ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Economic Growth and Employment

The fact that an economy, even when it experiences a higher growth rate in the capitalist segment, is saddled with an increasing unemployment rate, goes against the grain of conventional growth theory as indeed of the basic presumption underlying policymaking. In India, for instance, faced with growing misery in the midst of accelerating growth, the standard response has been that such "exclusion" will disappear if the growth rate can be further accelerated. The proposition that an acceleration of the growth rate in the capitalist segment will be accompanied by an increase in unemployment in the economy as a whole, which is the basic element underlying poverty (as Marx had recognised), cuts at the foundation of all such claims.

SPECIAL ARTICLE

Economic Growth and Employment

Prabhat Patnaik

The fact that an economy, even when it experiences a higher growth rate in the capitalist segment, is saddled with an increasing unemployment rate, goes against the grain of conventional growth theory as indeed of the basic presumption underlying policymaking. In India, for instance, faced with growing misery in the midst of accelerating growth, the standard response has been that such “exclusion” will disappear if the growth rate can be further accelerated. The proposition that an acceleration of the growth rate in the capitalist segment will be accompanied by an increase in unemployment in the economy as a whole, which is the basic element underlying poverty (as Marx had recognised), cuts at the foundation of all such claims.

Prabhat Patnaik (prabhatptnk@yahoo.co.in) recently retired from the Centre for Economic Studies and Planning, Jawaharlal Nehru University, New Delhi.

I
n all strands of economic theory, the concept of economic growth is discussed almost exclusively within the context of a closed and isolated capitalist economy. The usual setting is as follows: the application of labour to a set of produced means of production generates products which are distributed between the providers of labour and the providers of the means of production; those to whom these products accrue consume a part of what comes to them and “save” the rest; if these savings equal what is invested, then there is no problem of aggregate demand and full employment/full capacity output is realised; investment, less what is scrapped, adds to the stock of equipment available in the next period, which together with the available labour force determines the full employment/full capacity output of the next period.

The questions of whether autonomously-determined investment will necessarily equal full employment/full capacity savings (and in the absence of such an equality how investment will move over time in the demand-constrained system that will emerge); and of whether, even in the absence of a demand-constraint, there is any substitutability between equipment and labour such that both can be fully employed at the same time; and, related to this, what determines the movement of distributive shares in the economy; have occupied growth theorists. Differing answers to them characterise different strands of growth theory.

David Ricardo did bring in the point that land as a means of production was non-augmentable in supply, but his theory that in this situation cultivation necessarily proceeded from superior to inferior land was rightly criticised by Marx (who undermined Ricardo’s theory of rent). Besides, Ricardo assumed that agriculture in its entirety was carried out under capitalist conditions, in which case the fact of land not being augmentable merely meant that the economy moved to a stationary state. This left no scope for examining changes in social relations, which the non-augmenta bility of land in the context of capital accumulation might entail.

The general point, however, remains: growth theory looks exclusively at a capitalist economy existing in splendid isolation and uses the assumption of perfect augmentability of all inputs other than one primary input which is normally taken to be l abour. (In von-Neumann labour itself is perfectly augmentable.)

1

Within this paradigm where everything other than labour is produced, the relationship between output and employment growth is fairly straightforward. If we assume fixed coefficients which change over time through technological progress, and no deficiency of aggregate demand (with capacity rather than labour availability being the constraining factor for output), then the rate of growth of employment is determined as the difference

june 25, 2011 vol xlvi nos 26 & 27

EPW
Economic & Political Weekly

SPECIAL ARTICLE

b etween the rate of growth of output (which depends upon the savings and capital-output ratios) and the rate of growth of l abour productivity (which depends on the pace of technical progress). Here, a rise in the rate of growth, unless accompanied by a still greater rise in the rate of growth of labour productivity, will necessarily raise the rate of growth of employment.

If we take the Kaldor-Verdoorn formulation (Kaldor 1966), of the rate of growth of labour productivity being a function of the rate of growth of output, say Gp = a + bG where Gp is the rate of growth of labour productivity, and G the rate of growth of output, then the rate of growth of employment Ge which simply equals G-Gp, becomes an exclusive function of the rate of growth of output, i e, Ge = –a + (1–b)G. And with b<1, any rise in the growth rate of output raises the rate of growth of employment.

Self-limiting

When this perception is put together with another, quite plausible, perception, namely, that in a situation where the unemployment rate exceeds a certain threshold rate, real wages remain tied to a certain subsistence level (which need not of course be biological, but may be determined by historical factors), so that all gains in labour productivity accrue to the capitalists, we get a strong and apparently comforting conclusion, namely, that the unemployment rate cannot increase secularly, that it is necessarily self-limiting. This is because with wages tied to the subsistence level and productivity rising, the share of surplus in output rises, and with it the savings ratio in the economy (since consumption per unit of surplus is typically lower than consumption per unit of wage income). The rise in the savings ratio, since the capitaloutput ratio can be taken to be a constant, raises the growth rate. This, in turn, as we have just seen raises the rate of growth of employment. Since the rate of growth of the workforce is given, this continuous increase in the share of surplus and hence in the growth rate of output and employment, must eventually lower the unemployment rate and keep doing so until tightness develops in the labour market and wages start rising above the subsistence level. Two fairly standard conclusions therefore follow from this: first, the unemployment rate cannot increase secularly in a world where there are no problems of aggregate demand; and secondly, any rise in the rate of growth of output necessarily brings down the unemployment rate.

The foregoing can also be understood in terms of Ricardo’s argument on the impact on employment of the introduction of machinery, or what Hicks (1973) has called the “classical traverse”. This argument states that the introduction of machinery, even though it displaces labour to start with, raises the rate of profit in the economy and hence the rate of growth of output. Since we are talking of a one-shot introduction of machinery, which raises l abour productivity once for all (after which there is no further increase, i e, the subsequent rate of growth of labour productivity is zero) the rate of growth of employment after the introduction of machinery is the same as the rate of growth of output; and since the latter increases with the introduction of machinery ( because of the higher rate of profit), so does the former. With a

Economic & Political Weekly

EPW
june 25, 2011 vol xlvi nos 26 & 27

higher rate of growth of employment on account of the introduction of machinery, the initial labour displacement caused by machinery is more than made up after some time.

A criticism against this Ricardian argument is that it focuses only on a one-shot introduction of machinery, rather than on technological progress, i e, rather than on a sequence of introductions of newer and newer machinery which would keep raising l abour productivity. This is the criticism to which our argument of the preceding section is addressed. What it says is that as long as wages remain at the subsistence level, as in Ricardo, a sequence of increases in labour productivity keeps raising the rate of profit and hence the growth rate. As long as the rise in the growth rate of productivity is less than the rise in the growth rate of output (i e, b<1), the rate of growth of employment rises with the rate of growth of output, in which case tightness in the labour market is bound to arise sooner or later for any given rate of workforce growth.

3

But this entire reasoning has been shown to be untenable in practice. In India, for example, despite high growth being maintained for years there is no sign of any tightness in the labour market. This tightness of course cannot be inferred from the movement of the unemployment rate alone in a situation where there is so much underemployment, and disguised unemployment; some other proxy has to be found for gauging the tightness in the l abour market. One possible proxy is the wage rate of organised industrial workers, which in a situation of tightness in the labour market should be expected to show an increase. If we take this proxy, however, then we find that in India between the late 1980s and now there has been an absolute decline in the real wage rate of organised industrial workers, which suggests that the labour market, far from becoming tighter in the period of rapid economic growth, has continued to remain slack. How do we explain this phenomenon in the light of the above analysis?

Lack of Conformity

One obvious argument would be that even though the labour market has remained slack till now, further acceleration in the growth rate, which will necessarily ensue because of the rising share of surplus, will eventually make it tight. The theory outlined above in other words is right but its validation will be visible only after a lapse of time. But India’s rapid growth rate has been going on for almost a decade now, and if it still has made no impact on the labour market then there is no reason to believe that it will do so in the years to come, or that it will do so by b ecoming even more rapid in the coming years. What we find in other words is a lack of conformity between the theory given above and the real-world Indian experience.

Some may prefer to keep the theory as it is and simply postulate that an increase in the growth rate of output has such an extraordinarily large effect upon the growth rate of labour productivity that, far from increasing the growth rate of employment, it actually lowers the growth rate of employment, that in other words b>1.

Others may put forward a slightly different argument but to similar effect. This would run as follows: since the working

SPECIAL ARTICLE

p eople generally consume relatively simple goods, like foodgrains for instance, in whose case process and product innovations are of little significance and the rate of growth of labour productivity is relatively small, while the surplus earners consume commodities which are ever changing and where the rate of growth of labour productivity is relatively higher, the overall rate of growth of labour productivity should be a function not just of the rate of growth of output as Kaldor and Verdoorn had postulated, but of a distributional parameter as well. And since, with real wages given, the rate of growth of surplus per worker depends upon the rate of growth of labour productivity, this distributional parameter can be approximated by the rate of growth of labour productivity itself, so that we can say that Gp = a + bG + cGp. In this case, the effect of G on Gp is magnified from b to b/(1-c); there is in other words a sort of multiplier effect whereby a high productivity growth, by ensuring a high growth of surplus, contributes to further productivity growth. We can with greater plausibility think of b/(1-c) as exceeding 1, or, what comes to the same thing, b+c together exceeding 1, rather than b alone exceeding 1.

But this explanation, though not without significance, does not carry us very far. It simply makes the observed behaviour of the system explicable in terms of its obeying certain parameter values; but why it should obey those particular parameter values is left unexplained. This explanation in other words points in a c ertain direction but does not constitute a proper explanation. In what follows we shall attempt another explanation which can be additional to whatever has been said above, but which constitutes an explanation in a truer sense of the term.

4

To see the logic of this explanation let us set up an imaginary economy where there is no population (and workforce) growth, so that persistent unemployment can no longer be explained in terms of excessive growth in labour supply, as is commonly done. Let us also assume away any problem of aggregate demand. A part of the surplus is spent on activities where land is a major i nput (e g, golf courses); and to present the matter starkly let us assume that a part of the surplus is spent on acquiring land itself which generates no employment but satisfies buyers’ needs. The supply of land is given and non-augmentable; and that part of it which is not already being “consumed” by the surplus earners is under the control of peasants who use it for producing commodities, whatever they may be, which satisfy their own needs. The peasants in other words belong outside the capitalist economy. As for the workers who belong to the capitalist economy, their consumption needs are met exclusively from the output of the capitalist economy itself, as are the non-land consumption needs of the surplus earners. (We follow in other words the standard growth theory perception, except for the introduction of a d emand for land by surplus earners). Let us assume wages to be at the subsistence level to start with, and, hence, by implication, the existence of a certain level of unemployment.

Let the gross surplus produced in the capitalist segment of the economy in any period be denoted by M. The net surplus will be M-d.K where d.K denotes the amount of capital scrapped on account of decay (we assume that a proportion d of the capital stock

174 simply dies in each period but that all surviving capital stock, no matter of what vintage, is equally efficient irrespective of age). Let us denote this net surplus by S. A part m.S of this surplus is used, say at the end of the period, for “buying” land from the peasants. Let the price of land in terms of the capitalist sector’s product be p and let the amount of labour used per unit of land in the peasant economy be l. Then in any period t, employment in the peasant economy will go down by m.S (t-1).l(t)/p(t) owing to land purchase by surplus earners. Let n.S be the amount saved and invested by the surplus earners and let the capital-labour r atio in the capitalist economy be denoted by k. Then the increase in employment in the capitalist economy in any period t on a ccount of the reinvestment of savings by the surplus earners will be [n.S(t-1)]/k(t).

Sale Proceeds

In addition, however, we have to ask what the peasants who sell land in any period do with the proceeds of their land sale. Let us deliberately assume the most optimistic scenario for employment in period t; this is the scenario where the peasants selling land consume nothing from their sale proceeds but lend the entire sale proceeds back to the capitalists for undertaking additional investment. And let us assume for simplicity that the rate of interest they earn is nothing else but the rate of profit earned by the capitalists (which vastly exaggerates the benefits from investment to the peasants, since we are in effect assuming that the peasants after selling their land simply become capitalists). The total employment increase in the capitalist sector then will be [(m+n)S(t-1)/k(t)].

If m.S (t-1).l(t)/p(t) > [(m+n).S(t-1)/k(t)], then in the economy as a whole, i e, taking the capitalist and peasant economy together, there will be a net loss in employment in period t. This condition simply boils down to m.l(t)/p(t) > (m+n)/k(t), which means that the employment destroyed per unit of surplus through land acquisition exceeds the employment created per unit of surplus through investment. Now, even if the l.h.s. of the inequality is smaller than the r.h.s. to start with, if m and n are constants, labour requirement per unit of land in activities using land is unchanging and the price in terms of the capitalist sector’s product at which land is acquired is also more or less given, then sooner or later this inequality must be satisfied, i e, l.h.s. must exceed the r.hs., because k(t) is continuously rising through technological progress. Of course it may be argued that labour requirement per unit of land is not unchanging, but nobody can possibly argue that it is falling at a faster rate than labour requirement per unit of capital in the capitalist sector. Besides, land is typically acquired not from farms where rapid technological change is taking place, but from precisely the farms where such change is slow, i e, from small- and medium-sized farms.

Likewise as far as land price in terms of the capitalist sector’s product is concerned, what we have to consider here is not whether it is high or low to start with, but whether it rises at a rate faster than the amount of capital per unit of labour in the capitalist sector. And here what we are talking exclusively about is the price of land received by the peasants. There is no reason why the rate of inflation on land sales by the peasants, even when

june 25, 2011 vol xlvi nos 26 & 27

EPW
Economic & Political Weekly

SPECIAL ARTICLE

it exceeds the rate of inflation on the capitalist sector’s products (i e, even when p is rising), should do so by a margin that exceeds the rate of growth of capital per unit of labour in the capitalist sector. Even if perchance this were to threaten to become the case, we know that a paraphernalia of government intervention measures, supposedly “in the public interest”, is invariably put in place to keep under control the land price that is realised on sales by the peasantry.

Decline in Level of Employment

It follows that the condition for economic growth in the capitalist sector being accompanied by growing unemployment is hardly a stringent one. It requires not that the rate of growth of labour productivity in the capitalist sector should exceed the rate of growth of output in that sector, but that the rate of growth of l abour productivity in that sector should simply be positive, in which case with m,n, l, and p being more or less given, a time must come when economic growth under capitalism lowers the level of employment in the economy as a whole. (True, before this time has come, tightness may develop in the labour market, in which case the above reasoning will have to be somewhat modified, but that does not mean that the above conclusion about secularly growing unemployment will be necessarily invalidated, i e, that the tightness will persist; this, however, is a matter we need not pursue here and hence we assume that no tightness in the labour market develops in the interim, i e, before the inequality begins to hold.)

Putting it differently, unlike in the earlier discussion where we were talking about the generation of unemployment in an economy fully dominated by capitalism, i e, about unemployment b eing generated within the capitalist segment itself, here we always have the capitalist segment adding to its employment, but unemployment is generated outside of it, and that too (when the above condition is satisfied, as it must sooner or later be) to an extent that is higher than the employment generated within the capitalist segment. So, the capitalist sector in this universe is showing growth, including in employment, but the economy as a whole is experiencing growing unemployment, even out of a given workforce. (Of course this universe will not last forever, since at some point the peasantry will lose its entire land, but what happens in that limit need not detain us here. We are not talking about “steady states” anyway.)

A point about the above argument must be noted. Unemployment is rising not because the landowning peasantry is getting destroyed, but because traditional land-using activities, which also employ substantial quantities of labour, are disappearing. As a matter of fact of course both these things happen together, in the process of the so-called “primitive accumulation of capital”. Not only is the peasantry destroyed but “peasant agriculture” disappears. But the argument here is only about the d estruction of peasant agriculture, which will have an employment-constricting effect even if the landowning peasants from whom land is acquired flourish as creditors or equity-holders in the capitalist sector.

Economic & Political Weekly

EPW
june 25, 2011 vol xlvi nos 26 & 27

Questionable Justification

This distinction between peasants and peasant agriculture has a relevance. It is sometimes argued that as long as the compensation paid for land acquisition is adequate in the sense that the interest earned on it would exceed what the owners would have earned as income from keeping it in its original use, there should be no objection to land acquisition for the capitalist sector. This argument is questionable for two separate reasons. The first, which is not covered by our model, is the following: the income of the landowning peasantry itself is not some independently-given magnitude but depends upon its interaction with the capitalist sector. If the capitalist sector squeezes this income by turning, effectively, the terms of trade against it, then our condoning its acquiring land for a “pittance” on the grounds that the incomes of the peasants were already even worse than a “pittance” would be inexcusable. It would justify one injustice in terms of another.

The second reason, which the present paper highlights, is that the question of employment-generation is independently important and is different, as an issue, from the compensation for the land that is acquired. Even if the latter is sumptuous, it may nonetheless be the case that total employment in the economy over a period of time becomes smaller than it would have been in the absence of such land transfers.

Likewise, many well-wishers of the peasantry have demanded that the peasants who surrender land for acquisition by the capitalist sector should be given equity in the enterprises that use this land. This is a proposition that has much to recommend it by way of “fairness”. But even the practising of such “fairness”, it must never be forgotten, will not prevent the deleterious employment effects of land acquisition. In fact such “fairness” is assumed in the above model which nonetheless shows that employment over a p eriod of time will be lower than it would have otherwise been, a proposition that is the precise opposite of Ricardo’s on machinery.

6

On the basis of the above reasoning we can see why both the standard propositions about the relationship between growth and unemployment are erroneous. The proposition that the unemployment rate cannot increase secularly does not hold because in any economy where m.l/p > (m+n)/k, as growth occurs unemployment keeps increasing. Since we have assumed a constant workforce, the unemployment rate also keeps increasing.

The second proposition that a rise in the rate of growth of output under capitalism brings down the unemployment rate (for any exogenously given rate of growth of the workforce), also does not hold. This can be seen easily as follows. The growth path that the capitalist segment of the economy will be traversing in our model is not one of steady growth but one of increasing growth (since the real wage rate remains at the subsistence level and the share of surplus in output and hence the investment ratio keeps increasing). The capital-labour ratio in this segment not only increases but does so at an increasing rate along this path. This is because typically a rise in the rate of growth of output raises the rate of growth of labour productivity while keeping the capitaloutput ratio more or less unchanged (one of Kaldor’s “stylised facts” for which much empirical support can be adduced even in

SPECIAL ARTICLE

the more recent period (Patnaik and Chandrasekhar 1996)). Hence a rise in the rate of output growth lowers the rate of growth of employment per unit of capital stock.

In the notation given earlier this means that a rise over time in output growth raises the rate of growth k, which in turn implies, as we have seen, that if the inequality was not being satisfied to start with, it necessarily gets satisfied over time. When it is satisfied, as the rate of output growth in the capitalist segment keeps increasing, the magnitude of employment in the economy keeps falling and hence the unemployment rate keeps rising. What we observe over time therefore is increasing growth in the capitalist segment together with an increasing unemployment rate in the economy as a whole. The presumption that an increase in growth rate under capitalism is necessarily associated with a reduction in the unemployment rate breaks down. (It should be noted here for clarity that we are looking at increasing growth along a particular trajectory over time not comparing alternative trajectories, for which more explicit formulations would be needed).

Against the Grain

The fact that an economy, even when it experiences a higher growth rate in the capitalist segment, is saddled nonetheless with an increasing unemployment rate, goes against the grain of conventional growth theory as indeed of the basic presumption underlying policymaking. In India, for instance, faced with growing misery in the midst of accelerating growth (though official pronouncements never talk of “growing misery” only of “exclusion” which is never perceived to be growing), the standard response has been that such “exclusion” will disappear if the growth rate can be further accelerated. The proposition that an acceleration in growth rate in the capitalist segment will be accompanied by an increase in unemployment in the economy as a whole, which is the basic element underlying poverty (as Marx had recognised), cuts at the foundation of all such claims.

Basic Logic

The argument of the present paper has been developed on the basis of assumptions which have been kept deliberately strong in order to bring home its basic logic. But this logic is so simple and clear that it can be easily checked that the weakening of particular assumptions will make little difference to the argument, as long as the basic framework of the argument remains. Demonstrating this here is both unnecessary and tedious, and will be avoided. Likewise, the self-defined scope of this paper precludes any discussion of what an alternative employmentpromoting growth strategy will look like. It is of course implicit in the logic of the paper itself, but our concern here has been to elaborate this logic rather than propounding an alternative s cenario of growth and an alternative policy regime for achieving it.

References

Hicks, J R (1973): Capital and Time (Oxford: Clarendon Press).

Kaldor, N (1966): Causes of the Slow Rate of Growth of the United Kingdom (Cambridge University Press).

Patnaik, P and C P Chandrasekhar (1996): “Investment, Exports and Growth: A Cross-Country Analysis”, Economic & Political Weekly, 6 January.

REFLECTIONS ON EMPIRE
March 26, 2011
Resurrection and Normalisation of Empire – Rohit Chopra
Taming the Imperial Impulse: Realising a Pragmatic Moral Vision – Abdullahi Ahmed An-Na’im
Adam’s Mirror: The Frontier in the Imperial Imagination – Manan Ahmed
Indian Empire (and the Case of Kashmir) – Suvir Kaul
Imperial Democracies, Militarised Zones, Feminist Engagements – Chandra Talpade Mohanty
Rethinking News Agencies, National Development and Information Imperialism – Oliver Boyd-Barrett
Digital Imperialism through Online Social/Financial Networks – Radhika Gajjala, Anca Birzescu
Pandemic, Empire and the Permanent State of Exception – Cindy Patton
For copies write to:
Circulation Manager,
Economic and Political Weekly,
320-321, A to Z Industrial Estate,
Ganpatrao Kadam Marg, Lower Parel, Mumbai 400 013.
email: circulation@epw.in

june 25, 2011 vol xlvi nos 26 & 27

EPW
Economic & Political Weekly

Dear reader,

To continue reading, become a subscriber.

Explore our attractive subscription offers.

Click here

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top