CHAPTER 2: WAGE POLICY

2.1 Introduction

Wage policy is a major area of concern in India for three basic reasons. First, the government's influence on wage movements in the economy is very substantial. In the organized segment of the economy, this influence is quite pervasive. Apart from the direct role that it plays in wage setting in the public sector, the government also employs a large number of regulatory instruments to influence the wage setting process in the private sector. Besides, wage movements in the private sector cannot but be influenced by those in the public sector, given the dominance of the public sector in the organized segment of the economy. On the whole, it can be said that the role of collective bargaining is much less important than that of the government in wage determination in the organized sector. Even in unorganized sectors of the economy, the government seeks to influence wages, with varying degrees of success, through such mechanisms as minimum wage legislations. Thus alterations in the pattern of wage movements require changes in wage policy pursued by the government.

Second, the past pattern of wage movements does not seem to have been particularly healthy. In unorganized sectors, wage earners constitute a large proportion of the poor and wage movements in the past do not seem to have been very helpful in enabling them to lift themselves above the poverty line. In the organized sector, on the other hand, wage movements seem to have discouraged employment growth by encouraging growth of capital intensity.^{(Endnote 10)} There are thus reasons to think that the pattern of wage movements needs to be changed.

Third, the recent economic reforms as well as some recent developments in the world economy have brought the issue of wage flexibility in organized industries to the fore. As the economy opens up and the forces of globalization gather strength, cost adjustments will assume increasing importance and wage flexibility can clearly facilitate cost adjustments. The issues involved relate to appropriate linkages among wages, productivity and profitability of enterprises which determine the competitive strength of industries or enterprises. Given this context, wage policy in organized industries needs to be reviewed; in particular, possibilities of increasing the flexibility of the wage system need to be examined.

In what follows, the instruments and objectives of the government's wage policy are reviewed and the actual results in terms of wage movements are critically examined. This provides the basis for identifying some major weaknesses of the wage policy pursued in the past.

2.2 The Instruments of Intervention

The basic framework for government interventions in the wage determination process was set out in The Report of the Committee on Fair Wages, 1948. Following the recommendations of this Report, the government designed fairly elaborate methods of intervention into the wage determination process. Among the principal methods which are employed to this day are minimum wage fixing; guiding wage fixation/revision in particular industries and sectors; setting of indexation and bonus payment rules; and broad controls over wage differentials. Of these, only the minimum wage rules cover the unorganized sectors; all the others seek to regulate wages in the organized sector or a segment of it.

Fixing Minimum Wages

The principal instrument for fixing and enforcing minimum wages for the unorganized sectors is the Minimum Wages Act, 1948. Under this Act, minimum wages for an eight-hour work-day are fixed for certain types of jobs where, in the judgement of the government, the market conditions make the workers particularly vulnerable to exploitation. With anyone employing at least one hired worker being defined as an employer under the Act, the coverage is wide. Minimum wages under the Act can be fixed by either the central or the state-level authorities. To start with, the Act covered 13 occupations (agricultural labour and 12 non-agricultural occupations). Currently, 40 occupations are covered in the central sphere and, in the major states, the number of occupations covered ranges from 5 in Manipur to 79 in Orissa. In all, more than 200 occupations are currently covered under the Act.

Fixing/Revising Wages in the Organized Sector

An important recommendation of the Committee on Fair Wages was for establishment of Tripartite Wage Boards for such industries as deemed appropriate by the government. Accordingly, the central government and some state governments set up Wage Boards for several industries. Thus 27 Central Wage Boards were set up covering 19 industries; for a few industries, such Boards were set up more than once at different points of time. The basic task of the Wage Boards was to work out a wage structure for the industry concerned keeping in view considerations relating to the need to guarantee a decent minimum wage, the need to facilitate growth of the industry concerned and the need to provide incentive to workers to upgrade skills.

In the case of Central Public Sector Undertakings (PSUs), there are periodic wage settlements (the typical period of agreement being four years). While the individual Central PSUs enjoy a measure of autonomy and there is some scope for collective bargaining, actual settlements are required to be consistent with the fairly detailed guidelines issued by the Department of Public Enterprises. In principle, the state-level PSUs could have independent systems or guidelines for wage settlement, but in practice they tend to imitate the Central PSUs.

It should be noted, however, that the recommendations of Central Pay Commissions, specially appointed to fix salaries of central government employees, usually serve as reference points for the work of the Wage Boards and the Department of Public Enterprises. This has an important consequence. The fact that salaries of government employees serve as reference points for wage settlements in industries means that considerations relating to industrial growth and skill formation are virtually impossible to take into account in practice.

Setting Indexation Rules

Several government committees and study groups, apart from the wage-fixing bodies, have addressed the question of compensating employees for increases in cost of living. As a result of these discussions and deliberations, a few widely accepted principles have emerged.

There seems to be a consensus that full compensation for increases in cost of living should be allowed only at the lowest level of the wage ladder, i.e., at and around the minimum wage (in the organized sector); the extent of compensation should decline as the level of wage (or salary) increases. The consensus is based on two main arguments. The first is that full indexation would make the economy vulnerable to wage-price spirals. Second, it is argued that indexation is a device for neutralizing the effects of increases in the cost of essential items of consumption. Since the share of essential items in total consumption expenditure declines as income rises, the degree of indexation should decline too. However, whatever be the merits of these arguments, it is obvious that if the degree of indexation varies inversely with the wage level, wage differentials would tend to decline over time. This can generate dissatisfaction and hence erode incentives. But these aspects do not seem to have received much attention.

Given the principles, basically two systems of indexation have emerged. The first is the so-called Industrial Dearness Allowance system which involves payment of a fixed sum per point-increase in the Consumer Price Index to all employees irrespective of the wage level. The per point payment is usually so fixed as to provide full compensation at the lowest level of wage. And payment of fixed amount to all employees means that the extent of compensation declines as wage rises.

The second system is the so-called Central Government Dearness Allowance system. This stipulates declining percentage neutralization of the rise in the Consumer Price Index.

In the case of minimum wages for unorganized sectors, there is no systematic indexation. In one state (West Bengal), a system of full indexation of minimum wages for certain industrial occupations was introduced in 1965. In other states, efforts have been made to introduce indexation only since the mid-seventies.

Setting Bonus Payment Rules

There are provisions for statutory bonus payment governed by the Payment of Bonus Act, 1965 which applies to all establishments employing 20 or more persons. Under the Act, the minimum bonus is 8.33 per cent of the annual wage (i.e., the basic wage plus dearness allowance) of an employee subject to an assured minimum amount. This rule applies to workers whose monthly wage does not exceed a certain amount (currently Rs.3,500); if wages exceed this figure, the bonus is calculated as though wages were this amount. This minimum bonus is payable irrespective of whether or not an establishment earns a profit; it is payable even if an establishment suffers losses. Effectively, therefore, the minimum bonus is equivalent to wages for a thirteenth month. The maximum bonus that employees can claim under the Act is 20 per cent of annual wages. In addition to the statutory bonus, an employer can freely choose to pay incentive bonuses.

Other Related Aspects

Three other features of the wage determination process are worth noting. First, it is common for both manual workers and white-collar employees to have pay scales so that a series of annual increments to basic wages are automatically granted. An element of money wage growth, therefore, is purely a function of time. Second, workers in some establishments receive welfare benefits and other allowances. The welfare benefits are usually regulated through schemes launched by the central government in consultation with the National Social Security Authority (state governments also have supplementary schemes) while the allowances are either voluntarily made available by establishments or determined through collective bargaining. Third, the government has been keen to limit wage inequality in industry. It directly determines top salaries in public enterprises and has sought to control top salaries in private enterprises through appropriate provisions in the Indian Companies Act, 1956.

2.3 Wage Policy and the Growth of Wages in Unorganized Sectors

It is clear by now that the only influence that the government seeks to exercise on wages in unorganized sectors is by setting a floor in the form of minimum wages. Subject to these minima, wage determination is in principle left to market forces. Since casual employment predominates in unorganized sectors, the minimum wages are fixed for a standard (8-hour) day of employment. In what follows, an attempt is made to evaluate the principles underlying the fixation of minimum wages and the effectiveness of statutory minimum wages in influencing actual wage movements.

The principal sources of data available for analysis of minimum wages are the Reports (typically annual but sometimes biannual) on the working of the Minimum Wages Act, 1948. However, this series of Reports stops at 1984. Two other relevant publications are the Indian Labour Statistics, 1982-86 and the 1992 review of Wage Fixation in Industry and Agriculture in India. These two Labour Bureau publications, unfortunately, provide only bare statements of minimum wages. In particular, they carry no details in respect of cost of living allowances. For these reasons, the analysis here is confined to the period till 1984. While this makes the analysis incomplete, there are no reasons to suppose that the conclusions drawn here do not apply to the post-1984 period.

Table 2.1 brings together the available data on statutory minimum wages for selected industries in seven major states. The numbers reported here relate to daily wage rates for adult men for the lowest category of unskilled labour, and, where differentiated by `zones' within a state, the lowest of the rates specified. Also, they represent all-inclusive wage rates, i.e., the basic wage plus a cost of living allowance, wherever relevant.

A number of conclusions follow from these data. First, a most obvious fact is that the statutory minimum real wage rate for the same occupation varied across states, though not always very significantly. Such variations may be thought of as reflecting the existing variations in cost of living as well as in general labour market conditions. But if this is the case, it is legitimate to wonder about the presumed function of statutory minimum wages.

Second, within each state, there were significant variations in the minimum real wage rate across industries or occupations in the same period. It is perhaps possible to justify this in so far as job characteristics or locations (rural/urban) vary across industries. But, once again, the specific role of statutory minimum wages seems unclear unless it is assumed that market forces are blind to varying characteristics of industries -- a rather implausible assumption.

Third, minimum wage rates were revised very infrequently. And since most often they were not indexed, their values must have eroded in the years in between two revision-years to such an extent as to render them irrelevant. In the sample of states considered here, West Bengal was the earliest to specify, in 1965, an indexation formula, linked to the Consumer Price Index for Industrial Workers, which implied 100 per cent neutralization for price increase at a wage rate of Rs. 2.31 per day in 1960 prices. This formula was subsequently modified in 1979. The modified formula implied 100 per cent neutralization at wage rate of Rs.3.85 per day in 1960 prices. Very similar indexation rules were introduced in Gujarat in 1979, in Uttar Pradesh in 1981, in Madhya Pradesh in 1982 and in Maharashtra and Tamil Nadu in 1983.

Fourth, rather surprisingly, even when minimum wage rates were revised, the revised rates were often lower in real terms than the pre-revision rates. Thus in some industries, the statutory minimum wage (in real terms) declined in Uttar Pradesh between 1981 and 1984 and in West Bengal between 1970 and 1980.

Finally, in most cases, statutory minimum wages were below what might be called poverty-line wages, as the data in Table 2.2 show. This means that even if a worker had full-time employment (26 days in a month) and received the statutory minimum wage, the worker-family would still be living below the official poverty-line.

Source: Poverty-line wages are estimated by the author. Statutory minimum wages are from Table 2.1.

The story is only slightly different in the case of agriculture (see Table 2.3). As in the case of non-agricultural occupations, revisions were irregular, there was no proper indexation and revised statutory minimum wages were often lower than pre-revision wages in real terms. But quite often, statutory minimum wages were higher than the poverty-line wage (Rs. 2.8 per day in 1960 all-India prices).^{(Endnote 11)} Furthermore, they were often higher than those for non-agricultural occupations.

Statutory Minimum Wages and Actual Wages

The periodic Occupational Wage Surveys (OWS), carried out by the Labour Bureau, provide a useful data base for examining the link between statutory minimum wages and actual wages received by workers in different non-agricultural occupations. In respect of the industries or occupations taken up for survey, OWS provides estimates of minimum and maximum wage rates for each occupational category and separately for men, women, children and apprentices. In respect of occupations with a clearly specified pay scale, the minimum and maximum represent, respectively, the sum of basic wage plus dearness allowance at the minimum and maximum of the scale. In other cases they represent actual minimum and maximum earnings per day.

Table 2.4 brings together, for roughly comparable periods, statutory minimum wages for selected occupations and actual minimum daily wage rates for adult men. Clearly, these data are too sketchy to allow any firm conclusions. Nevertheless, they indicate rather strongly that statutory minimum wages were not effectively enforced. If they were strictly enforced, we should expect actual wages to be either equal to or greater than corresponding statutory minimum wages. But even the sketchy evidence shows that, more often than not, actual wages were below statutory wages.

As noted earlier, statutory minimum wages for agriculture tended to be higher than the poverty-line wage and were higher than statutory minimum wages for non-agricultural occupations. As such, it could not be said that statutory minimum wages, in the case of agriculture, were following rather than leading market-determined wages. But the data in Table 2.5 quite clearly suggest that enforcement was very weak in the case of agriculture.

The overall picture which emerges is rather disconcerting. Given the record of infrequent and sometimes downward revisions, lack of proper indexation and weak enforcement, statutory minimum wages had little impact on the movement of actual wages. Much more importantly, a clear conceptual basis for fixing statutory minimum wages has been missing. Objectives such as preventing exploitation are too vague to be useful for operational purposes. The idea of setting different minimum wages for different states (and for different occupations within a state) can be justified by asserting that local conditions have to be taken into account in setting minimum wages. But in the absence of other clearly defined criteria, this amounts to accepting the market-determined wage in a region as the basis for fixing the minimum wage. It is difficult to see the usefulness of such an exercise.

Note: The actual wages reported here are the occupation-specific minimum wages inclusive of dearness allowance.

Source: Anant and Sundaram, op. cit.

Source: Labour Bureau (Government of India), Rural Labour Enquiry, Reports for 1983 and 1987/88.

Movement of Wages in Unorganized Sectors

While government interventions were ineffective, market forces seem to have driven wages up in unorganized sectors. Figure 1 shows the movement of the real wage rate for male agricultural labourers during the period 1973/74 - 1990/91. The source of the raw data used to construct this series is the Ministry of Agriculture (Government of India). Wage data are collected from more than 1000 centres spread across 320 districts in the country. The methodology used to arrive at the estimates of real wage rate at the all India level is as follows. The simple average of the wage data from different centres in a district was taken as the wage rate for the district; the weighted average of the wage data for different districts in a state was taken as the wage rate for the state and the weighted average of the wage data from the major states was taken to be the wage rate at the all-India level. The weights were the shares of districts in the total number of agricultural labourers in the country (average of the figures available from 1971 and 1981 population census were used). Finally, the Consumer Price Index for Agricultural Labourers (CPIAL) was used to deflate the money wage rates.

The figure shows a clear upward trend in the real wage rate in agriculture throughout the period; the trend rate of growth was 3.3 per cent per annum. However, in the absence of indexation, there naturally were fluctuations which are explained by fluctuations in the relative price of food.^{(Endnote 12)} The real wage rate declines whenever food-grains prices rise faster than other prices but rises whenever

Figure 1

other prices rise faster than foodgrains prices. This indicates that there are informal mechanisms whereby money wages are adjusted with reference to general price rises. But foodgrains account for a very large proportion of expenditure of agricultural labourers and, as such, the real wage rate tends to be sensitive to relative prices of foodgrains.^{(Endnote 13)}

A caveat needs to be added, however. Though the wage series used here is the only one available, it is generally believed to be somewhat unreliable. The data in Table 2.6 offer some possibilities of verifying if such a belief is unfounded since the wage data, collected through periodically conducted large sample surveys, are generally thought to be reliable. A comparison of wage data from the two sources reveal the following. First, the wage rates reported by the Ministry of Agriculture are generally higher than those reported in the Rural Labour Enquiry Reports. Second, the rankings of states according to both sources are broadly similar. Third, the long-term (1977/78 - 1987/88) growth rates suggested by both sets of data are roughly similar.

Thus the trend growth rate of the real wage rate suggested by the data from the Ministry of Agriculture can be regarded as reliable. It should be noted in this context that the real agricultural wage rate grew rapidly in all the major states (see Table 2.7). The trend growth rate observed at the all-India level, therefore, is fairly representative and does not embody serious distortions.

RLE: Rural Labour Enquiry (Labour Bureau, Ministry of Labour).

What explains the rapid growth of the real wage rate in agriculture? It has not been possible to carry out detailed investigations needed to answer this question. The results of past research, however, indicate an answer.^{(Endnote 14)} These show that while the real wage rate can be sensitive to forces of demand and supply in the short run, growth of labour productivity exerts a positive influence on the real wage rate in the long run independently of forces of demand and supply. Since the available evidence does not indicate either a steady improvement or a steady deterioration in the conditions of employment, the growth of real wages must be attributed in large measure to the growth of labour productivity in agriculture which was substantial. The growth of rural workers' organizations may have been an additional factor.

A rising trend in the real wage rate in agriculture would also normally indicate a rising trend in real wage rates in unorganized industries. Unfortunately, time-series data on real wage rates in unorganized industries are not available. Sketchy data indicate that these indeed increased, but less rapidly than in agriculture. For example, data available from the National Sample Surveys show that between 1977-78 and 1987-88, while the real wage rate in agriculture increased at an annual rate of 3.9 per cent, the real wage rate of casual labourers in urban areas increased at an annual rate of only 1.8 per cent. This difference is hard to explain in view of the fact that too little is known about the factors influencing the real wage rate in the urban informal sector.

In spite of the growth, however, the real wage rate in agriculture remains substantially below the poverty-line wage. This is evident from Figure 2. Two qualifying remarks should be made here. First, as noted earlier, wage rates reported by the Ministry of Agriculture tend to be generally higher than those reported by the Labour Bureau (in the Rural Labour Enquiry Reports). It can be seen from Table 2.6 that, at the all-India level, the average wage reported by the Labour Bureau was about 66 per cent of that reported by the Ministry of Agriculture. Thus, use of the wage data reported by the Ministry of Agriculture seriously understates the gap between the poverty-line wage and the actual wage. Second, there is considerable variation in the real wage rate across states. In a few states (Punjab, Haryana and Kerala), the real wage rate is already above the poverty-line wage while in several other states, the real wage rate is far below the poverty-line wage.

Figure 2

It is very probable that real wage rates remain below the poverty-line wage in most unorganized industries too, though no direct evidence on this can be cited here. As observed earlier, statutory minimum wages in these industries have usually been below the poverty-line wage. And actual wages have often been below statutory wages. Besides, the growth of actual real wages in unorganized industries appears to have been much slower than that of real wages in agriculture. Given these facts and given that the agricultural wage still remains below the poverty-line wage, there can be little doubt that wages in other unorganized sectors have also remained below the poverty-line wage.

2.4 Wages in Organized Industries

As already mentioned, government interventions played a key role in determining wages in organized industries. The interventions included setting of minimum wage norms, direct determination of wages in public enterprises, indirect influence on wage determination in private enterprises through the establishment of Wage Boards, setting of norms for wage differentials and establishment of rules of indexation and bonus payment. Norms were also set for social security benefits to employees though these were generally less binding.

The main focus in this section is on the consequences of the government's wage policy for wage movements in organized industries. But to begin with, it is worth pointing out that government policy created or at any rate institutionalized labour market dualism. While statutory minimum wages, meant for the workers in unorganized sectors, were fixed without reference to levels of living of the workers concerned, in setting norms for minimum wages in organized industries, the government was much concerned with the notion of a "living wage" and dietary norms. The result was a fairly wide gap between statutory minimum wages for unorganized sectors and minimum wages for organized industries. A comparison of the data in Table 2.8 with those in Table 2.1 shows that statutory minimum wages were often less than 50 per cent of the minimum wages recommended by the Wage Boards. Besides, minimum wages in the organized sector were fully indexed while, as noted earlier, there was no proper system of indexation for those in unorganized sectors until very recently. The close link which exists between wages recommended by the Pay Commissions and those recommended by the Wage Boards are also quite clear.

Source: Anant and Sundaram, op. cit.

In order to study the effects of wage policy on wage movements, it is useful to establish the interrelationships existing among different components of the wage packet. Essentially, the wage packet consists of the basic wage, compensation for cost of living increases in the form of dearness allowance and bonus. Of these, as we shall see, bonus can be treated as a fixed percentage of the wage defined as the sum of the basic wage and dearness allowance and hence can be left out of account at this stage of analysis. As already explained, the degree of indexation is 100 per cent only for the lowest basic wage and declines as the basic wage increases. This means that for the average basic wage, the system of dearness allowance ensures that money wage growth is less than that of cost of living increase. The only other aspect to be considered is that it is common for employees to have pay scales so that they receive a series of annual increments to basic wages automatically.

This formulation is helpful in identifying, on an a priori basis, some basic characteristics of the wage system in organized industries. First, it is obvious that the money wage does not really have downward flexibility. Even when the rate of inflation is negative -- itself a rare occurrence -- the money wage does not necessarily fall since the annual increment can conceivably neutralize the effect of a fall in the dearness allowance.

Second, in the absence of revision of the basic money wage, downward flexibility of the real wage depends on the values of two parameters -- the average rate of annual increment and the average degree of indexation; the higher the values of these parameters, the lower is the downward flexibility of the real wage for a given rate of inflation. In organized industries in India, the average rate of annual increment is unlikely to exceed 2 per cent and the average degree of indexation is probably around 65 per cent. Under these conditions, it can be verified from expression (2) that the average real wage would fall whenever the annual rate of inflation is 6 per cent or more. Moreover, since the indexation is lagged by one period, accelerating inflation would tend to generate downward pressure on the real wage (just as decelerating inflation would always tend to increase the real wages).

What all this shows is that in Indian conditions, where the rate of inflation is rarely below 6 per cent, the average real wage tends to fall in the absence of wage revision and this is associated with declining wage differentials. Periodic wage revisions, which may or may not restore the differentials, are needed even to hold the real wage constant. This, it may be presumed, generates pressures for shortening the period of wage contract and for generous settlements during wage revisions.

Third, it needs to be noted that the relation between the real wage and the product-wage is governed by the agriculture-industry terms of trade; this is because agricultural products figure prominently in industrial wage-earners' consumption basket.^{(Endnote 15)} When the terms of trade are moving in favour of agriculture, the product-wage may be rising even if the real wage is constant or falling. On the other hand, when the terms of trade are shifting in favour of industry, the product-wage may be constant or falling even while the real wage is rising. Thus the product-wage tends to be flexible even when the real wage is rigid. However, the terms of trade tend to shift in favour of agriculture in times of high and accelerating inflation and in favour of industry in times of low and decelerating inflation. Hence low inflation and stable terms of trade are in the interest of both employees and employers.

It is to be noted that wage settlements occur with reference to the real wage and not the product-wage. Hence labour costs, from the employers' point of view, depend not only on wage settlements but also on macroeconomic policies which determine the rate of inflation and inter-sectoral terms of trade.

Some of these points are illustrated by the data on wages in industrial establishments employing 100 or more workers (Table 2.9). They confirm that both the real wage and the product-wage have downward flexibility though the money wage always moves upwards. They also suggest that wage growth occurs in spurts, i.e., through revisions.

A final point to be noted is that bonus and other non-wage payments (retirement benefits and welfare benefits) to labour tend to be fixed percentages of the wage (i.e., the basic wage plus the dearness allowance) as can be seen from Table 2.10. This means that the total cost of labour moves in almost exactly the same manner as the wage. As such, the observations made about the characteristics of wage movements are also applicable to movements in the total cost of labour.

An important fact highlighted by the data in Table 2.10 is that the bonus payment as a percentage of the wage was generally close to the statutory minimum (8.33 per cent). This shows that bonus is widely viewed as the wage for a thirteenth month by both employees and employers. It also suggests that incentive bonuses have not been common in Indian industry.

Source: Labour Bureau.

Wage Growth in Organized Industries

The linkage between the features of the wage system and the pattern of wage growth in organized industries can now be studied. Industries in this context refer to manufacturing industries as well as industries supplying electricity, gas and water; organized industries refer to industrial establishments employing 10 or more workers. The data presented in Tables 2.11 and 2.12 relate to employees and emoluments rather than workers and wages. But it should be clear from the preceding discussions that the behaviour of total payment to labour is expected to be exactly the same as that of the wage rate. These data suggest several conclusions.

First, the growth of real wages was significantly faster in the 1980s than in the 1970s even though the cost of living increased at a significantly faster rate in the later period. There also were fewer instances of wage decline in the 1980s. These conclusions hold irrespective of whether the real wage per work day or that per worker is considered.

Source: Labour Bureau.

It is worth noting that the discrepancies between the two estimates of real wage arose from the fact that the number of work days per employee varied from year to year. Figure 3 indicates that there was substantial under-utilization of labour in the 1970s; from 1984/85 onwards, however, labour utilization was at an optimum level (each employee worked 26 days in a month). Thus there was sustained improvement in labour utilization (perhaps indicating improvements in industrial relations and management practices) till the mid-1980s. It is well known that the growth in the number of employees decelerated quite sharply in the 1980s. One reason for this, it seems, was the improvement in labour utilization.

Second, the acceleration in the growth of real wages occurred in both sectors -- public and private. This should not come as a surprise. The characteristics of the wage system, analyzed earlier, suggests a linkage between wages in the public sector and those in the private sector. The wage-gap between the sectors of course widened throughout the period (Figure 4) as public sector wages always grew faster than private sector wages.^{(Endnote 16)} Unfortunately, we do not have data on workdays per employee separately for public and private sectors. But a perusal of the data in Table 2.11 tells us that the conclusions drawn above are most unlikely to be wrong.

Third, while the real wage growth was much faster than the product-wage growth in the 1970s, the reverse was the case in the 1980s. Obviously the reason is that the terms of trade were shifting in favour of industry in the 1970s and in favour of agriculture in the 1980s. The important thing to note is the critical role that inter-sectoral terms of trade and hence macroeconomic policies play in determining labour costs in Indian industry. Real wage growth is usually a poor indicator of changes in labour costs.

Figure 3

Table 2.12
Rates of Growth (Over the Previous Year) of Real Wage,
Product-Wage and Labour Productivity in
Public and Private Sectors of Organized Industry

Figure 4

These findings are neither surprising nor new. But, when viewed together with the features of the wage system discussed earlier, they highlight a major weakness of the wage system as it operates in India. The essential problem can be stated as follows: real wages tend to decline in the absence of wage revisions; wage revision process gets focussed on increasing real wages; wage revisions in public sector industries lead the way; and wage revisions in public sector industries are closely linked to pay revisions for government employees. The result is that the entire wage revision process is guided by political, fiscal and welfare considerations.

It should be clear that the wage system is such as to make wage growth exogenous, i.e., independent of productivity growth. This does not mean that there is no wage-productivity linkage in India's organized industries. For it is always possible to adjust productivity to wages by altering capital intensity and/or technology. But this method of adjustment makes wage-productivity relationship unstable and has undesirable consequences for employment growth. Since exogenous wage growth has to be countered by increasing capital intensity or by introducing new technologies, employment growth inevitably decelerates. The experience of the 1980s amply illustrates the point.

2.5 Concluding Observations

The basic framework of wage policy, developed in India, can be described as follows. The government sought to set a floor to the price of labour in the entire economy by statutorily fixing minimum wages for workers in the unorganized sectors. Subject to this floor, wage determination in unorganized sectors was left to market forces. On the other hand, in the organized sector, the government sought to exercise much greater control over wages. It not only set certain basic rules thus defining a wage system for the sector as a whole but also directly controlled wages in large parts of the sector. In industries where the public sector dominated, the government naturally played a central role in determining wages. In some other industries dominated by the private sector, it chose to play a major role by establishing Wage Boards. In all these industries, there was little space left for collective bargaining. Outside these industries, collective bargaining probably had a role to play, but this is unlikely to have been important for two reasons. First, enterprises in these industries tended to be relatively small and the degree of unionization of the workers in them was low. Second, even where collective bargaining was important, government-controlled wages are likely to have served as reference points for wage bargaining.

In seeking to control wages in the organized sector, the government clearly had several objectives in view: to ensure a decent minimum level of living for organized sector employees, to create conditions for systematic growth of wages, to sustain appropriate wage-productivity linkages, to incorporate appropriate incentive structures into the wage system and to limit income inequalities. Unfortunately, some of the objectives were mutually inconsistent even in theory, the instruments of interventions were not always appropriate and practical difficulties of implementation were many.

Statutory minimum wages have been largely ineffective in influencing wages in unorganized sectors. Irregular revisions, lack of proper indexation and weak enforcement rendered statutory minimum wages virtually irrelevant in practice. Besides, there was no well-defined basis for fixing minimum wages. Objectives such as prevention of exploitation are much too vague to be useful. It is not surprising that minimum wages varied widely across occupations in a state and across states for a given occupation. When norms vary so much, they cease to be norms.

A different set of minimum wages was fixed for the organized sector. Here there was a clearer basis, namely, a notion of minimum acceptable level of living which had no role in determining minimum wages for unorganized sectors. Thus there were double standards which legitimized and institutionalized labour market dualism.

The instruments of intervention, used to influence wages in the organized sector, gave the wage system some undesirable characteristics. The indexation rule, apart from being rather complicated, made average real wages vulnerable to inflation. Periodic wage revisions, therefore, emerged as the basic mechanism for protecting and increasing real wages. This made protection or increase of real wages virtually the sole concern of wage revision and created pressures for increasing its frequency. Indexation rule also has the effect of narrowing down real wage differentials. This either harms incentives for skill upgradation or creates pressures for compensation through the introduction of a variety of allowances or through the mechanism of wage revision. The overall consequence is an upward pressure on average real wages.

All this left very little scope for considering wage-productivity relationship in determining wages. The scope was further reduced by two other factors. First, there was little recognition of the fact that real wage and product-wage need not move together. And it is the product-wage which is relevant in considering wage-productivity relationship. Second, wage revisions in industries tend to be linked to pay revisions for government employees. Thus wage norms in sectors where productivity is a concern are influenced by pay norms in a sector where productivity cannot even be measured. In these circumstances, wage-productivity relationship can only be sustained through adjustments in productivity in response to exogenous changes in wages. Such adjustments involve changes in capital intensity and technology which harm prospects of employment growth.

These observations lead us to the issue of wage flexibility. The indexation rule certainly ensures downward flexibility of real wages, but this is linked entirely to the inflationary process and not to productivity and profitability of industries and enterprises. As such, it is of little help in facilitating industrial restructuring associated with structural adjustment. A further complication arises from the fact that there is no straightforward relationship between real wage flexibility and product-wage flexibility. This relationship is mediated by agriculture-industry terms of trade and it would be desirable to set stability of terms of trade as an objective of macroeconomic policy.

Endnote 9:
This chapter uses materials from Anant, T.C.A., and Sundaram, K., "Wage Policy in India: A Review", Background Paper, mimeo.

Endnote 10:
cf. ILO-ARTEP, op. cit., Ch. 2.

Endnote 11:
The poverty-line wage in agriculture is estimated on the basis of the following information and assumptions: poverty-line expenditure: Rs. 49.09 per capita per month in 1973-74 prices (Planning Commission); number of non-earning dependents per worker: 2; number of days of employment per worker per month: 20 (Labour Bureau).

Endnote 12:
This is shown by following regression equation:
ln (RWA) = 0.90 - 0.63 ln (RPF) + 0.034 t

        (27.79)  (03.15)   (11.92)

where RWA - real wage rate in agriculture, RPF - relative price of food defined as the ratio of the wholesale price index for food-grains to the wholesale price index for all commodities, t - time and figures in parentheses are T-ratios.

Endnote 13:
For a somewhat elaborate discussion of the point, see Ghose, A.K., "Rural Poverty and Relative Prices in India", Cambridge Journal of Economics, Vol. 13, 1989, pp. 307-331.

Endnote 14:
See, for instance, ILO-ARTEP, op. cit., Ch. 2.

Endnote 15:
For elaboration and evidence, see ILO-ARTEP, op. cit., Ch. 2.

Endnote 16:
For the period 1973/74 - 1990/91, the elasticity of private sector wages with respect to public sector wages is found to be 0.58. 

For further information, please contact South Asia Multidisplinary Advisory Team (SAAT) at Tel: 91.11.460.210, Fax: 91.11.460.2111 or by E-mail: delhi@ilo.org