of lower income groups; see above). Greater export growth, by expanding the growth of demand and output in the economy, would also create dynamic learning effects of the type emphasized in the models o f cumulative causation, which could raise the rate of growth by

Open-economy Macroeconomic Themes for India 63 encouraging more investment due to the more rapid change in technology. While these growth-inducing (and income-distributional) effects o f export expansion can be expected to be powerful, it is important to keep in mind that several models discussed in section 2 have the opposite implications. Structuralist full-capacity utilization models show that export growth may require a reduction in domestic absorp­ tion, and a worsening in domestic distribution, in the short run if the export expansion occurs at the expense of consumption, and even in the long run if it occurs at the expense of investment. Similar implications also emerge from Bagchi’s (1977) model which shows that export expansion through subsidies reduces output and worsens the distribution of income; however, apart from the problems noted in section 2.5, this result can be criticized because it is assumed that, and not shown why, export promotion is possibly only by shifting the distribution towards property income. Multisectoral models reveal. other problems. The mineral-exporter model implies a reduction in trade and production diversification due to the expansion in mineral exports. Sen’s (1977) model, developed to analyse the implications of export promotion by providing subsidies in India, implies that higher exports reduces investment in the home goods sector, which could reduce overall investment, creating long-run problems as well. Turning to the possibility of raising Indian exports, the arguments have centred around whether the constraints on export growth are external or internal. If the constraints can be shown to be primarily external, the possibilities are far weaker than if they are internal, especially policy induced. Early discussions (Patel, 1959) argued that given the heavy concentration of Indian exports in traditional items with an inelastic and slowly-growing world demand, the constraints were primarily external. Subsequent analysis (Cohen, 1964; Singh, 1964; Paul and Mote, 1971; Nayyar, 1976) instead stress internal factors, emphasizing supply bottlenecks and the internal pressure of demand; as noted above, some have placed the blame on domestic policies (Bhagwati and Srinivasan, 1975, 1984; Wolf, 1982). Aside from the detailed study of internal constraints in particular industries, proponents of this view have argued that the Indian share in world (and less developed country) exports is small and declining, so that the small country assumption is appropriate; econometric estimates by Lucas (1988), however, do not corroborate this. Others have tried to show that the profitability of export markets was lower than

64

A m it a v a K r is h n a D u tt

domestic markets and, except for a few markets, exporting was a marginal activity for manufacturing firms (Bhagwati and Srinivasan, 1975; Ray, 1987). Chandra (1986), however, reviews the evidence and argues that for many industries, exports were profitable compared to domestic sales; moreover, it is not clear that the lower profitability of export sales is the result of internal factors rather than external ones. Yet others have emphasized the poor technology (resulting in high costs and poor quality) and inferior marketing methods of Indian exporters. Recently, despite changes in the structure of Indian exports favouring non-traditional goods, the importance of external factors have again been stressed (Chakravarty, 1984; Ghosh, 1985; Nayyar, 1988): the slow growth of developed economies (recall the implica­ tions of North-South models stressing asymmetry), increasing protec­ tionism in the North, and the growth of competition from the NICs (recall the models of cumulative causation), have all been mentioned. Perhaps it is more fruitful to shift the debate from the internal-ex­ ternal dichotomy to one on the macroeconomic structure of the economy. If the economy is best described by neoclassical features, inefficiencies arising from domestic policies can be blamed. In this view, the liberalization of industrial and trade policy is appropriate. If capacity and other structural supply constraints are seen to create the constraints on growth, the pressures of domestic demand and the factors behind supply bottlenecks have to be emphasized, and measures should be taken to reduce domestic demand and attempt to remove the supply bottlenecks. If demand problems constrain the rate of growth, external problems as well as internal problems arising from high prices (due to inferior technology which imply high costs) and marketing problems should be emphasized. Chandra (1986) points out that given the level of excess capacities in Indian industries due to demand deficiency, the domestic demand pull is unlikely to restrict exports in general, although supply inelasticities in specific sectors (due to infrastructural constraints, for example) may have played an important role. In fact demand deficiencies in such models could imply low rates of growth of sectors exhibiting scale economies, low levels of learning and hence poor technology, and therefore low levels of exports. The appropriate policy would be to expand domestic demand (since this is easier to increase than foreign demand, and since foreign demand is a small component of total demand for Indian goods), and to remove sectoral infrastructural bottlenecks; greater long-run efficiency due to the technology-improving effects of growth

Open-economy Macroeconomic Themes for India 65 would increase exports as well. Structuralist models distinguishing between fix-price and flex-price sectors which allow for the exports of goods from both types of sectors, extended to incorporate scale economies and dynamic learning effects, could be used to pursue the implications of these lines of analysis.30

3.3. Stabilization policies We now examine the role of stabilization policies in balance of payments adjustment, and their general macroeconomic implications. We discuss, in turn, exchange rate policy and fiscal and monetary policies. (i) Exchange rate policy Up to 1971, when the rupee was pegged to the sterling and the Bretton Woods system was in place, India had a fixed exchange rate system. After the major international currencies began to float, following a brief period of pegging to the dollar, the rupee was pegged to the sterling, and from 1975 it has been pegged to a basket of currencies (the composition of which has not been announced) but the arrange­ ment allows managed exchange rate changes.31 Verghese (1984) has detected a steady nominal and real depreciation of the rupee against the currencies of India’s major trading partners, but the theoretical implications of this variability incorporating the effects of expecta­ tions regarding changes in exchange rates have not been examined, most of the literature treating the exchange rate to be in effect fixed. The main analytical issues discussed have been regarding the effects of a devaluation, and the desirability of multiple exchange rate sys­ tems. On devaluation, three main issues have been discussed. First, the direct effects of exchange rate changes on trade flows; second, the implications on the price level and hence the real exchange rate, and third, more complete macroeconomic effects. The effects on trade flows are important in themselves (for finding 30 It can be seen (hat this way of viewing (he controversy shows that the liberaliza­ tion view does not require the absence of external constraints on growth (since it aims to improve efficiency by competition rather than increasing growth by increasing exports), and that the domestic demand led growth view does not ignore domestic constraints such as efficiency and supply bottlenecks. 31 For a description of the exchange rate system see Verghese (1984).

66

A m it a v a K r is h n a D u tt

direct effects on foreign exchange requirements) and is also an essential component in broader macroeconomic assessments. Verghese (1984), without formal econometric analysis, finds exports uncorrelated with export growth, but is insightful about why simple regression exercises may be misleading (since it ignores many special features affecting exports). Wadhwa (1988a), using data for 1970-85, finds that India's aggregate real exports are highly elastic with respect to the nominal effective exchange rate (the same result is obtained at a more sectorally disaggregated level). However, the elasticities become statistically insignificant when exports are regressed on both world income and the exchange rate measure, which reveals the dangers of misspecified models. Lucas's (1988) more careful econo­ metric work reveals that exchange rate changes can have strong impacts on the demand for some manufactured exports, although the effects differ widely among industries. In interpreting these results it should be borne in mind, as Verghese (1984) points out, that the estimates obtained from a period when the exchange rate floated, may not provide an accurate measure of the impact of devaluation since the latter refers to an explicit policy change which will be interpreted differently by exporters from imperceptible and unpredictable chan­ ges in a period of managed floating. The effects on imports are much clearer, with devaluation having smaller effects (Verghese, 1984) since imports are confined to key intermediate and investment goods, and some necessary consumption goods (such as food) in a highly protected regime. Ghose et al. (1986) find that relative prices are not very relevant in determining total imports, given the binding foreign exchange constraint, but determine the composition of imports; moreover, relative prices do not substantially affect oil or investment good imports.32 The favourable competitive effect of a devaluation on exports can be eroded if the devaluation results in a rise in domestic prices by increasing the price of imported consumer goods, reducing real wages and causing an increase in money wages (an effect unlikely for India because of the low weight of imported consumer goods) or by increasing the prices of imported intermediate goods and causing a wage-price spiral (which is likely to be more important). Bhagwati 32 Verghese (1984) argues that the effect of the depreciation on other items in the current account balance are also negligible. For instance, remittances by migrants and expatriate workers consist of almost all savings of these unskilled and skilled workers who are not interested in gaining from exchange rate movements.

Open-economy Macroeconomic Themes for India 67 (1962), supporting devaluation for India, argues that it will not be inflationary. His focus is on demand-pull inflation and he examines the effects on the supply and demand for goods, but overlooks cost-push inflation which has generally been emphasized. It may also be argued that devaluation is not inflationary because it wipes out rents, or if it is part of a clean-up operation, removing import restrictions (Bhagwati, 1962; Ray, 1987), but this argument becomes less compelling the more liberal the system becomes. Ray (1987) argues for further devaluation if inflation occurs, ignoring the pos­ sibility of an explosive inflationary spiral. Regarding economy-wide approaches, the effects of devaluation have been examined by examining the actual effects of the 1966 devaluation, and by examining the hypothetical effects of a devalua­ tion in empirical models. Studies of the actual 1966 devaluation are complicated by the fact that they have to make allowances for other types of changes occurring in the economy at roughly the same time, such as trade liberalization (depending on whether one wants to examine the implications of the devaluation itself or of the whole policy package), and more impor­ tantly, adverse agricultural supply shocks. An early analysis by Bhagwati and Srinivasan (1975) argues that most of the contraction during the period is explained by the droughts; with appropriate corrections for this they find that the devaluation package had a favourable impact, mainly due to favourable efficiency effects: domestic output, investment, and exports increased, and the balance of payments improved. Sen (1982, 1986), on the other hand, takes a structuralist approach, and argues along the lines of sections 2.4 and 2.5 that the devaluation would have been contractionary had it not been for the drought of 1965-6. A simulation model in Sen (1982) shows that devaluation had a contractionary effect on the Indian economy in a structural model with reasonable parameter values. Sen (1986) argues that this contraction did not actually occur, however, since the drought caused an increase in the average propensity to consume in the face of a sharp reduction in real income (consistent with the permanent income hypothesis). The result of this contraction for the longer run would be ominous since it would lower production and rates of growth and private investment, causing a decline in long-run competitiveness along the lines discussed'earlier. Hypothetical exercises using computable models do not have to make adjustments to abstract from the effects of other exogenous

68

A m it a v a K r is h n a D u t t

changes. Sundararajan (1986) uses the monetarist model discussed in section 2.2 to analyse the implications of a devaluation, which has two effects: an effect on liquidity alpng monetarist lines and an effect on relative price along the lines of the elasticities approach. The relative price effect improves the trade balance along standard elasticity lines by increasing exports and reducing imports, but in the medium run this effect is eroded somewhat by an increase in domestic inflation due to the increase in real money supply. The liquidity effect, unlike the standard monetarist approach, initially reduces the demand for money by increasing the rate of inflation, and this reduces exports and increases imports initially; however, eventually the liquidity effects turn positive and improve the trade balance, dominating the weakening relative price effect so that the monetarist result is arrived at strongly. In the structuralist general equilibrium models of Taylor (1983) and Taylor, et al. (1984) described in section 2.5, when exports do not respond to changes in the exchange rate, a devaluation raises the costs of intermediate inputs and raises non-agricultural prices. The real wage falls, reducing aggregate demand, which reduces both output of non-agricultural goods and the prices of agricultural goods despite which there is an overall inflationary impact. The increased nominal values of exports and remittances and reduced volume of intermediate imports dominate the higher import cost, reducing the rupee current account deficit. The nominal value of investment rises due to the rise in non-agricultural prices, and the savings come from the government as its nominal export subsidies fall. The outcome of the fall in the government and trade deficit is a reduction in money supply, which has a contractionary effect by pushing up the interest rate. The devaluation also shifts the income distribution away from the agricultural sector. When exports respond to the exchange rate, the contractionary effects do not necessarily occur; an exchange rate elasticity of 0.5 raises export demand sufficiently to be expansionary for the non-agricultural sector. However, this expansion increases the demand for agricultural goods and raises agricultural prices; the income distribution now shifts towards agriculture. Taylor (1988) uses a simple structuralist equation to argue that given the plausible value of the price elasticity of export demand, devaluation is likely to be contractionary, at least in the short to medium run. Structuralist models thus imply that a devaluation has a favourable effect on the balance of trade, but there may be a contractionary effect

Open-economy Macroeconomic Themes for India 69 on the economy as a whole. To take advantage of the favourable effect but prevent contraction, it may be advisable to adopt dual and more complex systems of exchange rates. Thus, one rate could apply to exports and competitive imports, and another one to necessary imports of intermediate and capital goods; devaluation could affect the first rate and not the second, thereby preventing the contractionary effects achieving the same effect as the tax/subsidy policy recommended by Islam (1984), without its revenue effects. Kaldor (1984) argues for a different dual exchange rate system, which protects industries in need of protection forimport substitution purposes and promotes non-traditional exports, but does not raise the cost of imported inputs into exporting industries. A lower rate should apply to general items including essential inputs, while a higher rate should apply to the former categories. Sen (1987) combines these schemes, favouring a three-tier exchange rate system with one rate applying mainly to the price elastic non-traditional exports and all imports excluding vital intermediates, one to intermediates, and a third to the earning side of the current account and private transactions on the capital account; the first could then be directed towards increasing the competitiveness of India’s exports and providing the stimulus for import substitution, while the other two are used for minimizing the negative effects of devaluations; the distinction between the last two is because they move differently with respect to the agriculture-industry terms of trade and can be used as compensating mechanisms when there is an upward pressure on domestic prices. (ii) O ther stabilization policies Since devaluation is a politically unpopular tool (see Bhagwati and Srinivasan, 1975), greater reliance has been placed on other stabiliza­ tion tools to cure balance of payments disequilibria, although monetary instruments are inflexible because of the lack o f sensitive policy instruments (Joshi and Little, 1987). As in the case of devaluation, the effects of these can be studied by examining actual experiences, and those following the two oil shocks have been studied intensively. The current balance deficit as a result of the first oil shock was followed by the adoption of restrictive monetary and fiscal (reducing public investment) policies. Inflation was contained, but it is not clear to what extent this was the result of a good agricultural harvest. According to Joshi and Little (1987), the deflationary policy curtailed imports especially as a result

70

A m it a v a K r is h n a D u tt

of lower public investment, encouraged exports by reducing the pull of the home market, and had a favourable general effect on the balance of payments by containing inflation below the world rate and hence increasing the real exchange rate. The performance of the economy after the shock was favourable, but this has been explained by a number of fortuitous circumstances, including the increase in remit­ tances caused by the Middle Eastern oil boom. Before the second oil price shock there was also inflation caused by a poor harvest, but the policy was not contractionary. Money supply growth was generally accommodating, and public sector investment grew in real terms. Inflation remained high; exports did not perform well (handicapped by the higher rate of inflation), imports grew—because o f the expansionary policies but also because of continued import liberaliza­ tion—and industrial growth was low. The external deficit was financed by borrowing from the IMF, running down foreign exchange reserves, and by commercial borrowing. While these experiences can be taken to be the basis of an analysis of the impact of restrictive fiscal and monetary policies—and could be interpreted to be favourable for such policy—there is again the difficulty of separating the effects of the different exogenous changes such as the performance of agriculture, trade policy, and the behaviour of remittances. Turning to explicit models, in Sundararajan’s (1986) monetarist model, a reduction in credit expansion in the short run improves the trade balances, but over the longer run, by increasing the rate of domestic inflation, it reduces the demand for money, increases the absorption of'goods, and weakens the improvement in the trade balance. This can be taken as an argument against contractionist monetary policy in favour of devaluation, which was earlier described to have a favourable effect. The structuralist model of Taylor (1983) and Taylor, et al. (1984) shows a stagflationary effect of contraction­ ary monetary policy. Monetary contraction induced by decreasing non-monetary liabilities of the Reserve Bank raises the interest rate resulting in higher non-agricultural prices and a cut in investment demand. Prices rise in all sectors, and output decreases in the non-agricultural sectors; income distribution shifts away from wages. It is interesting to note that the trade deficit worsens, but what exactly causes this is not discussed. These results throw strong doubts on the' efficacy of contractionary monetary policy as a tool of stabilization in the Indian economy. The reduction in growth could also have

Open-economy Macroeconomic Themes for India 1 1 adverse long-run effects on growth and the balance of payments by slowing down technological change; cuts in government spending also imply lower spending on infrastructure which, as we have seen, is often singled out as an important constraint on growth in India, as well as the growth of exports. These adverse effects point to the importance of not using contrac­ tionary policy, of using devaluation with care (perhaps only with multiple exchange rates), of coping with the balance of payments problem without adjustment through policy, and of developing other policies for stabilization.

3.4 The Effects of Foreign Capital Inflows This section examines the effects of foreign capital inflows into the economy. First we look at the effects of inflow of foreign aid (abstract­ ing from interest payments) and then turn to the effects o f private flows and the payment for foreign capital. (i) Foreign aid The perception of effect of foreign aid will depend on the perception of the macroeconomic structure of the economy. For example, the undistorted neoclassical model implies welfare gains (this is not always the case in the presence of distortions), two-gap models imply higher rates of growth the exact magnitude depending on whether foreign exchange or savings constrains growth, and the effects in demand-constrained models depend on how aid affects investment, saving, and import propensities. The Indian literature has usually taken the two-gap model and examined two analytical issues, one which ascertains the true magnitude of aid flows (which is essential for examining the macroeconomic implications of a given amount of nominal aid into the economy), and another with the effects of aid on savings. A third issue, which implicitly'adopts a structuralist agriculture-constrained approach to growth, takes up the issue of the effects of food aid. Regarding the true amount of aid inflows it has been pointed out that these have to make adjustments for inflation, debt service charges, foreign exchange requirements when payments have to be made in hard currencies, and most importantly, aid tying—by source, project, and commodity (Bagchi, 1970a; Chandra, 1973). The analytical issues

72

A m it a v a K r ish n a D u tt

involved are discussed by Bhagwati (1967), and Chandra (1973) suggests that the true amount of aid may be nearer 60 per cent of reported figures. Regarding the impact of aid on the savings rate, several regressions (see Bhagwati, 1978) of savings on income, foreign capital inflows, and sometimes other variables suggest that greater capital inflows reduced the domestic savings rate. Chandra (1973) conjectured that in India the inflow of aid depressed the domestic savings effort, primarily that of the government, but regression estimates by Bhag­ wati and Srinivasan (1975) using total capital flows (rather than aid alone) failed to show any such effects for the fifties and sixties. While these estimates consider total capital flows and not just foreign aid, which is what the hypothesis is mostly about, Chaudhuri (1978) examines the effects of higher aid flows utilized (as a ratio of NNP) on the savings rate, using both gross and net aid, and found no evidence that aid reduced domestic savings; he also examined the effect of aid utilized on the government budget deficit and found that aid may have had some effect in reducing the deficit, or in increasing government savings. Weisskopf (1971) argues that the regression results do not take into account whether the level of domestic savings observed in each country reflected an ex-ante behavioural equation or simply an ex-post accounting relationship; using a two-gap ap­ proach he distinguishes between economies according to whether they were savings or foreign exchange constrained, and finds that India is savings constrained. A regression equation which could properly be thought of as a behavioural equation shows (using data for the 1950-65 period) that capital inflow has a statistically significant, negative effect on savings. A recent study by Bowles (1987) using data for 1960-81 also lends some support for the hypothesis, con­ firming that more aid Granger-causes a reduction in the savings rate for India. Given the difficulties connected with measuring the true amount of aid, as noted above, this author recommends caution against taking these results too seriously; moreover, the results appear to be extremely sensitive to the auxiliary hypotheses (whether exports or a lagged savings rate are included as additional variables). Even if foreign aid does depress the domestic savings rate, it may be argued that this does not signify a problem, since the inflow, by augmenting the amount of available resources for the economy can be expected to increase both consumption and investment, and therefore reduce domestic savings. The key issue is what happens to

Open-economy Macroeconomic Themes for India 73 investment, and here Chaudhuri’s (1978) results suggest that invest­ ment was positively related to aid flows which suggests that the inflows raised the rate of growth of the economy. A more sensible question to ask, as Grinols and Bhagwati (1976) do, is whether the savings rate which falls as a result of aid inflow rises in the future as income grows faster and catches up to what it was prior to the capital inflow, and how long it takes for the savings rate to recover its earlier value. Using an open economy version of the simple Harrod-Domar model (along the lines of the savings-constrained model of the previous section) for the case of India they find that the savings rate does recover; even assuming an unfavourable capital-output ratio of 5, recovery takes only a little more than eight years. However, Bhagwati and Grinols (1975) use the same model to show that a cessation of aid after a decade of its provision, together with a resource crunch that reduces the output-capital ratio by 3 per cent will have a significant negative effect on the economy compared to what would happen if aid were continued, implying that as a result of foreign aid an economy such as the Indian one may be put in a position in which it could be ‘blackmailed’ by the aid donors and thereby be restricted in its choice of policies and be weakened in its bargaining power in dealing with multinationals about the terms of investment (Bagchi, 1970a). As mentioned earlier, investigations of saving and growth effects have relied exclusively on the two-gap model; other models will have different conclusions. For example, Joshi and Little (1987) point out that the fall in aid in the late sixties was associated with a cut in public investment, and this, in demand-constrained models and the fiscal crisis approach, will have a retarding effect on growth. Regarding the impact of food aid, there is a large literature on the impact of PL 480 imports, with several writers arguing that such aid reduced the price of food and consequently food production. Aside from the fact that this ignores the effects of food aid on the demand curve due to the low cost of distribution of food, and does not adequately take into account government operations in foodgrains, Bhagwati (1985) criticizes this essentially partial-equilibrium microeconomic argument by pointing out, using a neoclassical framework, that such a shift may simply imply the allocation of resources towards other sectors. He feels this is normally not problematic unless one takes into account the possibility of a sudden cessation of aid and that the intersectoral mobility of resources back to food production

74

A m it a v a K r is h n a D u tt

is slow and that there are foreign exchange constraints which prevent food imports. However, the view has persisted in other forms: Bagchi (1970a) and Chandra (1977) argue that food aid led to the neglect of the agricultural sector and the failure to take politically difficult steps leading to land reform, all of which held down the rate of growth of agriculture, and exacerbated the agrarian constraint on growth. (ii) Private capital flow s Private capital flows raise several additional issues relating to the effects of payments on foreign investment and the long-run effects of debt accumulation.33 Taking into account the effects of payments, even a simple neoclassical approach introducing a distortion relevant for the Indian economy shows that foreign capital inflows under competitive con­ ditions result in welfare losses. Brecher and Diaz-Alejandro (1977) show that the inflow of foreign capital into an economy which imposes tariffs on its capital-intensive imports is immiserizing since the capital inflow increases the size of the distorted import sector. Other approaches stress that high payments (due to monopoly power) made for foreign capital which within a few periods converts net inflows into outflows, reduce the rate o f growth of the economy by adversely affecting domestic savings or the availability of foreign exchange (Chandra, 1973). Regarding the effects of foreign debt accumulation, Bhaduri (1987) examines the possibility of self-reliant growth using a simple savingsconstrained model. Self-reliance is defined as a situation in which the economy is not indebted to the rest of the world; dependence, on the other hand, is a situation in which the economy permanently becomes dependent on foreign borrowing. Bhaduri assumes that both exports and imports depend on domestic product, that savings depend on national product (domestic product less interest payments), that all resources are productively invested, and that the economy borrows from abroad at a fixed interest rate. The analysis shows that the economy cannot achieve self-reliant growth as long as its marginal 33 Multinational investment and foreign collaborations raise a variety of issues regarding the effects on exports (through no-export clauses), choice of techniques, technological change, domestic entrepreneurship, profit repatriation and transfer pay­ ments, income distribution, and more comprehensive socio-political factors, which would take at least a paper to cover, and we confine ourselves to a brief comments on the first two issues. See Chandra (1973) for a review of some of these issues.

Open-economy Macroeconomic Themes for India 75 propensity to import exceeds its marginal propensity to export. This emphasizes the fact that in addition to having a high marginal savings rate and productivity of investment, the economy needs to strengthen its incremental trade balance while it grows (although this is still not sufficient for self-reliance). Most developing economies are, however, very far from selfreliant in this sense; they would be happy if they did not find themselves in a situation in which their indebtedness did not get out of hand. The level of indebtedness has usually been monitored in terms of debt-service ratios as a proportion of exports, or debt-GDP ratios. Ahluwalia’s (1986) simulation model referred to earlier implies a marked deterioration of the debt-service ratio from 13 per cent of exports in 1983-4 to 30 per cent in 1990-1, resulting* from higher interest charges due to lower availability of concessional assistance and increase in the interest rate on other loans; the ratio would be even higher if the current account deficit grew faster, allowing a higher level of growth and capacity utilization. If the ratio is to be kept at the limit of 20 per cent, even with an export demand growth of 7 per cent (a fairly optimistic estimate), the growth of the current account deficit will have to be kept at a low (almost constant) which would constrain the economy to grow at less than what has been achieved in the past and below the potential. Taylor (1988), using stylized parameters and accounting equations, argues that while borrowing is in a decent shape with India having a good credit rating, if current deficit and export shares in output remain constant, the equilibrium level of the debt-export ratio would rise to 10.7 as opposed to a current value of less than 2, which would create borrowing problems in the future. The importance of concessional assistance and more favourable international lending conditions is obvious.

4. C O N C LU SIO N By way of conclusion we end with three remarks. 1. Despite the fact that open economy transactions are small for the Indian economy, trade and capital flows do have an important impact on the economy. A variety o f important issues relating to foreign exchange constraints, export and import policy, stabilization policy, and the effects of foreign capital inflows have been raised.

76

A m it a v a K r ish n a D u tt

and they deserve more attention. This is not to imply that India’s development problems are primarily external, or that the solutions to these problems lie in the external sphere. 2. Though there have been a variety of analytical contributions and some use of explicit models, the level of discourse on these issues has often been in the form of assertions and casual empiricism, rather than explicit theoretical and empirical modelling. While informal work has many advantages—the most important being creativity and the ability to generate new, interesting ideas— the proper under­ standing and evaluation of the issues requires more formal macroeconomic analysis. It is preferable to be explicit about one’s theoretical approach, so that the internal consistency and premises of the approach can more easily be assessed. The signs are good: in closed economy analysis formal models are increasingly becoming popular, and there is a large literature on models for other less developed economies which analysts working on India can draw from;34 protectionism in this sphere is likely to be dangerous. 3. However, the non-discriminating importation of all-purpose theoretical models developed abroad is not desirable. Alternative models and their implications should be carefully studied, and adopted for the Indian situation with careful modifications, appropriate for the questions being asked. It is hoped that this survey provides a useful sampling of what is available and also some impetus to the pursuit o f this goal.

REFERENCES (1985), Industrial Growth in India: Stagnation Since the Mid-sixties, Oxford University Press, Delhi. A h l u w a u a , M ontek S ingh (1986), ‘Balance-of-payments Adjustment in India, 1970-71 to 1983-84', World Development, 14 (8). A hm ad , J aleel (1968), 'Import Substitution and Structural Change in Indian Manufacturing Industry’, Journal o f Development Studies, 4 (3), A p ril. A r id a , P ersio (1986), 'Macroeconomic Issues for Latin America*, Jour­ nal o f Development Economics, 22 (1), June. A hluw alia , I sher J udge

34 See, for example, Arida (1986) and Taylor (1989) for surveys.

Open-economy Macroeconomic Themes for India 11 (1984), ‘Growth with Limited Supplies of Foreign Exchange: A Reappraisal of the Two-gap Model’, in M. Syrquin, L Taylor and L. Westphal (e'ds.), Economic Structure and Perfor­ mance* Academic Press, Orlando. ------ (1986), ‘Terms of Reference for the Country Studies’, World Development, 14 (8). ------(1989), ‘A Three-gap Model of Foreign Transfers and the GDP Growth Rate in Developing Countries* (unpublished). Catholic University, Rio de Janeiro. B agchi , A miya K umar (1970a), ‘A id M o d els and Inflows of Foreign Aid’, Economic and Political Weekly, Annual Number, January. ------ (1970b), ‘Long-term Constraints on Indian Industrial Growth, 1951-68’, in E.A.G. Robinson and M. Kidron (eds.), Economic Development in South Asia, Macmillan, London. ------ (1977), ‘Export-led Growth and Import-substituting Industrial­ ization*, Economic and Political Weekly, Annua! Number, February. B anerji, R anadev and J ames R iedel (1980), ‘Industrial Employment Expansion under Alternative Trade Strategies. Case of India and Taiwan, 1950-1970’, Journal o f Development Economics, 7. B a r d h a n , P r an ab (1984), The Political Economy o f Development in India, B a s il B la c k w e ll, O x fo rd . B ercsm an , J. and A lan S. M anne (1965), ‘An Almost Consistent Inter­ temporal Model for India’s Fourth and Fifth Plans’, Economic Week­ ly, 20 November. B haduri, A m it ( I 9 8 7 ) , ‘Dependent and Self-reliant Growth with Foreign Borrowing’, Cambridge Journal o f Economics, 11 (3), September. B hagwati , J agdish N. (1962), ‘The Case for Devaluation’, Economic Weekly, 14 (31), 4 August. ------(1967), ‘The Tying of Aid’, UNCTAD Secretariat, TD/7/Supp.4, United Nations, 1-57, reprinted in J. Bhagwati and R.S. Eckaus (eds.), Foreign Aid, Penguin, Harmondsworth, 1970. ------(1978), Anatomy and Consequences o f Exchange Control Regimes, Ballinger, Cambridge, Mass. ------(1985), ‘Food Aid, Agricultural Production and Welfare’, in S. Guhan and M.R. Shroff (eds.), Essays on Economic Progress and Welfare: Essays in Honour o f LG. Patel, Oxford University Press, New Delhi. B hagwati, J agdish N. and S ukhamoy C hakravarty (1971), Contribu­ tions to Indian Economic Analysis: A Survey, Lalvani, Bombay. B acha, E d m a r

78

A m it a v a K r is h n a D u t t

N. and P adma D esai (1970), India: Planning for Industrialization. Industrialization and Trade Policies Since 1951, Oxford University Press, Oxford. B hagwati, J agdish N. and E arl G rinols (1975), ‘Foreign Capital, De­ pendence, Destabilisation and Feasibility of Transition to Socialism*, Journal o f Development Economics, 2. B hagwati , J agdish N. and T.N. S rinivasan (1975), Foreign Trade Regimes and Economic Development: India, National Bureau of Economic Research, New York. ------(1983), Lectures on International Trade, MIT Press, Cambridge, Mass. ------ (1984), ‘Indian Development Strategy: Some Comments*, Economic and Political Weekly, 19 (47), 24 November. B iswas , B asudeb and S ukumar N andi (1986), ‘The Black Market Ex­ change Rate in a Developing Economy: The Case of India*, Indian Economic Journal, 33 (3), January-March. B ow les , P aul (1987), ‘Foreign Aid and Domestic Savings in Less Developed Countries: Some Tests for Causality’, World Develop­ ment, 15 (6). B recher , R ichard and C arlos D iaz-A leiandro (1977), Tariffs, Foreign Capital and Immiserizing Growth*, Journal o f International Economics, 1. B uffie , E dward F . (1986), ‘Devaluation, Investment and Growth in LDCs*, Journal o f Development Economics, 20 (2), March. C hakravarty , S ukhamoy (1974), ‘Reflections on the Growth Process in the Indian Economy’, in Hyderabad Administrative Staff College of India. ------(1979), ‘On the Question of Home Market and Prospects for Indian Growth*, Economic and Political Weekly, Special Number, August. ------(1984), ‘India’s Development Strategy for the 1980s’, Economic and Political Weekly, 19-26 May. ------(1987), Development Planning: The Indian Experience, Clarendon Press, Oxford. C handra , N irmal K umar (1973), ‘Western Imperialism and India Today*, Economic and Political Weekly, Annual Number, February, reprinted in Chandra, Retarded Economies. ------(1986), ‘Modernisation for Export-oriented Growth: A Critique of Recent Indian Policy’, Economic and Political Weekly, 19 June, reprinted in Chandra, Retarded Economies. ------(1988), The Retarded Economies. Foreign Domination and Class B hagwati, J agdish

Open-economy Macroeconomic Themes for India 79 Relation in India and Other Emerging Nations, Oxford University Press, Bombay. C hatterji, R uchira (1989), The Behaviour o f Industrial Prices in India, Oxford University Press, Delhi. C haudhuri, P ramit (1978), The Indian Economy. Poverty and Develop­ ment, 1978 edition, Vikas Publishing House, New Delhi. C henery , H o l u s (1961), ‘Comparative Advantage and Development Policy*, American Economic Review, 51. C ohen , B enjamin (1964), ‘Stagnation of Indian Exports, 1951-61*, Quarterly Journal o f Economics, 78, November. C orden , W illiam M. (1984), ‘Booming Sector and Dutch Disease Economics: Survey and Consolidation', Oxford Economic Papers, 36. D asgupta , A.K. (1954), ‘Keynesian Economics and Underdeveloped Countries*, Economic Weekly, 26 January. D ervis , K emal , J aime de M elo and S herman R obinson (1982), General Equilibrium Modelsfo r Development Policy, Cambridge University Press, Cambridge. D iwan , R omesh K. (1968), ‘A Test of the Two-gap Theory of Economic Development’, Journal o f Development Studies, 4 (4), July. D ornbusch , R udiger (1980), Open Economy Macroeconomics, Basic Books, New York. D utt , A mitava K rishna (1984), ‘Stagnation, Income Distribution and Monopoly Power*, Cambridge Journal o f Economics, 8(1), March. ------(1986), ‘Vertical Trading and Uneven Development*, Journal o f Development Economics, 20 (2), March. ------ (1988a), ‘Inelastic Demand for Southern Goods, International Demonstration Effects, and Uneven Development*, Journal o f Development Economics, 29 (1), July. ------(1988b), ‘North-south Trade Models: A Critical Survey* (un­ published), University of Notre Dame. ------(1990), Growth, Distribution and Uneven Development, Cambridge University Press, Cambridge. D utta , J ayasri (1988), ‘The Wage-goods Constraint on a Developing Economy: Theory and Evidence’, Journal o f Development Economics, 28 (3), May. E cka us , R ichard S. and Kmrr P arjkh (1968), Planning fo r Growth, MIT Press, Cambridge, Mass. F indlay , R onald (1980), ‘The Terms of Trade and Equilibrium Growth in the World Economy*, American Economic Review, June.

80

A m it a v a K r is h n a D u tt

Gersovitz, Mark (1982), ‘The Estimation of the Two-gap Model*, Jour­ nal o f International Economics, 12. G hose , D evajyoti, A soke K. L ahiri and W ilima W adhwa (1986), ‘Quan­ titative Restrictions and Indian Imports*, Journal o f Development Economics, 20 (2), March. G h o sh , J a yati (1985), 'Export Optimism and Import Liberalisation*, Economic and Political Weekly, 20 (22), 1 June. G rinols , E arl and J agdish N. B hagwati (1976), ‘Foreign Capital, Savings and Dependence’, Review o f Economics and Statistics, 58 (4), November. H arris , D onald J. (1972), ‘Economic Growth with Limited Import Capacity*, Economic Development and Cultural Change. H azari , B harat R. (1967), ‘Import Intensity of Consumption in India*, Indian Economic Review, 2 (2). I slam , S hafiqul (1 9 8 4 ), ‘Devaluation, Stabilization Policies and the Developing Countries: A Macroeconomic Analysis*, Journal o f Development Economics, 14. J oshi , V uay and I.M.D. L ittle (1 9 8 7 ), ‘Indian Macroeconomic Policies’, Economic and Political Weekly, 2 2 (9), 2 8 February. J ung , W oo S. and P eyton J. M arshall (1 9 8 5 ), ‘Exports, Growth and Causality in Developing Countries', Journal o f Development Economics, 18. K aldo r , N icholas (1970), ‘The Case for Regional Policies’, reprinted in Further Essays on Economic Theory, Duckworth, London, 1978. ------( 1984), ‘An Exchange Rate Policy for India*, Economic and Politi­ cal Weekly, 14 July. K alecki , M ichal (1 9 7 1 ), Selected Essays on the Dynamics o f the Capitalist Economy, Cambridge University Press, Cambridge. K a t r a k , Homi (1 9 8 0 ), ‘Industry Structure, Foreign Trade and Price-cost Margins in Indian Manufacturing Industries’, Journal o f Develop­ ment Studies, October. K rueger , A nne O. (1978), Liberalization Attempts and Consequences, National Bureau of Economic Research, New York. ------(1984), ‘Trade Policies in Developing Countries’, in R.W. Jones and P.B. Kenen (eds.). Handbook o f International Economics, Volume I, North-Holland, Amsterdam. K rugman , P aul (1981), ‘Trade, Accumulation and Uneven Development*, Journal o f Development Economics, 8. L all , S anj aya (1984), ‘India* s Technological Capacity: Effects of Trade, Industrial Science and Technology Policies’, in M. Fransman and

Open-economy Macroeconomic Themes for India 81 K. King (eds.), Technological Capability in the Third World, Mac­ millan, London. L it t le , Ian M.D. (1982), Economic Development: Theory, Policy and Industrial Relations, Basic Books, New York. L u c a s , R o b e r t E.B. (1 9 8 8 ), ‘Demand for India's Manufactured Exports*, Journal o f Development Economics, 2 9 (1 ), July. L u c a s , R o b e r t E.B. and G u sta v F. P a pa n ek (eds.) (1988), The Indian Economy: Recent Development and Future Prospects, Oxford University Press, Delhi. M a m m e n , T h um py (1984), ‘A Trade Gap Model and IMF Loan to India*, Indian Economic Journal, 31 (3), January-March. M c K in n o n , R o n a ld I. (1964), ‘Foreign Exchange Constraints in Economic Development and Efficient Aid Allocation*, Economic Journal, 74, June. M it r a , A sh o k (1977), Terms o f Trade and Class Relations, Frank Cass, London. M o h a m m a d , S h a r if (1981), ‘Trade, Growth and Income Redistribution: A Case Study of India*, Journal o f Development Economics, 9. M o o r e , B a s il J. (1988), Horizontalists and Verticalists, Cambridge University Press, Cambridge. N a m b ia r , R.G. (1983), 'Protection of Domestic Industry: Fact and Theory*, Economic and Political Weekly, 1-8 January. N a m b ia r , R.G. and R a jesh M ehta (1987), ‘Effect of Tariffs on Foreign Prices: The Case of India*, Economic and Political Weekly, 2 2 (2 4 ), 13 June. N ay yar , D eepa k (1976), India ps Exports and Export Policies, Cambridge University Press, Cambridge. ------(1978), ‘Industrial Development in India: Some Reflections on Growth and Stagnation*, Economic and Political Weekly, Special Number, August. ------(1988), ‘India*s Export Performance, 1970-85: Underlying Factors and Constraints', in Lucas and Papanek (eds), Indian Economy. O cam po , J ose A ntonio (1 9 8 6 ), ‘New Developments in Trade Theory and LDCs\ Journal o f Development Economics, 2 2 (1 ), Ju n e . ------ (1987), ‘The Macroeconomic Effect of Import Controls: A Keynesian Analysis*, Journal o f Development Economics, 27 (1-2), October. P a rik h , K ir it (1968), ‘Planning Without Policy?’, in S. Guhan and Manu Shroff (eds.). Essays in Economic Progress and Welfare, Oxford University Press, New Delhi.

82

A m it a v a K r is h n a D u tt

J. (1959), ‘Export Prospects and Economic Growth: India’, Economic Journal, 69, September. P a tn a ik , P r a b h a t (1987), ‘Recent Growth Experience of the Indian Economy: Some Comments', Economic and Political Weekly, An­ nual Number, 22 (29, 20, 21), May. ------( 1988), Time, Inflation and Growth: Some Macroeconomic Themes in an Indian Perspective, Centre for Social Studies, Calcutta. P a u l , S a m u e l and V a s a n t L . M o t e (1970), ‘Competitiveness of Exports: A Micro-level Approach', 80, December. P o l a k , I.J. (1957), ‘Monetary Analysis of Income Formation*, IMF Staff Papers, November. Raj, K.N. (1976), ‘Growth and Stagnation in Indian Industrial Development’, Economic and Political Weekly, 26 November. R a k s h i t, M ihir (1982), The Labour Surplus Economy, Macmillan, Delhi. R a k sh it , S usm ita (1985), ‘New Import and Export Policy: Some Economic Implications', Economic and Political Weekly, 20 (37), 14 September. R a m a n a , R . and A dity a B h a tta c h a r jea (1985), ‘Theoretical Presupposi­ tions for Liberalisation: A Critique*, Social Scientist, 146-7, JulyAugust. R a o , V.K.R.V. (1952), ‘Investment, Income and the Multiplier in an Underdeveloped Economy', Indian Economic Review, February. R ay, Alok (1987), ‘Economic liberalisation in India: Balance of Pay­ ments Implications’, Economic and Political Weekly, 11 July. S e n , A b h u it (1981), ‘The Agrarian Constraint to Economic Develop­ ment: The Case of India*, Ph.D. dissertation (unpublished). Univer­ sity of Cambridge. S e n , P ron ab (1982), ‘Imported Inputs and the Theory of Currency Devaluations’, Ph.D. dissertation (unpublished), Johns Hopkins University, Baltimore. ------(1986), ‘The 1966 Devaluation in India: A Reappraisal', Economic and Political Weekly, 21 (30), 26 July. ------ (1987), ‘Indian Experiences with Orthodox Stabilisation* (un­ published), Indian Council for Research on International Economic Relations, New Delhi. S e n , S una nd a (1977), ‘Strategy of Export-oriented Growth’, Economic and Political Weekly, 10 September. S in g h , A jit and J a y a ti G hosh (1988), ‘Import Liberalisation and the New Industrial Strategy: An analysis of their Impact on Output and P a t e l , S u ren dra

Open-economy Macroeconomic Themes for India 83 Employment** Economic and Political Weekly, Special Number, 23 (45-47), November. S in g h , M a n m o h a n (1964), India*s Export Trends, Oxford University Press, New York. S o h r a b u d d in , M o h a m m a d (1985), ‘Monetary Approach to the Balanceof-payments: Evidence from Less Developed Countries*, Indian Economic Journal, 33 (1), July-September. S r in iv a sa n , T.N. (1986), ‘Development Strategy: Is the Success of Out­ ward Orientation at an End?', in S . Guhan and Manu Shroff (eds.), Essays on Economic Progress and Welfare, Oxford University Press, Delhi. ------(1988), ‘International Trade and Factor Movements in Develop­ ment Theory, Policy and Experience*, in G. Ranis and T. Paul Schultz (eds.), The State o f Development Economics, Blackwell, Oxford. S u n d a r a r a ja n , V. (1986), 'Exchange Rate versus Credit Policy: Analysis with a Monetary Model of Trade and Inflation in India*, Journal o f Development Economics, 20 (I), January-February. T alele , C h aita ra m (1984), ‘The 1966 Devaluation of the Rupee: Em­ pirical Analysis from the Point of View of the Monetary Theory of the balance-of-payments’, Indian Economic Journal, 31 (3), Januaiy-March. T a y l o r , L a n c e (1981), ‘South-north Trade and Southern Growth: Bleak Prospects from a Structuralist Point of View*, Journal o f Interna­ tional Economics, 11. ------(1983), Structuralist Macroeconomics, Basic Books, New York. ------(1988), ‘Macro Constraints on Indian Growth* (unpublished), MIT, Cambridge, Mass. ------(1989), Stabilization and Growth in Developing Countries, Har­ wood, Chur. T a y l o r , L a n c e , H iren S a rk ar and J orn Rattso(1984), ‘Macroeconomic Adjustment in a Computable General Equilibrium Model for India', in Moshe Syrquin, Lance Taylor and Larry E. Westphal (eds.), Economic Structure and Performance: Essays in Honor o f Hollis B. Cheneryf Academic Press, Orlando. V erg h ese , S.K. (1984), ‘Management of Exchange Rate of Rupee since its Basket Link*, Economic and Political Weekly, part 1, 14 July, part 2, 21 July. W a d h w a , C h a r a n D. (1988), ‘Some Aspects of India's Export Policy and Performance*, in Lucas and Papanek (eds.), Indian Economy.

84

A m ita v a K r is h n a

D un

E. (1971), 'T h e Impact of Foreign Capital Inflow on Domestic Savings in Underdeveloped Countries’, Journal o f Inter­ national Economics, 2. W o l f , M a rtin (1982), India’s Exports, Oxford University Press, Oxford. W e issk o pf , T h om a s

Closed-economy Structuralist Models for a Less Developed Economy A m iy a K um ar B agchi

1. THE NEED TO SORT OUT THE SHEEP FROM THE GOATS The word ‘structure’ or structuralism has been used in many different ways in economic analysis, let alone the myriad other ways in which it has figured in linguistics, literary theory, sociology and social anthropology. It would be possible to devote a whole monograph to the discourse of structuralism in economics. I will confine myself to a rough attempt at classification in the context of structuralist ap­ proaches to macroeconomic imbalances, such as unemployment and inflation, and to long-term growth. I believe that we will gain in understanding if through a classificatory exercise we can sort out models in which long-term growth plays a central role and models that take the explanation of short-run movements in prices, incomes and employment as the basic problem; models that treat money as a passive variable and those in which substitution between money and other assets by economic agents seeking to better their positions plays an important part; models in which conflicts between different claimants to a share of the national income play a part and models in which such conflicts are either taken to be unimportant or are subsumed under other headings; and finally, models which incorporate expectations regarding the actions and reactions of other economic agents including governments and models which ignore such expectational variables altogether. Some of the groupings constitute virtually a null set, but even that finding will help our understanding of the work still needed in the macro­ economics of less developed countries, including India.

86

A m iy a K u m a r B a g c h i

Any structuralist model must be based on the idea of a structure which possesses a certain rigidity and which cannot be transformed into some other configuration solely or primarily through changes in relative prices. Structuralist models do not necessarily ignore changes in relative prices (some of them do) but consider them to be of secondary importance in either bringing back the economy to a position of stability (‘equilibrium*) in respect of incomes, prices and employment, or in putting the economy back on a reference path of growth from which it is supposed to have deviated. Just as the would-be gentleman in Moliere’s play was talking prose all his life without knowing it, there were many economists who were engaged in structuralist analysis before the term had been thought of in the context of macroeconomics (see in this connection Chenery, 1975). One early example of such structuralist analysis is Goodwin’s model of inflation, in an economy characterized by fixed input-output coefficients, and governed by either competitive rules or by producers setting prices on the basis of fixed mark-ups (Goodwin, 1952, 1953). He also considered the case in which some sectors were governed by competitive rules and some sectors followed cost-plus pricing rules. Although Goodwin called his models Walrasian or Walras-Leontief linear systems, they were fundamentally non-Walrasian in two dif­ ferent ways. First, the proportions of factors used in production generating the input-output coefficients did not change as absolute or relative prices changed. Secondly, in the sectors governed by cost-plus rules of pricing, the prices were determined not at the points at which supply and demand schedules intersected, but on the basis of costs of production alone, even though there might be excess supplies in the market at those prices. These multisector structuralist models of inflation seem to have exerted very little influence on later developments.

2. STRUCTURALIST MODELS OF PLANNING The influential two-sector Feldman-Mahalanobis model of growth is a structuralist model, (Feldman, 1928; Mahalanobis, 1953; Domar, 1957), the capital-output ratios are given, and only the allocation of the output of the capital goods sector between the two sectors is a decision variable. The Mahalanobis four-sector model constructed for the plan-frame of the Indian second five year plan was a fortiori a

Closed-economy Structuralist Models 87 structuralist model (Mahalanobis, 1955). For in that structure, not only the capital-output ratios but the labour-output ratios were taken to be given as well, so that a consistent outcome in terms of growth of national income and employment could be calculated by making particular assumptions about the allocation of investment (= output of the capitd goods sector) as between different sectors (Mahalanobis, 1955). These models were not only structuralist but they also assumed money to be just a veil with no implications for the long-term behaviour of the economy. Mahalanobis was realistic enough and imaginative enough to recognize that the implementation of his model required certain preconditions and was subject to certain limits. One such limit was the ability of the state to raise enough financial resources to support the investment programme. He also recognized that the capacity of the cottage and small-scale consumer goods industries to produce enough additional output, would be a precon­ dition for preventing the inflation that would result from a programme of accelerated investment in capital goods industries that would generate more consumer goods only with a considerable lag. There was a contradiction in the Mahalanobis plan-frame about the diagnosis of the cause of unemployment, in India and of the limits on total output. On the one hand, the basic cause of unemployment was taken to be the shortage of capital goods, and an explicit contrast was drawn with highly industrialized countries where unemployment occurred only when ‘the means of production [were] idle* (Mahala­ nobis, 1955, p. 71). In the very next paragraph, however, Mahalanobis postulated that production could be expanded in the cottage and small-scale industries if only the demand for such goods could be increased as had happened during the second World War. Mahalanobis wanted new demand to be created by raising the investment in ‘the heavy industries producing investment goods’ and the expenditure on health, education, and other social services. Mahalanobis foresaw some, though not all of the problems with this programme. First, under ordinary signals of the market, invest­ ment would continue to be in factory industries producing consumer goods. He wanted that diversion to be stopped so that the rate of expansion of the capital goods industries and the long-run rate of. growth of national income could be maximized. Secondly, factories, competing with consumer and small-scale industries might prevent

88

A m iy a K u m a r B a g ch i

the output of the latter from being sold (presumably, because of market dominance and the pursuit of various restrictive practices). He, therefore, wanted the further expansion of factories to be frozen (Mahalanobis, 1955, p. 72). It is not clear why, according to him, this could result in ‘some’ surplus factory capacity (sic) remaining idle temporarily (unless he assumed that factories were really higher-cost producers than cottage and small-scale industries and were using their financial and marketing clout to raise barriers against the latter). Thirdly, Mahalanobis recognized that the price of hand-made goods would be sometimes higher than the price of factory-made goods of comparable quality. A simple remedy is to levy suitable excise duties on factory-made goods to preserve price parity with hand-made goods at any desired level. This would, no doubt, raise prices to some extent but would at the same time supply additional resources for investment and hence for additional employment, increase of income, and national development (Ibid). This is the place at which the link between the Mahalanobis growth model and the possibility of inflation is recognized. The short-run impact of the Mahalanobis programme is illustrated in Figure 1.

FIGURE 1

Closed-economy Structuralist Models 89 In Figure 1, F x F indicates the marginal cost (including normal profit) of the factory sector. The capacity of the factory sector is fixed at the output OA (of cloth); since all technical coefficients are fixed the marginal cost is a straight line at given prices of factors of production. Ignoring quality differences we draw the marginal cost curve of the cottage and small-scale industries as C, C2. Before the planned investment programme starts, the demand curve for cloth D x D x intersects the aggregate marginal cost curve of the cloth industry at E. At that point, the factory sector produces its full capacity output OA and the cottage and small-scale industries produce AB amount of cloth. When the investment programme is stepped up and the capacity of the factory sector is frozen at output OAt the demand shifts to D2 D2 and intersects the aggregate marginal cost curve at G. At that point, the total output of the cottage and small-scale industries goes up from AB to AH, and the price of all cloth, including mill-made cloth, goes up to OF2 from OFv If the government does not interfere in the price mechanism, the factory sector will enjoy a supernormal profit (rent) equal to the rectangle F } F2 F3 F. Mahalanobis proposes to mop up this rent partially or wholly by imposing an excise tax which could range from any rate per unit above zero up ta Fx F2 per unit on factory cloth, and then reinvest the surplus in such a way as to raise the rate of growth of the economy and eventually allow part of that surplus to be reinvested in the factory consumer goods sector, and, thus, expand its output at some planned rate. Notice that the way we have presented the Mahalanobis scheme, a substantial part of the cottage and small-scale industries sector would also enjoy an intramarginal rent (equal to the area bounded by Cx Ft GF$ and F 3 Cj). Mahalanobis does not seem to have contemplated imposing a tax on that rent. This is only one illustration of the systematic way in which the private sector is treated as a passive agent in the Mahalanobis model—an agent which could be directed and regulated by the government any way it considered proper through some (largely undefined) fiscal and monetary policies. One major reason for the eventual irrelevance of the Mahalanobis plan-frame for economic policy-making in India was its failure to incorporate the reactions of the private capitalists—as savers, allocators of resources, investors, price-fixers, speculators and tax payers and tax evaders—consistently into its various calculations (Bagchi, 1970b). If Mahalanobis* assumptions about the easy taxability of the

90

A m iy a K u m a r B a g c h i

private sector and the deployment of the national surplus (including the taxes) in generating an increased flow, first of capital goods, and then of consumer goods were correct, then the rate of inflation in the economy would flatten out. The process of inflation in itself, however, would not perform any equilibrating or deflecting function, for the simple reason that in the Mahalanobis scheme all imbalances could be tackled through appropriate regulatory, fiscal or monetary inter­ vention. The Mahalanobis model is an example of a structuralist system without a full spelling out of the disequilibrium or destabilizing implications of structuralist rigidities in a world in which a powerful private sector orders its own economic priorities in the light of changes in prices or profits calculated in monetary terms. It is possible to classify the structuralist models of growth and instability as applied to less developed countries into two classes, those which take basically a closed or at least the behaviour o f an ‘insular* economy1 as their explanandum, and those which take the explanation of the functioning of an open economy as the desired objective. Within the first category, two sub-classes can be distin­ guished. The first sub-class consists of models which are growth oriented or growth driven. The Mahalanobis model, the two-gap models of trade, aid and growth, Rudra (1964) and Patnaik (1972) are members of this sub-class. In another sub-class would belong models or constructs which seek primarily to explain inflation, excess capacity or unemployment in the short period. Mitra (1977), Patnaik, Rao and Sanyal (1976), Rakshit (1982), Bose (1989), Bose (1993) and other papers which seek to explore different aspects of the so-called Keynesian models of the dual economy, can be clubbed together in this sub-class. It also includes most of the writings of the Latin American structuralist school (Seers, 1962; Olivera, 1964) and the avowedly structuralist macroeconomics of Taylor (1981, 1983). Unlike most of the Indian work in this genre, the latter explicitly take the openness of the economy as a basic feature. Interestingly enough, while labour market rigidities and their 1 McKinnon (1981) defined an 'insular* economy as one in which (he government can effectively insulate the economy against external disturbances through appropriate fiscal and monetary polices. In particular, international capital movements that can upset all calculations in the domestic economy are ruled out. Such movements are taken to be either on official account or as accommodating responses to imbalances in current account.

Closed-economy Structuralist Models 91 explanation have played a major role in the emergence of ‘New Keynesian Economics* in the context of developed market economies, there has been little application of that kind of structuralism to the explanation of the behaviour of less developed economies. In the case of the latter, labour market rigidities have figured almost exclusively as features to be somehow ironed out rather than explained even as a preliminary to severe clinical attention. Corresponding to the shortrun open economy models, there could theoretically also be growth driven, or growth oriented, structuralist models of open economies. But while some attempts have been made to construct models of economies constrained by exogenously given export earnings (Raj and Sen, 1961; Chenery and Bruno, 1962) models of global economic growth, and models of growth of two economies linked by trade, structuralist models of growth of open economies are still rather thin. The pioneering work of Kaldor linking economies of scale and economic growth has been followed up by the more recent work of Paul Krugman, Paul Romer, Avinash Dixit and Robert Lucas in this area, but I will not try to bring out their import for the growth of developing economies in this paper. Finally, of course, there are models of working of less developed economies which are not structuralist in the usual sense. No serious attempt has been made (fortunately) to fit these economies into the , mould of the so-called new classical macroeconomics of the LucasSargent variety. However, there are models which put money centre stage and trace the implications of the existence of a universally acceptable asset which acts as the price leader in all investment decisions and, hence, as a critical, if volatile, link between the present and the future for all economic agents. Such models must also take the formation of expectations seriously. But in order to explain why economies do not explode or contract at very high speeds, money must also be linked to the institutions which give particular shapes to expectations and operate under certain rules of the game which have been evolved along particular paths traversed by these economies. Thus, history must enter integrally into the analysis of economic behaviour. It is welcome that hysteresis and path-dependence have now become part of the discourse of macroeconomists. Institutions and history are generally fraught with conflict of expectations, over shares of income, and over retaining or changing the rules of the game. Hence, conflict theories of inflation or stagnation have also entered the picture (see, for example, Rowthom,

92

A m iy a K u m a r B a g ch i

1977 in the context of developed capitalist economies and Patnaik, 1988 in the context of less developed countries; see also Skott, 1989). We will, however, leave these out in our sketch of more narrowly ‘structuralist* models.

3. SECTORAL IMBALANCES IN THE GROWTH PROCESS AND THE DYNAMICS OF INFLATION IN A CLOSED ECONOMY Rudra (1964, 1972) grew out of Ashok Rudra’s dissatisfaction with the failure of Manne and Rudra (1965) to deal with the consumption loop of a dynamic consistency model of growth and the implications of differing rates of growth of different sectors of the economy. Manne and Rudra (1965) made strategic use of the finding that the Indian input-output table of the early sixties could be rearranged so as to assume an almost block-triangular form. This meant that the subsec­ tors of the economy could be partitioned into three blocks such that there was one block (the ‘universal intermediates*) which provided current inputs to the two other blocks* i.e. ‘agriculture* and agro-based industries on the one hand, and ‘industry* comprising mainly mining, metals and machinery making industries on the other. But these last two blocks did not have any transactions with each other in terms of the provision of current inputs. Given this structure, it was concluded that agriculture and industry could virtually grow independently of each other. The rate of growth of universal intermediates would be a constraint on the growth of the other two sectors, but in respect of current input requirements, industry (except for agro-based sub­ sectors) could grow at a fast rate while agriculture virtually stagnated. It was recognized that slow agricultural growth would mean that reasonable targets of growth in consumption would not be attained and that limitations on the purchasing power of people engaged in agriculture would create demand deficiency problems for industries producing consumer goods. But the exploration of the consequences of these imbalances was not considered to be within the purview of the model. Rudra (1964) sought to plug this loophole by fitting a linear expenditure function on to the Manne-Rudra framework of a dynamic input-output model. He also allowed the foreign trade sector to act as a buffer and worked out the implications of different rates of growth

Closed-economy Structuralist Models 93 of different sectors for increases in imports and, hence, for deficits in external trade. In working out the latter he implicitly took the exchange rates and total exports to be given. In this he was generally following the practice of the builders of two-gap models of trade and aid. But he was a pioneer in his attempt to work out the implications for changes in relative prices of different rates of growth of agriculture and industry. At the time he worked on his model, the fourth five year plan (which was never put in place in actuality) was on the anvil in official circles. Among the targets that were being considered were a rate of growth of 12 per cent per year in industry and 5 per cent per year in agriculture. According to Rudra’s calculation if these rates of growth were actually realized then over the five year period of the contemplated fourth plan, that is, between 1966-7 and. 1970-1 the prices of foodgrains in relation to those of non-food products would rise by 109 per cent, and the prices of food products other than grain relatively to those of non-food products would rise by 274 per cent. The inflationary potential of unbalanced growth of agriculture and industry was, thus, starkly revealed. Rudra’s model was not a macroeconomic model of the actual working of the Indian economy. But as he insisted, it was not just a ‘demonstration model' either. What the model sought to do was to portray the likely consequences of stepping up investment in industry without worrying about the rate of expansion of the agricultural sector. Patnaik (1972) built a structuralist model in which exogenously determined rates of agricultural growth act as a triggering mechanism for generating cycles in the economy. These cycles are characterized by an inflationary upswing, a fall in the capacity utilization of the industrial sector, and a deflationary downswing. The fall in the capacity utilization of the industrial sector is caused by (a) a fall in the demand for industrial goods on the part of wage earners whose real wages are squeezed by the rise in prices of agricultural goods, (b) a fall in public expenditure caused by the government’s sensitivity to an unacceptable rate of inflation, and (c) a fall in private investment induced by a decline in capacity utilization and the associated decline in profitability. The structuralist features of this model are the exogenously given output or rate of growth of agriculture, the fixed money wages of labour and the cost-plus pricing rule followed by the industrial sector. The special feature of the model which distinguishes it from most other structuralist models is the recognition of the government as a

94

A m iy a K u m a r B a g c h i

major spender in the economy and its sensitivity to threats of inflation. There is no specific role of organized labour in the model seeking to defend some incompressible real Wage or adjusting money wages in step with inflation or with growth in incomes in other sectors of the economy. But the government in effect acts as a protector of a minimum real wage, not so much because of its concern for the welfare of organized sector workers as because of its anxiety, let us say, not to allow the urban cost of living to go beyond certain limits. The reaction of the private businessman to a drop in capacity utilization caused on the one hand, by an erosion in the workers* purchasing power and the resulting trimming of their demand for manufactured wage goods and by the slashing of government expen­ diture on the other also acts as a brake both on employment in the industrial sector and on the fall in the real wages of industrial labour. It can be seen that although Patnaik (1972) is couched in terms of the calculus of differential rates of growth of agriculture and industry and the disproportionality crises resulting therefrom, it can also be treated as a short-term model of instability. This change in interpreta­ tion is especially plausible because Patnaik does not explicitly work out the impact of a rise (or decline) in prices of agricultural goods on the investment behaviour of private businessmen. There are, in fact, several difficulties in rationalizing long-term, stable models of capitalist growth, or stable growth paths, the deviations from which are supposed to trigger short-run fluctuations but are not allowed to affect the long-term growth path itself. One problem emerges out of how much capitalists are supposed to foresee these patterns of fluctuations. In the Patnaik model, economic agents adapt passively to changes in relative prices or levels of effective demand that occur because of exogenous shocks originating in agriculture. If, however, businessmen foresee some of these develop­ ments, they may resort to speculation or intertemporal and intersec­ toral arbitrage in the short run, and in the longer term, they may invest more in producing the commodity with a higher rate of expected return. Secondly, the behaviour of capitalists in substituting between different types of assets, and in particular, between money and other assets is left out of most growth models. Attempts were made by Tobin (1955,1965) and following him, by Johnson (1966,1967) and by Levhari and Patinkin (1968) to incorporate money in a neoclassical growth model. The strategy for incorporating money was to treat all

Closed-economy Structuralist Models 95 of it as outside money, created as a result of governmental operations, or to assume that the gains and losses of debtors and borrowers within the private sector balance exactly, so that the effects of 'inside money* or credit can be ignored. Then the attribute of money as net wealth was taken into account in influencing the saving or consumer behaviour of the private sector, and in governing choices between money and other assets. Then results were derived to see whether, between different economies with different combinations of physical capital and monetary assets, saving behaviour, rates of price change, etc. would differ systematically, with the assumption that such differences would be permanent. Such comparisons were supposed to yield comparative dynamic results. Some stability analysis was also carried out by finding out whether a permanent change in preferences, saving behaviour, or the stocks of physical and monetary assets would permanently change the characteristics of the long-term growth path. However, all these results were dependent on the assumption that at any moment all the economic agents know all their behavioural attributes, and all the agents or productive factors they deal with are given for all future time. This assumption would have to be made, in the case of stability analysis, from the moment a change has occurred in any of the relevant variables or behaviour or production functions. Even if this change has occurred, everybody would have to believe that no other change will ever occur again. Not only are many of these assumptions rather implausible; it is also difficult to find room in such economics for money as a real store of value and hedge against uncertainty. Money in such an economy then becomes simply an instrument created arbitrarily by that deus ex machina, the government. For all these reasons, it is highly problematic to theorize stable long-term growth paths to any degree of plausibility in a capitalist economy, in which asset-holding decisions profoundly affect invest­ ment behaviour and, hence, productivity and growth. When it comes to the growth of a less developed economy in an era of unrestricted capital flows across national borders, any credible story of its growth must start with a clear outlining of the way in which it is inserted into the global economy. The description of its insertion or location must include among other things, a clear conceptualization of the invest­ ment and competitive behaviour of its capitalist class, and the degree to which its own institutions (which include the government) can

96

A m iy a K u m a r B a g c h i

support the competitive strategies of that class, when such strategies are worth talking about in the first place (Bagchi, 1988; Porter, 1990). Money and credit have to be treated as strategic variables in outlining such a story. It is difficult to see how they can be handled by any of the standard models of growth.

4. STRUCTURALIST MODELS OF SHORT-RUN DISEQUILIBRIUM The main stream of neoclassical economics did not admit the pos­ sibility of persistent unemployment or persistent under-capacity working of the economy. If there was disequilibrium in any market, prices would ultimately adjust to clear it. If this did not happen, then these must be either due to market failures in the sense that appropriate markets had not developed in particular cases, or due to rigidities in the system. The policy-makers’ task would be to complete those markets or to remove those rigidities. Although there was some fleeting recognition that unexhausted economies of scale, incomplete information, or the asymmetric distribution of information among different economic agents, and the failure of economic agents to anticipate the actions of other agents with some control over prices or other variables could lead to the jamming of market processes and to coordination failures, theoretical work of an acceptable degree of rigour2 did not emerge in this area until the end of the sixties and the seventies. The fact that coordination failures lay at the heart of Keynes* General Theory or of Marx’s notion of the anarchy of capitalism did not carry much weight with academic economists until they had verified Keynes’ and Marx’s findings by using tools that the latter day theorists considered to be sophisticated enough. Hence, it would be true to say that most of the mainstream literature on macroeconomics before the seventies was devoted to the explora­ tion of various supposed rigidities interfering with the smooth work­ ing of Walrasian general equilibrium and thereby producing disequilibria characterized by unemployment. Since rigidities figured prominently in these constructs, they can all be regarded, in a sense, 2 Radner ( 1968), Akerlof ( 1970), Diamond ( 1982) and Weitzman ( 1982) arc some of the important landmarks in the reconsideration of mainstream economics in the light of uncertainty, the asymmetric nature of the distribution of information, the requirement of density of market interchanges and the presence of economies of scale.

Closed-economy Structuralist Models 97 as structuralist models. We will not try to summarize even a fraction o f this literature, but will instead refer to certain landmarks. All these take the institutional features of advanced capitalist economies as the basis for their assumptions. A distinguishing mark of the structuralist models of inflation in contrast to pure monetarist or early Keynesian models (following Keynes, 1940) is that they make strategic use of disaggregation of the economy along lines of goods markets displaying different kinds of supply, demand and price behaviour, or along lines of goods, factor, and asset markets with differing responses to emerging excess demands or excess supplies. The distinction between agricultural and industrial goods in terms of the competitive rules under which they operate goes back to Kalecki (1943). This distinction has figured very prominently in the Indian structuralist models and in the earlier Latin American structuralist models.3 The distinction between factor and goods markets as the key to the understanding of inflationary phenomena was put forward by Hansen (1951). Hansen distinguished between the state of excess demand or excess supply in the factor market, meaning mainly the market for labour and the state of excess demand or supply in the market for goods, and called these the ‘factor gap* and ‘the goods gap’ respec­ tively. Hansen's analysis was useful in drawing attention to the fact that in an inflationary situation, the rates of price rise in goods and factor markets might be different, and, therefore, make differing contributions to the eventual moderation of price rise or to its acceleration (see in this connection, Bronfenbrenner and Holzman, 1965). However, the question was left open as to whether inflation could persist even if there was excess supply in the factor markets as defined in ordinary partial or general equilibrium analysis. It took 3 The word ‘structuralism’ came into vogue in economics following the work of Osvaldo Sunkel and Celso Furtado, which in turn was affiliated to the general ECLA school, following in the footsteps of Raul Prebisch and Hans Singer, emphasizing the need for structural change in less developed countries on the one hand, and structural barriers (meaning, institutional rigidities and sectional conflicts) against industrializa­ tion on the other. The Latin American structuralist school was introduced to English speaking economists by Seers (1962; see also Seers, 1964 and Olivera, 1964). While the Latin American structuralist school has carried on a rich tradition of discussion couched in the language of institutions, class or sectoral conflicts or dependency, the analytical building blocks of their theoretical edifices were not always clearly delineated. This is one reason I have confined my summary primarily to Indian work.

98

A m iy a K u m a r B a g ch i

economic theorists a long time to accept that this could happen, and that this was not entirely due to the irrationality or wickedness of trade unionists pushing forward wage claims despite the prevalence of widespread unemployment. Trade unionists did not possess all the information about the way prices were changing, and could not possibly work out the real value of their pay packet. Moreover, there was no unique price index by which the money wage could be deflated so as to reflect losses or gains for members of the relevant group of wage-eamers. Moreover, in a world in which prices were fixed by firms with perceptible market power, the managers or owners of firms could always remain one step ahead of wage-eamers by fixing prices so as to absorb the money wage claims made by the wage earners. In this world of costly and asymmetrically distributed information, and of economic agents with different degrees of purchasing power, it made sense for unions to bargain primarily in terms of money wages rather than some ideal index of real wages (see, in this connection, Sheffrin, 1989). Furthermore, the labour market is naturally segmented by skills, sectoral peculiarities, length and nature of experience, location, and a host of other characteristics. In such segmented markets, it is natural for wage earners to look over their shoulder and argue their case in terms of fairness, and keeping in view their position relative to other wage earners in the same industry, or the same or similar job categories. This is one of the reasons why Keynes stressed wage bargaining in terms of nominal or money wages rather than in terms of real wages. The appropriate index number for deflating nominal wages would vary according to the position of the workers in the scale, and according to their geographical location (Keynes, 1936; Solow, 1990). The above arguments would apply to wage bargaining in any economy governed wholly or partially by the capitalist rules of the game, and that means it could include most of the less developed countries as well. But there is another set of arguments which runs in terms of the firms* requirement of a stable, experienced and specialized labour force. Their applicability to the less developed countries, with dense populations and abundant labour supplies, is less convincing than in the case of advanced capitalist countries, but they are not entirely dismissible either. All these arguments have been sometimes clubbed together as the ‘efficiency theory* of wages, but

Closed-economy Structuralist Models 99 several strands can be distinguished between them (Akerlof and Yellen, 1986). ' The first strand would stress the need of a modem, relatively capital-intensive enterprise for a stable labour force not given to absenteeism, and amenable to factory discipline. In order to eliminate these transaction costs, and inculcate a spirit of team work, an enterprise would be willing to pay wages that are higher than are earned by labour which is hired on a daily basis. Secondly, some of the operations would be learned only on the shop-floor. Once an operative has acquired such on-the-job skills, the firm may be willing to pay him (her) higher wages than somebody with no skills or even somebody who has been trained in a formal institution. The reason for this last type of preference is that the firm may trust its own screening procedure more than the screening procedure of more ‘bookish* institutions. Thirdly, there are particular ways of doing things, and linking up different operations in different firms, so that the pattern of division of tasks may be idiosyncratic to a particular firm. Such idiosyncratic skills may fetch a higher premium if the controllers of firms perceive them as adding more net value than apparently comparable skills acquired in other firms or through formal training. Finally, when skills and team work are perceived to be cumulative the firm may put a premium on seniority within its work force, and the firm and the senior employees enter into contractual relationships of a long term nature even if those contracts may not actually be written down (this is ArthurOkun’s ‘invisible handshake*; Okun, 1981). For all these reasons, in any private enterprise economy with factories and other enterprises employing a labour force on a more or less permanent basis, there would be a tendency for money wages and (at some distance) real wages too to display rigidity in a downward direction. The resistance of the wages of the more skilled workforce would also be transmitted, at least partially, to the resis­ tance of the less skilled workers, because of trade union solidarity, and where trade unions are absent, through the struggle to maintain some kind of stability of skill-related wage differentials. The ‘structuralism* of the macroeconomics of less developed countries has, however, not been couched primarily in terms of the failure o f labour markets to behave in accordance with the rules of a Walsarian auction market but in terms of the differences in market behaviour as between the agricultural and industrial sectors.

100

A m iy a K u m a r B a g ch i

5. THE AGRICULTURE-INDUSTRY CONTRAST AND PROBLEMS OF EFFECTIVE DEMAND AND INFLATION Almost all the structuralist models which have been developed for India and for other less developed countries start with the Kaleckian distinction of pricing (and quantity) behaviour between industry and agriculture. The supply of agricultural products, and especially of food, is taken to be either completely exogenous to the model or determined according to competitive rules of the game. Even in the latter case, however, if the price elasticity of supply is allowed to be positive, it plays only a small role in the short run, for by the nature of the agricultural production process, the supply cannot respond strongly to changes in price. (However, price changes may be allowed to affect the rate of release of stocks and, hence, the effective supply in some models). Given such an inelastic supply schedule in the short run, it is the demand for agricultural products that then determines their price. In the industrial sector, however, it is assumed that imperfect competition prevails. Following the Kaleckian simplification, it is then assumed that every industrial firm possesses some degree of monopoly power, and it uses that power to charge a mark-up over costs. This assumption can be easily rationalized whenever there is a single dominant firm that can ignore the reactions of other firms in setting its prices. It can also be rationalized quite easily in an industry with a few dominant firms which explicitly or implicitly follow certain market-sharing rules. A third case in which the assumption acquires plausibility is the one in which there is a recognized price leader in the industry. The assumption of stable mark-up pricing rules becomes much more tricky in situations in which there is monopolistic competition between a large number of firms. In this case, the demand curve of every firm is only a conjectural entity that can be derived, if at all, only on the basis of certain stable (and usually realized) expectations regarding the behaviour of other firms in the neighbour­ hood (the ‘neighbourhood’ being defined in terms of physical or locational nearness, nearness in terms of product quality, or nearness in terms of market penetration). It may be claimed that within a Marshallian short period (during which the effective capacity of the firms remains unchanged) it is reasonable to assume that the com­ petitive structure of the different industries will remain unchanged,

Closed-economy Structuralist Models

101

and in combination with an assumption about the constancy of effec­ tive capacity, a stable average mark-up for industrial products can be rationalized. However, for the purpose of construction of structuralist models, it is necessary to assume that the mark-up remains unchanged over an agricultural production cycle. In cases in which such produc­ tion cycles are longer than the Marshallian short period, the assump­ tion regarding the stability of the mark-up becomes much more debatable. In order to illustrate the way such ‘dualistic’ structuralist models work, we use an elegant model set out by Bose (1989). Bose first distinguishes between profit-eamers and wage-eamers. Profit-eamers in his model include big farmers, traders and industrialists. Following the Marx-Bortkiewicz-Kalecki tradition, Bose shows that if luxuries are consumed entirely by the rich who are also the investors in the economy, then the level of investment determines the total amount of profits, and profits determine the amount of luxuries consumed and produced. In fact, given the level of investment, the total amount o f income in the luxury-cum-investment sector is also determined, so that the aggregate money value of wages in this sector is also known. However, in the particular class of dualistic models considered, the wage-good has two components, food and, say, clothing. These two goods are produced in two sectors of the economy with different supply conditions and different price-formation rules or mechanisms. It is the significance of this difference that is sought to be brought out in Bose’s model. After having eliminated luxury consumption as a causal factor in the story, Bose also proceeds to get rid of the consumption of food by the rich as well as the poor in the agricultural sector, leaving only an agricultural surplus entering into urban consumption as a variable influencing prices and money and real incomes in the system. The total profit of the rich in the rural sector is calculated as the value of the total marketed supply of food less the payment to small farmers at a fixed price for a fixed amount procured from them. In this model, the money wage rate of the workers in the industrial sector (which is the whole of the non-agricultural sector as well) is taken to be given. The real wage rate naturally changes with changes in prices of food and cloth, the two commodities consumed by workers. But neither the money nor the real wage of agricultural labour is specified in the model. If there is a Lewisian condition which operates, viz. that the real wage in the industrial sector must not fall

102

A m iy a K u m a r B a g c h i

below a certain subsistence level determined by the earnings ob­ tainable in the agricultural sector, then that will affect the dynamics of the economy when a rise in prices threatens to depress the industrial sector real wage below this floor level. With a slight change in notation, the basic model of Bose (1989) can be wriiten as: (1) (2)

(3)

(4)

X = X: the agricultural surplus, supposed to be constant. Dx = A + g (Px , W), the urban demand function for food, where A is a constant denoting the (fixed) aggregate food consumption of the urban rich, Px is the price of food, and W is aggregate money wages and g(-) is a function with g x < o , g2 > o (food demand is assumed to be price inelastic). Ey - B + h (Px , IV), the expenditure function for clothing, where B is a constant (denoting the fixed aggregate clothing expendi­ ture of the rural poor), and /((•) is the expenditure function of wage-eamers for clothing with h x < o, o < h2 < 1. Sy = k W y , where A: is a constant such that k > 1, Wy the money wage-bill in the clothing industry and, hence, Sy is the value of supply of the clothing industry for any wage-bill Wy

A is the aggregate amount of the demand of the rich for food, and B is the aggregate value of the demand of the poor for clothing. Thus variations in prices of clothing and food can affect A and B also. But initially, we can ignore this complication. The total wage bill, W, is split into two parts, Wf which is paid out to workers in the investment and luxury goods sectors, and is as autonomous as investment itself, and Wv which is paid out in the sector producing clothing. By equating (1) and (2), and varying W, we can get the locus of equilibria of Px and W in the food market denoted by XXx (Figure 2). Similarly, by equating (3) and (4), for each pair of W and Px we can get the locus of equilibria in the clothing market, denoted by YYV XXx is upward sloping with the assumption that g x ^ o, and g2 > o. YYXis downward sloping since h x < ot and o < < 1 < *• Bose then carries out a number of comparative static exercises. In particular, he finds that an increase in Wt that is in employment in the investment and luxury sectors together or in other words, in autonomous expenditure, will raise food prices. Because of the fixity of money wages in the clothing sector, this will mean a decline in real wages. Hence, with inelastic demand for food in real terms, this will lead to a decline in effective demand for clothing and, therefore, to a decline in output and employment in the clothing sector. In an

Closed-economy Structuralist Models

103

F igure 2

extreme case, in which the elasticity of demand for food is zero, every rise in W will be matched by a corresponding decline in clothing sector employment, leaving total employment in the non-agricultural sector unchanged. If we take into account the fact that a rise in Px will also mean a rise in the incomes of those poor peasants who sell their output through traders at some price, say, p lx , then the result may be modified. However, if the elasticity of demand for clothing is also rather low at low levels of income, a rise in income of the rural poor may lead them to consume more food, thus, depressing the total marketed surplus X, and raising Px further and, hence, again cutting down employment in the clothing sector. Bose (1989) establishes that the condition for a rise in the price received by poor farmers (say, from the government carrying out a price maintenance policy) to raise total employment in the industrial sector is the MarshalI-Lemer-Robinson condition for stability in the foreign-exchange market, namely that the sum of elasticities of demand for the two goods involved, namely food and clothing, must be more than unity. Bose's prototype model can obviously be extended and modified

104

A m iy a K u m a r B a g c h i

to cover numerous stylized facts. One outcome it can handle easily is the case in which the aggregate price level rises as a result primarily of a rise in food prices, and employment in the industrial sector falls. The further consequence of that development will depend on assump^ tions about the behaviour of capitalists controlling the industrial sector. A fall in clothing employment will naturally lead to a fall in profitability. This may trigger a fall in investment and hence in autonomous employment. However, if capitalists assume that the rise in food price or the fall in employment is only temporary and will be reversed soon, this may not happen, and investment may rise accord­ ing to earlier plans. If the expectations about the reversal of price rise is realized then the inflation rate will come down, and ‘structural’ differences will not upset the orthodox story. However, differences in the behaviour of different types of income earners, differences in the pricing behaviour of different commodities, the resistance of wage-eamers to real income erosion, government policies supporting prices in the agricultural sector or the credit of owners and managers of land and capital, can repeatedly lead to situations in which price rises, sectoral or aggregate falls in employ­ ment, and growth or fall in real incomes in the agricultural and industrial sectors together or separately can occur in a bewildering array of combinations. I will not try to summarize the results of a number of models which have taken some of these behavioural and structural assumptions as their building blocks. An early exercise by Patnaik, Rao and Sanyal (1976) tried to controvert the usual view that an increase in food prices, however brought about, would ultimately lead to a decline in the real demand for food and, hence, supply its own corrective. The triggering mechanism for price instability was the spending behaviour of landlords and the characteristics of the goods or services they spent their money on. If the profits of increased food prices accrued primarily to landlords whose marginal propensity to consume was high and if they spent on goods and services which were highly labour-intensive (such as the hiring of more retainers for the household), then the effect of a rise in landlord profit could be to raise the demand for food further and, hence, accelerate the inflation in food prices. If money wages in the industrial sector move up more or less proportionately with food prices, and if the industrialists follow a mark-up pricing rule, then food price inflation will transmit itself to a rise in other prices.

Closed-economy Structuralist Models

105

This story will be modified if the government follows a policy of procuring food from the farmers at a declared procurement price and combines it with a policy of releasing it through the ration shops. If the procurement price is higher than the prevailing market price, then the act of procurement itself will give an upward boost to the food prices and have a depressing effect on industrial employment. If, however, the government releases the food through ration shops to industrial workers at a subsidized price, and the amount so released is larger than would otherwise be available in a laissez faire market for food, then there will be a positive expansionary effect (Das, 1989). Furthermore, if the government operations in the market for agricul­ tural crops are combined with a policy of building up agricultural infrastructure and selling intermediate inputs such as fertilizers at subsidized prices, then agricultural output will show an upward trend. If this trend rate is high enough, and if the government can credibly release larger food stocks through ration shops over time, then the speculative pressure on food prices as a result of sudden harvest deficiencies may also decline. However, all these government operations increase private sector liquidity. Growth of industrial output is also backed by credit expansion, often supported by the government itself acting to expand the monetary base through the printing of notes to defray budget deficits. In this case, the story of movements in food prices and food stocks has to be linked up with questions of the rate of expansion of the monetary base, accrual of liquidity to the private sector, and resulting expansion in monetary demand giving rise to inflation and output growth in different proportions. (Theoretically, it can be a process of pure inflation with a zero or even negative output growth. But such a sequence is almost certain to lead to hyperinflation). SanyaU Mukherjee and Patnaik (1989) have analysed several sequen­ ces in which access to credit can act as a constraint on the stock building activities of both agriculture and industry, and on the investment activities of the private sector. Thecredit limit is fixed by the total amount of outside money created through government operations earlier and by the amount of money creation in the current period, and by certain parameters of allocation of credit between different activities. (Sanyal, et al., assume, for simplification, that government expenditures are entirely financed by the printing of currency notes; but the main results would go through with a more general specification). An important upshot of this kind of formulation

106

A m iy a K u m a r B a g c h i

is that private investment and stock-building activities may easily be constrained by the credit limit acting along with the allocational parameters, and thereby influence the effective demand for the industrial sector. An interesting part of this exercise is the assumption that in­ dustrialists regularly expect their demand curves to be shifted outward by the government engaging in money creation, adjusting agricultural prices upward as costs and prices go up, and effectively increasing the credit limits of the industrial sector for stock-building as well as investment activities. The political economy of this set of activities is that all the groups benefit from such an upward revision of expectations without inflationary forces getting out of hand provided that the industrialists know that there is a limit beyond which they cannot raise their prices in any single period. However, this happy inflationary scenario can be disturbed badly once the government fails to live up to the expectations of the major players or once one of the major player groups decides that it can gain more by stepping out of line. With the above sequence, we have already moved out of the confines of one-period structuralist models. In fact, the analysis of any sequence of changes in output, prices or capacity utilization cannot be conducted just by putting together a series of single-period models. In the structuralist models, the workers* demands are con­ strained in several ways: they spend all their income on consumption, a major part of their income is spent on food and their demand for food is inelastic with respect to price changes. This is why I had called the short-period demand problem in a less developed economy a poverty-related demand problem (Bagchi, 1988). This poverty-related demand problem remains valid so long as this kind of poverty endures. However, in their investment, production and stock-building decisions, the capitalists have to tackle yet another kind of problem, which I called the medium-term demand problem. This arises from the continual shifting of consumption patterns of the top 10 to 15 per cent income earners because of differing Engel elasticities of different goods, changes in tastes, often in imitation of affluent lifestyles abroad, technological changes in materials entering into consumption (such as polyester fibres entering into clothing), and variations in the numbers who can afford newer types of consumer goods (Bagchi, 1988). Because of all these variations, there will occur unanticipated changes in capacity utilization in existing industries, and opportunities

Closed-economy Structuralist Models

107

for earning quick profits through stock-building, assembly, finishing or investment activities. This kind of medium-term structuralism can be linked with the medium-term financial scenarios outlined by Sanyal et al. (1989) to produce richer stories of short- and mediumterm fluctuations in a closed economy with a multiplicity of structural constraints.

6. CONCLUDING REMARKS: OPENING OUT IN SPACE AND STRETCHING OUT IN TIME The early structuralist models of growth such as the Mahalanobis two-sector or four-sector models or the Raj-Sen model (Raj and Sen, 1961) assumed the economy to be closed or they treated the foreign trade sector to be essentially exogenous. In particular, exports or the supply of foreign exchange through aid and trade were taken to be exogenously given and to be unaffected by the working of the model itself. Some time later, two-gap models of trade and aid built by Chenery and his associates (Chenery and Bruno, 1962; Chenery and MacEwan, 1966; Chenery and Strout, 1966), explicitly incorporated structuralist features. The levels and time patterns of export earnings were taken to be exogenously given as were the import requirements of production, or the endowment of skills (in the phase in which the development of the economy was supposed to be constrained by inadequate supplies of skills). The structure of these and similar models was soon criticised on theoretical and political economy grounds. In the Chenery-MacEwan or Chenery-Strout formulations, foreign aid was supposed to be available to cover the larger of the two gaps— viz. (a) the deficiency of domestic saving in relation to domestic investment, and (b) the gap between the foreign exchange needed to support a programme of development and the actual qet foreign-exchange earnings. It was pointed out by several economists (Vanek, 1967; Fei and Ranis, 1968; Bagchi, 1970a) that such an assumption was highly unrealistic and that furthermore, the adjust­ ment of the economy to either assured flows of foreign aid or failure o f the promised aid to materialize, would knock it off the planned path of development, so that the elaborate calculations of various gaps could be not only irrelevant but positively misleading. An elaborate critique of such models would be highly anachronistic in a world in which foreign aid has become one of the dirtiest words

10 8

A m iy a K u m a r B a g ch i

in the economists* vocabulary. The trend now has shifted towards construction of optimal or strategic models of foreign borrowing. Such models seem to have little to do with the solution of problems of countries caught for years in a foreign debt trap. At the same time, the need for diagnostic models of adjustment of open less developed economies to exogenous shocks or to imbalances and rigid structures of demand, supply or import dependence is much greater than ‘it has ever been’. I will not try here even to summarize the theoretical frameworks of open underdeveloped economies that have been advanced by a number of economists from Latin America, USA. and elsewhere (Taylor, 1983 contains a bibliography of most o f the important references in this area). The structural constraints of an open, underdeveloped economy may be located in the foreign trade area itself. The M arshall-LemerRobinson condition for the stability of the foreign-exchange market (namely, that the sum of the price elasticities of home demand for the foreign good and of foreign demand for the home good should be at least unity) may be violated (for a survey of the literature see Ray, 1990). Such a violation is, however, also linked to domestic structural rigidities. The foreign demand for the home good may be rather price-inelastic because the economy produces goods with low income elasticities of demand which cater to poorer people. The low price-elasticity o f home demand for the foreign good generally arises out of the inability of the home economy to produce import substitutes at a competitive price. If the Marshall-Lemer-Robinson condition is augmented as it should be, with the two elasticities of supply of exports and imports, then the failure of the foreign exchange market to clear may be found to be due to the low elasticity of supply of the export good of the underdeveloped economy. When the exportable of the particular country concerned is also a basic consumer good (such as food) then devaluation, and the accompanying rise in the domestic price of the exportable may trigger an inflationary spiral. Or, in the polar case, when the major importable is a basic consumer good, then also similar inflationary consequences may follow from an act of devaluation. These illustrations are enough to show that even if we neglect the consequences of capital mobility and changes in the portfolio of assets of domestic and foreign firms, as between titles to domestic capital assets and foreign capital assets, an adequate analysis of the open economy problems would require the stretching of the time period of

Closed-economy Structuralist Models

109

analysis beyond the short period of the typical non-Walrasian dis­ equilibrium model (see, for an array of such models, Benassy, 1986). For example, the analysis of the rigidities of supplies of exportables would require an investigation of the proximate causes o f such rigidities. Or the analysis of the non-decreasing dependence of the underdeveloped economy on imports would require an analysis of the causes of underinvestment or unsustainable, costly investment in import substitutes and the reasons for the failure of the economy to absorb importable technologies quickly and efficiently and to diffuse the appropriate techniques widely. When we take the capital mobility and asset-shifting behaviour of capitalists seriously, we realize that an investigation of the sizes of the supply and demand elasticities alone can throw but a murky light on the reasons for the balance of payments disequilibria of the less developed economies. We have to examine the degree of autonomy of the local capitalist class, the strategic choices of transnational corporations actually engaged in, or likely to enter the space of the domestic economy, the place of the economy in the (generally politically determined) credit rating of the transnational banks and other funding agencies, and the degree of interference in the policies of the country by the IMF, World Bank, the USAID and other multilateral agencies controlled by the G -7 countries. The borders of macroeconomics and political economy then become increasingly blurred. While rigorous analysis of carefully specified models will continue to provide the backbone of structuralist or more widely, new Keynesian or new Kaleckian macroeconomics, they cannot obviate the need for situating such models in a wider political economy framework.

REFEREN CES G.A. (1970),4The Market for “Lemons”: Quality Uncertainty and the Market Mechanism’, Quarterly Journal o f Economics, August, 84,488-500. A kerlof , G.A. and J. Y ellen (eds.) (1986), Efficiency Wage Models of the Labour Market, Cambridge University Press, Cambridge. B agchi, A.K. (1970a), *Aid Models and Inflows of Foreign Aid*, Economic and Political Weekly, Annual Number, January, 5, 3-5. ------(1970b), ‘Long-term Constraints on India’s Industrial Growth A k e r lo f ,

110

A m iy a K u m a r B a g c h i

1950-1968*, in E.A.G. Robinson and M. Kidron (eds.), Economic Development in South Asia, Macmillan, London, 170-92. ------(1988), ‘Problems of Effective Demand and Contradictions of Planning in India*, in A.K. Bagchi (ed.). Economy, Society and Polity: Essays in the Political Economy o f Indian Planning, Oxford University Press, Calcutta, 227-66 B enassy , J.P. (1986), Macro-Economics: An Application to the Non- Wal­ rasian Approach, Academic Press, San Diego. B o se , A. (1989), ‘Short Period Equilibrium in a Less Developed Economy’, in Rakshit, 1989, 26-41. ------(1993), ‘Price-income fluctuation and agricultural shocks in a semi­ industrialized economy*, in B. Butta et al. (eds.) Theoretical Issues in Development Economics, Oxford University Press, New Delhi. B ronfenbrenner , M. and F.D. H olzman (1965), ‘A Survey of Inflation Theory*, in Surveys o f Economic Theory, Money Interest and Wel­ fare, Macmillan, London, 1,46-107. C henery , H.B. (1975), ‘The Structuralist Approach to Development Policy*, American Economic Review, 65 (2), May, 310-16. C henery , H .B . and M. B runo (1962), ‘Development Alternatives in an Open Economy: The Case of Israel', Economic Journal, March, 72, 79-103. C henery , H.B. and A. M ac E wan (1966), ‘Optimal Patterns of Growth and Aid: The Case of Pakistan*, Pakistan Development Review, Summer. C henery , H.B. and A. S t r o u t (1 9 6 6 ), ‘Foreign Assistance and Economic Development’, American Economic Review, 5 6 (4), September, 6 7 9 -7 3 3 . D a s , C. (1989),

‘Food Policy in

a D ual

Economy*, in Rakshit, 1989,

66-93. D iamond , P.A. (1982), ‘Aggregate-demand Management in Search Equilibrium', Journal o f Political Economy, 10, 881-94. D om ar , E.D. (1957), ‘A Soviet Model of Growth*, in E.D. Domar, Essays in the Theory o f Economic Growth, Oxford University Press, New York. Fei, J.C.H. and G. R anis (1968), ‘Foreign Assistance and Economic Development: Comment*, American Economic Review, 58 (4), Sep­ tember, 897-912. F eldman , G.A. (1928), ‘On the Theory of the Rates of Growth of the National Income, Planovoe Khozyaistvo (in Russian), 11 and 12;

Closed-economy Structuralist Models

111

English translation in N. Spulbcr (ed.), Foundations o f Soviet Strategy fo r Economic Growth, Indiana University Press, Bloomington, 1964, 174-99 and 304-31. G oodw in , R.M. (1952), ‘A Note on the Theory of the Inflationary Press’, Economia Interazionale, reprinted in Goodwin, 1983, 57-74. ------(1953), ‘Static and Dynamic Linear General Equilibrium Models’, in Netherlands Economic Institute (ed.), Input-Output Relations, reprinted in Goodwin, 1983, 75-120. ------(1983), Essays in Linear Economic Structure, Macmillan, London. H ansen , B. (1 9 5 1 ), A Study in the Theory o f Inflation, Allen & Unwin, London. J ohnson , H.G. (1966), ‘The neoclassical One-sector Growth Model: A Geometrical Exposition and Extension to a Monetary Economy*, Economica, August, 3 3 , 265-87. ------(1967), ‘Money in a Growth Model*, excerpted from H.G. Johnson, Essays in Monetary Economics, Unwin, London, 161-78 in Sen, (1970) Growth Economics, Penguin Books, UK, 253-71. K alecki, M. (1 9 4 3 ), Studies in Economic Dynamics, Allen & Unwin, London. K eynes , J.M. (1936), The General Theory o f Employment, Interest and Money, Macmillan, London. L evhari , D. and D. P atinkin (1968), ‘The role of Money in a Simple Growth Model*, American Economic Review, 58 (4), September, 713-53. M c K innon , R.I. (1981), ‘The Exchange Rate and Macroeconomic Policy: Changing Postwar Perceptions*, Journal o f Economic Literature, 19, 531-7. M ahalanobis , P.C. (1953), ‘Some observations on the Process of Growth of National Income*, Sankhya, 12, part 4, reprinted in Mahalanobis, 1985, 13-18. ------(1955), ‘The Approach of Operational Research to Planning in India*, Sankhya, 16, parts 1 and 2, reprinted in Mahalanobis, 1985, 51-139. ------(1985), Essays on Planning, in P.K. Bose and M. Mukherjee (eds), Statistical Publishing Society, Calcutta. M anne , A.S. and A. R udra (1965), ‘A Consistency Model of India’s Fourth Plan’, Sankhya, Series B, 27, 57-144. O kun , A.M. (1981), Prices and Quantities: A Macroeconomic Analysis, Blackwell, Oxford.

112

A m iy a K u m a r B a g c h i

J.H.G. (1964), ‘On Structural Inflation and Latin American “Structuralism” \ Oxford Economic Papers, New Series, 16 (3), November, 321-32. P atnaik , P . (1972), ‘Disproportionality Crisis and Cyclical Growth’, Economic and Political Weekly, Annual Number, 7 (5-7), February, 329-36. ------(1988), Time, Inflation and Growth, Orient Longman, Calcutta. O livera ,

P atnaik, P., S.K. R a o and A. S anyal( 1976), ‘The Inflationary Process: Some Theoretical Comments’, Economic and Political Weekly, 11 (43), 23 October, 1696-1701. P orter , M.E. (1990),

The Competitive Advantage o f Nations, Macmillan,

London. (1968), ‘Competitive Equilibrium under Uncertainty’, Econometrica, 36, 31-58. R aj , K.N. and A.K. S en (1961), ‘Alternative Patterns of Growth under Conditions of Stagnant Export Earnings’, Oxford Economic Papers, New Series, 13(1), February, 43-52. R akshit , M.K. (ed.) (1989), Studies in the Macroeconomics o f Develop­ ing Countries, Oxford University Press, Calcutta. Ray, A. (1990), ‘The Theory of Devaluation for Less Developed Economies: Where do we Stand?, in B. Dutta, S. Gangopadhyay, D. Mookherjee and D. Ray (eds.), Essays in Economic Theory and Policy, Oxford University Press, Bombay, 239-58. R ow thorn , B. (1 9 7 7 ), ‘Conflict, inflation and money’, Cambridge Jour­ nal o f Economics, March, 1,1. R udra , A. (1964), ‘Relative Rates of Growth: Agriculture and Industry’, Economic Weekly, November, reprinted in P. Chaudhuri (ed.), Read­ ings in Indian Agricultural Development, Allen & Unwin, London, 1972, 19-33. S anyal , A., A. M ukherjee and P . P atnaik (1989), 'Stocks, Credit and the Commodity Market: A Preliminary Exercise’, in Rakshit, 1989, 111-28. S eers , D. (1962), ‘A Theorj of Growth and Inflation in Underdeveloped Economies based on the Experience of Latin America’, Oxford Economic Papers, New Series, 14 (2), June. ------(1964), Normal growth and distortions: some techniques of struc­ tural analysis, Oxford Economic Papers, New Series, 16(1), March,, 78-104. Sheffrin, S.M. ( 1989), The Making o f Economic Policy: History, Theory and Politics, Blackwell, Oxford. R adner , R .

Closed-economy Structuralist Models

113

Skott, P. (1989), Conflict and Effective Demand in Economic Growth, Cambridge University Press, Cambridge. S olow , R.M. (1990), The Labour Market as a Social Institution, Blackwell, Oxford. T aylor , L. (1981), ‘IS/LM in the Tropics: Diagrammatics of the New Structuralist Macro Critique*, in W.R. Cline and S. Weintraub (eds.), Economic Stabilization in Developing Countries, The Brookings Institution, Washington, D.C. ------(1983), Structuralist Macroeconomics, Basic Books, New York. T obin , J. (1955), ‘A Dynamic Aggregative Model*, Journal o f Political Economy, 63, 235-52, reprinted (without the Appendix) in Sen, (1970), Growth Economics, Penguin Books, UK. ------(1965), ‘Money and Economic Growth*, Econometrica, October, 33, 671-84. V anek , J. (1967), Estimating Foreign Resource Needs fo r Economic Development, McGraw-Hill, New York. W ejtzman , M.L. (1982), 'Increasing Returns and the Foundations of Unemployment Theory*, Economic Journal, 5, 48, 565-76.

The Macroeconomics of Imbalance and Adjustment C .P . C h a n d r a s e k h a r

In macroeconomics textbooks, the term ‘adjustment* merely describes the diverse, autonomous mechanisms through which macroeconomic equilibrium could be restored when a system is in disequilibrium. But the form in which the term has come to dominate the vocabulary of development economics in recent years refers essentially to the policy changes considered necessary to correct for balance of payments disequilibria, defined as arising from inflationary causes. In fact, the literature often uses the term without discrimination to refer to two separate processes. First, policy realignments within a given trade regimes warranted by developments in the international or domestic economy that are not ‘temporary*, such as a secular shift in the terms of trade or an ‘excessive* accumulation of internal debt over time.1 Second, changes in internal policy necessitated by an autonomous shift to a more open trade regime consequent to the removal of quantitative restrictions on imports, a reduction in tariffs, and the adoption of a liberalized exchange rate system. Since the early seventies most developing countries have faced external or internal ‘shocks* of the kind necessitating the first o f these processes of 1 ‘Structural adjustment policies may be defined as policy responses to external shocks, carried out with the objective of regaining the pre-shock growth path of the national economy. Regaining the growth path, in turn, will necessitate improvements in the balance of payments following the adverse effects of external shocks, since a country's balance of payments position constrains its economic growth. . . A broader definition will also include adjustments to internal shocks which may find their origin in inappropriate policies, such as the excessively expansionary fiscal measures taken in Mexico after 1972, or in political events, such as the April 1974 revolution in Portugal* (Balassa, 1982: 23).

The Macroeconomics o f Imbalance and Adjustment

115

adjustment. Forced, at that point, to turn to the International Monetary Fund for interim financing, they have had to accept policy advice that recommends a more open trade regime, which, in turn, warrants adjustment of the second kind. Adjustment in either of these forms involves correcting for external imbalance without reference to the maintenance of internal balance’, defined in conventional open-economy macroeconomics as the achievement of full employment. Within that framework, an economy afflicted by both unemployment and a balance of payments deficit, would worsen the external deficit by increasing domestic absorption, when it uses fiscal policy to correct for internal imbalance. However, external balance, it was argued, could simultaneously be achieved by resorting to devaluation.2 This presumes that the post-devaluation impact of the multiplier, determined by the effects of devaluation on the trade balance, on the one hand, and on the value of the multiplier, on the other, is unambiguously positive in the wake of currency depreciation. But devaluation, if it is to be 'effective' in a situation of cost-plus pricing (with fixed mark-ups when there is excess capacity) in the manufacturing sector, must involve distributional shifts. Either wages or domestically produced inputs must fall in real value, in the wake of an increase in the cost of imported intermediates, to prevent domestic inflation from neutralizing the effects of the devaluation on relative prices. Such distributional shifts are likely to result in a fall in the value of the Keynesian multiplier, just as much as the devaluation would result in an increase in net exports. And in a situation where the export response is weak and the share of imports in variable costs is high, a devaluation could result in a domestic recession if the decline in the value of the Keynesian multiplier is greater than the increase in the value of the multiplicand (/ + X - Af). due to the beneficial trade balance effects of the devaluation (Kaldor, 1978; Krugman and Taylor, 1978; Taylor, 1983; Patnaik, 1991).

B A LA N C E O F PA Y M EN TS SU STA IN A BILITY If both internal and external balances are difficult to ensure simul­ taneously, any model of adjustment must target one of them on a priority basis. In the post-War period stretching till the early seventies. 2 Sec for example Rivera-Batiz and Rivera-Batiz (1989), 166-75.

116

C P . C h a n d r a se k h a r

persistent external imbalances were rare. This was not because there was no material basis for such imbalances. Underdeveloped nations were at the bottom of the international industrialization league table and, therefore, externally vulnerable. Their dependence on imports of manufactures and exports of primary commodities often resulted in trade and current account imbalances when growth was sought to be accelerated within an open trade regime. But current account deficits in the balance of payments were restricted, since they could only be financed with controlled, 'official*, bilateral and multilateral flows that fell within the broadly defined category of foreign aid. The development problem consisted in ensuring a semblance of internal balance at the exogenously-given level of access to foreign aid, which defined the current account deficit that could be sustained. Hence, a degree of protection and a less import-intensive growth strategy were seen as necessary to pre-empt any unacceptable deceleration in growth. That is, countries adopted an interventionist trade and indus­ trial policy regime to ensure that 'enforced* external balance did not lead to persistent internal imbalance reflected in unemployed resour­ ces. The premise of policy was that any level of the current, account deficit had alternative trajectories in terms of growth and employment associated with it. Development strategies attempted to maximize employment o f domestic resources, including labour, given the level of access to foreign savings. In recent years, however, capital flows other than those through official bilateral and multilateral channels have accounted for a growing share of total flows to developing countries. Commercial loans and autonomous flows of foreign direct and portfolio investment are now the main mechanisms for recycling funds from surplus to deficit countries. The sustainable deficit on the current account of the balance of payments, defined as that which can be financed by autonomous (as opposed to exceptional) flows, becomes difficult to define in this context. The factors influencing the volume of such financial flows to any particular developing country are nebulous. These autonomous flows are far more volatile than flows of develop­ ment aid were in the past. And given the uncertainty surrounding a country’s export performance, the extent of financing consistent with a country’s future debt-servicing capacity is a matter for speculation as well. In other words, while at any given point in time developing countries are in a position to finance a much higher current account

The Macroeconomics o f Imbalance and Adjustment

117

deficit than they could in the past, there is no a priori definition of the current account deficit that is ‘sustainable* over time. If in the effort to approximate internal balance the current account deficit widens to a degree which undermines ‘investor confidence’, financial flows diminish or dry upt necessitating exceptional financing and a process of adjustment. Further, the capricious nature of ‘investor confidence* implies that even a relatively innocuous level of the current account deficit relative to GDP often proves unsustainable. Once established as ‘unsustainable', it is the deficit that must adjust to the uncertain, but ‘autonomously’ given, level of foreign exchange access. This reduction of the current account deficit could be effected through a curtailment of investment and growth and, thus, reduce the demand for foreign exchange. Alternatively, there could be direct curbs on imports into the system, aimed at regulating the trade and current account deficit to levels that are strategically acceptable.

THE IMF'S PERCEPTION However, IMF-type adjustment is far more specific. In its framework, a sustainable current account is defined to include a trade regime that is ‘liberal’ or ‘open*. ‘A viable balance of payments has two aspects. First, it implies that balance o f payments problems will not be merely suppressed but eradicated, and second, that the improvement in the country's external position will be durable.* (Emphasis added. Khan and Knight, 1985: 2). This definition of viability forecloses a protec­ tionist response to balance of payments difficulties and is based on the presumption that the failure to transit to a non-interventionist and open policy regime only suppresses balance of payments difficulties, rendering even a comfortable balance of payments position unsus­ tainable. Implicit in the IM F's reasoning is, of course, the notion that trade liberalization is a means for enhancing the supply of foreign exchange as well. Thus, increased access to foreign exchange is attributed partly to an increase in exports stimulated by enhanced international com­ petitiveness in the wake of liberalization, and partly to increased access to capital from the international financial system. The problem is that while the evidence for a direct relationship between liberaliza­ tion and export expansion is spotty, that for a definitive causal link between liberalization and capital inflow is scant. And even the

118

C . P . C h a n d r a se k h a r

meagre evidence for such relationships suggests that they are sig­ nificant at all only in the medium or long run. Despite the lack of theoretical and empirical grounds for such associations, the IM F s prescription merges the stabilization and long-run adjustment components of the balance of payments problem in developing countries. Most of those economies were till recently characterized by an industrial and trade policy regime fashioned to partially insulate the system from the ravages of international ine­ quality and stimulate growth based on the domestic market. Whenever external shocks or the process of accelerating domestic growth resulted in an unsustainable balance of payments, stabilization in the shbrt run involved a reduction in absorption. However, the process o f growth itself was not premised on access to foreign markets or foreign savings. As opposed to this, in the prevalent IMF-Bank perception, when faced with ‘shocks*, external or internal, these economies have to both stabilize immediately as well as quickly transit to an openeconomy regime that precludes a strategy attempting to dilute the linkage between the rate of growth of output and imports. This has two implications. First, at the prevalent level of production, imports tend to be higher than before liberalization, i.e. there is a tendency for the import intensity of domestic production to rise. Second, any attempt to raise the rate of growth of output translates into an increase in the rate of growth of imports. This implies that, given exports, associated with each sustainable level of the current account is a sustainable level of output. Not surprisingly, IMF-type adjustment in developing countries carrying the vestiges of interventionist regimes to different degrees, calls for a combination of trade liberalization and rather stringent 'demand-side* policies that correct for macroeconomic imbalances reflected in an unsustainable current account deficit. To the extent that trade liberalization does not enhance foreign exchange access in the short run, demand management must cope with the needs of stabilization as well as those created by the enhanced imports that liberalization results in. In other words, the IMF*s definition of sustainability limits the notion of 'adjustment* in a double sense. First; it makes external balance, or the equation of the demand for foreign exchange to its exogenously-given supply, the dominant target of the adjustment effort. And, second, it forces changes in trade policy that render the

The Macroeconomics o f Imbalance and Adjustment

119

process of adjustment dependent largely on a direct reduction in domestic absorption. This differentiating feature of an IMF-type package is missed in descriptions of its framework as one in which an 'unsustainable* current account deficit is treated as symptomatic of an overheated economy where aggregate demand has been allowed to exceed supply. To quote an IMF study (Khan and Knight, 1985:2): T y p ic a lly th e n e e d fo r a sta b iliz a tio n p ro g ra m m e , w h e th e r s u p p o rte d by th e F u n d o r o th e rw ise , a rise s w h en a c o u n try e x p e rie n c e s a n im b a la n c e b e tw e e n a g g re g a te d o m e stic d e m a n d (a b so rp tio n ) a n d a g g re g a te su p p ly , w h ic h is re fle c te d in a w o rse n in g o f its e x te rn a l p a y m e n ts p o sitio n .

In actual fact, stabilization through an IMF-type strategy involves a reduction in absorption to a far greater degree than typically needed in less developed countries faced with internal or external shocks.

READING THE IDENTITY How is this reduction in absorption to be ensured? Total domestic absorption consists of the sum of consumption, investment, and government expenditure and identically equals national income minus the current account surplus (or, if we ignore other current account flows, the trade surplus). Reduced absorption therefore would imply reduced national income, a higher current account surplus (or lower deficit) or both.3 IMF-type stabilization theory obviously reads this identity from left to right, ‘explaining’ the deterioration or improve­ ment in the current account balance (CAB) in terms of autonomous changes in domestic absorption. But there are three issues that need to be clarified here. First, which of the components of domestic absorption can be treated as being independent of income and there­ fore exogenously given? Second, what are the instruments that the

3 Nominal domestic absorption is given by A =C+/+G

Defining national income as Y = C + I + G +X - M

we have A = Y-(X-Af)= Y-CAB

See, for example, Dombusch (1980).

120

C . P . C h a n d r a sek h a r

government can use to ensure a reduction in these exogenous vari­ ables? And, finally, what is the mechanism through which the reduc­ tion in absorption affects current account flows on the balance of payments? As a first case, we can reasonably assume that consumption cannot be treated as an autonomous component of domestic absorption. This implies that it should be investment or government expenditure that generates an imbalance or restores balance. Accounting for taxes to finance government expenditure and rearranging terms in the national income identity, the current account deficit can be read as being equal to the sum of the deficit on the government's budget and the excess of private investment over savings.4 That is, given the level of consumption associated with any level of disposable income, domes­ tic absorption exceeds domestic supply when the government's budget is in deficit or private investment exceeds savings or both, necessitating a current account deficit to close the demand-supply gap. On the other hand, if foreign savings serve as a substitute for domestic savings, an increase in the current account deficit is accom­ panied by a widening of the domestic investment-savings gap. Consider now an economy which is running a current account defi­ cit of a magnitude which it can finance with the prevailing degree of access to foreign finance that it possesses. If, in such an economy, there is an autonomous increase in private investment or government expenditure or both, the current account deficit would remain un­ changed only if there is an equivalent increase in private savings or tax receipts or the two in combination. And if there are any constraints to an increase in domestic supply, the current account deficit would remain unchanged only if price increases ensure distributional changes 4 Subtracting taxes from both sides of the identity A = Y -{X -M ) = Y-CAB

we have C +I +(G -T ) =(Y -T ) -C A B

or CAB = [ ( Y - T - Q - I ) + { T - G )

Since Y - T equals disposable income and 7*- G the budget surplus CAB = (S - 1) + Budget surplus

or the current account deficit CAD = ( I - S ) + Budget deficit

The Macroeconomics o f Imbalance and Adjustment

121

that ‘force out’ savings or tax revenues from the economy.5 This, of course, would happen only if the government prevents access to foreign supplies, which in the IMF’s view amounts to ‘suppressing’ balance of payments difficulties. If they are not suppressed, the increase in absorption would result in an increase in the current account deficit rather than inflation. Allowing the macroeconomic imbalance to spill over onto the balance of payments provides, according to the IMF, a clear indication of the extent of adjustment required by the system, given its capacity to ‘earn’ foreign exchange.

THE MECHANISM We need to look at what is the mechanism which ensures the reduction in ‘autonomous* government expenditures or private investment required by the adjustment effort? And how would the effects of this reduction transmit themselves onto current account flows of the balance of payments? In the theory which underlies IMF adjustment strategies, the decision variable which influences the deficit on the government’s budget or the excess of investment over savings in the system is credit creation by the central bank. But identifying the credit-creating role of the central bank as the instrument of interven­ tion does not simplify matters altogether. An essential feature of an open economy is that money supply depends not only on the policies o f the central bank and the behaviour of domestic residents and financial intermediaries, but also on trends in the balance of payments. Defined in terms of the assets side of the central bank*s balance sheet, 5 In fact, the issue is more complex because the price level has a logic of its own, independent of the relation between aggregate demand and supply. Few can deny that the prices of most commodities are not determined purely by demand and supply interactions, but on a cost-plus basis in which margins above costs are determined by a host of factors. While the reasonable and convenient assumptions of prime cost-plus pricing and constancy of prime costs till some level of capacity, allows for quantity adjustments to dominate so long as the degree of monopoly remains constant, there is nothing to prevent an increase in the degree of monopoly in contexts where demand is buoyant and expectations are that demand would exceed capacity constraints in the future (Kalecki, 1971: 43-61; Bhaduri, 1986: 62-85). That is, well before supply constraints are realized, the multiplier could operate through both quantity and price adjustments. The imbalance which determines the rate of inflation is not necessarily, and definitely not only, excess aggregate demand, but also the contested terrain of production and pricing, where distributive shares are determined.

122

C . P . C h a n d r a sek h a r

high-powered money consists of net foreign assets (NFA) and central bank credit (CBC). With high-powered money (H) defined as H = NFA + CBC the change in high-powered money 8H = 8JVFA+8CBC

(1)

Hence, the change in money supply (Ms) is given by 8MS= *i(5//) = tfSNFA + 8C BQ

(2)

where \i is the money supply multiplier. Thus, if an improvement in the current account balance, that is, an increase in net foreign assets held by the central bank, is to accompany a reduction in central bank credit, these changes should either not affect the supply of money or do so in a fashion that does not neutralize the expected effect of the change in central bank credit on net foreign assets. The simplest version of the orthodox monetarist position overcom­ es this difficulty by considering an economy in full employment that satisfies the small country assumption and is therefore a price taker in international commodity and capital markets. With real income given at full employment and interest rates given from international markets, the demand for money in such a system is given and stable. Now, assuming =1 we have 8NFA = 8MS - 8CBC

(3)

Since money market equilibrium requires that the supply o f money is equal to the given demand for money, 8NFA = 8Md - 8CBC That is, net foreign assets register a decline whenever credit creation by the central bank increases relative to money demand. This is a situation where the public finds itself loaded with excess money balances. Since prices in the domestic market are determined by international prices, this implies an increase in net real wealth or the real value of money balances held with the public. The monetary approach to the balance of payments argues that this has a 'real balance effect*, or spurs an increase in expenditure. And given full employment of domestic resources, increased expenditure results in

The Macroeconomics o f Imbalance and Adjustment

123

increased consumption of foreign goods and a consequent balance of trade deficit. Since that deficit results from an excess of credit creation relative to money demand, it is assumed that it can be corrected through some sort of domestic credit ceiling.

THE FISCAL DEFICIT The inadequacy of this simple framework should be clear. It is based on the simplifying, but far too stringent, assumptions that countries are at full employment and function as price takers in international markets. Yet the idea that curbs on credit provide a mechanism to correct for balance of payments aisequilibria has persisted, and has even been elaborated. In IMF-type adjustment strategies, the war­ ranted curbs on credit take the form of a simultaneous restriction of the fiscal deficit of the government and the credit drawn by it from the central bank to finance that deficit. The emphasis on the fiscal deficit is not without design. In terms of the conventional national income identity, it reflects one of the two components into which any excess of ex ante investment over ex ante savings can be decomposed, with the total excess of investment over savings equal to the deficit on the current account or the decline in net foreign assets, given by

bNFA - §MS- hCBC Now, assuming the government’s budget deficit is financed by

G -T = § C B C g + $DCg

(4)

i.e.

&NFA = m s - ( G - T - 5DCg) - 5CBCng

(5)

where

5CBC = 5CBCg + hCBCng

That is, a decline in the central bank*s foreign assets position occurs when the sum of the government's deficit financed by the central bank and additional central bank credit to the private sector is greater than the increase in money stock. Assume that the adjustment programme emphasizes a reduction in the deficit of the government financed by the central bank. If the size of the fiscal deficit remains

124

C . P . C h a n d r a se k h a r

the same, this would involve an increase in the government's bor­ rowing from the open market, which could ‘crowd out’ lending to the private sector. To foreclose that possibility, IMF-adjustment strategies target a reduction in the fiscal deficit of the government. However, the argument still requires that the either the supply or demand for money is fixed in some fashion, allowing equilibrium in the money market to determine the other. If the full employment and small country assumptions do not hold, the demand for money is not a ‘given’. If we assume that Ms is given from ‘outside’ and willingly held by the population at the prevailing interest rate and real income configuration, and that central bank credit to the private sector either remains unaffected or decreases, then the reduction in the fiscal deficit implies an unambiguous improvement in the balance of payments. The problem is that in a world far more complex than the simple monetarist model, no such assumption is valid. In which case, equi­ libria which suggest an inverse relation between the fiscal deficit and the change in net external assets of the central bank are not the only ones compatible with money market equilibrium. No more can purely monetarist explanations like the ‘real balance effect' prove adequate. To illustrate this, we can better use the identity depicting the balance sheet of the consolidated banking system rather than the central bank alone, namely 8NFAh = 5Af, - (5DCg + hDCp)

(6)

Now since the government's fiscal deficit is financed either by domestic or external borrowing G ~ T - &DCg + $ECg Hence.

8VM* = ( 7 - G + fiE C p+ ( 8 ^ - 8 0 0 , )

(7) (8)

It should be clear that a reduction in the fiscal deficit would be accompanied by a reduction in the borrowing of the government either from the central bank or from the public. If the assumptions of textbook monetarism with regard to full employment output and prices do not hold, this could affect the demand for money in two ways: first, the reduction in government expenditure could adversely affect the level of real income in the system; and second, to the extent that any reduction in expenditure results from a cut in the government’s borrowing from the public, it could reduce the rate of interest. These two mechanisms have contrary effects on the demand

The Macroeconomics o f Imbalance and Adjustment

125

for money, with the former reducing it and the latter tending to increase it. If the interest elasticity of demand for money is low, the net effect would be a decline in the demand for money. This implies that, if money market equilibrium holds, money supply would be lower than it would have been without the cut in the fiscal deficit. In addition, lower interest rates would raise domestic credit to the public (though not necessarily to an extent that nullifies the decline in income arising from the cut in public expenditure). Both these effects, by re­ ducing the second term on the RHS of (8), namely (&MS - 8DCp), could more than neutralize the increase in the first term, namely ( 7 - G + 5ECp, consequent to the fiscal deficit cut, and therefore result in a decline in net foreign assets with the banking system. This illustration merely suggests that even assuming that the money market is in equilibrium, we have multiple equilibria involving alternative values for the different terms of the identity. That is, the monetarist effort to focus on one among many possible equilibria, read the monetary identity from right to left, and have the budget deficit determine the change in net foreign exchange assfets of the banking system is fraught with problems.

MODIFIED APPROACH It is possibly for this reason that actual policy-oriented discussion within the IMF initially adopted a modified version of the monetarist argument when discussing the adjustment mechanism. In Polak's celebrated defence of monetary stringency (Polak, 1957) as a mechanism for balance of payments adjustment, he defines macroeconomic imbalance as an increase in income that raises induced imports while leaving autonomous exports untouched, making adjust­ ment a process of aligning growth rates to levels 'warranted* by the rate of growth of exports. That adjustment is ensured through a sharp reduction in credit creation by the banking system. In Polak’s original model, the process of adjustment was rather straightforward, being based on the simple quantity theory of money notion that the ratio of money to income is constant. Any explanation of the changes in the quantity of money was also, therefore, an explanation of changes in income. As before, taking ‘money* to be nothing more than the formal liabilities of the banking system, changes in the quantity of money

126

C . P . C h a n d r a se k h a r

can be equated to changes in the (net) foreign assets of the banking system p lu s changes in domestic credit by the banking system. Of these two, only the latter is treated as an autonomous ‘decision variable’,6 making income dependent on the extent of credit creation. Changes in foreign assets, on the other hand, are seen to be the result of an exogenously given level of exports, and a level of imports determined by the level of income. This essentially means that to reduce the current account deficit, the system would have to reduce the level of income without reducing exports, which it can do only if it reduces new income resulting from internal credit creation.7 The assumption here is, of course, that the supply of credit (and therefore the level of nominal income) is fully controlled through controls on expansion of credit by the banking system. Implicit in such an assumption is a definition of money that arbitrarily restricts it to the debt obligations of the central bank and the formal liabilities 6 ‘Credit expansion is a function of the banking system. It may be difficult, perhaps in some circumstances humanly impossible, for the system to withstand demands for credit from the government or other insistent borrowers; and in such circumstances the desire to make public development expenditure, or to construct private factories, may be considered from many points of view, the cause of the expansion of the economy. But for purposes of monetary analysis and monetary policy, there is a clear gain in clarity if the responsibility is pinpointed on the credit expansion. The economic development could also have been financed by higher taxes or a foreign loan. The factories may have been built by restriction of consumption or by the repatriation of capital. In all these situations, the desire to spend for a particular purpose would not have led to a payments problem. In a real sense, the credit expansion is the cause of the payments problem' (Polak, 1957: 13). 7 The equations in the original Polak model read as follows

Y(t) = H ' - 0 + SOC(0 + X(t) - M(t) M(t) = m Y(t- 1) where Y is the money value of national income; 6DC is the increase in credit by die banking system; X and M stand for exports and imports; and m the propensity to import at the margin. The first of these equations is based on the assumption of an income velocity of circulation of money equal to 1, which gives 6MI(r) = n O - K ( / - l ) where 6Af, is the increase in the quantity of money. That combined with the batance sheet identity 5DC(0 + X(r) - M ( i ) = 5W,(r) yields the first equation.

The Macroeconomics o f Imbalance and Adjustment

127

o f the commercial banking system, and excludes all forms of ‘near­ money*. Not only have financial instruments that play such a role been important in the financial system of the less developed countries, but their presence has increased with the process of 'financial liberalization’ that accompanies IMF-type adjustment packages. Their exclusion, implying the lack of financial assets that can be held as an alternative to money and goods is, however, in keeping with the implicit notion that money is kept almost exclusively for transactions purposes in developing countries. The latter in turn provides the basis for the assumption that people adjust their holdings of money in proportion to changes in monetary transactions, resulting in a constant ratio of money to income, needed for the relationship between changes in money supply and nominal income.8 There is one other serious difficulty with Polak’s formulation of the problem. Since the relationship specified is between money supply and nominal income, the model is unclear on the extent to which the process of adjustment is mediated by a decrease in prices as opposed to a decrease in real output. For, unlike conventional monetarist models, it does not assume full employment, allowing for Keynesiantype adjustments in response to a curtailment of credit expansion. And since simultaneous price adjustments are not ruled out, restric­ tions on credit creation affect the level of nominal income via both output and prices. This implies that the import equation in the model has to relate imports to the nominal value of income, and that the import propensity ‘m’ captures the increase in imports due to both an increase in real income and on account of that resulting from the substitution away from domestic goods because of a price increase. The fallacy o f treating the effects of substitution away from domestic goods in favour of imports because of an increase in domestic prices on par with and equivalent to an increase in import demand due to an increase in real incomes should have been obvious even to Polak. But having chosen to retain the monetarist framework (and con­ clusions) while giving up its unrealistic assumptions, he could not make the requisite distinction between two qualitatively and quan­ titatively diverse processes. 8 ‘The assumption of a constant velocity of circulation eliminates another compli­ cation, namely the extension of creditor the purchase by the banking system of existing domestic assets which does not lead to expenditures on goods and services by the bor­ rower or seller of the assets. An operation of this nature would be equivalent to an increase in the quantity of money without an increase in income* (Polak, 1957: 16).

128

C . P . C h a n d r a se k h a r

However, some noteworthy features of this early formulation of the IMF*s position on adjustment are that: (i) it relies explicitly on a transmission mechanism that involves a fall in nominal income and therefore imports; and (ii) it does not explicitly distinguish between regimes which are interventionist-and those which are not, or make the case that a shift in policy in one direction or another would change the pace of expansion of exports or the volume of capital inflow. It merely states that for any level of exports generated within any trade regime, there exists a level of credit creation and income growth that is compatible with a sustainable balance of payments position.

D EM A N D M A N A G EM EN T This is significant because, as noted above, a defining feature of more recent IMF writing (Khan and Knight, 1985) is that it treats trade liberalization as a part of stabilization, on the one hand, and a mechanism which ostensibly partially endogenizes foreign exchange access, on the other. This implies that in the course of the traverse to the ‘appropriate’ policy regime, adjustment would involve both trade reform and demand management.9 Once the economy was on the growth path characteristic of the appropriate regime, adjustment to specific shocks would be based on demand-side policies similar to those suggested by Polak. However, a more open trade regime aimed at shifting production in favour of tradables or enhancing the international competitiveness of domestic firms, results in higher imports in the short run, while promising returns in terms of higher exports, if any, only in the medium term. This perverse effect of a change in the trade regime in the short run, introduces an obvious bias in favour of demand-management policies in all adjustment packages recommended by institutions 9 ‘In the context of developing countries, economic stabilisation usually refers to a programme of comprehensive economic measures designed to achieve broad mac­ roeconomic goals such as a sustained improvement in the balance of payments, a reduction in domestic inflation, or some combination of the two. Naturally the im­ plementation of such a programme presupposes the existence of serious economic problems which necessitate a basic redirection of economic policy. In general terms, the fundamental objective of a stabilisation programme is to achieve “a suitable relationship between resource availabilities and needs that causes minimum strain on the internal price level and produces a desired balance of payments result" * (Khan and Knight, 1982: 709).

The Macroeconomics o f Imbalance and Adjustment 129 like the World Bank and the IMF, to which developing countries turn when strapped for balance of payments finance. As a result, in the more recent analytical support for IMF-type adjustment strategies, demand management is even more central, since it is needed to neutralize the perverse effects of ‘structural reform* aS well. And, as indicated earlier, within demand manage­ ment, the focus has shifted from credit creation in general to the fiscal deficit on the government’s budget financed with credit from the central bank or the banking system. In practice, that reduction has both direct and indirect effects on demand: direct, because of the decline in government expenditure on private sector goods and services; and indirect, because o f the multiplier effects on income of government expenditure reduction. In most developing country environments, where government expendi­ ture significantly influences the expansion of the home market, this is likely to have an adverse effect on private investment as well. That is, there is a strong likelihood that an increase in government investment and borrowing normally ‘crowds in* private investment and borrowing, and that a reduction in the former would reduce private investment as well. As a result, the overall deflationary impact of a fiscal deficit reduction is substantial. The curtailment of demand consequent to such deflation would, in turn, affect the volume of imports into the system, permitting a reduction in the current account deficit, for any exogenously given rate of increase in exports. To the extent that there is any autonomous increase in imports in the short run, either because of a change in the trade regime accompanying stabilization, or because o f income distributional factors that transfer incomes away from income groups whose import intensity of consumption is far less than average, the extent of the recession needed to ensure a given increase in net foreign assets would be greater. This, of course, is the real side of the picture, where the transmis­ sion mechanism is a fall in output consequent to fiscal adjustment. However, the monetary identity we have at hand suggests an alterna­ tive scenario. Assume, for example, that for ‘extraneous reasons* autonomous flows of portfolio and credit capital to the private sector increase sharply, or that public sector assets are sold to foreign investors as part of a privatization programme. If these foreign1 exchange receipts do not get spent on imports, but are ‘sold* to the domestic banking system, they would increase the net foreign assets

130

C . P . C h a n d r a se k h a r

with the banking sector, which would release an equivalent volume of rupee resources to the agency acquiring the foreign assets in the first place. This implies an increase in money supply or a positive 5Af, on the right hand side of identity (8) relating to the consolidated banking system, equal to the increase in net foreign assets on the left hand side. But here, of course, the direction of the causality is reversed, inasmuch as it is the inflow of foreign capital and the consequent increase in net foreign assets with the banking system that is resulting in the increase in money supply. To the extent that the accretion of rupee resources with the private or public sector results in an increase in domestic expenditure, the inflow of foreign resources results in an increase in domestic income as well. That is, when the identity is read from left to right, the narrative could be one in which an autonomous increase in foreign capital inflows raises money supply and domestic income. This suggests that any increase in net foreign assets can a priori involve both processes: a deflation that reduces import demand and an inflow of foreign capital that raises domestic income. The causality can run from the RHS to the LHS when adjustment is ensured through a recession, or from the LHS to the RHS when adjustment is supported by a large inflow of foreign capital. If the latter tendency dominates, the process of 'adjustment* appears benign from the point of view o f economic activity. That is, if at all a process of adjustment based on a reduction in the fiscal deficit is to prove non-recessionary, it can happen only if there is an inflow of foreign resources either because of privatization of government assets through sale to foreigners, or because of other developments that encourage foreign capital inflow when fiscal adjustment is in progress. It is for this reason that the IMF has made the sale of public assets to foreigners and financial liberalization important elements of a 'macroeconomic adjustment* package. It is this autonomous, foreign capital inflow requirement that is underplayed in the argument that 'adjustment* need not have recessionary consequences because ‘supply-side policies* spur private investment.

D EV A LU A TIO N It is, of course, true that in a typical IMF-type package, fiscal adjust­ ment is combined with a devaluation that is directly aimed at the

The Macroeconomics o f Imbalance and Adjustment

131

current account deficit. Operating with the conventional small country assumption and taking the tariff structure as given, the textbook neoclassical framework implicitly treats the real exchange rate as ‘the relative-price variable in a simple supply-and-demand analysis in which the quantity of real dollars demanded and supplied is expressed as a function of its real price’10 (Helmers, 1988). If at the prevailing exchange rate there is a demand-supply gap, it is because capital inflows are ensuring balance of payments equilibrium and equating aggregate spending to total income and the net inflow of funds from abroad. The balance o f payments disequilibrium resulting from any reduction in access to foreign exchange can be corrected through devaluation because (i) it stimulates an expansion in the supply of tradables by rendering the production of both importables and export­ ables more profitable, while reducing demand for them through a rise in relative domestic prices; and (ii) it reduces the supply and stimulates the demand for non-tradables and eliminates any excess supply that arises because of reduced spending consequent to reduced access to foreign exchange. Thus, in this framework, a devaluation is essentially an expenditure-switching policy that operates on both the demand and supply sides. There are of course ‘exceptions* where expenditure reduction is also inevitable. The most obvious is when it is the government which is the recipient of foreign capital inflow and is forced to cut back expenditures to ensure equilibrium at the new exchange rate. But as discussed at the beginning of this essay, even when such a cut back is not warranted, a devaluation tends to be contractionary if the decline in the value of the Keynesian multiplier resulting from distributional changes is greater than the increase in income resulting from the rise in the trade balance associated with the devaluation. Assessing these factors in the light of the possibility that the supply responses expected to be stimulated by the devaluation may be slow in coming, a policy of currency depreciation could also largely be seen as an effort at demand management aimed at deflating a system to correct For balance of payments disequilibrium, however generated. Besides the difficulties inherent in the small country assumption, there is an additional problem with this approach. Balance of pay­ ments deficits only partly reflect differences in flow s of imports and 10 The demand and supply curves reflect alternative equilibrium situations in the foreign exchange market.

132

C . P . C h a n d r a se k h a r

exports (or income and expenditure).'Increasingly, such deficits may become expressions o f a stock-flow disequilibrium between total foreign debt and current income. The significance of this is that exchange rate movements designed to affect flow variables are limited in their impact depending on the weight of inelastic elements such as amortization of and interest payments on external debt. At the same time, if the exchange rate is market determined, then the very nature o f the stock-flow disequilibrium itself affects expectations and thus also impinges on the currency market and consequent changes in the exchange rate (Chandrasekhar and Ghosh, 1993). In sum, though the overall strategy of IMF-type adjustment involves expenditure-reduction through fiscal and monetary contraction, expenditure-switching brought about by exchange rate devaluation, and trade liberalization measures with longer-run efficiency aims, contraction is at the core of the programme. And, unlike the approach embodied in the Polak model, the currently dominant strategy is doubly contractionary, inasmuch as it has to neutralize the perverse effects of liberalization on the current account in the short, and possibly medium, term. The approach is one of deflating the system to a degree where, at the prevailing import intensity of production, the demand for foreign exchange is in keeping with the access to foreign exchange within an unequal world order. However, the perception appears to be that, openness, which provides international agents greater freedom to produce for domestic markets and own domestic assets, would substantially increase that access.

MODELS OF ADJUSTMENT WITH GROWTH In fact, to meet the criticism that the IMF’s package is extremely adverse from a growth point of view, two kinds of arguments have been put forward by supporters of conventional adjustment program­ mes. The first is that it is possible conceptually to design ‘growthoriented adjustment programmes* (Khan and Montiel, 1989), which ‘simultaneously eliminate the macroeconomic imbalances in the economy and raise the growth rate*, by combining the monetary approach to the balance of payments with a simple version of the open-economy neoclassical growth model. The link between these two apparently conflicting models, one of which uses contractionary policies to correct balance of payments disequilibria and another

The Macroeconomics o f Imbalance and Adjustment

133

which relates the rate of growth to the rate of investment,11is ensured by taking the rate of growth in capacity output in the latter as deter­ mining the rate of growth of output in the former. For a given rate of growth of output delivered by the growth module, there exists a rate of price increase compatible with money market equilibrium, since both increases in real output and an increase in prices raise the demand for money in flow terms. With exchange rate changes taken as ex­ ogenously given,12 the rate of price increase influences the relative profitability o f exports and the relative costs of imports, and thereby the trade deficit. And that deficit in turn determines the quantum of foreign savings available to the system and therefore the level of investment and growth. This implies that one can read both positive and negative relationships between the rate of growth of real output and the rate of price increase. In the monetary model, increases in prices and in real output are negatively associated because both in­ crease the flow excess demand for money. And in the growth model, increases in the domestic price level, which increase the trade deficit and the corresponding foreign savings, are associated with increases in real output. Thus, the rate of growth and rate o f price increase are simultaneously determined in an equilibrium which tallies with the identities and the relationships characterizing conventional models. There are a number of obvious inadequacies in this conceptual union, the most striking being that it treats the flow of foreign savings as endogenously determined. By doing so, the model implicitly assumes that if a system chooses to realize a rate of growth which requires a level of foreign exchange access greater than currently available, prices change through monetary adjustments to ensure the realization of that rate of growth. That is, the model assumes that the burden of adjustment falls only on prices—a proposition that it implicitly sets out to establish. In practice, the problem of adjustment arises because exogenous­ 11 Since the rate of investment is assumed to be equal to the rate of saving, 'an increase in the private savings rate, government savings, or in the exogenous com­ ponent of the current account deficit would increase real output growth, in each case by increasing aggregate saving and, therefore, investment* (Khan and Knight, 1989). 12 The effects of a devaluation on growth, working through its effects on foreign savings, is ambiguous. 'Devaluation simultaneously increases the real value of the initial level of savings and, by discouraging imports and encouraging exports, reduces that level. The first effect increases real investment, whereas the second decreases it* (Khan and Knight, 1989: 288).

134

C . P . C h a n d r a se k h a r

ly-determined flows of foreign capital are inadequate to finance the current account deficit. As Polak puts it, the merged model is not needed to show that ‘if a country can afford an increased current account deficit, it can use the resulting resources to raise its rate of growth’ (Polak, 1990). If at all the model has any usefulness, it is to solve, under rather stringent restrictions on the ‘exogenous* variables, for the growth in capacity output associated with each level of external financing. That is, it is not so much a model of adjustment-withgrowth as one defining the prospects for growth, given the adjustment requirement set by the access to international finance.

THE ALTERNATIVE PERSPECTIVE In contrast, the second defence of IMF-type adjustment theory against the criticism that it is unduly contractionary is more sophisticated. It sets out to argue that given the changes in international financial markets, an appropriate adjustment package that includes major reforms in the financial sector, can ensure external financing of a magnitude that either minimizes the extent of adjustment through growth reduction or even overwhelms that reduction with parallel growth-inducing effects. Adjustment policies that correct for balance o f payments disequilibria also create an appropriate macroeconomic environment that contributes to enhancing foreign financial inflows. That is, efforts at reducing the deficit in the government's budget or liberalizing trade policies have an indirect effect on the net foreign assets of the central bank. By spurring investor confidence they result in inflows of direct and portfolio investment that amount to an accumulation of such assets. And to the extent they do, they are reflected, on the liabilities side of the central bank’s balance sheet identity, in an increase in currency with the public or deposits of the commercial banking system with the central bank, that imply an increase in high powered money and therefore in money supply. In terms of the identity relating to the consolidated banking system, net foreign assets increase on the left hand side and money supply on the right hand side. And in terms of the reduced fonn of the national income identity, the argument reverses the causality that adjustment conventionally implies in a developing country. It moves from an increase in the net foreign assets of a country to an increase in the

The Macroeconomics o f Imbalance and Adjustment

135

excess of ex-ante investment over domestic savings, which implies a rise in the real output. As argued earlier, it is with this implicit reasoning in mind that recent versions of the IMF-adjustment package have emphasized the role of financial liberalization. In practice, there are two aspects to such liberalization. The first involves *a substantial reduction of government intervention in setting interest rates and allocating credit’ (Cho and Khatkhate, 1989). The second involves ‘policy actions that increase the degree of financial openness, i.e. the ease with which residents can acquire assets and liabilities denominated in foreign currencies and non-residents can operate in national financial markets (including the enjoyment of market access by foreign banks)’ (Akyiiz, 1993). Such openness is often accompanied by a gradual shift to convertible currency regimes. Clearly the two forms of liberalization are related. Foreign banks and institutional investors, which are to be encouraged to undertake investment, would prefer an environment in which regulation of their activity by domestic governments is minimal. In fact, while there are adequate grounds to argue that financial liberalization is no route to efficiency in resource allocation in developing countries (Stiglitz, 1993), from the point of view of external financial agents, liberalization is a major opportunity for profit in the developing countries. First, emerging financial markets in these countries, though volatile, offer extremely high returns in a period when the debt overhang and slow growth in the developed countries has affected financial interests adversely. That makes risk-discounted returns in the developing countries much better than in the developed. Second, privatization programmes have put up for sale resources of substantial value that can be acquired relatively cheap in a context of currency depreciation. Third, these are markets in which the pent-up demand for credit is substantial and innovative financial instruments have not been experimented with in the past. Fourth, real interest rates tend to be relatively high in developing countries undertaking adjustment programmes involving monetary stringency. And, finally, the sharp increase in the external debt of most developing countries since the mid-seventies has created new market opportunities for swaps and other such mechanisms aimed at dealing with that debt.

136

C . P . C h a n d r a se k h a r

THE CASE FOR FINANCIAL LIBERALIZATION In the event, there has been a sharp increase in the volume of transborder financial flows to developing countries in recent years.13 Capi­ tal today is not just fluid, but is also drawn from across the globe to be invested in a few selected areas deemed fit or creditworthy by a handful of transnational financial institutions (Patnaik, 1993). It is not just corporations but nations that are now routinely ‘rated* for their creditworthiness, based on a host of nebulous economic and political criteria. And even though multinational commercial or merchant banks do not exert exclusive control over financial flows, the lead given by them is followed by a host of individual rentiers. Liberaliza­ tion, including financial liberalization, becomes a necessary (but not sufficient) condition for attracting such flows reflective of a new feature of the international environment, namely the centralization of finance. There are however a number of problems here. To start with, foreign capital flows are of three kinds: direct investment aimed at productive capacity creation; portfolio flows which enter under conditions of easy repatriability and are therefore extremely unstable; and commercial credit from foreign institutions and non-residents. If the first of these flows is realized within a relatively open trade regime, the capacity created tends to be neutral between production for the domestic and external market, and therefore tends to enhance exports as well.14 On the other hand, flows of purely ‘financial capital* have 13 Between 1989 and 1992, while global flows of foreign direct investment (FDI) declined from $234 billion to $150 billion, those to the developing countries rose from $29 biljion to $40 billion, fuelled by a $50 billion privatization drive in these countries. Second, portfolio flows in the form of investments in bonds, equities, certificates of deposit and commercial paper, rose from less than $10 billion in 1990 to $37 billion in 1992. All told, flows of this kind have risen from around $35 billion to more than $75 billion in the first three years of the nineties. 14 The rise to dominance of finance capital has been accompanied by substantial changes in the nature of foreign investment, especially in terms of its relationship with trade. In colonial times, foreign direct investment in the colonies was closely linked with the needs of colonial trade, and was directed in the main into areas such as plantations, the extractive industries, shipping and insurance. Such investment only strengthened the ‘enforced bilateralism’ in trade that colonialism implied. With the onset of decolonization and the adoption of protective, import-substituting strategies by independent post-colonial states aiming to industrialize rapidly, foreign investment often became a substitute for trade. With tariff and non-tariff barriers foreclosing-the

The Macroeconomics o f Imbalance and Adjustment

137

little impact on real output and tend to be more speculative, respond­ ing adversely to any instability either of the real economy or financial variables like the rate of inflation and the exchange rate. ‘Capital flows exert a considerable influence on exchange rates and financial asset prices, and are themselves influenced by expectations regarding rates of return on financial assets denominated in different currencies. This means not only that domestic policies have a new channel of influence on exchange rates, trade, balance of payments and, hence, the level of economic activity (namely through their effects on capital flows), but also that these will all be influenced by financial policy abroad and by events at home and abroad that alter expectations* (Akyiiz, 1993). In such a world, the ‘national space’ available to the state as its area of control, within which it acts to promote develop­ ment, is substantially eroded. For, the entire range of fiscal, monetary, or external policies have to be adjusted keeping in mind the implicit requirements set by the fluidity of finance, subordinating national requirement to the caprices of international capital. Dependence on fluid finance also implies that to keep financial stocks within the country and maintain consistent flows over time, countries have to encourage, within the reformed financial sector, a relatively high interest rate by opting for a tight money policy. High interest rates are not only a prerequisite for capital inflows that ensure balance of payments viability and exchange rate stability, but also keep inflation down and dampen disincentives for financial-asset holders. To meet this danger of the loss of national space inherent in the developing country markets of leading international firms, they found the need to jump those barriers, by establishing production facilities that could service local markets. More recently, not only has the revolution in transport and communications and the changes in technology, that have segmented production processes, expanded world trade and capital flows over time, but over the last two decades or more there has been a rapid dismantling of protective regimes and relaxation of regulations on foreign investors across the globe. This has affected the character of foreign investment as well, since after allowing for national peculiarities and variations in political structures, any production site worldwide is becoming a potential site for production for world markets. For the transnational firm, which over the years has increasingly detached itself from dependence on home country resources, it offers the opportunity of locating itself in environments where it can overcome the disadvantages that specific macroeconomic developments such as inflation or microeconomic features like high wage levels create, and substantially enhance its international competitiveness (see ESCAP, 1993).

138

C . P . C h a n d r a sek h a r

conventional adjustment package it is often argued that current account liberalization is safe (and indeed necessary) as long as controls on capital flows remain. The literature thus advises develop­ ing countries not to lift controls on capital flows, or at least to delay lifting controls until current account balances or surpluses are assured. However, such advice is likely to have no real relevance given de facto openness to capital flows. Indeed, the argument that opening of only the external current account is a low-risk strategy is based on a rather specious distinction between current and capital accounts that has very little meaning in practice. As long as there is no way of ensuring that exporters repatriate their earnings, and no other incen­ tives are provided for them to do so, then ‘capital flight* could simply take the form of exporters keeping their money abroad in anticipation of future devaluations. To the extent that structural trade imbalances persist, this will also depress expectations about the value of the currency and act as a barrier to capital inflow. These in turn imply that expectations become self-fulfilling in the absence of large-scale intervention by the central bank which would involve depletion of foreign reserves.

THE POLITICAL ECONOMY OF ADJUSTMENT Adjustment packages involving reduced government expenditure, high interest rates, and convertible currencies, while in the interests of metropolitan financial capital, are obviously inimical to both metropolitan and developing country industrial capital. Reduced government expenditure while imparting a degree of stability to finan­ cial markets, slows the rate of expansion of the home market in developing countries, which provides the base for productive invest­ ment. High interest rates are a disincentive for productive investment, aggravating the recession associated with adjustment. And volatile exchange rates that cannot be directly controlled, substantially in­ crease uncertainty, while yielding little additional benefit relative to well-defined dividend and capital repatriation laws. The background to this bias in favour of financial interests of the adjustment strategies being adopted by developing countries needs elaboration. Consider, for example, the evidence on ‘acceptable* inflation rates and targets, on the one hand, and ‘acceptable’ levels o f the current account deficit, on the other. While the rate of inflation

The Macroeconomics o f Imbalance and Adjustment

139

deemed acceptable varies across countries, the 3-4 per cent target set for Germany by the Bundesbank has gradually caught the imagination o f policy-makers in the developed industrial nations and, more recently, in the developing countries. This figure is, of course, remark­ ably low when compared to the rate of inflation which prevailed in most countries during the post-War years. Over the period 1965-80, World Bank figures indicate that inflation averaged 19.2 per cent per annum in ‘low-income countries excluding China and India*, 20.9 per cent in middle income countries and 7.6 per cent in high income economies. The resistance to inflation in most countries comes mainly from two sources, though all sections are concerned about the uncertainties that episodes of hyperinflation involve. First it comes from those, like wage workers, peasants and the salariat, who are unable to adjust their incomes fast enough to keep pace with inflation. Second, it comes from financial interests, who are concerned about the effects of inflation on the value of financial assets and on the real rate of interest. There is reason to believe that, till recently, the influence of international and domestic financial interests on policy in the develop­ ing world was limited. In fact, in many developing countries, control over financial institutions rested directly or implicitly with the state. What determined the limit to profit inflation then, was the political feasibility of eroding real incomes beyond a point, which defined the ‘inflation barrier’ faced by these governments. Depending on the play of autonomous forces influencing the rate of price increase and the relative weight in influencing policy of those whose incomes were eroded by inflation, the government’s maneuverability in terms of deficit-financed expenditure was circumscribed by this barrier. In periods when the threat of ‘excessive’ inflation was low, the govern­ ment stepped up its expenditure, only to retreat when that threat became real. But given the correlation of political forces in these countries during these years, the acceptable rate of inflation was much higher than the targets commonly associated with 'structural adjustment* packages today.

ACCEPTABLE DEFICITS Changes with regard to what is acceptable have been true of the current account deficit as well. In fact, till the early seventies, the

140

C . P . C h a n d r a sek h a r

limits to the deficit on the current account that developing countries could incur were set by their limited access to international finance. When they exceeded those limits, and were forced to turn to the IMF for balance of payments finance, it was not the deficit on the government's budget which was the focus of IMF attention (for that was largely kept in course by the inflation barrier). Rather, it was the controls and regulations characteristic of import-substituting regimes and, of course, what were considered their ‘overvalued’ exchange rates. Besides deregulation in the industrial sector, trade liberalization and devaluation (not convertibility) were the main elements of stabilization packages, ostensibly aimed at improving the export com­ petitiveness of these economies, so that they could earn the foreign exchange needed to finance their imports. The emphasis on these elements of the package also served the interests of metropolitan industrial capital, which dominated policy­ making in the developed world during the post-War boom. Liberaliza­ tion opened up markets, which earlier could be accessed only by jumping tariff barriers and establishing capacity in the domestic tariff area of the developing countries. With foreign investment rules being uncomfortably stringent, this was hardly the best option. Further, devaluation, by reducing the foreign exchange value of domestic assets, permitted cheap access to crucial resources in countries that were prepared to allow foreign investment in areas like the extractive industries and agricultural production. The period immediately after the first oil shock saw a dramatic change in this scenario. The international payments system now reflected huge current account surpluses with the oil producers, relatively small surpluses with the developed countries as a group, and massive deficits in the developing world. Since oil surpluses were held in the main as deposits with the international banking system controlled in the developed world, the private financial system there became the powerful agent for recycling surpluses. This power was indeed immense. Expenditure fuelled by credit in the developed and developing world generated surpluses with the oil producers, who then deposited these surpluses with the transnational banks, who could offer further doses of credit. This power to the finance elbow was all the more significant because a slowdown in productivity growth in metropolitan industry had already been bringing the post-War indus­ trial boom to a close—a process that was hastened by the contrac­ tionary response to the oil shocks.

The Macroeconomics of Imbalance and Adjustment

141

From the point of view of the developing countries, these develop­ ments in the wake of the oil shocks appeared positive. Needing liquidity to finance their post-shock deficits, they found it easier to negotiate with a relatively atomistic banking system that could impose no conditions rather than the centralized multilateral financial insti­ tutions like the IMF. Banks flush with funds were keen to lend, and the possibility that the rather high current-account deficits they were financing were unsustainable was not considered. No level of the current-account deficit was unacceptably high. What mattered was that the nature of the international financial system hitherto had kept the volume of commercial borrowing by these countries relatively low. Thus, this congruence of interests—of the developing countries to borrow and the banks to lend—resulted in the fact that the currentaccount deficit was for almost a decade-and-a-half no constraint on growth in the underdeveloped world. This had two implications. To start with, it appeared to render import substitution as a strategy meaningless, since that strategy was prompted in the first instance by the external vulnerability that underdeveloped countries faced ever since the Great Depression. In a world awash with liquidity which could be easily accessed, that vulnerability hardly seemed to be a problem, let alone a factor that should be the prime determinant of the nature of growth strategies. Secondly, it appeared to remove all constraints on the maneuverability of the state in developing coun­ tries. Governments could incur massive deficits on their budget, since imports financed with international borrowing could ensure that any excess demand would not spill over in the form of inflation. That is, so long as there are no limits on the current account deficit, the inflation barrier is inoperative as well. The fall-out of this scenario is now history. Right through the seventies and eighties governments in one developing country after another combined more liberal growth strategies with huge budget deficits financed by international borrowing, since that partly neutral­ ized the adverse effects on domestic growth that liberalization had. In fact, during those years many developing countries actually recorded rather creditable rates of growth, which were often attributed to liberalization rather than the irresponsible pump-priming by domes­ tic governments, which the irresponsible lending practices of the international banking system encouraged. The boom obviously could not last, as it became clear that none

142

C . P . C h a n d r a sek h a r

of these borrowers were in a position to meet their debt-service payments, without resorting to further borrowing. This together with the evidence of the colossal overexposure of the international banking system in many developing countries set afoot the deceleration in the flow of liquidity that came to be called the ‘debt crisis*. The banks o f course could not pull out, because that would have spelt closure for many of them, as much of developing-country debt would have had to be written off rather than rescheduled. But they needed an agent to discipline governments whom they had in part tutored to be undisciplined. And the hitherto neglected policeman, the IMF, came in handy. For the developing countries involved, the situation was a virtual trap. The current-account deficit on the balance of payments which was hardly an indicator of the health of borrowers in the past was now found ‘excessive*. It was no more just a constraint on growth, but a problem in itself, since it could not be financed. And given the context in which it took the place of inflation as the bindiog barrier to growth, it could easily be identified as the result of profligacy rather than of external vulnerability. This changed attitude, which was in keeping with the IMF*s perspective on the matter was strengthened by the fact that an attempt to reduce the deficit by a retreat to a more insular regime was found near impossible. To start with, the borrowing splurge had meant that the interest payment component had risen to be such a large share o f outflows that the curb on imports had to be massive to ensure the necessary correction on the current account, which leaves amortiza­ tion uncovered. Secondly, the period of liberalization had created powerful lobbies interested in the government persisting with that policy, which rendered a retreat ‘politically infeasible*.

T H E IN TERESTS O F FIN A N C E In the event, most governments turned to the IMF for balance of payments finance, and had to accept in return a package that involved curtailing central bank credit to the government, intensifying trade reform, dismantling regulations on national and foreign firms and agents, devaluing and moving towards a convertible currency, privat­ izing the public sector, and in the new phase, reforming the financial sector. All of these were in keeping with the requirements set by the

The Macroeconomics o f Imbalance and Adjustment

143

rise to dominance of international finance. Trade liberalization and deregulation are inevitable elements of a strategy that provides the basis for international investments aimed at world-market oriented production that can be 'facilitated* with finance. Crucial resources in the hands of the state or the domestic private sector, like for example hydrocarbon resources, are rendered eligible for acquisition by foreign interests, so that real assets can serve as collateral for the financial transactions of foreign firms. And business conditions are made acceptable to financial agents through a liberalization of the financial sector and a gradual shift towards convertibility for capital account transactions. In sum, the current IMF-type adjustment package is Wide in scope and targeted towards the requirements of finance capital. As a result, neither fiscal nor monetary policy can be controlled in keeping with domestic objectives. In addition, real growth in the developing world has to be curtailed substantially. However, this appears to be im­ material so long as the objective of a larger share of a smaller cake being diverted to international financial interests is realized. It is in this sense that adjustment in its present form is nothing more than the macroeconomics of the international financial oligarchy.

REFERENCES A kyOz , Y ilmaz (1993),

4On Financial Openness in Developing Countries*, in UNCTAD, 1993, 110-24. B alassa , B ela (1982), 'Structural Adjustment Policies in Developing Economies*, World Development, 10, 1, 23-8. B haduri , A m it (1986), Macroeconomics: The Dynamics o f Commodity Production, Macmillan, Delhi. C handrasekhar , C .P . and J ayati G hosh (1993), ‘Economic Discipline and External Vulnerability: A Comment on Fiscal and Adjustment Strategies’, Economic and Political Weekly, 10 April, 28, 15, 66773. C ho , Y oon -J e and D eena K hatxhate (1989), Lessons o f Financial Liberalization in Asia: Comparative Study, The World Bank, Washington. D ornbusch , R udiger (1980), Open Economy Macroeconomics, Basic Books, New York.

144

C . P . C h a n d r a sek h a r

•Economic and S ocial C ommission for A sia and the P acific (1 9 9 3 ),

Economic and Social Survey o f Asia and the Pacific, 1992, part u, United Nations, New York. Helmers, F. Leslie, L.H. (1988), ‘The Real Exchange Rate*, in Rudiger Dombusch and F. Leslie L.H. Helmers (eds.), The Open Economy, Tools for Policymakers in Developing Countries, World Bank/Ox­ ford University Press, New York. K aldor , N icholas (1978), ‘What is Wrong with Economic Theory’, in Further Essays in Economic Theory, Duckworth, London. K alecki , M ichal (1971), Selected Essays on the Dynamics o f a Capitalist Economy, Cambridge University Press, Cambridge. K h a n , M ohsin S. and M alcolm D. K night (1982), 'Some Theoretical and Empirical Issues Relating to Economic Stabilisation in Develop­ ing Countries’, World Development, September, 10, 709-30. ------ (1 9 8 5 ), Fund-Supported Adjustment Programs and Economic Growth, Occasional Paper No. 41, International Monetary Fund, Washington. K h a n , M o h sin S. and P eter J. M o n tie l( 1 9 8 9 ) , ‘Growth-oriented Adjust­ ment Programs: A Conceptual Framework’, IMF Staff Papers, 36, 2 , 2 7 9 - 3 0 3 , International Monetary Fund, Washington. K rugman , P aul and L ance T aylor (1978), Contractionary Effects of a Devaluation*, Journal o f International Economics, 8, 4 4 5 - 5 6 . P atnaik , P rabhat (1991), ‘Devaluation, Dual Exchange Markets and Existence of an Equilibrium in a Flexible-rates Regime, Economic and Political Weekly, 28 September, 26, 2251-62. ------(1993), ‘Post-War Capitalism and the Problem of Transition to Socialism’, Indian Institute of Advanced Study, Simla, mimeo. P olak , J acques J. ( 1957), ‘Monetary Analysis of Income Formation and Payments Problems’, Staff Papers, November, 6,1-50, International Monetary Fund, Washington. ------(1990), ‘A Marriage between Fund and Bank Models: Comment on Khan and Montiel, IMF Staff Papers, March, 37, I, 183-91, International Monetary Fund, Washington. R ivera -B atiz * F rancisco L. and Luis R ivera -B atiz (1989), International Finance and Open Economy Macroeconomics, Macmillan, New York. SnG Lrrc, J oseph E. (1 9 9 3 ), T he Role of the State in Financial Markets’, paper presented at the Annual Bank Conference on Development Economics, The World Bank, Washington.

The Macroeconomics o f Imbalance and Adjustment

145

(1983), Structuralist Macroeconomics: Applicable Models fo r the Third World, Basic Books, New York. UNCTAD (1993), International Monetary and Financial Issues fo r the 1990s, Research Papers fo r the Group o f Twenty-Four, 2, United Nations, New York. T aylor , L ance

i

State Intervention in the Macroeconomy J ayati G hosh

I There can be no macroeconomics without a conception of the state. Whether the analysis makes it loom large as a major player in the economic arena, or reduces its role to one of behind-the-scenes ac­ tivity or even treats it as something which should be excluded from the domain of ‘positive economics* as far as possible, there is in­ evitably an underlying theory about the way in which the state and civil society are formed and interact. Rarely, however, are such theorizations made explicit in economic writing, which often renders the conceptions opaque, especially when discussing macroeconomic tendencies and even macroeconomic policies. Despite this frequently observed tendency to brush aside the implicit conception of the state, it remains crucial because it informs and affects the way in which macroeconomic processes themselves are viewed. Even when it is accepted that states in different societies have always been and continue to be involved in economic activity, whether through direct intervention or indirect setting of the rules of the game, a host of questions persist. These relate not just to the type of intervention which is desirable (which is the focus of most of the relevant literature) but also the nature of the state, the forces that act upon it and the interaction between the state and the civil society and the economic processes that result from this interaction. The ways in which such questions are dealt with determine the underlying theory of the state in different macroeconomic writings. Broadly, the various explicit and implicit conceptions of the state in economic literature fall into four categories: the idea that the state is an interloper, infringing upon territory which would be better left to the workings of markets and other social and contractual

State Intervention in the Macroeconomy

147

mechanisms between individual agents, which is seen in an extreme form in the more libertarian of neoclassical writings; the instrumen­ talist notion of the state, as a deus ex machina detached from society, whose activities are designed to further certain explicit goals, as in the perception of Keynes and latter day development economists who see it as an agent of growth; the view that the state is essentially an intermediary, mediating between (or succumbing to) the claims of different groups, classes and lobbies, which can be gleaned from writings such as those of the ‘rent-seeking* literature; and finally, the incorporative conception of the state, in which it is basically em­ bedded in society, being crucially affected by institutions and proces­ ses within it (including economic ones) and in turn affecting them by its own activities. This categorization does not really correspond to those which are based on ideology or even approach to economic analysis, but rather cuts across them. This is because analytical discussion on the role of the state is all too often obscured by positions on what should be its role, which are informed in turn by relatively stark distinctions drawn between the ‘the role of the state’ and ‘the role of markets’. In macroeconomic analyses of developing economies in particular, the approach has been critically affected by prevailing ideological and normative positions. Thus, much of the recent literature on state intervention in the economy is either explicitly hostile to it or defensive about its thwarted possibilities. Yet even a generation ago, the necessity of and the benefits from government economic invol­ vement were axiomatic among economists in developing countries. Indeed, the latter part of the twentieth century can feasibly be described as one in which the role of the state has dominated discussions of economic development. The problems and dilemmas which are increasingly identified with state involvement in a developing economies have beenaccompanied by numerous theoretical arguments, which stem from varying per­ ceptions of the rationale for government intervention and the nature of such involvement. This chapter is concerned with a critical appraisal of some of these theories, and their applicability to the Indian experience o f the past several decades. In what follows, ‘state* and ‘government* are often used interchangeably, even though the two are obviously not strictly synonymous and the concept of ‘state* is broader and richer than the administrative functions implied in ‘government*. This is because the concentration specifically on

148

J a y a ti G h o sh

macro-economic policy allows for this type of coalescing. Several sets of theories are considered in the next section, ranging from neoclassical approaches to the economic role of the state through developments such as rent-seeking activities, to the discussions on ‘soft’ and ‘hard’ states, to some Marxist theories. In the third section, analyses specifically relevant to explaining the Indian macroeconomic experience are considered.

II The neoclassical economic literature on the role of the state has as its starting point the question: why does the government exist at all for purposes of the economy? Indeed, in an economic paradigm in which the historical process is given no recognition, this is an obvious question since if markets function efficiently and easily as in the basic model, there is no economic role for the state at all. The answer to the question in turn must come from the possibility of market failures or other limitations of market functioning, in which case the role of government can be defined simply as that which nudges the markets to complete and perfect functioning and Pareto optimality through various policy devices. But there is a prior, and more crucial role for the state in this theory, that is, the minimalist role of government as the guarantor of private property rights. Some theorists have argued that this should be the only macroeconomic function of the state, as the minimum necessary for social cooperation between individual agents (Nozick, 1974). Without explicit state sanction and authority, obviously private property cannot exist. And the establishment of individual rights to initial endowments (including essentially inheritance) and to the gains from trading these endowments, is accepted within the theory as a clear necessary condition for the use of markets to obtain a Paretoefficient allocation. With the establishment, and state backing of private-property rights, the theory states that an economy in which markets span all current and future goods can achieve a Pareto-efficient allocation. Therefore, 'efficient* economic functioning in this sense is predicated upon the government as the enforcer and protector of private property rights, both in terms of initial endowments and the retention of 'gains from trading’. It also follows from this that state reallocations of initial private resource endowments (or assets)

State Intervention in the Macroeconomy

149

for redistributive purposes can disturb the efficient functioning of markets, and that state activity should be correspondingly restrained in this regard. Thus, activity of the state beyond the requirements of protecting private property and gains from trading are seen as examples of ‘interloping* behaviour, which would have negative welfare implications. Quite apart from the obvious philosophical problems with the use of the Pareto principle in such normative economics, the point is that the final allocation is inevitably tied directly to initial endowments, and unequal distribution of endow­ ments through the property rights system can still ensure a stable outcome. Also, this basic and fundamental requirement of the state within this paradigm of economic activity can involve some contradic­ tions given the way that markets function. Thus, protection of private property rights may actually impede ‘efficient market functioning* in certain cases, as the growth of monopolies makes evident. Conversely, most markets require state recognition and backing (if not actual overseeing) for their very survival. Further, the large area of economic activity not covered by market relations as generally understood, whether it be common property resources, household production and reproduction, etc. remains untouched within this approach. The minimalist function of the state decries all other aspects of state economic activity on the assumption that markets exist to allocate all valued goods and services efficiently. This perception is too far removed from reality to be widely acceptable among theorists, so that the focus of most neoclassical attention has been on the more obvious cases of market failure, and the role the state can play in rectifying the effects of such failures. The major areas include public goods, externalities, industries characterized by increasing returns to scale, and cases of incomplete information. More recently, the possibility of aggregate unemployment has also been considered by some as an instance of ‘market failure*. Even as regards these, however, it is sometimes argued that market institutions are not being allowed to operate to their full potential, and that state intervention should be indirect, helping markets to work better, rather than direct. The underlying principle, of course, stems from the ‘interloper* theory that in general government intervention is the main originator of ‘distortions* from the free market ideal, and that even when markets have failed, state action can actually create worse economic alterna­ tives from a welfare point of view. This perception permeates almost all the mainstream literature. More recent attempts to describe how

150

J a y a t i G h o sh

state policy should be directed towards making markets more 'peoplefriendly’ are essentially variants of this view despite their rather different terminology. Within the neoclassical paradigm, the basic reason for ‘market failure* can be traced to the lack of cooperative behaviour among self-seeking, utility-maximizing agents in situations where such cooperation is essential for beneficial trading. The possibility of free riding on public goods is one such case, which means that the production of public goods would always remain sub-optimal with sole reliance on market institutions (Samuelson, 1954). The role of government in this case is to provide public goods which reflect the ‘true* preferences of the members of society, in a way which is not possible for competitive mechanisms. A similar function for the government is to compensate for externalities in production. Some have argued, basing themselves on Coase’s famous theorem, that even in the presence of externalities, government intervention is not necessary since individuals can privately strike a mutually advant­ ageous and Pareto-efficient bargain to determine the level o f the external activity. However, since Coase's conclusion depends on a very special (and unreal) world of two agents, costless bargaining and complete information, as well as no ‘free riding', it is not considered a convincing alternative to government intervention which goes beyond ‘minimal state’ requirements of protecting private property, even by most neoclassical theorists (Coase, 1960; Cooter, 1982; Inman, 1987). Indeed, it has been pointed out (Lemer, 1962) that it is also dangerous to assume that market institutions spring up auto­ matically whenever and wherever there is ‘need* for them. Thus state policy is suggested as relevant to fill such gaps in market formation. While the possibility of increasing returns to scale in production has been recognized for some time, this potential source of market failure has only recently come to be accepted within the neoclassical tradition as a reason for economic intervention by the state. The case here is made on the grounds that pricing decisions will be taken so that price equals average rather than marginal cost, thus, implying a shortfall in potential output. Pricing at marginal cost (which would be Pareto-efficient) would occur only if the firms could cover long-run losses through some system of lump-sum charges on all consumers, which in turn is only possible if the demands and consumers for the product are known before the production facilities are installed. The role of government here is to extract from both producers and

State Intervention in the Macroeconomy

151

consumers the *true* costs and benefits of producing the commodity. This is essentially a limited intervention which is really more that of providing an information network which can transfer resources between agents in industries with increasing returns, and does not affect the substantive nature o f production in these sectors. Incomplete information creates market failures essentially because it can give rise to two sets of problems. The first includes those related to moral hazard, eg. whereby producers can supply goods/services of inferior quality, while the second relates to adverse selection, typified by the possibility of purchasing a lem on’ in the market for used cars (Akerlof, 1970). It has been argued that market institutions can be evolved to overcome these problems, as in contingent contracts, professional associations or importance given to seller reputation. But these are at best limited in scope and cannot reliably overcome such difficulties. Where these problems are severe, the need arises for government intervention primarily to ensure that more information is available and supplied. (Holmstrom, 1984). Information here has the characteristics of a public good, in which case the first set of arguments for government intervention hold. More recently it has been argued that problems associated with learning* constitute an important source of market failure which can necessitate state intervention in the process of development. Thus, (following from Arrow, 1962) problems relating to innovations, their introduction and adaptation, the ability to alter product-mixes in response to changing economic environments, etc. have been seen as areas in which state action can be effective. Recent theory brings out a host of problems with laissez-faire which stem not only from incomplete markets and imperfect information, but also from strategic interactions, principal-agent problems and transactions costs, all of which reinforce the justification for state involvement (Stiglitz, 1987). These issues form the subject matter of public economics, which is the major form of explicit neoclassical consideration of state economic activity (Samuelson, 1954). The focus is on the technical reasons for market failure in specific areas, which in turn means that the inquiry is directed towards a limited set of questions: What are the goods/services for which the market cannot achieve the Paretoefficient outcome? What is the mechanism through which collective choices can/should be made in order to determine the decisions of the state? What is the nature of individual agents* responses to the state intervention consequent upon these public collective choices?

152

J a y a t i G h o sh

This restricted focus reduces the explanatory power of the theory when assessing the actual patterns of state involvement in the economy, whether in industrial or developing countries. The discus­ sion can at best be descriptive or make welfare-theoretic propositions along the lines mentioned above. Since there is no conception of the form of the state, the actual policies undertaken by the agents of the state are seen only in relation to the ability of the state to nudge economic activity back to some abstract Pareto-efficient ideal. The fact that both historically and currently it is impossible to find an economy in which the state (directly or indirectly) is not the main regulator of economic activity, whatever be the importance or effec­ tiveness of markets in that system, cannot be explained at all within the confines of this theory. Furthermore, explanations of'state activity must come largely from ‘outside’ the theory: thus, externalities which lead to market failures are themselves external to the theory rather than emerging from the very nature of the production process. Similarly, shifts in public expenditure or in the methods and extent of revenue raising can come about in this paradigm only through changes in individual preferences or in productive technologies which too are externally determined. Since there is no concept of cjass employed (or indeed any other socio-economic categorization), and the only groupings considered are shifting coalitions which are based merely on the accidental and temporary convergence of individual goals, there can be no analysis of the motivations or effects of state economic policies on groups of people with different relations to the process of production. In fact, the basic implication of state policies in capitalist or mixed economies—that they affect different classes differentially, and that such effects may be intended— is missed out altogether within this relatively static and ahistorical paradigm. In addition, the micro economic focus of public economics has meant a neglect or lack of understanding of the macroeconomics of state intervention. : The major lacunae of this approach in providing an understanding of state economic behaviour cannot be filled without a change of para­ digm. However, a recognition of these limitations has led some neoclassical economists to attempt to fill at least two major perceived gaps: the question of how state interventions affect particular groups and the implications thereof (nowadays referred to as ‘the new polit­ ical economy*) and an attempt to explain state involvement in the macroeconomy, particularly in determining the level of employment

State Intervention in the Macroeconomy

153

and activity. Some of this is related to the concept of the ‘political business cycle’ and associated analyses. The ‘new political economy*, despite its grandiose name, is essentially an extension of micro economics. Most of the literature on ‘rent-seeking* which stems from state economic involvement and/or control is based on micro-theoretic considerations, usually with examples from trade and commercial policy. This literature is very much within the category of analyses stemming from a perception of the state as ‘intermediary*, in which state actions particularly in the economic arena result from the activities of lobbies and pressure groups within society. The early contribution which has influenced much of the subsequent discussion (Krueger, 1974) was devoted to describing different forms of rents stemming from various govern­ ment regulations, and a consideration of the welfare implications of import quotas in particular, theoretically and with reference to India and Turkey. Krueger argued that government controls such as import quotas can generate economic rents for the licence-holders, which are so large as to inspire competition for these licences. Private resources which could have flown into productive channels are, thus, diverted towards competition to secure these licences so as to avail of the rents, and this leads to a welfare loss for the society. However, Krueger*s argument that quantitative restrictions such as quotas necessarily involve a greater welfare loss than tariffs because of rent-seeking, was contested by Bhagwati and Srinivasan (1980) who pointed out that import tariffs in turn may lead to ‘revenue-seeking* and that such activities may actually be welfareimproving. This is because, in a welfare economics context, resource diversion to ‘unproductive* activities is wasteful in a first-best scenario, but the very presence of tariffs (or quantitative restrictions) creates distortions that result in a second-best situation, in which rent-seeking and revenue-seeking may increase welfare or reduce it depending on particular assumptions. This result, emerging from the ‘interloper* paradigm, reinforces the orthodox mainstream position that state regulation of free market functioning involves so many distortions that parallel markets or corruption may actually be cor­ rective mechanisms. (Incidentally, it is interesting to note that even to date, mainstream theorists almost inevitably draw on the example o f import tariffs to illustrate ‘bad* state intervention.) Krueger*s article described various forms of government inter­ vention which generate rents and therefore imply a diversion of

154

J a y a t i G h o sh

productive resources to the unproductive activities of competing for such rents in unofficial markets. Among her list of such government measures were the following: Fair trade laws result in firms of less-than-optimal size. Minimal wage legislation generates equilibrium levels of unemployment above the optimum associated with deadweight losses . . . Ceilings on interest rates and conse­ quent credit rationing lead to competition for loans and deposits and/or high-cost banking operations. (Krueger 1974, 301-2.) Rent-seeking activities associated with each of these lead to the economy performing below the production possibility frontier even beyond the shortfall resulting from the deadweight loss associated with monopolies. Subsequent work, (eg. Buchanan and Tullock, 1963; Buchanan, 1970; Brennen and Buchanan, 1977; Buchanan et al., 1980) has essentially developed and elaborated upon these and other similar processes, highlighting in particular the importance of political lobbying in the USA. Similarly, Bhagwati (1982) describes various other forms of ‘directly unproductive profit-seeking* (DUP) activities associated not merely with regulations and quotas, but with other forms of govern­ ment activity including tariffs. The essential propositions of this argument are, thus, threefold: firstly, that people will spend resources to capture property rights from the government, giving rise to rentseeking behaviour; secondly, that interest groups will spring up to defend their positions; and thirdly, that since many agents within the government are self-interested, there will be conflicting interests within the government that add to waste. As a result, there will be acts of omission and commission such that ‘in many countries, there could be little question but that government failure significantly outweighed market failure*. (Krueger, 1990: 10) The argument about the implications of rent-seeking activities for welfare can be countered within a micro economic perspective (as in Bhagwati and Srinivasan, 1980), but there are further and more substantive problems with it. The idea that resources can be diverted from productive investment to unproductive channels through bribery, parallel market dealings, etc. is both plausible and well known. The economic ramifications of state patronage and clientelism cannot come as a surprise to any minimally aware resident of either developed or developing countries. However, the concentration on government regulations alone seems to be strangely limiting. Citizens

State Intervention in the Macroeconomy

155

o f India, for example, need no reminding that the greatest scope for the private appropriation of such rents comes not from state regula­ tions but from the awarding of government contracts, whether in defence or infrastructural expenditures (considered by all to be neces­ sary prerogatives of the state). Competition for government jobs or direct transfers, or even for benefits from licences, typically involve much smaller amounts of rent than those to be garnered through manipulation of such 'necessary* government expenditures. In recent times, many more pecuniary advantages of clientelism derived by favoured groups or individuals have accrued in circumstances of reduction of state economic involvement, such as schemes of privatization which allow for the sale of public enterprises to favoured buyers at below market values. Deeper consideration of the issue also suggests that such pecuniary benefits which come from manipulation of markets (if necessary, through other markets) need not always stem from government activity, but can be created by private agents themselves. Thus, deals across monopolies or oligopolies for sales/purchases/credits, bribery to ensure the tying of contracts, etc. all form part of the ‘normal* functioning of capitalists everywhere across the globe, certainly not merely in ‘corruption-ridden* developing countries. Thus, rent-seek­ ing competition can occur even in areas not marked by significant government intervention. (The activities of the Mafia, and their forms o f economic control, both in the USA and in Italy, form a crude but telling example.) In all capitalist societies today, vast sums and creative resources are expended in activities such as advertising and related areas, which are clearly unproductive from a ‘classical* standpoint, essentially to gamer economic rents from control over markets. Thus, to view ‘rent-seeking’ in simplistic terms as stemming from government regulations per se rather than relating them to the political and economic configurations within the society and the weapons for control amongst different groups within it, is not very useful. It also leads to the mistaken conclusion that doing away with government controls, for example, and moving towards a marketbased system, would eliminate rent-seeking behaviour and the as* sociated waste. Certainly in the Indian context, isolating certain government activities (such as import quotas) as rent-seeking, and ignoring the entire range of other activities and processes of both government and ‘civil society* which generate rents and related corruption, can be extremely misleading. This approach can also lead

156

J a y a t i G h o sh

to simplistic monocausal explanations for all economic ills in such societies, (as in Deepak Lai, 1983) whereby ‘irrational dirigisme* becomes the only culprit for the distorted, unbalanced and slow growth of economies like that of India. A further problem with this line of reasoning relates to the macroeconomic implications of this resource diversion into ‘unproductive’ activities. Consider Krueger's description of the extreme case, in which, with perfect (government) restrictions, regulation would be so all-pervasive that rent-seeking would be the only route to gain. In such a system, entrepreneurs would devote all their time and resources to capturing windfall rents. (Krueger 1974: 302.) Obviously, such a system, in which productive activity would be zero, could not sustain even simple reproduction, and is, therefore, logically unviable. Krueger explains that this simply represents the other end of the continuum from the regulation-free, and, therefore, rent-less economy. Her discussion of the additional ‘deadweight loss' from rents suggests .that these resources virtually disappear from the economy, a simple withdrawal from the circular flow, as in hoarding or capital flight. But if this is perceived to be unrealistic, and in fact these resources are somehow spent within the system, then their aggregate effect on the economy is not clarified by this argument, unless full employment is assumed. Is consumption the final form of these resources which are spent on lobbying, bribery, etc. and if so, is the argument that rent-seeking activities actually lower the investment rate in the economy over time? Or do they get reflected in higher costs of productive outlays, thus, leading to an increase in incremental capital-output ratios? To all this, there is the (Keynesian) response to be made that all expenditure, ‘unproductive* or otherwise, constitutes an increase in effective demand, which would encourage increased production and, therefore, investment in economies with slack. Indeed, in economies with unemployment or unutilized capacity, rent-seeking expenditures may actually lead to a rise in the equilibrium level of production and consequently (by influencing expectations) lead to increases in the rates of growth of future investment and production. Of course, the argument takes on another dimension if it is assumed that any otherwise fully-employed resource is diverted to rent-seeking activities, since this would lead to inefficiency and slack. In under­

State Intervention in the Macroeconomy

157

developed countries such resources are not difficult to enumerate: scarce foreign exchange, skilled professional services, etc. But, as pointed out earlier, such resources may be occupied in ‘rent-seeking* for reasons quite unconnected with government regulations as such, related instead to macroeconomic features such as the pattern of income distribution. The rent-seeking literature is based on the insight that different government policies affect various groups, strata and individuals in a society in different ways, and, therefore, energy and resources are expended in attempts to influence these policies. But the obvious fallacy is to link this with macroeconomic growth tendencies and potential. Any comparative study would bring out the fact that, in rough empirical terms, there appears to be no relationship between rent-seeking and associated corruption, and the level of income and rate of its growth (Eisenstadt and Lamarchand, 1981). To take only one obvious and well-known case, the experience of Japan amply testifies that an elaborate network of bribery, kickbacks and corrup­ tion involving a nexus of politicians, bureaucrats, businessmen and ‘underworld* gang leaders can still be part of an economic system that produces enviable rates of income growth and export perfor­ mance. Nor does the argument sufficiently capture the extent and ramifications of ‘waste’ and associated economic slack in capitalist systems, which arise from much more than simply government regulation and intervention. In a similar vein it has been pointed out (Shapiro, 1988) that the state can also extract ‘rents* from the private sector, in particular one in which oligopolies are dominant. This is why treating such waste as emanating only from ‘government failure* is misleading. This branch of the ‘new political economy* has another important limitation: the absence of any real sense of politics in the economic forces at work, apart from a cursory look at some lobbies and interest groups which benefit from very specific (and limited) government regulations. It has been pointed out by radical critics that this kind of ‘political economy’ posits a peculiar counterfactual alternative which combines a non-politically organized society with an ideal liberal state which is completely neutral with regard to income distribution. (Shapiro and Taylor, 1990; Bowles and Eatwell, 1983; Whynes and Bowles, 1981.) Indeed, this is a deficiency in much o f the ‘state-asintermediary* literature: despite the focus on influences on govern­ ment policy, there is no systematic attempt to identify the elements

158

J a y a t i G h o sh

of the ability of different groups and classes to affect the direction of state intervention or to see how the policies themselves act on the relative power positions of these groups. Much as in the other public economics literature described earlier, both the functioning of markets and the nature of state intervention are seen as somehow distinct from and independent of the nature of society at large and the social and institutional forces that shape interactions and the relative power of different groups. Proponents of the ‘incorporative* approach, by contrast, would argue that a holistic appraisal of the totality of the state’s intermeshing with and impact on the economy would move beyond micro economic confines, as well as embed both state and markets in the pattern of society itself and its changes. The absence of a macroeconomic perspective, which leaves the rent-seeking literature full of unanswered questions, is sought to be remedied in the other stream of ‘new political economy’ writing ranging from some of the work of James Buchanan and others, to the Italo-Swedish school and work on ‘political business cycle*. This concentrates on the effects of aggregate government expenditure and taxation, and in particular the level of the government budget deficit, for the functioning of the macroeconomy. The discussion is couched in terms of ‘endogenizing the government’ as a variable in macroeco­ nomic modelling, through certain assumptions about the govern­ ment’s expected behaviour over the cycle and with respect to elections. Much of this writing (in particular Buchanan’s work) has argued that the context of political democracy is one which systemati­ cally biases the fiscal and monetary policy instruments so that they are applied non-symmetrically in favour of government deficits. Fiscally this operates both on government expenditure (as various groups and lobbies demand transfers, subsidies and public employ­ ment) and on resource raising (as others, or even the same groups, resist higher taxation). Governments striving to be popular, thus, inevitably tend to run deficits in their budgets, spurred on by the eco­ nomic justification provided by Keynes and his followers. But these deficits, resulting from the short-run appeasement of various groups, create both short-run and long-run problems for the macroeconomy. In this argument, the government deficit serves to retard capital formation in a variety of ways. The first is through bringing about inflation, which disturbs the institutional framework of production and generates more instability, uncertainty and, therefore, risk. Inter­ temporal planning, (eg. of medium term investments), becomes more

State Intervention in the Macroeconomy

159

difficult and prone to error because markets may send out 'distorted' signals, and therefore shifts in relative prices may alter patterns of resource use in unsustainable ways. In an inflationary situation, the entrepreneur's returns from productive activity fall relative to the private gains to be made from successful adjustments to inflation. All this implies the alteration of the economy's basic structure of produc­ tion, with long run implications for investment, output and produc­ tivity. Further, with inflation the fiscal drag allows the state to tax fictitious (or purely nominal) profits and, thus, increase its share of value added at the expense of private producing sectors. In addition, debt-based deficits put pressure on domestic credit and interest rates, leading to crowding out of private investment which over time leads to capital shortage as the capital stock shrinks because of lack of investment. The result may be a situation characterized by under­ utilization of capital, resulting not from a surplus o f capital itself, but from a shortage of complementary capital because of inadequate past investments. This does not imply a deterrent to inflation, but actually has stagflationary consequences because of emergent bottlenecks in production. The solution to this vicious circle, proposed by Buchanan and Wagner (1977) is constitutional provision forcing governments to balance their budgets, on the grounds that ‘budgets cannot be left adrift in a sea of democratic politics’. (Buchanan and Wagner 1977: 175.) This argument of course has strong contemporary overtones, especially in the context of the American economy. And this approach does highlight some of the more obvious negative implications of budget deficits and the inflationary process in capitalist economies. But despite its more overt consideration of the state as an integral part of the society, subject to pressures and pulls from within the system, it also is limited in concentrating on the deficit alone as the important variable in determining macroeconomic processes. This is a strand common to most policy discussions of government involve­ ment across the world today, in which the budget deficit, defined simply as the excess of state expenditure over income, is considered to be the only relevant indicator of state performance and is seen inevitably as ‘bad’ for the economy in normative terms. The more recent literature on these lines takes its cue from such work as well as the argument of Nordhaus (1975). Again in the context o f electoral democracy (and the related desire of governments to win elections), Nordhaus argues that politically determined policy choices

160

J a y a t i G h o sh

involve higher inflation and lower unemployment than would be ‘optimal' in terms of the impact on future generations. This in turn tends to be associated with a ‘political business cycle’ with greater unemployment and deflation in the early years of a regime followed by inflationary booms as elections approach. Similarly, on the external front, the official concern with BOP deficits and reserve losses will be greater at the start of elected regimes and less towards the end. The pattern is echoed by Lindbeck (1976) as follows: a rapid increase in unit labour costs leads to a fall in profits and declining investment and employment in tradeables as well as a reallocation o f resources to non-tradeables, especially in the public sector. After some time, a more restrictive economic policy involving devaluation or cuts in wage-led inflation restores profit and investment incentives in the tradeables sector, possibly to be followed by a new phase of rising wages and falling profits, etc. This political-economic business cycle of profit-investment-devaluation-demand management is said to be typically of longer periodicity than the ordinary business cycle. Some empirical substantiation of this cyclical process in advanced capitalist economies in the postwar period has also been attempted. (Estey and Caves, 1983; Hibbs and Fassbender, 1981.) The argument appears to be plausible on a very broad level, but it has been criticized on the grounds that it is neither theoretically convincing nor supported by evidence (for a cogent critique, see Alt and Chrystal, 1983). Indeed, it is a poor theory of electoral behaviour that links voting simplistically with economic booms and slumps. In turn, patterns of government spending and balance of payments man­ agement have not followed electoral cycles in the manner described, nor have ‘regime changes* always resulted from unpopular economic circumstances such as recessions with high unemployment. Social tolerance of inflation— or indeed of levels of unemployment— varies across countries and over time according to the relative strength o f different groups within it, and cannot always be captured by theories invariant to such power configurations. Indeed, as Lindbeck had noted , ‘there are instabilities in both the market system and the politicaladministrative system, and . . . these interact in a complex way* (Lindbeck, 1976:11). The problem is that often these interactions are so complex that they do not lend themselves easily to the kind of simple modelling attempted by such theories (Boddy and Crotty, 1975). Extensions of this argument have gone into the area of public debt.

State Intervention in the Macroeconomy

161

usually to the effect that political incentives and constraints create a bias towards both high debt and high inflation. Where changes in the political regime through elections are common, public debt can be used strategically to influence policies of future governments with different perspectives. Thus, Persson and Svensson (1989) argue that ‘right wing* governments find it optimal to issue debt so as to force ‘left wing* successors to curtail public spending. Alesina and Tabellini (1992) obviously basing themselves on the Italian experience, show that differing political regimes will issue more debt, the greater the disagreement across regimes about how to spend public money and the greater the instability (in terms of frequent elections or changes of government) in the political system. Although it is not made explicit, these contributions conform to the idea that the state responds automatically to certain demands— essentially populist in nature—which bias it towards greater expen­ diture and larger deficits. This is another variant of the ‘intermediary* description o f the state, in an extreme form which assumes that the only relevant goal of the state is to stay in power. In this conception, there is little possibility o f independent state action which is not influenced by some group within society or seeking to influence them in order to win elections. This is seen as a starkly alternative possibility to that of the independent state which undertakes ‘good* economic policy decisions regardless of the political outcome. This is a dichotomy which is not very different from the concept popularized by Myrdal (1970) of ‘soft* versus ‘hard* states. (The quintessential example of softness for Myrdal was the independent postwar Indian state which he saw as unable to insulate itself from pressure from below.) Later terminology has developed what is broadly the same distinc­ tion, in terms of ‘autonomous* and ‘subordinate* states (see Rodrik, 1992). The autonomous state is defined (in analogy from microeco­ nomics) as the Stackelberg leader vis-a-vis the private sector, whereas the subordinate state is the follower. The argument is, that the subordinate state is ineffective in intervention to remove market failures because it sets its policies after the private sector has selected the level of activity which the policy is geared at regulating; thus, it lacks incentives to reward or punish desired behaviour or the lack of it. As a result, compared to the autonomous state, the subordinate state 'systematically underprovides economically desirable interven­ tions, and systematically overprovides politically-motivated (and

162

J ayati G ho sh

economically harmful) interventions’ (Rodrik, 1992: 334). This in turn suggests that the same policy, (eg. an import tariff or and export subsidy) would have completely different content and implications depending on whether it is implemented under subordinate or auto­ nomous states. Criticisms of subordinate ‘populist’ states range from arguments that they encourage corruption and are ineffective, to the even more dramatic claim that they are responsible for such major repercussions as the build-up of external debt in Latin American countries. (See, for example Dombusch and Edwards, 1990.) In a peculiar sense, this is still ‘political economy’ without the politics, since it envisages the possibility of the state being ‘auto­ nomous’ of the rest o f society or particular groups within i t The latter involves essentially an ‘instrumentalist’ notion of the state which sees it as actively pursuing certain goals without having to succumb to any internal pressures. A related idea is that ‘good’ policy advice can nudge states into a state o f greater autonomy in this sense. A major lacuna of this approach is that it completely ignores the distributive effects o f the government’s expansionary and contractionary policies. It is well known that those who benefit from periods o f fiscal-led expansion are rarely those who bear the burden of the subsequent adjustment which rather falls on poorer and weaker groups, and only the most naive of observers would find this result purely accidental. Thus, an ‘incorporative’ notion of the state would quesdon this argument on the grounds that treating macroeconomic policies as ‘populist’ or soft versus hard stabilization policies, simply obfuscates the point that government policies primarily have distributional implications, and reflect power equations within society. One extreme form of ‘subordinate’ state in the sense defined above is the focus of Janos Komai’s discussion of ‘hard’ and ‘soft’ budget constraints, for firms as well as the public sector as a whole (Komai, 1974, 1980, 1982). This concept, which has recently acquired much popularity, was developed with reference to socialist economies which Komai saw as essentially ‘resource-constrained’ unlike ‘demand-constrained’ capitalist economies. Soft budget constraints in this context are applicable at the micro level, for the production unit or firm, and can be identified ultimately by the simple fact of survival of the firm despite continuing losses. Thus, ‘general insol­ vency or financial fai lure almost never occurs in the sense of true arid final bankruptcy* (Komai, 1980B: 518) because the state helps the firm out of trouble. This can be done in a variety of ways, ranging

State Intervention in the Macroeconomy

163

from subsidies and softer credit to tax rebates/reductions and centrally adjusted input or output prices. Related aspects of the soft budget constraint are that bank credit is generally freely available for working capital requirements, and the proliferation o f inter-firm credit allows for a further looseness o f the budget even beyond the centrallydetermined limit. But the essential point is that all this is possible because ‘the state is a universal insurance company which compen­ sates the damaged sooner or later for every loss. The paternalistic state guarantees automatically the survival of the firm.’ (Komai 1979: 806) This is an intriguing notion of state involvement, in which the state is all-encompassing in shaping the economic system, but totally passive in directing it according to any priorities, since all firms equally face soft budget constraints. It is, thus, an instrumentalist’ state, but a rather inefficient one, since other than emphasize ac­ cumulation it appears to have no objectives for its instrumentality. Therefore, the state has no specific targets in either aggregate production and consumption or its distribution, and is completely subordinate to firms once the initial parameters have been fixed. The state in this model cannot hope to affect even the aggregate level of employment and output, which is determined instead by bottlenecks in supply resulting from shortages which are actually created by the system. The very softness of the budget constraint is what results in a resource-constrained system and the emergence and multiplication o f shortages. The major shortage Komai describes is that of labour, which can affect any one sector or industry, affect production there and consequently translate into a materials shortage. According to Komai, once this institutional character of the system is established, state macroeconomic policies— either fiscal or monetary— have little role in affecting the level o f economic activity or the phenomenon of shortage. Since money plays only a passive role with soft budget constraints, restricting money supply is never an obstacle to the expansion of a firm. Similarly, ‘there is no unequivocal or necessary causal relationship between the balance in the state budget and the intensity of general shortage in a resource-constrained economy’ (Komai 1980B: 528). Nor need there be any Phillips-curve type trade-off between shortage and inflation. Since the shortage depends upon institutional conditions, it is compatible with stable, falling or rising price levels. Also, full employment of labour is seen as a consequence of the soft budget constraint just as conversely it is

164

J a y a t i G h o sh

argued that ‘a genuine hard budget constraint actually keeps reproduc­ ing unemployment’ (Komai, 1979: 815). In this conception, the state practically does not exist in the economy other than as an all-encompassing financier with no strings or conditions. The obvious question is, what determines the macroeconomic equilibrium in such a system? Since there is no limit to demand from enterprises (and indeed Komai describes a permanent investment hunger, in which 'demand for investment resources is almost insatiable’ (Komai, 1980A: 195)), the system will always tend to overinvest and overspend relative to aggregate resource avail­ ability, and, thus, generate shortages. Komai, therefore, speaks of a ‘socially acceptable limit* of shortages, exceeding which would cause the state to step in with a hardening of budget constraints. But if the state can do so at that point, then indeed it can, within the same institutional constraints, do so at any earlier point as well, including the point, at which no shortage is manifest. Implicit in this argument, therefore, is the concession that the state can affect both the macroeconomic equilibrium and the level of shortages, if it is sufficiently responsive to the social unhappiness felt over the extent o f shortages. If so, it can also prioritize resource flows, according varying degrees of 'softness* or 'hardness* of individual firms* budget constraints depending upon their perceived importance for social and economic welfare. This very possibility undermines Komai’s basic argument: if the extent of shortage/resource constraint is ultimately determined not by the institutional conditions summarized by the presence o f the ‘soft budget constraint’ but by state response to social tolerance limits, then the nature of the state and what determines its actions become relevant. Here Komai’s theory provides no illumination at all, since it treats the state as a ‘black box’ which simply doles out soft finance to firms. Much of the more recent literature described above arose as a reaction to the perception^of state involvement in the ‘development* literature of the fifties and sixties which was not only explicitly ‘instrumentalist’, but largely tended to ignore or simply to assume away some of the complexities and difficulties associated with state intervention which are now common knowledge. The notion of the developmental state, which underlay numerous Third World in­ dustrialization attempts in the aftermath of the second World War, found justification not only in various theoretical developments of the period but also the actual historical experience o f important

State Intervention in the Macroeconomy

165

capitalist industrialized countries.such as nineteenth century Ger­ many, France and later Japan. There emerged a consensus among economists studying the process o f development that late indus­ trialization was not possible without active state involvement and participation, based on both theory and experience. The need for some kind of planned development directed by the state seemed obvious for an economy with low savings (Nurkse, 1953), inadequately developed physical and social infrastructure, the presence o f many externalities in production (Rosenstein-Rodan, 1943; Scitovsky, 1954), the likelihood of strong forward and backward linkages in production and investment (Hirschman, 1958), and the need for major 'catching up’ in terms of technology (Gerschenkron, 1962). In addi­ tion, social realities underlined the need for state economic involve­ ment. In many post-colonial countries, the industrial class itself looked to the state for assistance in its own development and in meeting the challenge o f larger multinational producers. The unequal distribution of income in most societies implied a low base for mass consumer demand which, if left to be determined entirely by market forces, would distribute investible resources in socially inoptimal ways. The requirements o f changing the production structure of the economy towards greater diversification and higher productivity ac­ tivities meant some degree of control over external trade as well as conscious directing of internal economic forces. These obvious needs, it was more or less assumed, would be met by the activities of the ‘developmental state’ which was seen as somehow external to the society, certainly above the fray in terms of the social conflicts be­ tween groups and classes over distribution of resources and incomes. It was this unrealistic assumption that reduced the persuasiveness of any argument in favour o f government intervention once ‘govern­ ment failures’ became widely observed and discussed (Hirschman, 1982). Also, state intervention became closely associated with importsubstituting industrialization, which was seen to involve too many problems in terms of resulting tendencies to autarky and increased domestic inefficiency. While it may be true, that the very success of state intervention is what has created conditions for criticisms about it subsequently, it is also undeniable that the ‘instrumentalist’ percep­ tion of the state in these writings made them over-optimistic about the ability of states to achieve stated economic goals without due reference to social and political realities, and in particular the ability of important groups and classes to influence government decisions

166

J a y a t i G h o sh

and be affected by them. The handful o f countries that are now accepted in the neoclassical development literature as successful examples of the ‘developmental state* (Japan, South Korea, Taiwan) have been countries marked by a strong degree of cultural homo­ geneity, a very strong, indeed authoritarian state, an unabashedly ‘capitalist* and usually mercantilist motivation which meant that government policies were very closely linked with the aspirations of the domestic business class, and an initial degree of equality in income distribution created by sweeping land reforms at the start o f the process. (Amsden, 1989; Westphal, 1990; Wade, 1990.) Thus, by ignoring political and social constraints and circumstances, attempts to propagate the instrumentality of the developmental state laid themselves open to later criticisms o f the actual effects of state involvement, which so often have belied expectations. Despite these limitations, there have been attempts very recently to resurrect the concept of the ‘developmental state’ in the Latin American context (Hamilton, 1981). It has been argued that the state has to be ‘redesigned’ (Fishlow, 1990)— almost as if it is really only a question of engineering or design, rather than a necessary involve­ ment of governmental activity in the sooty environs o f political struggles. Political scientists too have contributed to this tendency, by highlighting the concept of state autonomy in its role as an actor and as a force which has an independent effect in social dynamics. Thus, states, conceived as organisations claiming control over territories and people, may formulate and pursue goals that are not simply reflective of the demands or interests of social groups, classes or society (Skocpol, 1985: 20- 21.)

This being so, the goals need not be ‘developmental’ at all, but would be influenced by the pattern of ‘state structures* themselves, which would in turn be crucial in the ability o f the state to implement different types o f economic strategy. The rather simplistic distinctions which were made in the past between the activities of the state, the workings o f markets, and the functioning o f civil society generally, are increasingly rejected in the recent literature. (See, for example, Bean, 1989; Sen et al., 1990.) The rejection o f the idea of the pure ‘developmental state* has in turn brought about a reaction wherein the state is seen essentially as an

State Intervention in the Macroeconomy

167

intermediary between conflicting groups. Thust Killick (1990: 360) argues that decision making in the face of major divisions becomes a balancing act rather than the search for optima; a process of conflict-resolution in which social tranquillity and the maintenance of power is a basic concern rather than the maximization of economic growth. Accepting this constraint, Amadeo and Banuri (1991) put forward the case for a 'pragmatic state’, which would attempt to develop given the need to resolve internal conflicts and competing claims for resources, basing itself on the Keynesian concept of ‘governance’. A peculiar feature of these approaches is that despite their greater sensitivity to the issue of interactions between state and society, the state is still conceived as somehow separate from society, rather than an integral part of it which both creates and responds to various social pressures. Thus, the goal o f the Latin American state is supposed to v be 'development*, a goal which it pursues while simultaneously mediating conflicts between various groups (in this case, capitalists, unionized workers, landed groups and financiers). That the very conception o f 'development* is dependent on the nature of the state itself, and, thus, cannot be maintained as some abstract goal, just as the nature of 'mediation* between classes cannot imply a uniform impartial judging of rights, is missed out in this argument. These are difficulties which are sought to be addressed in the 'incorporative* view of the economic role of the state, which sees the state—and indeed markets— as socially embedded and an integral part of the way the polity and society themselves evolve. Much of the literature that has been here classified as 'incorporative* is part of the Marxist tradition (O’Connor, 1973;Gough, 1975;Foleez, 1978; Williams, 1982; Crouch, 1979), but it need not all be so. As Alt and Chrystal (1983: 240-1) point out: When economists do include government as an actor in their models, it is sometimes as a disembodied and disinterested puppeteer pulling strings from on high, with nothing more in mind than to improve the lot of the puppets. If it is not this, then it is an evil witch entering an idyllic scene where everybody is as happy as could be already. Neither of these is a sensible conception of government when tiying to make constnictive comment about the role of government in any real economy. Government is an institution within the system made up of actors of the system.

168

J a y a t i G h o sh

Marxist economists writing on planned development have tended from the start to be more careful about the social environment within, which policies are designed and implemented. Thus, they have rarely been guilty of a purely instrumentalist view of the state even when dispensing policy advice. As early as 1938, Oskar Lange while putting forward the case for socialist planning, could admit that ‘the real danger of socialism is that of a bureaucratization o f economic life and not the impossibility o f coping with the problem of allocation of resources’ (Lange, 1938: 109). However, as he pointed out, the same or even greater danger exists under monopoly capitalism. Similarly, there were major difficulties with the idea of the ‘mixed economy* in which the government woul4 attempt to control and plan production and investment without removing private enterprise and private ownership of the means of production. Lange felt that such experi­ ments were unlikely to be successful, since [t]he great economic power of corporations and banks being what it is, it would be they who would control the public planning authorities, rather than the reverse. The result would be planning for monopoly and restrictionism, the reverse of what was aimed a t . . . To retain private property and private enterprise and to force them to do things different from those required by the pursuit of maximum profit would involve a terrific amount of regimen­ tation of investment and enterprise (Lange, 1938: 119). Similarly, Maurice Dobb, while emphasizing that growth in under­ developed countries necessarily required a substantial volume of state investment along with a large public sector of nationalised industry and public services, struck a note of caution: There are important political and social factors as well which will affect the result. One of these is the question of ownership and motivation, and the type of social organisation, especially in agriculture . . . There is also the question of the human factor and its motivation (Dobb, 1962: 56). An even more perceptive awareness of the politics of government economic policy in developing countries is evident in the works o f Michal Kalecki. For advanced capitalist economies, Kalecki had already pointed out that full employment was really an unattainable ideal even with state intervention, except for very transient periods, since the presence o f a reserve army of labour was integral to the ability to control labour under capitalism (Kalecki, 1971). In non­ socialist developing countries, this was compounded by other fac­ tors—technical, institutional, social—which created the existence o f

State Intervention in the Macroeconomy

169

slack and involved underutilization of resources. Much of this was dependent on the nature of the state in such developing countries, and which class interests its actions reflected. Thus, Kalecki too accepted that there would be difficulties in implementing plans in a mixed economy, not only because of powerful institutional factors impeding the growth of some sectors, but because of the constraints on directing investment in the private sector as well as the difficulties of raising sufficient resources through taxation. (Kalecki, 1976: 29-30.) He coined the term ‘intermediate regimes* for those governments in which a coalition of lower-middle class and rich peasantry constituted the ruling class, and pointed out the contradictions of such a ruling class with the upper-middle class allied with foreign capital as well as feudal landowners from above and the small landowners and landless peasants along with the poorer urban population from below. Kalecki felt that attempts to resolve such contradictions would lead to the regime ultimately serving the interests of domestic big business. (Kalecki, 1976). This argument has been criticized on the grounds that few regimes in developing countries correspond closely to this description. However, the very method of seeing state policies as reflective of the interests o f a dominant coalition, which may itself be affected by the working out of these policies to be superseded by others, and the whole complexity of the interaction, show this to be a typical example of an ‘incorporative* approach. The ‘incorporative’ approach, thus, stresses that both the nature of the state and the functioning of markets in a particular society are shaped by social and historical processes: indeed, they are totally enmeshed in a complex web of socio-political relations and class configurations. Thus, there is no such thing as a ‘market* which can be abstracted from the social context and historical conjuncture which determine how various markets evolve and function. By the same token, there cannot be such a thing as an ‘autonomous’ state which is independent of the civil society and is guided only by abstract welfare principles. Where ‘government failures’ expressed as corrup­ tion, ‘distortions*, etc. abound, there markets too function in peculiar ways distorted by the social reality and quite unlike the textbook free market ideal. Consequently, the mere attempt to replace state regula­ tion with the so-called free market can never be a solution to perceived economic ills— it will, given the context, simply replace one form of distortion with another, possibly more pernicious. However, while governments and markets are both seen to be

170

J a y a t i G h o sh

shaped by social forces, there is a fundamental difference between the two in their economic implications. Markets are inherently inequalizing, not only because of their basis in private property, but because the very process of generalized exchange relations creates gainers and losers and thereby spurs differentiation. The point has been made that a capitalist economy cannot function without con­ tinuous and sizeable levels of unemployment of workers, essentially to ensure worker discipline. (Marx, 1971; Kalecki, 1974; Patnaik, 1991). State economic activity on the other hand, can vary in its distributive effects depending upon its nature, and need not rely on unemployment as a coercive tool. A point of significance emphasized in this approach relates to the importance of history. This is more than a blanket statement, for it means that the evolution and subsequent nature of the capitalist class and its relationship with other classes in society have to be given critical consideration in a study of state economic policies, as well as the mechanisms by which commercialization and exchange processes generally are spread and the ways that relations with wider interna­ tional markets and capitals are shaped. The degree of homogeneity or heterogeneity of the society as well as the potential for exerting social control given by the culture and prevailing institutional condi­ tions, have also to be taken into account. This is related to the idea that the distribution o f resources reflects the distribution o f power and in turn shapes it. This also means that in a multi-layered and very heterogenous society the imperial face of the state is more directly observable vis-a-vis particular groups. The difference o f this argu­ ment from the 4new political economy* modelling of lobbies and pressure groups should be noted: rather than certain activities/regula­ tions being the culmination of simple lobbying or rent seeking activities, the state itself is seen as being part o f the social equations within society and all its actions reflect changing economic social and political power configurations. And in this sense the state is no different from the other prevailing institutions— including market institutions— in the society. This approach also points to a basic requirement of the state: the need for legitimation; and the periodic crises of legitimation which force the ruling elites to seek additional support from different social groups, if necessary through altering state policies (Habermas, 1973). In electoral democracies this finds periodic expression though the ballot-box, but this is by no means the only or even most significant

State Intervention in the Macroeconomy

171

expression of this need, Indeed, all states require a degree of legitimation for sheer survival, and this affects economic actions, including those often classified as 'populist* as well as changes in the degree of accountability and responses to particular pressure groups. Mystification has also long been recognized as an easy means by which a state can combat attacks on its legitimacy. In turn the process of legitimation tends to exclude weaker groups by definition, which suggests why already underprivileged groups can be effectively economically disenfranchised (to use Krishna Bharadwaj*s term, in Bharadwaj, 1994). These internal requirements and processes are shaped by the broad international contours within which states, especially in developing countries, can act. These contours are determined by the nature of interaction with global capital, with economic and geo-political pressures and bargaining strength inter­ nationally, and the like. The emphasis on historicity is important not only in understanding changing patterns of state interaction with the macroeconomy, but even to comprehend why the explanations of such interactions have themselves altered over time. Thus, the dominance of particular ideologies and their related analytical perspectives in certain periods is crucially related to the nature of the economy and the changing power equations between classes. It can be argued that monetarism is the economic ideology o f finance capital. Indeed, this dominance of finance capital in recent times is significant in explaining the proliferation of economic theories of the state of the ‘interloper’ and ‘intermediary* variety, and of arguments for the conscious reduction of state economic activity in favour of market operations.

Ill The early macroeconomic writing on India contained a view of the state that was implicitly instrumentalist in nature. However, for the nationalist writers of the late nineteenth century, this had a very different connotation from that of the ‘developmental* state. In their argument the colonial state in India had a negative instrumentality as far as the economy was concerned, since its goal was not the economic development of the colonized country but material advantages for the colonial power (Naoroji, 1906; Dutt, 1906). The economic actions of the state were directed towards the extraction of resources from the

172

J a y a ti G h o sh

country. Thus, trade policies (the so-called 'laissez-faire') encouraged both exports of raw materials and imports of manufactured goods from Britain, and constrained the emergence of new industries in India in an unprotected environment, even while traditional industries were hit by import penetration and declining demand. Public investment in physical and social infrastructure was not only niggardly in per capita terms and by international standards, it was also primarily oriented towards resource extraction and market access (in terms of expansion of railways and region-specific large-scale irrigation). Land settlement policies not only contributed to severe taxation of the peasantry but also to commercialization which generated conditions for the increased export of agricultural produce. While in the early phase of colonial rule straightforward plunder and the exaction of tribute involved a drain of resources from the economy, by the late nineteenth century the state exchequer bore a substantial amount o f ‘Home Charges’ which, along with other items of state expenditure, amounted to a unilateral transfer to Britain which further im­ poverished the country. For the first half of the twentieth century it was argued (Bagchi, 1974) that while tariff protection did allow the emergence of some import-substituting industries in the inter-war period, this policy was still essentially directed towards serving Britain’s economic needs, and the colonial state’s industrial and credit policies actually were biased against domestic producers. Thus, in a variety of ways the state was seen to be the instrument o f economic exploitation rather than growth. Since in this construct the colonial state simply reflected the interests of a hegemonic outside power, internal political and social conflicts were not given significance even while the basic international relationship (between ruler and ruled) was emphasized. The dichotomy in macroeconomic perception was highlighted by the debate between ‘nationalists’ and ‘imperialists’, in which the latter argued that the colonial state had actually been instrumental in developing the economy and making for material progress (see Morris, et al., 1969). The immediate post-colonial literature on the macroeconomy also fell into an ‘instrumentalist* perception of state involvement, but from a completely different perspective. The need to develop the newly independent country was so overriding that the economic literature of the time assumed a broad consensus in favour of industrialization, a mixed economy and certain obvious developmental goals. It, thus ignored the possibilities of internal conflicts and competing claims

State Intervention in the Macroeconomy

173

for resources in the mechanics of state action, and concentrated instead on the technicalities of the matter of rapidly increasing production and incomes. The macroeconomic literature of the fifties, for example, was completely ‘technocratic* and 'instrumentalist* in its discussion of state involvement. Not only was the necessity for state involvement taken for granted (along the lines o f development literature of that period throughout the world), but it was also assumed that the state was (unlike its colonial predecessor) inherently benevolent in its desire for material betterment of the people, and essentially oriented to the development of the whole economy rather than the favouring of particular classes or groups (see Wadhwa, 1973; Bhagwati and Chakravarty, 1971). TTiis meant that the issues that dominated the discussion were those of the type of trade and industrial strategy (import-substituting versus export-oriented), the investment allocation in the plans, etc. The Nehru-Mahalanobis strategy that became the hallmark of the fifties and sixties assumed a ‘develop­ mental state’, so much so that even in the framing of the five-year plans themselves it was taken for granted that successful land reforms and other state-directed institutional changes relating to agriculture would increase output in that sector without additional expenditure by the government. The subsequent evolution of the economy as well as both the successes and failures of state intervention, have tempered that idealistic vision of the developmental state untainted by the domin­ ance of class interests and above the fray of social antagonism. The instrumentalist view of the state in India, just as in other countries, has come under attack for its naivete in accepting official statements at face value, and for its perception of the state as being outside civil society. However, even in more recent times, some versions o f the instrumentalist view can be found, albeit more sensitively drawn. Thus, Chakravarty (1987) in what is otherwise an excellent account of the experience of development planning in India, concentrated essentially on the economic and technical constraints facing state policy, with an implicit assumption that the state itself essentially pursued its stated developmental goals, only coming up against difficulties posed by the political and economic structures outside it. While this is useful in highlighting the precise economic difficulties associated with the planning process, it does leave the picture incom­ plete, since many of the constraints which appear to be ‘external* to the state are actually related to the nature of the state itself. Thus, for

174

J a y a t i G h o sh

example, the fiscal crisis of the state by the mid-seventies was clearly closely related to the inability to raise resources from taxation. Similarly, the pattern o f expenditure and the shift from direct public investment to subsidizing o f private investment could be seen as a result of the pressures emanating from the self-perceived requirements of the domestic large business class. Chakravarty was obviously sensitive to these issues (see Chakravarty, 1992), but conducted the analysis without explicit reference to them. Thus, the ‘rent-seeking* argument was criticized on the grounds that 'distortions are them­ selves structural properties of the economy being studied’ (Chakra­ varty, 1987: 44) without mentioning the nature of the state and the interests it serves are themselves some of these 'structural properties*, while the inability to raise tax revenues, is consigned to the realm of 'fiscal sociology’. The focus of Chakravarty*s analysis is, thus, on the design of government economic and planning policies, keeping in mind structural characteristics and what are seen to be political constraints. The creation and design of policies, however, appears to occur in a social vacuum, while the implementation faces the usual and much-discussed social and political constraints. A similar idea imbues other recent considerations of the macro­ economy in which the interaction of the state with the economy is largely seen as a problem of correct formulation of policy and design of intervention, as well as subsequent implementation. Thus, Basu (1992) considers the difference between ‘good’ and ‘bad’ policy advice given to governments which can supposedly be influenced into ‘correct’ economic policy decisions, much in the same way as Rodrik (1992) had talked of economists being able to nudge states into becoming ‘autonomous*. Jalan (1991) speaks of 'redefining* the role of government in India. He argues that 'the primary failure in several developing countries, including India, has been in implementation and in the tacit assumption that the state had an unlimited capacity to intervene.* (Jalan, 1991: 87.) But the point is that it was the nature of the state itself and the other political, social and institutional realities which created a certain role in the first place, and meant that this role had very definite implications for growth and distribution. In these characterizations, the conception of the state is certainly more sophisticated than implied in a simplistic ‘instrumentalist11 position, but the basic perception is very much still a part of this tradition— that the state can be almost a supra-institutionaJ form, above the nittygritty of internal and external politics and struggle over resources,

State Intervention in the Macroeconomy

175

and that ‘correct* economic understanding can allow it to formulate and implement policies to achieve developmental goals. The simpler version o f this position, as we have seen, informed the Indian planning literature of the fifties and sixties. The reaction to this, from the mainstream neoclassical position, came fairly swiftly. An early critique of the strategy of development came in Bhagwati and Krueger (1966) followed by Bhagwati and Desai (1970) and numerous otherwritings(suchasSrinivasan, 1985;Ahluwahlia, 1987; Butta-Chaudhuri, 1990), until it actually came to dominate economic policy discussion in the late eighties. The early works fell squarely in the ‘state-as-interloper* tradition, identifying ‘government failure’ as the root cause of all the major developmental problems o f postIndependence India. Thus, the combination of the strategy of planning with import-substituting industrialization and regulation of private industry was seen to have resulted in a ‘high-cost economy’ replete with inefficiencies, corruption, delays, distortions due to perverted market signals, stifling of entrepreneurial zeal through bureaucratic control, etc. In addition the specific strategy based on export pessim­ ism along with substantial dependence on public sector industrial investment came under fire not only for discriminating against agri­ culture and labour-intensive production but also because it created a gargantuan state sector whose performance was seen to be well below desired levels. The arguments— and indeed the counter-arguments to this position— are by now so well known that it is almost unnecessary to repeat them. It should be pointed out, however, that as a theory of state economic behaviour this analysis is extremely lacking, since the basic premise is that all state intervention in undesirable and leads to. inefficiency-creating distortions. The fact that there is some truth in many of the allegations does not render the whole argument an acceptable one. As pointed out earlier, this view implicitly posits an ideal-type counterfactual o f perfect market functioning which cannot be realised in any actual economy. Furthermore, simply castigating all state intervention as ‘bad’ does little to reveal why and how governments actually do affect the economy and why certain policies are undertaken with particular effects. Thus, as far as explaining the state’s role in macroeconomic performance in India is concerned, the ‘interloper’ theories in their various forms are of rather limited applicability and do little to shed light on either the mechanics o f state intervention or the interaction of government decisions with the economy and society.

176

J a y a ti G h o sh

Another type o f reaction to the 'developmental state’ perception in India comes from the view that the state is 'intermediary’, completely subordinated to the pressures and pulls of interest groups. This view is best expressed by Bardhan: The Indian public economy has thus become an elaborate network of patronage and subsidies. The heterogenous interest groups fight and bargain for their share in the spoils of the system and often strike compromises in the form of 'log-rolling' in the usual fashion of pressure-group. politics. (1992: 325.) Since none o f the classes or interest groups is individually strong enough to dominate the state’s resource allocation process, the result is a proliferation o f subsidies and transfers to placate all of them. This in turn involves a state that is quantitatively enormous and has a large body of regulatory practice, but is nonetheless rather weak in shaping the economy. In Bardhan’s view, therefore, the state is totally subordinate to society: rather than the autonomous state of instrumen­ talist perception which treats it as outside society, in this case the state is so totally embedded and influenced as to have no capacity for independent action other than mediating between and placating the various vested interests groups. Indeed, insofar as the state can be said to have any autonomy at all in this view, ‘the autonomy of'the Indian state is reflected more often in its regulatory and patronagedispensing role than in a developmental role’. Once again, there is much that is descriptively accurate in this argument, in particular the discussion of how specific government actions benefit particular groups even while they are associated with macroeconomic stagnation (Bardhan, 1984). However, this argument also is subject to several of the criticisms that can be levelled against the 'rent-seeking* school. Thus, it is really only a post facto explanation of observed failure rather than a theory o f state and macro-economic interaction. It has already been pointed out that there is no necessary correlation between growth rates and the proliferation of corrupt networks and grants of subsidies, either for developed or developing countries. Indeed, similar patterns of patronage and clientelism as those described by Bardhan for India can be found among fast growing developing countries such as South Korea and recently Indonesia and Thailand, as well as less successful economies like Pakistan and many Latin American countries. Thus, as an explanation of poor macroeconomic performance, this general castigation of the negative

State Intervention in the Macroeconomy

111

role of interest groups in determining state actions is simply inade­ quate. This reflects a fundamental flaw in the analysis, which is not really political economy except in a rather superficial sense. Essen­ tially, this amounts to an application of the theory of coalitions to the macroeconomy; there is no insight into why particular groups are dominant or are able to influence the state, why and how the relations-between these groups, the rest of society and the state change over time, what determines ideological and policy regime changes within such a system. Thus, for example, in the Indian mixed economy the shift in emphasis from ‘planning’ to liberalization’ cannot be explained by this formulation. In contrast to this totally subordinate state, there is a view of relative autonomy of the Indian state that comes from its sheer size in the economy as well as its ideological position in the society. The state’s dominance is related to its material condition, to its overwhelming control of investment and employment in the organized sector, and to its ideological advantage, as the presumed defender of the collective interest and socialist purposes and as the enemy of private and partial gains. (Rudolph and Rudolph, 1987: 13.) Thus, ‘the resources it controls and the state’s strategic and bargaining advantage make it possible for the state in India to be not only relatively autonomous and self-determined but also selfinterested. (ibid.: 399). This autonomy of the state as a ‘third actor’ in Indian politics is in turn seen to contribute to a centrist-oriented social pluralism in which class politics is marginal and subordinate to other social formations such as caste, religious and language communities and regional nationalisms. This is an ‘incorporative’ view of the state, but one in which the macroeconomic processes are not really explained. Thus, as this powerful ‘third actor’, what has prevented the state in India from achieving its stated developmental and distributional goals? And why, over time, have both the rate and pattern of growth as well as its sectoral, regional and class-wise distribution varied? The Rudolphs’ position is certainly a striking commentary on domes­ tic political processes, but it is unable to explain the impact o f the Indian state on the macroeconomy, or the changing nature o f the relationship. Marxist and other radical assessments of the Indian economy have tended to take an 'incorporative' position on the analysis of the state.

178

J a y a t i G h o sh

While some o f the planning literature contained analyses o f this type, explicit consideration of the nature of the Indian state and the ways in which it could operate against the stated goals of modernization within a ‘socialistic pattern of society’ could be found within a few decades of the initiation of the planning process. Thus, by the mid-seventies doubts began to be expressed about the ability o f a ‘mixed economy’ to function efficiently, since the plan targets depended crucially on private sector response, and the pattern of income distribution dictated a likely pattern of private resource allocation behaviour completely at variance with the official goals. The limits to (and problems with) regulation and control over the private sector were clearly recognized. (See, for example, the papers in Kurien, 1974.) This literature gave great emphasis to the importance o f institutional change prior to the planning process, since asset redistribution, in particular land reform, was seen as essential in altering initial endowments in a way so as to ensure balanced growth. The inability to achieve land reform, or any other asset redistribution of significance, or indeed to control the acquisitive tendencies of the elites, was seen to reflect the nature of the state itself. Some early writings described the nature of the state as one of an (uneasy) alliance between the rural landed classes and the big industrial bourgeoisie, influenced also by the interaction with metropolitan capital. (Thus, Patnaik, 1974.) However, since the relationship between state and society was seen as a dialectial one, the nature of the state was not viewed as immutable, but constantly changing along with the shifting configuration of class forces. Mitra (1977) used the idea that the different dominant classes had divergent economic interests to point out how shifts in the intersec­ toral terms of trade could be instigated by state action. The rich farmer lobby was able to ensure high farm output prices through influencing the government procurement price which operated as a floor, and this meant that terms of trade shifted in favour of agriculture. This not only worsened the material condition of both industrial workers and the rural workers and small peasants who were net purchasers of food, it also contributed to industrial stagnation in various ways. While numerous difficulties exist with this argument, most notably that in a context of oligopolistic industry such political manipulation of farm prices would affect the level of agricultural prices but not the terms of trade (since industrialists could choose to raise their prices in order to maintain profit margins), the notion of the state implicit in this

State Intervention in the Macroeconomy

179

argument is a useful one. Thus, this analysis highlights not only the ability of particular classes to affect specific sectoral policies (such as in this case the farm procurement price, which is also seen to differ according to region and crop depending on the political strength of the concerned farmers), but also the impact that this has on the macroeconomy. And in turn, how the behaviour of the economy affects the government’s decisions in the next round. As an explana­ tion of industrial stagnation between the mid-sixties and late-seventies, this argument may not be satisfactory, but it does involve a more complex representation of the interaction of state with economy. Other writers have explored this interaction also with the specific aim of explaining macroeconomic trends. Thus, Patnaik (1988) models the growth process with the explicit consideration of the state interacting with capital and labour, in a context in which there is an inflation barrier posed by the political sensitivity of the state to the impact of inflation on the more vocal social groups. This inflationary barrier forces the state to cut back on its own real spending since it is also politically constrained from raising more tax revenues. In turn this contributes to the decline in overall growth given the close association between public and private investments. The fiscal crisis o f the state—brought on by political economy reasons relating to its very nature—thus, leads to a more general economic stagnation. In later work, Patnaik (1994) has explained the expansion of the eighties in terms of a resurgence of government expenditure, this time based on the accrual of internal and external debt. This has also rendered the economy as well as the polity more vulnerable to external pressures and more dependent upon acceptability by international capital. Chandra (1988) has analysed state regulatory practice with reference to the external sector and multinational corporations, and has discussed both capital inflow and technology transfer in the context of a state which has a contradictory relationship with external capital, simultaneously dependent and striving for some autonomous space. Bagchi (1982, 1986, 1991) has considered the various facets of intermeshing of government decisions relating to the economy and the configuration of class forces and social institutions, to try and explain why the Indian case differs in significant ways from both the Latin American and East Asian stereotypes. Such recent work points to the directions that future research in the area can take. The need for an incorporative approach to the relationship between the state and the economy appears to be

180

J a y a t i G h o sh

especially crucial in the current context. This is because the sheer spread and mobility o f capital has forced a degree of globalization, and that too at a time when internal state policies increasingly reflect not just the desires o f a burgeoning middle class whose aspirations are now closely wound up with the advanced countries, but also the requirements o f international finance. At the same time, the supposed ‘withdrawal* of the state from the economic arena is essentially a reflection of a change in the manner of association rather than the fact or extent of it. Despite the announcements of liberalization, the Indian state has become even more of a centralized and centralizing force conditioned by the changing needs of large domestic and international capital given the more fluid and particularistic social configurations, even as the overt form of the state appears politically weaker and economically more fragile. The reasons for such tenden­ cies and their likely outcomes could perhaps best be analysed in terms of the close, yet continuously shifting relationship between the Indian state, the various classes and groups in society and the external forces at work. In such analyses, some significant contributions of the incorpora­ tive approach, especially in its Marxist versions, could be usefully borne in mind. The first is the point already mentioned, about the historicity and shifting nature of dominant ideologies, which are closely intermeshed with changes in social and material conditions. Despite the veneer o f objectivity which theories may possess, they are actually crucially affected by these changes. Currently, for example, the growth of finance capital cannot be separated from theoretical developments in the social sciences. Secondly, any analysis must be sensitive to the complexity of the relations between state, society and economy, which are multilayered and involve interaction in diverse ways and with various implications. Thus, macroeconomic policies which are reflective of particular socio­ political conditions may in turn alter these conditions, leading to pressures for another set of policies, and so on. All these are variants of the most crucial point: that any analysis must be aware of the dialectical nature of such processes, and the constant possibility of changes not only in the policies but in the nature of state and society as well. This in itself argues against rigid and simplistic theoretical formulations.

State Intervention in the Macroeconomy

181

REFERENCES I.J. (1987), Industrial Growth in India: Stagnation Sincc the Mid-sixies, Oxford University Press, Delhi. A kerlof , G. (1970), ‘The Market for lemons: Quality uncertainty and the market mechanism*. Quarterly Journal o f Economics, 84, 488500. A lesin a , A . and G. T a b e l u n i (1992), ‘Positive and Normative Theories of Public Debt and Inflation in Historical Perspective*, in European Economic Review, 36, 337-44. A l t , J. and K.A. C h r y st a l (1983), Political Economics, Wheatsheaf Books, Brighton A m a d e o , E . and T. B a n u r i ( 1 9 9 1 ) , ‘Policy, Governance and the Manage­ ment of Conflict*, in T. Banuri (ed.) op. cit. A m sd e n , A . ( 1 9 8 9 ), Asia *s Next Giant: South Korea and Late Industrial­ ization, Oxford University Press, New York. A rrow , K . (1962), ‘The economic implications of learning b y doing’, Review o f Economic Studies, 29, 155-75. B ag ch i , A .K . (1 9 8 2 ), The political economy o f underdevelopment, Cambridge University Press, Cambridge, UK. ------(1986), Public Intervention and Industrial Restructuring in China, India and Republic o f Korea, ILO-ARTEP, New Delhi. ------(1991) ‘From a Fractured Compromise to a Democratic Consensus? Planning and Political Economy in Post-colonial India, Economic and Political Weekly, Annual Number, 26, 11 & 12. B anuri, T. (ed.) (1991), Economic Liberalization: No Panacea, Claren­ don Press, Oxford. B a r d h a n , P . ( 1 9 8 4 ) ,The Political Economy o f Development in India, A h lu w a lia ,

B a s il B la c k w e ll, O x f o r d .

------(1992), ‘A Political-economy Perspective on Development', in B. Jalan (ed.) The Indian Economy: Problems and Prospects, PenguinViking, New Delhi. B asu, K. (1992), ‘Markets, Laws and Governments*, in B. Jalan (ed.) The Indian Economy: Problems and Prospects; Penguin-Viking, New Delhi. ------ (1980), ‘Revenue-seeking: A Generalization of the Theory of Tariffs*, Journal o f Political Economy, 88, 1069-87. B h a g w a ti , J. (1982), ‘Directly Unproductive Profitseeking Activities’, Journal o f Political Economy, October, 988-1002. B h a g w a ti , J. and P. D esai (1966), India: Planning fo r Industrialization,

182

J a y a t i G h o sh

Industrialization and Trade Policies since 1951, Oxford University Press, Delhi. B h a g w a ti , J. and S. C h ak ravartv (1971), Contributions to Indian Economic Analysis: A Survey, American Economic Association. B h a g w a ti , J. and T.N. S rinivasan (1976), ‘Foreign Trade Regimes and Economic Development: India’, Macmillan, London. B h a rad w aj , K . ( 1 9 9 4 ) , Accumulation, Exchange and Development: Es­ says on the Indian Economy, Sage Publications, Delhi. B o d d y , R. and J. C r o ttv (1975), ‘Class Conflict and Macro-policy: The Political Business Cycle*, Review o f Radical Political Economy, 7, 1, 1-19. B ow les , S. and J. E atw ell (1983), ‘Between Two Worlds: Interest Groups, Class Structure and Capitalist Growth*, in D.C. Mueller (ed.) The Political Economy o f Growth, Yale University Press, New Haven. B r en n a n , G. and J. B u chanan (1980), The Power to Tax: Analytical Foundations o f a Fiscal Constitution, Cambridge University Press, New York. B u ch an an , J. (1977), ‘Political Equality and Private Property: The Dis­ tributional Paradox*, in G. Dworkin, et al. (eds.) Markets and Morals, Hemisphere Publishing Co., Washington, D.C. B u ch a n a n , J. and G. T u llock (1962) The Calculus o f Consent: The Logical Foundations o f Constitutional Democracy, University of Michigan Press, Ann Arbor. B u c h a n a n , J. and W a g n e r , R (1977), Democracy in Deficit: The Political Legacy o f Lord Keynes, Academic Press, New York. B u ch a n a n , T o l u s o n and T u llock (eds.) (1980), Towards a Theory o f Rent-seeking Society, A&M University Press, College Station, Texas. C h ak ra va rty , S. (1987) Development Planning: The Indian Experience, Oxford University Press, New Delhi. ------(1992), 'Development Economics and the Indian Experience*, in K. Basu and P. Nayak (eds.) Development Policy and Economic Theory, Oxford University Press, New Delhi. C h a n d r a , N.K. (1988), The Retarded Economies: Foreign Domination and Class Relations in India and Other Emerging Nations, Sameeksha Trust and Oxford University Press, Bombay. C o a se , R. (1960), ‘The Problem of Social Cost*, Journal o f Law and Economics, 3, 1-44.

State Intervention in the Macroeconomy

183

Cqqter, R. (1982), ‘The Cost of Coase', Journal o f Legal Studies, 11, 1-33. Crouch, C. (ed.) (1979), State and Economy in Contemporary Capitalism, Croom Helm, London. D atta -C haudhuri, M. (1990), Market Failure and Government Failure, Journal o f Economic Perspectives, 4, 3, 25-39. D eane , R (1989), The State and the Economic System: An Introduction to the History o f Political Economy, Oxford University Press, Ox­ ford. D obb , M. (1963), Economic Growth and Underdeveloped Countries, Lawrence and Wishart, London. D ornbusch , R. and S. E dw ards (1990), ‘Economic Crises and the Macro­ economics of Populism in Latin America’, Journal o f Development Economics. Durr, R.C. (1906) An Economic History o f India, Routledge and Kegan Paul, London. E istcnstadt , S.N. and R. L amarchand (eds.) (1981) Political Clientelism, Patronage and Development, Sage Publications, London. E stey , D. and R. C aves (1983), Market Structure and Political Influence: New Data on Political Expenditures, Activity and Success’, Economic Inquiry, 21, 24-38. F ishlow , A. (1990), ‘The Latin American State’, in Journal o f Economic Perspectives, 4, 3, 61-74. Foley, D. (1978), ‘State Expenditure from a Marxist Perspective’, Jour­ nal o f Public Economics, 9, 221-38. G erschenkron , A. (1962), Economic Backwardness in a Historical Perspective, Harvard University Press, Cambridge, Mass. G ou gh , I. (1975), ‘State Expenditure in Advanced Capitalism', New Left Review, 92, 53-92. H abermas , J. (1973), Legitimation Crisis, Heinemann, London. H am ilton , N. (1981), ‘State Autonomy and Dependent Capitalism in Latin America’, British Journal o f Sociology, 32, 305-29. H ibbs , D. and W. F assbender (1981), Contemporary Political Economy, North-Holland, Amsterdam. H irschman , A.O. (1958), The Strategy o f Economic Development, Yale University Press, New Haven. ------(1982) Shifting Involvements: Private Interest and Public Action, Princeton University Press, Princeton. H olmstoom , B. (1984), ‘The Provision of Services in a Market

184

J a y a t i G h o sh

Economy’, in R.P. Inman (ed.) Managing the Service Economy: Problems and Prospects, Cambridge University Press, Cambridge, UK. I nm an , R.P. (1982), ‘The Economic Case for Limits to Government’, American Economic Review, 72, 176-83. ------(1987), Markets, Government and the ‘New* Political Economy*, in A J. Auerbach and M. Feldstein (eds.) Handbook o f Public Economics, North-Holland, Amsterdam. J alan , B. (1991), India's Economic Crisis: The Way Ahead, Oxford University Press, New Delhi. K alecki , M. (1971), ‘Political Aspects of Full Employment*, in Selected Essays on the Dynamics o f the Capitalist Economy, Cambridge University Press, Cambridge UK. ------ (1976), Essays on Developing Economies, Harvester Press, Brighton Sussex. K illick , T. (1990), A Reaction Too Fart Overseas Development Institute, London. K ornai , J. (1974), The Economics o f Shortage, North-Holland, Amster­ dam. ------(1979), ‘Resource-constrained versus Demand-constrained Systems*, Econometrica, 47, 801-19. ------(1980), The Economics o f Shortage, Parts A and B, North-HollandElsevier, Amsterdam. ------(1982), Growth, Shortage and Efficiency, Basil Blackwell, Oxford. K rueger , A. (1974), ‘The Political Economy of the Rent-seeking Society*, American Economic Review, 64,291-303. ------ (1990), ‘Government Failures in Development*, Journal o f Economic Perspectives, 4, 3, 9-23. K urien , K .M . (ed.) (1975), Indian State and Society: A Marxian Ap­ proach, Orient Longman, Bombay. L al , D. (1983), The Poverty o f Development Economics, Institute of Economic Affairs, London. L ange , O. (1938), On the Economic Theory o f Socialism, University of Minnesota Press. L erner , A.P. (1962), ‘The Economics an d Politics of Consumer Sovereignty*, American Economic Review, 62, 258-66. L indbeck , A. (1976), ‘Stabilization Policy in Open Economies with Endogenous Politicians*, American Economic Review, 66, 1-19. M a r x , K. (1971), A Contribution to the Critique o f Political Eoonomy, Lawrence and Wishart, London.

State Intervention in the Macroeconomy

185

M itra , A. (1977)* Terms o f Trade and Class Relations, Frank Cass,

London. M orris , M.D., et al. (1969), The Indian Economy in the Nineteenth Century: A Symposium, Indian Economic and Social History Asso­ ciation, New Delhi. M yrdal , G. (1970), ‘The “Soft” State in Underdeveloped Countries’, in P. Streeten (ed.) Infashionable Economics: Essays in Honour o f Lord Balogh, Weidenfeld and Nicolson, London. N aoroji, D. (1906), Poverty and UnBritish Rule in India, Natesan, Madras. N ordhaus , W. (1975), ‘The Political Business Cycle', Review o f Economic Studies, 42, 164-90. N ozick , R. (1974), Anarchy, State and Utopia, Basic Books, N e w York. N urkse , R. (1955), Problems o f Capital Formation in Underdeveloped Countries, Oxford University Press, Oxford. O ’C onnor , J. (1973), The Fiscal Crisis o f the Statet St. Martin’s Press, New York. O lson , M. (1965) The logic o f collective action, Harvard University Press, Cambridge, Mass. P atnaik , P . (1974), ‘Imperialism and the Growth of Indian Capitalism’, in K.M. Kurien (ed.) Indian State and Society: A Marxian Approach, Orient Longman, Bombay. ------ (1988), Time, Inflation and Growth: Some Macro-economic Themes in an Indian Perspective, Orient Longman, Calcutta. ------(1991), Economics and Egalitarianism, Oxford University Press, New Delhi. ------(1994), ‘International Capital and National Economic Policy: A Critique of India’s Economic Reforms’, Economic and Political Weekly, 19 March. P ersson , T. and L. S vensson (1989), ‘Why would a stubborn conservative run a deficit’, Quarterly Journal o f Economics. R akshit , M. (ed.) (1989), Studies in the Macro-economics o f Developing Countries, Oxford University Press, Calcutta. R odrik , D. (1992), ‘Political Economy and Development Policy*, European Economic Review, 36, 329-36. R osenstcin -R odan , P.N. (1943), ‘Problems of Industrialization of East­ ern and South-eastern Europe’, Economic Journal, 53, 202-11. R udolph , S. and L.H. R udolph (1987), In Pursuit o f Lakshmi: The Political Economy o f the Indian State, University of Chicago Press, Chicago.

186

J a y a t i G h o sh

P. (1954), ‘The Pure Theory of Public Expenditure', Review o f Economics and Statistics, 36, 387-9. ScrrovsKY, T. (1954), ‘Two concepts of external economies'. Journal o f Political Economy, 62, 143-51. S en , A martva , N. S ter n , J. S tiglitz and S . F ischer (1990), 'Roundtable Discussion on Development Strategies: The Roles of the State and the Private Sector*, Proceedings of the World Bank Annual Con­ ference on Development Economics, Washington, D.C. S hapiro , H. (1988), ‘State Intervention and Industrialization: The Origins of the Brazilian Automobile Industry*, Ph.D. dissertation, Yale University, New Haven. S hapiro , H. and L. T aylor (1990), ‘The State and Industrial Strategy’, World Development, 18, 6, 861-78. S kocpol , T. (1985), ‘Bringing the State Back In: Strategies of Analysis in Current Research*, in P.R. Evans, D. Rueschemeyer and T. Skoc­ pol (eds.) Bringing the State Back In, Cambridge University Press, 3-37. S rinivasan , T.N. (1985), ‘Neo-classical Political Economy, the State and Economic Development*, Asian Development Review, 3, 38-58. S tiglitz , J.E. (1987), ‘Learning to Learn, Localized Learning and Tech­ nical Progress', in P. Dasgupta and P. Stoneman (eds.) Economic Policy and Technological Performance, Cambridge University Press, Cambridge. T aylor , L ance (1988), Varieties o f Stabilization Experience, Clarendon Press, Oxford. W ade , R. (1 9 9 0 ), Governing the Market: Economic Theory and the Role o f Government in East Asian Industrialization, Princeton University Press, Princeton. W adhw a , C.D. (ed.) (1973), Some Problem o f India's Economic Policy: Selected Readings on Planning, Agriculture and Foreign Trade, Tata McGraw Hill, Bombay. W estphal , L. (1990), ‘Industrial Policy in an Export-propelled Economy: Lessons from South Korea’s Experience', Journal o f Economic Perspectives, 4, 3, 41-59. W hynes , D.K. and R.A. B owles (1981), The Economic Theory o f the State, Martin Robertson, Oxford. W illiams , M. (1982), ‘Industrial Policy and the Neutrality of the State', Journal o f Public Economics, 19, 73-96. S a m u e ls o n ,

Macroeconomic Character of the Indian Economy: Theories, Facts and Fancies* V. PANDIT

1. INTRODUCTION The purpose of this paper is to discuss some salient relationships which characterize the Indian economy from a macroeconomic view­ point. It is not our intention to evaluate policies though the ensuing discussion is bound to have serious policy implications. The explicit focus is only on the empirical judgements that are analytically relevant, and in so far as such judgements can be made with some measure of confidence. References to actual policies will be made only to illustrate specific points. Quite naturally, we shall draw upon the sizeable relevant empirical research that has been undertaken on various macroeconomic aspects of the economy over the past two decades. Issues relating to trade and commercial policy are not taken up because limitation of space will not permit us to do justice to these. An econometric model illustrating some of the linkages discussed in the paper is given at the end.

2. ISSUES AND METHODOLOGIES There are three issues to which we should address this discussion in view of their importance to macroeconomic policy and the underlying macroeconomic adjustments in an economy. Each of these is nested, as it were, within the one which precedes it. First, there is the general * This paper is part of a wider on-going work on macroeconomic modelling for India at the Institute of Economic Growth. 1 am grateful to Prof. 1C. Krishnamurty for useful comments on an earlier draft.

188

V . P a n d it

question of an appropriate analytical framework. The question can indeed be posed in a number of different ways. But let us be more precise by restricting the choice as between the neoclassical general equilibrium and the Keynesian analysis. In empirical terms, one needs to make a judgement whether the level of economic activity is, broadly speaking, either supply-constrained or demand-constrained. The answer may not necessarily be unequivocal. It may be specific to the particular period or phase of its development or to the particular sector or set of activities one is considering. Apart from this question of demand adjusting to supply, or vice versa, there is the latent issue of what linkages exist and types of adjustments they permit. Second, following from the point, we need to identify the major determinants o f the different macroeconomic aggregates like con­ sumption expenditure, intended investment, prices, trade flows, and the like. This identification along with certain other elements deter­ mines (i) what the possible policy variables can be, (ii) the kind of linkage that exists between different phenomena and therefore the critical equilibrating variable, and (iii) the nature of equilibrium and the implied adjustments that may come about in response to policyinduced or purely exogenous changes. These are relevant to a proper understanding of the economic process as also to a correct formulation of policies consistent with the accepted socio-economic objectives. Third, it is necessary to have some idea, however approximate, of the magnitudes of critical parameters that govern the different linkages and adjustments. These include the familiar propensities and elasticities associated with consumption, investment, production, trade flows, etc. Some knowledge of the magnitudes of these para­ meters is necessary to evaluate the strength of different adjustments and the extent of their relevance. Before we consider the three sets of question in some depth and detail, let us first clarify the approach we adopt. Most of the government policies can broadly be classified into two types. First, there are policies which are microeconomic in character in the sense that their consequences in the short and the medium run are more distributional than aggregative. Such policies may have serious consequences as regards industrial composition of output or demand, distribution of income across different classes of households and workers, and the like. On the other hand, there are policies which may influence aggregates without having clearly discernible effects

Macroeconomic Character o f the Indian Economy

189

on the distribution or composition of these aggregates in the short or even the medium run. In the real world, however, such distinctions are somewhat blurred. All policies have in varying degrees, and with different lags, both distributional as well as aggregative consequences. Our purpose in magnifying the distinction is to ensure analytical clarity. We shall almost entirely focus on issues which are relevant in the context of the latter policies—commonly labelled as 'macroeco­ nomic* policies. Even the simplest laws in economics are subject to a variety of assumptions and qualifications. This is also true of empirically observed interrelationships between different economic phenomena. Any inferences, either qualitative or quantitative, have therefore to be based on more than the simple ‘roles of thumb* or ‘back o f the envelope' calculations. This is true as much of developing economies as it is of mature industrial economies. Discussions of macroeconomic policy have therefore got to be carried out in the framework of a model, which is amenable, in principle, to quantification. Certain issues and policies are, however, in practice not easy to quantify. In this paper we shall consider only quantifiable phenomena and with reference to an analytical model or framework.

3. THEORETICAL BACKDROP Before considering various questions relating to the macroeconomic mode of functioning we shall have to digress briefly into macroeconomic theory. It is necessary to recall the salient aspects o f alter­ native theories in order to ensure clarity and smoothness in the discussion that follows. Theories relevant to macroeconomics can broadly be classified into two alternative paradigms depending on the role that is assigned to prices and on how markets are assumed to function. The simplest versions in the two sets which are used for pedagogical discussions can be referred to by their parent models as Walrasian and Keynesian. The latter includes several vintages and extensions of the basic theory expounded in the General Theory.* The former is not identified with any single author but draws its basic elements from the general 1 Some of the terms frequently used are neo-Keynesian, Keynes-Wicksellian, or Keynes-Kaleckian, etc.

190

V . P a n d it

equilibrium theory of Walras. Later vintages of this are the neoclassical theory and the more recent new classical economics. The last two are also referred to as Monetarism Mark I and Monetarism Mark II in macro-theoretic literature (Tobin, 1980). For the present discussion we shall simply label the two alternative paradigms as classical and Keynesian, ignoring the differences that exist within each fold. The classical theory emphasizes the role of prices in bringing markets to equilibrium. All prices, wages, and interest rates are assumed to be flexible in the sense that these can freely and quickly move in either direction as and when required for equating supply and demand. Further, equilibrium is defined as the situation in which excess demand in each market is zero, i.e. all intended supplies and demands are equated and markets clear. Thus, the basic Walrasian premises of the classical theory are (a) prices as sole equilibrating variables, (b) market clearing as necessary for equilibrium, and (c) price flexibility, or absence of market rigidities. These assumption o f the classical theory ensure that all resources, particularly labour, are fully employed in the sense that, for example, anyone seeking employment at the prevailing wage rate will find it. The level of activity and of output are at their full employment levels. It is important to clarify here that full employment does not mean that the entire labour force is employed nor that the level of production is at it§ technically-feasible maximal limit. On the other hand, full employment means that firms are actually producing what they intend to produce at the prevailing commodity and factor prices. They may voluntarily leave some capacity unutilized. In macro-theoretic terms, we say that both households as suppliers of labour as well as firms as suppliers of goods are on their respective supply schedules. Correspondingly both firms as well as households are on their demand schedules for labour and goods, respectively. This is in a nutshell, an account of the working of a frictionless smoothly-behaved economy. Prices play their designated roles, all markets clear, and labour is fully employed. All adjustments take place through prices which are flexible. All markets and all agents interact symmetrically, there being no hierarchy of decisions. A couple o f points regarding money and government expenditure before we leave this topic. The Cambridge cash balance theory which assumes that people hold a fixed proportion of their nominal income as cash is an integral part o f the classical macro-theory in the form of the celebrated quantity theory of money. This implies, under certain

Macroeconomic Character of the Indian Economy

191

additional, assumptions, that all nominal prices would vary propor­ tionately with the supply of money leaving the relative prices to be determined by the demands for, and the supplies of goods and services in different markets. In consequence, money is only a veil having no influence on the real variables like output, investment, or employ­ ment. There is a similar impotency associated with the government expenditure. Any increase in government expenditure only displaces private expenditure because total expenditure must be equal to total income which is fixed at the full employment level. Only structural shifts affecting productivity, preferences, and the like can alter income or employment. Keynes* departure from the classical or Walrasian paradigm has been in many directions. Early evaluations of the General Theory pointed at three things, namely money wage rigidity, liquidity preference, and consumption function. This has been common know­ ledge for a long time and we need not elaborate on it.2 But let us look a bit deeper at these three things in terms of the present slate of Keynesian economics. Way back in the fifties, Tobin had argued that even if money wages were flexible, unemployment could not be eliminated. More recent interpretations of the Keynesian methodology demonstrate that the critical feature o f this methodology is that the price system, whether flexible or not, fails to perform the job that is assigned to it by the Walrasian general equilibrium theory. Thus, we may replace 'money wage rigidity’ by ‘price-ineffectiveness*.3 Next, we note that though Keynes took expectations to be ex­ ogenous, they play a vital role in shaping the equilibrium in a Keynesian economy the way it is. Again, in terms of the recent recasting of the Keynesian theory, it is expectations which cause the difference between notional (ex-ante) and constrained (ex-post) supply and demand schedules. If all expectations are realized the distinction between the two disappears. In the General Theory, expectations were important for the liquidity preference phenomenon as well as for the marginal efficiency of investment schedule. Thus, the crux of the liquidity preference phenomenon, which was given prominence in the Keynesian system, lies in the role it accords to 'expectations’. 2 See, e.g. Branson (1992). 3 This discussion draws upon Clowcr (1965), Leijonhufvud (1968), and Malinvaud (1977), as well as some related subsequent work.

192

V . P a n d it

About the consumption function, the important point to note is that the dependence of intended expenditures on expected or perceived income is an explicit quantity adjustment hypothesis. The price level and the rate of interest play only minor roles in this relationship. What this means is that a large part of the adjustment, at the individual level as well as in the aggregate, takes place through quantity rather than price signals that economic agents produce. For example, when employment is feared to go down households will cut their consump­ tions plans. Similarly, when expectations of the aggregate or, more precisely, effective demand are low producers will reduce their pro­ duction targets and cause unemployment rather than cut prices in order to stimulate demand. Thus, the consumption function is a major component of the 'quantity adjustment' process. Two more points that are relevant for the subsequent discussion need to be underlined at this stage. First, both classical as well as Keynesian theories are tailored for market economies. This is par­ ticularly true of the former which assumes perfectly competitive markets. Second, neither of the two imposes a physical resource constraint on the level of output. In the Keynesian case, more is not produced because the perceived demand is low. On the other hand, in the classical case, more is not produced because it is not profitable to do so. Physical constraint on resources like capital is absent in both cases. It follows from this discussion, that the choice of an appropriate framework to analyse macroeconomic phenomena must be guided by the following somewhat interrelated characteristics of the economy. (a)

(b) (c) (d) (e)

The role played by the price system in bringing about equi­ librium, i.e. the extent to which economic agents may respond in terms of quantity rather than price signals, The extent to which markets can be assumed to be cleared in the Walrasian fashion, e.g. absence of false trading, The organizational structure of markets, i.e. the prospects of price making, Availability and costs of information, and the consequent mode of expectation formation, and The severity of resource constraints.

Macroeconomic Character o f the Indian Economy

193

4. MACRO-ADJUSTMENTS IN LESS DEVELOPED ECONOMIES Let us now return to the first of the three sets of issues raised at the beginning. It has already been said that the question is whether the level of activity in a less developed economy (LDE) can be treated as demand-constrained o r supply-constrained. While this is clearly a matter of empirical verification, it is necessary to hammer out. on an analytical basis, the framework within which alternatives can be tested. Until about the early seventies, a widely shared consensus seemed to be that the Keynesian methodology was not of much relevance to LDEs. Rao (1952) was perhaps the earliest to present theoretical reasons favouring this proposition.4 Brahmananda( 1978) goes further to claim that the developmental problems of these economies should be dealt with in terms o f classical theory. The distinction between empirical judgements and analytical issues was seldom made in this context. International organizations, notably the International Monetary Fund (IMF) have provided advice to LDEs on the basis of rules-of-thumb associated with the classical theory, eg. the Quantity Theory of Money. The consequent policy prescriptions, usually referred to as IMF conditionality,3 have in varying degrees em­ phasized the need for: (a) Restrictions on money supply and credit expansion. (b) Strictly maintained balanced government budgets. (c) Avoidance of quantity restrictions and price controls. (d) Flexible exchange rate or devaluation to correct chronic trade deficits. (e) Real wage hikes linked with increased productivity. (0 Limited investment and production activity by the government. This approach to economic policy in LDEs has been forcefully rejected by structuralists on grounds which we shall consider in the next section. Arguments against the relevance of Keynesian economics to LDEs have, in varying measures, presumed some characteristics of these economies: 4 In a much earlier paper A.K. Dasgupta had taken the same position but on different grounds. 5 For a recent discussion of IMF conditionality, see Bacha (1987).

194

V. P a n d it

(a)

Predominance of the primary sector particularly food-producing agriculture, and large share of subsistence activities. (b) Low income and high propensity to consume. (c) Disproportionately high share of food and other basic neces­ sities in consumption. (d) Absence of competitive conditions in markets for goods and factors. (e) Absence of assets other than money, goods, and precious metals. (0 Constraints on production imposed by inadequacy of resources, particularly capital. It has been argued that these characteristics, some of them on their own and some in conjunction with others have the following implica­ tions for the working o f LDEs. First, the level of output can be taken as exogenously given by the availability of resources along with fluctuations caused by natural factors like weather. Second, since capital and other resources are relatively scarce the rental on them is high. Consequently, they are likely to be fully utilized. Third, a large part o f the economic activities are insulated from market signals through either prices or quantities. Fourth, a high mpc implies a high Keynesian multiplier so that even a small increase in autonomous demand would choke off whatever unused productive capacity may exist, setting the economy up against severe supply bottlenecks. Fifth, a large part of the increased demand is directed towards food and other necessities whose output cannot be increased in the short and the medium run, so that an investment-saving gap sets off an inflationary process. Sixth, since there are no active capital and money markets, increased money supply drives up the demand for goods through hoarding or other means depending on how additional money is induced into the system. This results in proportional? increase in prices a la quantity theory over the medium run. Seventh, investment and saving must move closely to avoid an inflationary gap. This implies a validation of Say's law, with money failing to serve as a store of value. It also ties down the rate of growth to the rate o f saving given the productivity o f capital.

5. A C R ITIQ U E O F T H E C O N V EN TIO N A L V IEW This view has been questioned on various grounds by Chakravarty

Macroeconomic Character of the Indian Economy

195

( 1979), Pandit (1977) and Rakshit (1982)— particularly in the context of India. A criticism of the above-mentioned IMF-style approach to the analytics of development problems had come much earlier from a number of Latin American economists led by Prebisch (1962). Their arguments have also been reformulated in clearer analytical terms in recent years, notably by Taylor (1979, 1983, and 1987). The general point in these writings is that the economic problems of LDEs cannot be cast in terms of simple propositions of either classical or neoclas­ sical economics. The conventional wisdom has been questioned on three sets of arguments. First, some of the empirical characteristics of LDEs are often exaggerated even though they may not be altogether false. Second, a distinction needs to be made between the Keynesian methodology on the one hand, and the actual policy recommendations following from the naive oversimplified Keynesian model. Third, the conclusion that simple theoretical constructs associated with the classical theories are relevant to LDEs is not warranted even if the alleged characteristics do hold. Let us look at some of these argu ments more closely. To begin with, we are dealing with economies that are less developed but not primitive. While it is plausible that agricultural output is predetermined in the short run, this is not so in the medium run. Farmers in LDEs have been shown to respond to incentives and to allocate resources rationally.6 Regarding the inflexibility of output, the extreme view presupposes four things. These are: (i) the resource constraint is binding not only in the short run but also over the medium run, (ii) capital is fully utilized in all sectors and all activities (iii) No factor substitution is possible in any of the sectors, (iv) the consump­ tion basket permits no substitution. It would be fair to say that while these things might have a measure of validity in varying degrees, they are prone to be exaggerated. A number of studies have, for example, demonstrated the pos­ sibilities of substitution between factors not only in agriculture but also in manufacturing.7 Several countries including India have ex­ perienced low capacity utilization even in consumer goods industries. In many activities, considerable scope exists for high labour intensity so that output can be raised with relatively small quantities of capital. 6 This has been well established during the sixties by Dharm Narain, Raj Krishna, and Jere Behrman. 7 Behrman (1972) shows this in case of Chile.

196

V . P a n d it

Since most countries are engaged in trade and many large economies like India and China have the capacity to produce a wide variety of capital goods, the resource constraint cannot be stringently binding except in the very short run. Moreover, while food shortage continues to be an acute problem in parts of Africa, this is not a universal phenomenon. There may be problems of fair distribution of what is available but that is a different matter. With regard to saving propensity and the multiplier we merely note that the mps has increased considerably in most LDEs since the late sixties. For India it has increased from about 15 per cent during the sixties to nearly 30 per cent in the eighties. It is, however, true that in segments of these economies, e.g. in the rural sector and the small scale and family enterprise sector, saving-investment decisions are coterminous. But this is not true for the economy as a whole. Turning now to the two methodological questions, we note that the Keynesian model o f macroeconomic analysis is more flexible and comprehensive. It is thus capable not only of extension in several ways but also of accommodating rigidities where these exist. More­ over, it assigns a role for monetary and fiscal policies which are admitted to be important for LDEs. For these reasons, even as one is groping for an appropriate framework, the Keynesian system serves a better starting point than the classical system. Many of the features mentioned above can be built into a Keynesian system. Once this is effectively done, the naive Keynesian policy conclusions need not follow. In any case, irrespective of whether we accept or reject the Keynesian model for LDEs, there is practically no case for the validity of the classical theory for such economies. Perfectly competitive markets for commodities and factors, price flexibility and costless information anything but exist in such economies. It has been fairly convincingly argued that the functioning of LDEs is quite likely to be on Keynesian lines.8 Lysey (1980) goes on to argue that develop­ ment policy in many LDEs has failed on the employment front because of its neglect in stepping up demand. We may recall the earlier discussion of the Keynesian methodology in terms of the works of Clower (1965),. Leijonhufvud (1968), Malinvaud (1977), and others. The central, thrust is: (a) Owing to 8 See Chakravarty (1979), Pandit (1977), and Rakshit (1982, 1989). Pandit also shows empirically for some LDEs that the multiplier model is more reliable for the prediction of the level of economic activity than the velocity model.

Macroeconomic Character o f the Indian Economy

197

poor information systems and the resulting expectations we must think in terms of a non-Walrasian equilibrium in which markets need not clear, (b) The price system may fail to perform its equilibrating role, and equilibrium generation may instead rely on quantity signals. The non-existence of the Walrasian ‘auctioneer*, causes the failure of the Walrasian tatonnement process, the bed-rock of classical/ neoclassical economic systems. Returning to LDEs we observe that one very visible characteristic of these economies is segmentation of commercial activities,9 and the related market imperfections. Segmentation is caused by geographical features, group affiliations, and community linkages of economic agents and several other factors. Market imperfections are the result of inequitable distribution o f income, wealth and economic power generally. All these tend to erode the capability of the price system as an equilibrating force. Consequently, economic agents are prone to rely more on quantity than on price adjustments. Finally it must be reiterated that a straightforward application of the General Theory to LDEs is not being justified. Instead, the argument is for the search o f an alternative paradigm which should preferably start from the Keynesian rather than the classical structure. One important feature of the intended system should be to make allowance, in principle, for resource-constrained output, which is indeed still a problem for LDEs in may circumstances. Some of these things will be concretized subsequently in the light of experience with the Indian economy.

6. T H E LEV EL O F EC O N O M IC A CTIV ITY IN INDIA Let us now turn to the empirical literature relating to the macroeconomic structure of the Indian economy. It must be noted at the outset that considerable research has been undertaken in this context over the last two decades or so. Most of it employs econometric methodology to examine the important relationships between dif­ ferent macroeconomic aggregates. While some of the studies focus only on specific sectors or phenomena many of them specify and estimate economy-wide macroeconometric models. Given the volume of this literature it would not be possible to go into all the details nor 9 Sec Rakshit (1987) and Pandit (1985).

198 V.

P a n d it

make extensive references except where necessary.10 On the other hand, we shall focus on certain selected issues and try to sift the robust and the important results. In doing so a greater emphasis will be laid on the relatively more recent work. Following from our earlier discussion, the foremost question is the role played by demand and supply." Most of the studies have more or less followed the Keynesian format by modelling private consump­ tion and private investment in order to determine aggregate demand. But most of them have at the same time also specified production functions for different sectors. This gives rise to the difficult problems of specifying the appropriate adjustment or closure rule. Open eco­ nomies would usually provide larger channels of adjustment between supply and demand* but these cannot be depended upon in the case of the Indian economy. This is partly because the external trade relative to GNP has not been very large nor well spread across sectors and activities, and partly because a good deal of the trade is controlled. In any case, most models also do not have an extensive treatment of the external sector.12 A basic problem has been that most researchers have shied away from using one or the other pure models13 and tried to accommodate both sets of factors without succeeding in identifying a workable closure rule. This is because of the absence of reliable data on variables like capacity utilization, unemployment rate, unintended inventory accumulation, and free market interest rates, etc. which could have been used to track divergence between supply and demand. Since the price level generally fails to be an argument in either the supply function or the demand function, its role as an equilibrating variable in the commodities market in a neoclassical fashion is ruled out. Even if this was not done, there would be the problem of specifying equilibrium in the money market—an issue to which we shall come later. 10 For recent and detailed reviews of this literature see Krishna, Krishnamurty, Pandit, and Sharma ( 1989) and Pandit and Pandit ( 1989). For earlier reviews, also see Bhaduri ( 1982), Krishnamurty and Pandit ( 1984), and Desai ( 1973). These papers also provide extensive bibliographies. 11 In this discussion ‘supply* denotes resource-constrained or capacity output unless otherwise specified. 12 Two recent exceptions are Pandit ( 1986) and Krishnamurty, Pandit, and Sharma ( 1989). *3 Exceptions to be noted in this context are the models by Madhur ( 1986) and Brahmananda ( 1978) which are distinctly monetarist.

Macroeconomic Character of the Indian Economy

199

In an early study, this author (Pandit, 1973) attempted to model the level of economic activity without bringing in production func­ tions. The model performed well in tracking down variations caused by reasonably-sized perturbations around the actual path. Similar results were confirmed in a later study in which the simple reducedform multiplier model was tested against a similarly simple velocity model. The former performed distinctly better.14 While these exer­ cises have consistently lent support to the Keynesian view of the level of economic activity in India, the underlying issues did not appear to have been settled. Further examination and refinement was called for. Two directions of such refinement are distinctly visible. First, there is the need for disaggregation of the economy into sectors which might prima facie be subject to different adjustment processes. Second, a complete explanation should include the mode of adjust­ ment not only for outputs but also for prices.

7. A SUGGESTED DISAGGREGATIVE TREATMENT There are three major considerations in deciding the type o f disag­ gregation one should adopt. First, the availability of reliable data; second, the nature of the adjustment process, and third, the extent to which prices and quantities are meaningfully conceptualized and measured. The second and the third are somewhat related. As far as data availability is concerned, we could consider about ten sectors, but to keep the problem manageable we need to stop at five for reasons explained below. One is broadly concerned with splitting the economy into sectors which are more or less homogeneous as regards price and output adjustments. This will depend partly on whether activities are publicly or privately controlled. However, this distribution is not available according to sectors. In the third place, one has to pay some attention to the concept and statistical measurement of prices and outputs. It is this last consideration that prompts us to separate public administra­ tion and defence from other services. The same reason suggests separation of services like transport from those like banking and insurance. 14 Pandit (1977). The question can be restated in a Friedman-Meiselman (1963) sense as one of relative stability o f the two parameters. Following this approach, Pandit (1980) confirms the earlier conclusion.

200 V. P a n d it Following these considerations it appears advisable to disaggregate the economy into five sectors, namely15 (a) (b) (c) (d) (e)

Agriculture and related activities Manufacturing and construction Economic infrastructure which consists of power generation, transport, communications, mining, water supply, etc. Economic services including real estate, banking, insurance, trade, hotels, etc. Public administration and defence.

This much of disaggregation is, on the one hand, meaningful and manageable and, on the other, reliable in terms of data. With regard to the concepts o f output and price, there is no problem in (a), (b), and (c). Measurement also poses no hard problems. In the case o f (d), concepts are clear but measurement is difficult. For (e), there are difficult conceptual as well as measurement problems. That is why these activities have been set apart from those in (d). (c) is separated from (b) because of the critical role of the former in development and constraints on the output of this sector. Further, it is now widely agreed that for sector (a), output is fixed in the short run but flexible in the medium and the long run. Also, prices in this sector are market clearing and thus competitively determined.16 For sector (b), also, there is general consensus that output may possibly be demand-constrained and that prices may mostly be subject to the Kaleckian mark-up on unit cost. For sector (c), output is likely to be resource constrained and prices administered subject to unit production costs. In the case of (d), output can safely be assumed to be capable of adjusting to demand and therefore related to the level of economic activity in the commodity-producing sectors including infrastructure. Modelling of price behaviour is hard but not impossible. For (e), both output and price are hard to disentangle. The nominal value added in this sector is merely determined by expendi­ ture. Any unscrambling of this value into prices and quantities is quite

15 This classification has been adopted in Pandit (1985a). On second thoughts, it might be preferable to put construction along with economic infrastructure. See also Sarkar and Rao (1982). 16 This may no longer be as valid a judgement as it has been till the late seventies due to the government setting procurement prices for all major crops. See Pandit (1985a) and Patnaik (1992). See also footnote 19 below.

Macroeconomic Character o f the Indian Economy 201 arbitrary. Finally, we also note that whereas (a) is nearly completely privately controlled, (b) and (d) are mixed, and (c) is predominantly in the public sector and likely to remain so even in the wake of economic reforms. The last one, (e) is entirely constituted of govern­ ment activities. As mentioned earlier, the public sector should preferably not be lumped with private sector activities but data availability is a serious problem in this context. There would additionally be differences within each sector so that inaccuracy arising from ignoring these differences is a cost one has to bear for manageability o f the model. We do, however, believe that in our scheme the inaccuracy should not be too large. The foregoing discussion thus suggests that in modelling both outputs as well as prices, we need to adopt a differential approach across sectors because of differences in (i) the extent to which the resource constraints may or may not be binding (ii) forms of the market structure, (iii) the extent to which activities are publicly or privately managed, and (iv) nature of the product and the associated technology and organization of production. Clearly, disaggregation of the economy into only agriculture and non-agriculture as was done in early studies is not at all adequate. In our treatment of the manufacturing sector, we postulate that unlike other sectors this one may be either demand-constrained or resource-constrained in the sense that demand may either exceed or fall short of the capacity output. Clearly, the actual output ex-post must be equal to the smaller of the two. Thus, the level of activity in this sector and the rest of the economy might switch between two regimes, generating a pattern of cyclical growth. This methodology was actually used to analyse the Indian economy in a study prepared for the Committee to Review the Working of the Indian Monetary System of the Reserve Bank of India.17 The results were much better than these from the methodologies of any aggregate demand or ‘only aggregate supply*. Capacity output was taken to be constrained by the stock of capital, raw materials availability, and the supply infrastructural facilities.18

17 See Pandit (1985a, 1986a). The committee is popularly referred to as the Chakravarty Committee after its chairman. 18 See the appendix.

202 V.

P a n d it

8. PRICE ADJUSTMENTS Let us now turn to the next most important issue, namely the adjust­ ment of sectoral and overall price levels. Here again the conventional approach for LDEs has been based on the quantity theory of money and the associated monetarism. Policy makers as well as professional economists in LDEs have mostly followed this approach in dealing with price stability. For India, such views have been articulated by Gupta (1974), Minhas (1987), and Brahmananda (1983). As men­ tioned earlier, it has also served as the framework for advice tendered by international agencies like the IMF over the last three decades for dealing with inflation as well as external disequilibrium. We have, however, argued so far that the assumptions needed to validate the QTM are unlikely to hold for LDEs, particularly India. The fact that money supply often turns out to be highly correlated with the price level ex-post does not necessarily prove the existence of the ex-ante causal relationship between the two variables. An early negation of the QTM approach by Raj (1966) is noteworthy. This is not to deny that money supply has a considerable effect on prices. It is only intended to suggest that (i) money, as a demand-pull factor, is not the only cause for changes in prices, (ii) money affects prices not only through demand but also through supply, in so far as money supply (and credit) positively influence output, so that the net effect is smaller than the partial effect, (iii) the total impact o f monetary changes on the price level takes about two years to work out. Price behaviour in India for the period 1950 through 1975 was examined in some detail in Pandit (1978). Since most o f the later studies have used a similar framework let us look at it closely. In agriculture, we assumed the price level to be market-clearing with predetermined total output. In a reduced form, the price level (or the rate of change in it) can thus be related to the level of output, per capita income, and some proxy for price expectation.19 In the remain­ ing sectors, prices are cost determined in a manner the genesis of which goes back to Kalecki (1971). This is also consistent with the Keynesian formulation. We use money supply per unit o f real GDP as an additional explanatory variable for all sectors to capture the liquidity situation in the economy which determines the extent to 19 As mentioned in fn, 16, the government's procurement price also has to be accommodated in this relationship for the more recent periods. See Pandit (1985a) and Patnaik (1992). See also Pandit (1993).

Macroeconomic Character o f the Indian Economy 203 which the mark-up factor i s influenced by demand pressure. A similar formulation has in recent years been suggested to be appropriate for LDEs and widely popularized by Taylor (1983, 1987). For India this has been used in recent models by Pandit (1985a), Bhattacharya (1984), Krishnamurty (1984), Sarkar and Rao (1985), and Ahluwalia (1979). However, a polar view has recently been put forward by Balakrishnan (1991). Using more sophisticated econometric techni­ ques now available, he concludes that the link between monetary expansion and inflation is tenuous. In more recent formulations (Pandit 1985a, 1986a), the important cost factors turn out to be (i) wage rate adjusted for productivity, (ii) raw material costs, (iii) interest costs associated with working capital,20 (iv) energy cost, and (v) the rate of profit. The last variable, which is exogenous, captures the mark-up factor. For the economy as a whole, elasticity of the price level with respect to money supply is about 0.35 and is spread over about two years. The model performs quite well in explaining the sample period variation and in short-run forecasting of sectoral and general price levels.

9. SAVING AND INVESTMENT Let us now consider the saving-consumption functions which play a pivotal role in the quantity adjustment process. It should, however, be noted that in assigning this role much depends on how the rest of the system is specified. Alternative systems can be built such that the saving-consumption functions have an altogether different role.21 A mere recognition of a Keynesian-type saving function does not, by itself, imply any specific kind of adjustment. Since the two functions are obverse of each other, whatever is said of one in the subsequent discussion applies more or less to the other.22 The simple Keynesian consumption function has been considerably modified and extended during the fifties and the sixties giving rise to the permanent income, life cycle, and relative income hypotheses as alternatives to Keynes' simpler absolute income hypothesis. The main 20 That the rate of interst affects prices via costs is important in the context of monetary policy for inflation control. 21 See, for example, Friedman (1970). 22 For more extensive discussion, refer to Pandit and Pandit (1994) and Pandit (1991).

204 V. P a n d it thrust of these extensions has been to distinguish between expected and unexpected incomes, to introduce new variables like wealth, to take account of demographic characteristics affecting household decisions, and to recognize ratchet effects in saving-consumption behaviour. However, most of these micro-theoretic hypotheses are derived from assumptions unlikely to hold for countries like India where a predominant proportion of households live at or below the subsistence levels of income. It is, therefore, not surprising that empirical evidence has not found much use for these theories. Researchers have typically used the simple Keynesian consumption function with some modifications like the introduction of lagged effects. It is nevertheless conceivable that the sophisticated theories might be relevant to some cross-section of the population like upper income groups in urban areas. Recent studies on the aggregate consumer behaviour show that besides disposable income, which is the dominant determinant of private consumption expenditure, other determinants include the share of agricultural output in GDP. This variable is presumed to serve as a proxy for the rural/urban income distribution, the relevance of which rests on the assumption that the mps for the rural sector is lower than that for the urban sector. However, using cross-section data, Pandit (1991) shows that the difference between the two propensities at comparable income levels is not as large as it is made out to be. It is perhaps the case that results based on aggregative data also include the impact of income distribution. Lack of data has prevented quantification of the effect of functional or size distribution of income on savings. For the same reason, it has not been possible to include wealth as an explanatory variable. Both the variables are in principle and in fact, considered to be important. Investment behaviour has been harder to model not only in India but also elsewhere. Investment activity, unlike consumption, is much more than a routine compulsion. It is subject to the influence of several factors that are hard to quantify. For an economy of continental size, only partially subject to market forces, with a predominantly large household and informal sector and split between public and private management, modelling o f investment is bound to be particularly hard. Aggregates in such an economy are not likely to reveal much. All the same, studies at disaggregative levels do reveal some inter­ esting features.

Macroeconomic Character of the Indian Economy 205 Most of the empirical research, in varying combinations, has focussed on the following issues. (a) (b) (c) (d)

impact of output reflecting either anticipated demand or resource availability, role of credit, and finance in general, as promoter of capital formation, relevance of cost of capital, expected inflation, and profitability, effect of public investment in either crowding-in or crowdingout (i.e. promoting or displacing) private investment.

A substantive, though somewhat dated, study on fixed investment behaviour at the industry level, was undertaken by Krishnamurty and Sastry (1975). Another study based on flow of funds is by Swamy and Rao (1975). In addition, investment behaviour has been examined as part of economy-wide models in studies mentioned earlier.23 Most studies deal only with private investment. Many of them further disaggregate between the private corporate and the household sectors, on the one hand, and between fixed and inventory investment, on the other. There is some evidence in favour of the accelerator hypothesis but this is not sufficiently robust. However, either the level of output or change in it turns out to be an important determinant of private corporate investment in almost all studies. Specification of finance as a determinant of private investment is relatively recent. Bhattacharya (1984) and Pandit (1984, and 1985a) find commercial bank credit to be important in this context. This is consistent with the fact that credit is rationed and the interest rate fails to reflect fully the extent of excess demand in the organized capital markets. Other factors which have varyingly been found to be important are the advance rate of commercial banks and the rate of profit.24 Since the household sector which includes household enterprises, small businesses, and other informal small-scale activities function under constraints and motivations different from those relevant to the corporate sector, it has to be dealt with differently. One important factor relevant in the context of investment behaviour is the limited access of enterprises in this sector to the organized capital and credit markets. This implies that they have to depend largely on their own 23 See Pandit and Pandit (1994) for more references to the earlier literature. 24 Pandit (1983) finds the impact of expected inflation on private corporate investment to be significant.

206

V . P a n d it

resources to finance capital formation. This suggests that investment and savings are partly tried up. In addition, inflation rates in excess o f the nominal rates o f interest may not only induce higher inventory investment but also a larger shift away from holding financial assets. Since all this is more likely to occur in agriculture a higher share of agriculture in total output means a higher degree of the household sector's self-investment activity. In this manner, we have coterminous or joint saving-investment decisions of a classical type taking place in a part of the economy. This is quite well corroborated by common experience with the behaviour of peasantry, small traders, urban household enterprises, etc. Decisions to undertake investments induce them also to plan saving for such investments. Finally, before we close this section the effect of public investment on private investment has to be understood in two respects. One is whether the ultimate effect of the former on the latter is one of displacement. The second is about the channels through which the impact is transmitted. Under the classical theory, as well as its modem versions, public investment must displace private investment or, more generally, expenditure. The impact works through the rate of interest. It may also through the price level depending on how such investment is financed. That output is always at the full employment level is critical in this process. On the contrary, the net effect of increase in public investment on private investment will be positive in a Keynesian world, via the multiplier. The critically important considerations in this context are (i) the extent to which public and private investments are competitive or complementary, (ii) the state o f capacity utilization in both consumergoods as well as capital goods industries, (iii) the ability to import investment goods, i.e. the foreign exchange resource position, and (iv) the manner in which public investment is financed. In most LDEs, and this is certainly true of India, public investment has typically taken the form of infrastructure facilities like irrigation, power generation, transport, communications, and manufacture of intermediates like steel, manufacture of heavy machinery, electricals, etc. To the extent this is so, public investment should stimulate and facilitate private investment. Moreover, extension of activities like transport and power creates new capacities, particularly in backward regions. Since the foreign-exchange resources have been waxing and waning no general statement is likely to be valid. With regard to financing, the important point to note is that credit in the formal capital

Macroeconomic Character o f the Indian Economy 207 market issued at administered low rates is usually rationed. Govern­ ment with its pre-emptive power has the first claim. This may drive up the cost of credit for small investors who have to depend on informal credit sources. Also, if public investment is financed by borrowing from the central bank (Reserve Bank of India) there will be some rise in prices, eroding the real value of outlays which are usually set in nominal terms. We shall come back to this issue in the next section. In a study relating to Korea and India, Sundarajan and Thakur (1980) show that in India public investment has crowded-out private investment. However, the model employed for this exercise is rather naive and deficient in several ways so that the conclusion cannot be taken to be definitive. On the other hand, Krishnamurty (1984) finds a clear case for the crowding-in effect of public investment. In another study, Krishnamurty, Pandit, and Sharma (1989), considering more recent time series data and using more comprehensively specified relationships show (i) a generally crowding-in phenomenon, (ii) weakening of this phenomena since the mid-sixties presumably because of the reduced infrastructure-generating role of such invest­ ment and, (iii) crowding-out through financial channels, and due to price effects. It would appear that the crowding-in that is still taking place will continue for some more time.25

10. M O N EY , IN FL A TIO N A N D RESO URCES As mentioned earlier, monetary policy in LDEs has generally been linked to the objective of price stability d la the Quantity Theory of Money. The presumption has been that a dichotomy exists between the real and the monetary sectors at the macro level. But at the micro level, several linkages are invoked to show why ‘excessive* increase in money supply is harmful to the economy. In a proper assessment of monetary policy in relation to inflation and resource mobilization in the Indian case, we need to take note of the following features of the economy borne out by empirical research. First, both money supply as well as bank credit are related to the same monetary base (or reserve money) so that they are subject to the same controls and movements.26 Consequently, any restraint on 25 See also Bhattacharya (1984), Ahluwalia (1985), and Srivastava (1981) in this context. 26 Gupta (1986).

208

V . P a n d it

money supply also restrains credit in the economy and vice versa. Second, it is not possible for the monetary authority to control exactly the supply of money or credit because the monetary base is partly determined by the government’s fiscal operations and partly depend­ ent on other endogenous factors including the portfolio choice of the public. Third, the ex-post supply of money and credit are also to some extent responsive to changes on the demand side and, therefore, entirely exogenous. Fourth, the supply of goods and services is not independent of the availability of credit. The same is also true of capital formation. Consequently, credit availability, with different lags, affects the level o f economic activity through demand as well as supply. Finally, the contribution of the interest cost to the unit production costs in the non-agricultural sectors is not negligible, as it is generally taken to be. These features of the Indian economy imply that a restrictive monetary/credit policy can lead to recession by forcing a downward movement in both aggregate demand as well as aggregate supply. To the extent that a credit squeeze is sought to be achieved by directly or indirectly raising the interest rates, there would be a tendency towards an increase in prices. It is shown elsewhere27 that setting the nominal interest rate at some fixed margin above the rate o f inflation induces a significant measure of inflation. Thus, monetary squeeze coupled with high interest rates is likely to be stagflationary— though not as much as Taylor (1983) suggests in the context of the relatively more inflation-prone Latin American economies. To recall what has been mentioned earlier, we note that a 1 per cent increase in money supply eventually raises the general price level by only about 0.35 per cent. Further, it takes more than a year, almost two, for the full impact to be realized. Let us now turn to the general problem of resource mobilization for growth. Three phenomena relevant in this context are (i) the impact of inflation on investment and its composition, (ii) the effect of inflation on household savings, and (iii) thechoice between alternative modes of financing public investment. Regarding (i), it is clear that in so far as nominal interest rates are not fully indexed with respect to the rate of inflation, an increase in the latter lowers the real cost of funds. While this might affect total investment positively, it is more 27 Pandit (1985a, 1985b).

Macroeconomic Character of the Indian Economy 209 likely that the composition of capital formation will change in favour of inventory accumulation. This is borne out by many studies. As far as (ii) is concerned, inflation is likely to diminish household savings. Further, as several studies show, inflation will induce households to shy away from holding financial liabilities issued by the public and the corporate sectors. This adds to the problems of resource mobilization—particularly for the public sector, which typi­ cally offers low nominal rates of interest. Next, there is the question of how public investment should be financed.28 Broadly there are four options, namely (i) taxation, which under the present socio-political set-up means indirect taxes, (ii) bor­ rowing either domestically or externally, (iii) deficit financing, i.e. borrowing from the Reserve Bank, and (iv) raising the prices of public sector goods and services in the hope of raising the investible surplus of public sector enterprises. In the subsequent discussion, we consider only the last two choices. In analysing the choice between the two alternatives, namely deficit financing and public sector pricing, we have to bear in mind the following. First, associated with raising the prices of any subset of prices is the well-known cascading effect so that ultimately all prices will increase and by more than what the initial increases were intended to be. Second, since the prices of inputs rise for all enterprises, the net increase in their surpluses is much less than targeted for by each enterprise. Third, since all prices increase, the real value of whatever resources are generated get considerably eroded. Fourth, in so far as the prices of manufactures—particularly capital goods—rise more than the other prices, real investments will be even smaller. In the case of deficit financing, on the other hand, while there is bound to be a price increase it would be more uniformly spread over all the sectors. Also, deficit financing, working directly through fiscal channels, might have some positive effect on output. The question here is one of comparing the inflationary consequences arising from alternative choices—all leading to a given step-up in the growth rate. Such a trade-off analysis, carried out for the Chakravarty Committee by Pandit and Bhattacharya (1985), shows distinctly the superiority of deficit financing over public sector pricing to generate resources.29 We are not going into a comparison of the productivity of investment in the two sectors. It is presumed that public sector investment is desirable. 29 See. also, Mukherji, Pandit, and Sundaram (1992), and Pandit (1993).

210

V . P a n d it

11. POLICY REFORMS AND MACROECONOMIC MODELLING The introduction of a new policy regime in mid-1991 has given rise to a number of questions which we need to address before we con­ clude. In a nutshell, one needs to ponder over whether the accumulated ‘wisdom’ based on the available macroeconomic models is of any relevance in dealing with the contemporary policy issues. The ques­ tion is legitimate because results based on past data which were generated under a different policy regime may no longer be valid. Moreover, the processes of adjustment which are bound to undergo changes require to be recast in a manner consistent with the changed policy scenario. What, if any at all, is the new agenda for macroeconomic analysis? One may be strongly persuaded to hold the mistaken view that empirical macroeconomic analysis has no role, at least for a while, till we have the relevant data sets and a better understanding of how the economy is likely to function under the new regime. We strongly feel, however, that the need to continue macroeconomic analysis is as compelling as it appears to be difficult. Perhaps the difficulties, which no doubt are there, tend to be overstated. To be specific, let us consider a macroeconomic system consisting of seven segments dealing with the following phenomena: (i) (ii) (iii) (iv) (v) (vi) (vii)

sectoral productivity relations and overall output, consumption and saving behaviour, price formation and inflation, monetary and fiscal linkages and responses, capital formation—quantum and composition, financial and capital markets, and trade flows, BOP, and exchange rate adjustments

In our view, the segments dealing with (v)—(vii) will undergo rapid changes which may be either directly introduced by the new policies or indirectly induced by such policies. Changes in (iii) and (iv) would be slower and those in (i) and (ii) may not even be visible for quite some time. If this view is correct, then it would follow that segments of the existing models or those re-estimated with the existing data30 30 1982.

As far as we are aware, there does not yet exist a model based on data after

Macroeconomic Character o f the Indian Economy 211 would continue to be valid for a while. Some of the segments will have to be checked for possible slow changes while others may need to be overhauled substantially. From an econometric point of view, three things can be suggested. First, one may utilize the existing models for policy analysis by judiciously modifying some of the important parameters o f these models. As suggested by Krishnamurty (1992), it may not be inap­ propriate to vary the estimated parameters using a standard error scale appropriately so as to generate alternative scenarios and examine their economic plausibility. Second, the degrees of freedom constraint for testing structural change may be overcome using the alternative methodology based on recursive residuals.31 Third, some of the critical relationships may be re-estimated using a variable coefficients model incorporating both random as well as systematic variations in parameters.32

12. AN OVERVIEW In view of the ground that this paper has covered, as well as the diversity and complexity of the issues that have been touched upon, it is difficult and perhaps unwarranted to conclude with any neat assertions. It might, all the same, be useful to have an overview and draw attention to some general themes that emerge, just to ensure that the full implications are understood. At the outset, it should be stated that we have tried more to question established presumptions and conventional approaches than to assert clear-cut alternatives. This is intended to provoke a useful debate incorporating analytical reasoning as well as empirical evidence. We firmly believe that ad hoc theorizing couched in neat algebra is as much if not more meaningless as haphazardly put-together facts, figures and statistical relationships. The scope for a healthy interaction between theoretical and empirical research relating to macro­ economics o f the Indian economy is, I believe, yet to begin. It appears to us that the following features of the Indian economy which are analytically plausible and empirically robust can be high­ lighted on the basis of the preceding discussion. First, while the 31 See, e.g. Greene (1993). 32 Pandit (1980), illustrates an application of such models to stabilization problems in LDEs.

212

V . P a n d it

quantity adjustment mechanism and the expenditure functions in an aggregative sense have generally been stable, they may not hold outside certain bounds. This is because the economy may indeed frequently be faced with significant resource constraints, implying the need for a careful modelling of both capacity output as well as aggregate demand, and allowing for a switching of regimes in response to external shocks and policy changes. Second, in so far as (i) the money-price relationship is not as overwhelmingly direct and strong as it is presumed to be, (ii) credit supply promotes production as well as investment, and (iii) the impact of interest costs on unit production costs in some sectors is significant, monetary policy needs to be very carefully tuned. It may on occasion, not only be inadequate to curb inflation but also, cause recession and adversely affect growth. Third, since prices in the industrial sector seem to be cost-determined, resource mobilization through price hikes of public sector products and services may be an inferior alternative to deficit financing for promoting public investment. Fourth, the household sector which is still the mainstay for resource mobilization appears to be sensitive to inflation and interest rates. Further, saving-investment decisions are jointly taken in a significant part of the household sector covering agriculture, rural enterprises, small urban establishments, and other family-based ventures. While this weakens the multiplier effect in the economy globally, significant demand-induced output increase may nevertheless exist within each specific segment of the economy. This appears to be important for rural development and poverty removal in backward areas. Finally, we must stress that economies are always changing, sometimes suddenly but more often imperceptibly, so as to become visible only after a passage of time. Simple observations, quantitative models, as well as analytical constructs are therefore valid only within small neighbourhoods o f the ‘present state*. Observations and inferen­ ces discussed relate to the Indian economy observed during the period of the early sixties through to the mid-eighties. These may not remain valid after another decade has elapsed or if the system is subjected to sharp changes. This is true as much of analysis based on assumed parameters as of that based on estimated relationships. Hence, the need for continuous review, updating of facts, and recasting of analytical tools and empirical results can hardly be overemphasized.

Macroeconomic Character o f the Indian Economy 213 REFERENCES I.J. (1979), Behaviour o f Prices and Outputs in India: A Macroeconomic Approach, Macmillan, Delhi. ------(1985), Industrial Growth in India, Oxford University Press, Delhi. B acha , E.L. (1987), ‘IMF Conditionality: Conceptual Problems and Policy Alternatives*, World Development, December. B alakrishnan P. (1991), Pricing and Inflation in India, Oxford Univer­ sity Press, Delhi. B ehrman , J.R. (1972), Short-run flexibility in a developing economy, Journal o f Political Economy, 80. ------(1975), 'Econometric Modelling of National Income De-termina­ tion in Latin America, with Special Reference to the Chilean Experience’, Annals o f Economic and Social Measurement, 4. B haduri, A. (1982), ‘Existing State of Macroeconometric Modelling in India’, mimeo, ESCAP, Bangkok. B hattacharya , B .B . (1984), Public Expenditure, Inflation and Growth: A Macroeconomic Analysis fo r India, Oxford University Press, Delhi. B rahmananda , P.R. (1977), Determinants o f Real National Income and Price Level in India, Bombay University Press, Bombay. ------ (1983), Growthless Inflation by Means o f Stockless Money, Himalaya, Bombay. B ranson , W. (1992), Macroeconomics: Theory and Policy, 3rd edition, New York. C hakravarty , S. (1979), ‘Keynes, Classics and Developing Economies, in C.H.H. Rao and P.C. Joshi, (eds.), Reflections on Economic Development and Social Change, Essays in Honour of Professor V.K.R.V. Rao, Allied Publishers, New Delhi. C lower , (1965), ‘The Keynesian Counter Revolution*, in F. Han and F.P.R. Brechling, (eds.) The Theory o f Interest Rates, Macmillan, London. D esai , M.J. (1973), ‘Macroeconometric Models for India: A Survey*, Sankhyat Ser. C, December. F riedman , M. (1970), ‘A Theoretical Framework for Monetary Analysis’, Journal o f Political Economy, 78. F riedman , M. and D . M eiselman (1963), Relative Stability of the Mul­ tiplier and the Velocity*, Commission on Money and Credit. G reene , W. (1993), Econometric Analysts, 2nd edition, Macmillan, New York. A hluw alia ,

214

V . P a n d it

S.B. (1974), ‘Food Shortages, Demand-Pull and Inflation in India*, in S.L.N. Simha, (ed.)* Inflation in India, Vora, Bombay. ------(1986), Monetary Planning fo r India, 2nd edition, Oxford Univer­ sity Press, Delhi. K alecki , M. (1971), ‘On Costs and Prices', in M. Kalecki, Selected Essays on the Dynamics o f the Capitalist Economy, Cambridge University Press, New York. K eynes , J.M (1936) The General Theory o f Employment Interest and Money, Harcoutt, Brace and World, London. K ha n , M o h sin S ., P eter J. M ontiel , and N adeem K . H aque , (eds.)(I991), Macroeconomic Models fo r Adjustment in Developing Countriesy International Monetary Fund, Washington, D.C. K lein , L.R. (1965), *What Kind of a Macroeconomic Model for Develop­ ing Economies*, Indian Economic Journal, Econometric Annual. K rishna , K .L ., K . K rishnamurty , V. P andit , and P .D . S harma , (1989), 4Macroeconometric Modelling in India: A Selective Review*, Development Papers, No. 9, United Nations, E S C A P , Bangkok. K rishnamurty , K. (1984), ‘Inflation and Growth: A Models for India*, Indian Economic Review, 19 (1), reprinted in Krishnamurty and Pandit (1985). ------(1992), 'Status of Macroeconometric Modelling in India*, Institute of Economic Growth, (mimeo.). K rishnamurthy , K. and V. P andit (1984), ‘Macroeconomic Modelling of the Indian Economy: An Overview*, Indian Economic Review, 19 (1), reprinted in Krishnamurty and Pandit (1985). ------(1985), Macroeconomic Modelling o f the Indian Economy, Hin­ dustan Publishing Corporation, Delhi. K rishnamurthy , K ., V. P andit , and P.D.. S harma (1989), ‘Parameters of Growth in a dEveloping Mixed Economy: The Indian Experience*, Journal o f Quantitative Economics, 5 (2). K rishnamurty , K . and D.U. S astoy (1975), Investment and Financing in the Corporate Sector in India, Tata McGraw Hill, Delhi. L euonhufvud , A. (1968), On the Economics o f Keynes and Keynesian Economics, Oxford University Press, New York. M ukherji, B., V. P andit , an d K. S undaram (1992), ‘P ric e s, Q u a n titie s a n d Resource M o b ilisa tio n : G en eral Equilibrium in a Miniature M ix e d E c o n o m y ', Indian Economic Review, Special Number in G upta ,

m e m o ry o f S u k h a m o y C h ak rav a rty . L ysey , F.J. (1980),

‘Investment and Employment with Unlimited Labour:

Macroeconomic Character o f the Indian Economy 2 15 The Role of Aggregate Demand*, Journal o f Development Economics, 7 (4). M alinvaud , E .M . (1977), The Theory o f Unemployment Reconsidered, Basil Blackwell, Oxford. M adhur , S. (1986), A Structural Model fo r India, (mimeo.)* Policy Group, New Delhi. M [ nhas , B.S. (1987), ‘Budgetary Deficits and Inflation in India\ Indian Economic Review, 22 (2). P andit , B.L. (1983), ‘A Note on Inflation and Corporate Investment*, Indian Economic Review, 18 (2). ------ (1985), 'Saving Behaviour and Choice of Assets of Indian Households*, Indian Economic Review, 20. ------(1991), The Growth and Structure o f Savings in India, Oxford University Press, Bombay. P andit , B.L. and V. P an dit (1994), ‘Behaviour of Households and Firms in a Macro-theoretic Context: A Selective Review of the Literature*, in S. Jha and K. Parikh, (eds.) Methods o f Planning and Policy Analysis fo r Mixed Economies, Allied Publishers, New Delhi (forthcoming). P andit , V. (1973), ‘National Income and the Price Level in India, 195051 to 1965-66*, Sankhya, Ser. B, 35. ------(1977), ‘Multiplier, Velocity and Underdevelopment*, Manchester School, 45 (2). ------ (1978), ‘An Analysis of Inflation in India*, 1950-75, Indian Economic Review, 13. ------(1980), ‘Money, Expenditure and Stabilisation in LDCs: A Variable Coefficient Formulation', Sankhya, Ser. D. ------(1984), ‘Macroeconomic Adjustments in a Developing Economy*, Indian Economic Review, 19 (1), reprinted in Krishnamurty and Pandit, (1985). ------(1985a), Monetary Policy Growth and Stability, report prepared for the Reserve Bank of India, Committee to Review Working of the Monetary System in India (Chakravarty Committee). ------ (1985b), ‘Supply, Demand and Monetary Policy in an LDC\ (mimeo.), Discussion Paper no. 8501, CADE, University of Pennsyl­ vania. ------(1986a), ‘Growth Performance of the Indian Economy: Supply or Demand*, Delhi School of Economics, Working Paper no. 277, presented to the annual meeting of the American Economic Associa­ tion, New York, December 1985.

216

V . P a n d it

------(1986b), ‘Price Adjustments and India’s Trade Performance: A Macroeconomic Analysis*,, paper presented to the Fall Meeting of the Project LINK. Bangkok. ------(1993), ‘Controlling Inflation: Some Analytical and Empirical Issues’, Economic and Political Weekly, 28 (1 and 2). P andit , V . and B .B . B hattacharya (1987), ‘Resource Mobilisation, Growth and Inflation: A Trade-off Analysis for India*, in M. Dutta, (ed.) Asia Pacific Economies: Promises and Challenges, Jai Press, New York. P atnaik , P. (1992), ‘Terms of Trade and the Rate of Inflation*, in K. Basu and P. Nayak* eds. Development Policy and Economic Theory, Oxford University Press, New Delhi. friEBiscH, P. (1962), ‘Economic Development or Monetary Stability: The False Dilemma', Economic Bulletin fo r Latin America, Santiago. R a j , K.N. (1966), ‘Price Behaviour in India*, Indian Economic Review. R akshit , M.K. (1982), The Labour Surplus Economy, Macmillan, Delhi. ------(1989), Towards an Integration o f Monetary and Real Models fo r Mixed Economies, International Conference on Methods o f Planning and Policy Analysis for Mixed Economies, Indira Gandhi: Institute of Development Research, Bombay. R a o , V.K.R.V. (1952), ‘Investment, Income and the Multiplier in an Underdeveloped Economy*, Indian Economic Review, 1 (old series). S arkar , H. and S .V . S ubha R ao (1982), ‘A Computable General Equi­ librium Model for India*, Indian Economic Review. S rivastava , D.K. (1981), ‘Policy Simulation with a Macroeconomic Model of the Indian Economy*, Journal o f Policy Modelling. S undarajan , V . and S . T hakur (1980), ‘Public Investment, cRowding Out and Growth: A Dynamic Model Applied to India and Korea*, IMF Staff Papers, Washington. S wamy , D.S. and V.G. R ao (1975), ‘The Flow of Funds in the Indian Manufacturing Sector*, Sankhya, Series B , 37. T aylor , L. (1979), Macro-Models fo r Developing Countries, McGraw Hill, New York. ------(1983), Structuralist Macroeconomics, Basic Books, New York. ------(1987), ‘Macro Policy in the Tropics: How Sensible People Stand*, World Development, 15 (12). ------(1988), ‘Macroeconomic Policy in India*, Indian Economic Review, 23. T obin , J. (1980), Asset Accumulation and Economic Activity, Yrjo Jahnsson Lectures, Basil Blackwell, Oxford.

Appendix

AN ILLUSTRATIVE MODEL In this appendix we give a model of the Indian economy extracted from Pandit (1985b) to illustrate some of the issues raised in the paper.

The notation is given at the end. The economy is divided into five sectors as follows: (a) (b) (c)

(d)

Agriculture consisting of (i) Agriculture proper, (ii) Forestry and logging, (iii) Fisheries. Manufacturing consisting of (i) Large and small scale industry, (ii) Construction. Infrastructure consisting of (i) Power generation, gas and water supply, (ii) Mining and quarrying, (iii) Transport, storage and communication. Services consisting of (i) Trade, hotels and restaurants, (ii) Banking and insurance, (iii) Real estate, ownership of dwelling and business services, (iv) Other miscellaneous

services (e)

and. Government consisting of (i) Public Administration, (ii) Defence. The estimated equations (and identities) are as follows:

(1)

lo g P /a = 0.683 + 0.327 (logM l +Af 2/2 X ) (2.14) (2.60) - 0.574 log(F_,/A O + 0.714 logPa* (2.19) (5.71)

218

(2)

V . P a n d it

log Pnfc = 5.041 + 0.815 (log M_t + M_2/ X ) (7.50) (6.49) -

0.536 log ( Q n f / X m ) + 0.136 logPn/c , (2.01) (1.14)

+ 0.140 © (5.62) OLS: J? 2 = 0.989, 4 = 1 .9 7 (3)

log Pfm —— 0.078 + 1.036 log Pfa (0.33) (20.42) OLS: tf 2 = 0.959, d= 1.50

(4)

log Prm = - 0 .1 9 3 + 1.012 log Pnfc + 0.031 log P im l (6.46) (42.01) (1.39) OLS: R 2 = 0.999,