PROTOCOLOS Y CONVENIOS DE COLABORACIÓN

The criteria suggested by Dobb (1960) and by Sen (1968) are rather similar. For brevity, we shall consider only the proposed criterion as developed by Sen.

In the Sen model the economy is divided into two sectors: one is modern, the other is backward. The modern sector is again subdivided into two parts: one sector (A) is producing machinery with only labour;

the other sector (B) is producing corn by using machinery and labour. In the backward sector corn (i.e. a consumer good) is produced by labour alone. Labour productivity in the modern sector A is given by the capital intensity of the technology applied there where the capital intensity is given by the total number of man-years necessary in sector A to turn out enough machinery for one unit of labour in sector B. Sen assumes that wages in the modern sector are determined by the corn output produced by sector B. But since it takes some time to set up the modern sector, wages in the modern sector would have to be paid out of the

‘surplus’ in the backward sector.

Sen then distinguishes between the following aims: (i) maximization of current output (i.e. corn); (ii) maximization of the rate of growth of output; (iii) maximization of the undiscounted flow of output over a finite period of time. The choice of capital intensity will differ according to the nature of the objective. Sen describes how a conflict can arise between the current output maximization principle and the criterion to maximize the rate of growth of output (see Figure 4.1). Let the vertical axis measure output (corn) and the horizontal axis measure employment of labourers. In the southeast corner the combination of labour and capital is measured. The curve OQ is the production function and OW is the given wage line which shows wage bills at various levels of employment. At point E, we obtain maximum output (EL) and employment of OL which would work with OK amount of capital; the degree of capital intensity is given by the tangent of the angle OLK. Labour productivity is given by the tangent of the angle P_1nOL. Note that at E we find that the capital turnover and the SMP criteria are satisfied. But if the objective is to maximize the reinvestible surplus to obtain a higher rate of growth, the point to choose is E_1, where AB is tangent to OQ and parallel to OW. Maximum surplus is shown by E₁S₁. It is easily observed that if the Galenson-Leibenstein criterion (i.e. MRIS) is followed, the point E₁ rather than E would be chosen, capital intensity would be higher at E₁ than at E (i.e. the tangent of the angle OL₁K is greater than that of OLK) and labour productivity would also be higher (i.e. the tangent of the angle P_2nOL is greater than that of P_1nOL). Given similar diagrams, it is also possible to show that maximization of current output may lead to the emergence of a negative surplus. However, maximization of output need not mean maximization of employment as can be shown by the difference between OL and OL₂. In fact, employment expansion beyond OL₂ leads to the emergence of negative ‘surplus’ (‘eating’ the railways!).

The moral of the above exercise is to show clearly that there is a basic conflict between maximizing consumption at present (by using the capital turnover or SMP criterion) or in future (by applying the MRIS criterion) (see Figure 4.2). Let Pp′ be the production possibility frontier for an LDC. If all the resources are allocated to the production of investment goods, OP of capital goods would be produced; on the other hand, if all the resources are spent on the production of consumer goods, Op′ of consumer goods would be produced.³ Obviously, society will choose to produce some combination of both the goods. However, lines Og and Og₁ represent different growth rates (4 per cent and 6 per cent respectively). Assume that the economy is growing along Og. But if a higher growth rate (say Og1) is regarded as desirable, it requires a cut of consumption goods by cd which would allow resources to be released for the production of more

investment goods to take the economy to F; given a rise in the production of investment goods, growth rate will be higher, i.e. 6 per cent instead of 4 per cent. But present consumption must be sacrificed to obtain the higher growth path, though the choice of a higher growth path at present will ensure higher consumption in the future as well (see also the appendix to Chapter 11).

As one of the solutions to the dilemma, Sen has proposed that since the choice of investment criterion depends upon the time horizon of output generation, the time preference and the social welfare function (assuming that such a function is available to the planners), the best way of looking at the problem would be to derive the alternative time series of consumption obtained by following different criteria. The point can be shown clearly in Figure 4.3. Let the vertical axis measure the growth of output of consumer goods and the horizontal axis measure time. Output can be produced by either technique K or technique L. Technique Figure 4.1

Figure 4.2

K produces less output now than technique L but after B the rate of growth of output under technique K is such as to compensate for the initial loss of output by the year T₁. It is assumed that the area AA′B=BCC′.

Although the use of technique L generates more output in the current period, it yields progressively less and less in future, given the slope of the curve AL. Now if the society is prepared to wait until period T₂ (say for thirty years), consumption sacrificed at present could be made up by that time and after T₂ society would enjoy higher output and consumption by choosing technique K. But if the social welfare function is such that society values present output and consumption more than future output and consumption, then the society may well choose technique L.

Evaluation of the reinvestible surplus criterion

The choice between maximization of output and maximization of employment is more complex than the previous analysis would suggest. Output is a heterogeneous concept; so is employment (Stewart and Streeten 1972). Since both output and employment change over time and since present output and employment may affect future levels, both intra- and intertemporal weighting is very important. It is possible to state that generation of more output (with given capital and technology) will require more labour and to that extent the conflict between the objectives to maximize employment and output is more apparent than real. However, the conflict between the two objectives would be more real if it is assumed that a new technology is chosen. The example given by Stewart and Streeten (1972) can be cited. Suppose £100,000 is the amount of money available for investment in a textile industry. Let the capital-output ratio be 2.5 if the advanced technology is used; if the capital cost per work place is given as £1,000, then additional output would be worth £40,000 and additional employment would be 100. But if a traditional hand-spinning technique is used, where the capital-output ratio is 5.0 and the cost per work place is given as £100, then additional employment would be 1,000 but the value of additional output would be only £20,000. Note that in this case, although the capital-labour ratio is lower, the capital-output ratio is higher than for the more capital-intensive modern method. This is because a large-scale capital-intensive technique can economize Figure 4.3

on capital since economies of scale occur and this leads to a fall in capital cost in relation to output (Kaldor 1965; Amin 1969).

It is also necessary to point out that output is likely to rise with extra employment (unless labour is wholly unemployed in which case they should not be employed, at least not in the organized profit-maximizing modern sector, in the first place) and the level of employment will be largely given, inter alia, by the level of wages. In the Dobb-Sen model (see Figure 4.1) it is assumed that wages are fixed and this helps to explain the dichotomy between employment and output maximization. But if the real wages are allowed to fall, the conflict between the two objectives will be minimized. Evidence suggests that more labour-intensive methods like traditional spinning could also save more capital per unit of output in comparison with modern factor methods (Bhalla 1964). So long as indigenous materials could be used by the unemployed labour without involving a diversion of resources, an increase in employment will also lead to a rise in output. (See Appendix 4 for a simple proof.)

It is usually assumed that higher consumption rather than saving will follow the use of labour-intensive rather than capital-intensive techniques of production and this will lower the growth rate for the economy as a whole. A conflict then occurs between macro and micro concepts of efficiency (Meier 1976). But such an argument rests on the following premises:

1 wages are independent of the choice of techniques;

2 all wages are consumed and all profits are saved;

3 fiscal policy is inadequate to raise taxes to obtain the desired savings ratio and real wages are unlikely to be reduced even when inflation takes place in many LDCs.

Given these premises, the impact of the choice of increasing capital intensity on growth and employment within the neoclassical theory can easily be analysed using Figure 4.4. The horizontal axis measures the capital-labour ratio (C/L) and the vertical axis measures output per unit of labour (Q/L). The production function is given by OP and it shows that, for any output, present employment is maximized by using the most labour-intensive technology and this is reflected in a move towards the origin in the diagram.

However, employment growth will be given by output growth at a given capital intensity (assuming away technical progress or even assuming neutral technical progress). Given the assumption (2) above, output-and employment-maximizing technologies will be determined by the level of wages. Let wages be OW output-and given the assumption (1) the rate of growth of the economy is shown by the slope of WR to the production function. The highest growth rate is reached at E but this is not the point at which maximum employment is obtained since we have moved further from the point of origin. If the techniques of production to optimize the growth rates of employment and of output are the same, the conflict between them disappears. The conflict will be aggravated if real wages rise.

Evidence suggests that wages are related to labour productivity, the scale of activity and the choice of technology, and that small enterprises which usually adopt more labour-intensive methods (i.e. lower C/L) also offer a lower level of wages (Dhar and Lydall 1961; Shetty 1963; Okita 1964). Thus the choice of technique is not independent of the scale of operation and the level of wages. Also it is not wholly realistic to assume that all profits would be saved. A part of them may be frittered away in conspicuous consumption. Besides, foreign and multinational enterprises repatriate some profits, interest and royalties which clearly reduces reinvestible funds. The implicit assumption in the Sen-Dobb model is that all profits would be saved because profits would accrue to the public sector and the means of production would be owned by the state.

This may be questioned because most LDCs have mixed economies. Further, the inability of the government to raise savings by manipulating fiscal and wage policies is viewed with scepticism when the same government

is able to sacrifice present consumption and employment by choosing certain techniques (Stewart and Streeten 1972). Moreover, choice of techniques could be influenced by the degree of competition. In a competitive economy, to maintain or maximize profits, producers will try to minimize costs and look for an optimum combination of factors of production, while in a protected economy, the producers are under no such compulsion to reduce costs by introducing technical progress or by choosing the optimal proportions. It is worth mentioning that despite labour abundance and capital scarcity in many LDCs, neither domestic nor foreign enterprises have shown much interest in taking advantage of the existing factor-cost ratios.

Some other points are also worth emphasizing in this connection. First, the problem of investment allocation cannot be viewed only in static terms. It is important to observe the dynamic optimal growth paths. It has been shown (Srinivasan 1962) that such an optimal sustainable per capita consumption growth path exists for an economy consisting of two sectors— one sector producing consumer goods, the other capital goods. Others have also tried to analyse the theoretical properties of such growth paths (Findlay 1966; Dixit 1968; Bose 1968; Uzawa 1962).

Second, the choice of optimum technology is a rather complex issue. Different sectors may require different intensities which would be optimal; e.g. an optimal technology for agriculture may require the use of labourintensive technology whereas such an optimality may be reached in choosing a capital-intensive method in generating power and electricity. Thus micro concepts of optimality should not be confused with the macro objective of maximizing output growth rate by utilizing all the inputs which could be obtained.

To achieve overall consistency in resource allocation, investment planning or programming is necessary.

The use of input-output tables and programming models would be quite useful in such cases.

Third, different technologies embody different types of externalities which are not usually considered within a static analysis. Such externalities could arise because of the economies of scale in industries. Due regard to the externalities in a dynamic analysis may well influence the choice of techniques. Finally, the pattern of income distribution would be different with different technologies. A labour-intensive technology (say, a seedfertilizer revolution in the agriculture of LDCs which could be labour using) could raise total and per acre output, but because of higher employment and higher marginal propensity of agriculturalists to consume, the saving-income ratio may fall, leading to a fall in the growth rate. It is, of course, assumed that higher savings and investment, rather than higher consumption, leads to a higher rate of growth. Mirrlees Figure 4.4

(1975), on the other hand, has argued that growth can be increased by increased consumption. This line of argument is not new altogether as it has been shown before (Leibenstein 1957) that an increase in consumption in LDCs will mean better nutrition, greater efficiency and higher productivity of labourers in LDCs. The empirical tests of such models have hardly been carried out either to accept or to reject such theories.

Empirical evidence suggests that choice of techniques does exist in many LDCs in manufacturing, metal working and textile industries (Bhalla 1975). Similarly, cost of production and thus the choice of technique is influenced not only by prices but also by the scale of production. Also, relative factor-price differences between rural and urban areas in an LDC may influence the choice (Stewart 1975). Again, substitution possibilities between different types of labour and between labour and working capital, as well as the choice of products, can affect the final choice of technology.

4.6