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

A+| A| A-

Labour Market Flexibility: An Empirical Inquiry into Neoliberal Propositions

There have been proposals to make the Indian labour market more flexible by amending the Industrial Disputes Act, 1947 and Contract Labour Act, 1970. But the Indian labour market has already achieved a substantial degree of flexibility by the contractualisation of factory workers. This paper critically investigates the claims made in favour of introducing greater flexibility in the labour market. The analysis is done through an empirical inquiry into the proposition that casualisation of labour leads to higher output and employment growth. We find that employment and output growth do not have a statistically significant dependence on labour market flexibility in Indian organised manufacturing.

SPECIAL ARTICLE

Labour Market Flexibility: An Empirical Inquiry into Neoliberal Propositions

Atulan Guha

There have been proposals to make the Indian labour market more flexible by amending the Industrial Disputes Act, 1947 and Contract Labour Act, 1970. But the Indian labour market has already achieved a substantial degree of flexibility by the contractualisation of factory workers. This paper critically investigates the claims made in favour of introducing greater flexibility in the labour market. The analysis is done through an empirical inquiry into the proposition that casualisation of labour leads to higher output and employment growth. We find that employment and output growth do not have a statistically significant dependence on labour market flexibility in Indian organised manufacturing.

This paper is a revised version of that written by Atulan Guha and Prasenjit Bose, with the same title, presented at the International Conference on Development in Open Economies: Industry and Labour organised by the Academy of Third World Studies, Jamia Milia Islamia in Delhi in 2006. I would like to thank an unknown referee for c omments. Also, I would like to thank Satyki Roy for his comments on the revised draft.

Atulan Guha (atulanguha@yahoo.com) is with the Institute for Studies in Industrial Development, Delhi.

S
everal Indian economists and policymakers have taken the view that distortions in the labour market on account of the extant labour laws in India are obstacles to higher e conomic growth and expansion of employment. The argument put forward by them posit “rigidities” in the labour market as disincentives to private investment as well as the adoption of labourintensive technologies. Following the same line of argument the Second National Commission on Labour had called for greater labour market “flexibility”. In 2005 the prime minister’s office (PMO), prepared a note (“PMO Note on Labour Market”) where the proposals for introducing greater labour market flexibility were clearly stated: Amendments to Chapter VB of the Industrial Disputes Act (IDA), 1947 and the Contract Labour Act, 1970. The attempt is to introduce two kinds of changes in the laws governing the labour market. First, to give greater freedom to the e mployers to retrench or lay-off permanent workers by diluting the provisions under Chapter VB of the IDA (i e, to introduce hire and fire) and second, expanding the scope of employing contractual or casual labour in greater number of jobs as well as industries. These measures, it is argued, would impart the required “flexibility” to the Indian labour market, which would enable higher growth and employment generation, especially in the manufacturing sector.

Although, none of the above mentioned acts have been amended to achieve labour market flexibility, the enforcement of both the laws, especially the contract labour laws, have been poor. As a result, there is an increasing presence of temporary workers in regular work. This is actually providing a means through which the factory owners have reduced the scope of permanent employment. This tendency has not only liquidated the rigidities in the labour market due to the Contract Labour Act, but has also liquidated partially the rigidities in the IDA. So without changing these two laws, a substantial degree of labour m arket flexibility has already been achieved in the Indian labour market. To capture the extent of these labour market flexibilities we have indexed them as the ratio between workers employed through contractors and the total number of factory workers. The rising trend of contracting out the production process to the informal sector by the formal sector is another way of liquidating the rigidities in the labour market, which arise out of IDA, 1947. Though our labour market flexibility index does not capture this, the index reasonably indicates the current d egree of labour m arket flexibility.

The purpose of this paper is to critically investigate the claims made in favour of introducing greater flexibility in the labour

Economic & Political Weekly

EPW
May 9, 2009 vol xliv no 19

market. This is done through an empirical inquiry of the proposition, popular in neoliberal policy circles, that casualisation of l abour leads to higher output and employment growth. The focus of the study is on the organised manufacturing sector in India. In Section 1 we discuss the debate on labour market flexibility in the international context. A brief overview of the Indian setting is undertaken in Section 2. Section 3 proceeds with our empirical analysis of the impact of greater labour market flexibility, measured in terms of greater casualisation of the workforce, on output and employment in Indian manufacturing. The findings are s ummarised in the concluding section.

1

The policy debate on labour market flexibility, although quite an old one, was triggered in the recent times by the Jobs Study (OECD 1994), which revived the orthodox classical argument while explaining the high unemployment rates witnessed in Europe since the 1980s. Holding “rigid” labour markets of Europe responsible for the incidence of high unemployment, the OECD study recommended greater labour market flexibility in terms of (a) increasing the flexibility of working time, (b) making wage and labour costs more flexible by removing minimum wage regulations,

(c) reforming employment security provisions, and (d) reforming unemployment and related benefit systems. The IMF World Economic Outlook (2003) said,

A wide range of analysts and international organisations – including the European Commission, the Organisation for Economic Cooperation and Development, and the International Monetary Fund – have argued that the cause of high unemployment can be found in labour market institutions. Accordingly, countries with high unemployment have been repeatedly urged to undertake comprehensive structural reforms to reduce “labour market rigidities” such as generous unemployment insurance schemes, high employment protection, such as high firing costs, high minimum wages, non-competitive wage-setting mechanisms and severe tax distortions.

It was also claimed in the same document that “…well-designed labour reform could produce output gains of about 5% and a fall in the unemployment rate of about 3 percentage points”.

The World Bank’s view on this has evolved over time. The World Development Report (1990) was of the view that labour market policies – minimum wages, job security regulations and social security – are usually intended to raise welfare or reduce exploitation. But they actually work to raise the cost of labor in the formal sector and reduce labor demand…and thus (depress) labor incomes where most of the poor found.

But a World Bank publication titled Unions and Collective B argaining: Economic Effects in a Global Environment (2003) notes that

workers who belong to trade unions earn higher wages, work fewer hours, receive more training, and have longer job tenure on average, than their non-unionised counterparts… On the other hand, temporary lay-offs can be more frequent in unionised firms. At the macroeconomic level, high unionisation rates lead to lower inequality of earnings and can improve economic performances (in the form of lower unemployment and inflation, higher productivity and speedier adjustment to shocks).

The International Labour Organisation (ILO) disagrees with the view that labour market rigidity has been the major cause of

46 unemployment and greater labour market flexibility is the solution. ILO’s Jobs Report (1996-97) says, “jobless rates appear to have risen independently of levels of labour market regulations… trade union power was reduced in many countries, together with unemployment benefits and in some cases minimum wages, producing little if any positive employment effect”. Baker et al (2002, 2004) has noted that the findings of the time series models in the OECD job study, that show labour market institutions adversely affecting aggregate outcome of the economy, are not robust. They have shown that some modest changes in the measure of institutions, countries covered or time period of analysis changes the finding of negative relationship between labour market rigidities and economic outcomes.

2

The Constitution of India has listed labour as a subject in the Concurrent List whereby both the central and state governments have the right to enact legislation, subject to certain matters being in the exclusive domain of the central government. There is a plethora of labour laws in India both at the central as well as the state levels. An estimate puts the total number of acts and rules concerning labour in the range of 25,000 to 30,000 (Debroy 2005). Despite the plethora of laws, their efficacy remains questionable. The Second National Commission on Labour stated,

It can be said that our labour laws have not flowed from any vision of a harmonious and just social order that takes into a ccount the needs of an efficient and non-exploitative society, or a vision of the rights, duties and responsibilities of the different social partners to themselves, to each other, and to the totality of the community. They have been criticised as being ad hoc, complicated, mutually inconsistent, if not contradictory, lacking in uniformity of definitions and riddled with clauses that have become outdated and anachronistic, in view of the changes that have taken place after they were introduced many years ago (Chapter 1, p 12).

The note prepared by the PMO in 2005, proposed certain amendments in two important central acts in order to make the labour market more flexible. It proposed that the present filter number of 100 under Chapter VB of the IDA, 1947 be raised to

300. Currently Chapter VB of the IDA says that industrial establishment, viz, factories, plantations and mines employing not less than 100 workers have to seek prior permission from the appropriate government department to effect lay-off, retrenchment or closure. In their proposal to amend the Contract Labour Act, 1970 the PMO note proposed that certain activities may be kept outside the purview of Section 10 of Chapter 3 of the Contract Labour (R&A) Act, 1970 which prohibits the use of contractual labour in industrial activities. The proposed activities where contract labour was sought to be allowed are:

  • (1) sweeping, cleaning, dusting and gardening, (2) collection and disposal of garbage and waste, (3) security, water and ward, (4) maintenance and repair of plant, machinery and equipments, (5) housekeeping, laundry, canteen and courier,
  • (6) loading and unloading, (7) information technology, (8) support services in respect of an establishment relating to ports/ dockyards, airports, railway stations, interstate bus terminals, hospitals, educational and training institutions, guesthouse,
  • may 9, 2009 vol xliv no 19

    SPECIAL ARTICLE

    club and transport, (9) export-oriented special economic zones and units exporting more than 75% or more of their production, and (10) construction and maintenance of buildings, roads and bridges.

    It was further proposed that this list can be amended from time to time. In addition, Section 31 of Chapter 3, which provides discretionary power to the government to allow contract labour in case of emergency, is proposed to be amended to allow the government to use this discretion invoking public interest even without an emergency.

    These proposals to amend the labour laws have been inspired by the works of economists who have argued strongly in favour of introducing greater labour market flexibility in India since long. Fallon and Lucas (1991), for instance, argued that employment growth in the organised segment of the manufacturing sector would have been higher by 17.5% in the absence of rigid provisions on job security. Dutta Roy (1998), however, contradicted the finding and argued in a study based on Annual

    Table 1: Correlation Coefficient of Labour Market Flexibility with Employment Growth and Output Growth

    Employment Growth Rate Output Growth Rate
    Growth in labour market flexibility -0.0738 0.1087
    (0.1671) (0.0415)

    Values in parenthesis give the level of significance.

    S urvey of I ndustries (ASI) data for the period 1960-61 to 1993-94 that job security regulations (both the 1976 and 1982 amendments) have not been responsible for the slowdown in employment growth. Three primary survey based studies on labour flexibility in India by Sudha Deshpande et al (1998), Sharma and Sasikumar (1996) and Lalit Deshpande et al (2004) also have not been able to find any evidence that support the hypothesis that firms that have more than a 100 employees face greater hurdles in laying off workers when compared to firms that employ less than 100 workers. An interesting and common finding of these studies is that firms, irrespective of size, were found to increase employment mainly by increasing the share of non-permanent workers. This shows that despite the existence of labour laws like the Contract Labour Act, contractualisation or casualisation of labour has already occurred significantly, as far as the Indian labour market is concerned. In other words, manufacturing industries have already achieved

    Table 2: Correlation Coefficient of Explanatory Variables in Regression Equation 1

    substantial labour market flexibility through increase in the share of non-permanent workers in total employment.

    3

    In this section we discuss the impact of labour market flexibility, caused by increasing contractualisation of labour force, on output and employment for the organised manufacturing sector in India in the 1990s. Sunanda Sen et al (2006) have found using ASI data that the ratio of non-permanent workers to total number of workers for most of the manufacturing sectors at the threedigit level have gone up during the 1990s. Their data show that the proportion of contractual workers to total number of workers in all the sectors taken together (in other words, for organised manufacturing as a whole) has gone up from 9.89% in 1992-93 to 23% in 2000-01. In our analysis, we try to study the impact of this increase in labour market flexibility that has already taken place on output and employment.

    Our analysis has used the data at three-digit level of NIC 1998 sectoral level for the period of 1994-95 to 2003-04. The total number of sectors is 44. The total number of sectors at the three-digit level of NIC 1998 is slightly more. But to get a balanced panel we are limited to 44 sectors. The variables required for the analysis are output, net fixed capital stock, workers’ wage bill, total input cost, number of workers, number of workers employed through contractor and wholesale price index (WPI) numbers.

    For these variables, excluding net fixed capital stock and number of workers employed through a contractor, we have used the ASI data provided by Economic and Political Weekly Research Foundation (EPWRF). This database has provided time series data of more than 25 years at the three-digit sectoral level of NIC 1998 classification. They have prepared these time series data, excluding WPI, by using ASI data of the Central Statistical Organisation (CSO) and the concordance table between NIC 1998, NIC 1987 and NIC 1970. The WPI series is from the CSO.

    The variables, output wage bill of workers, total input cost, number of workers and number of workers employed through contractor, are same as that of value of output, wages to workers, total inputs, number of workers and number of workers employed through contractor respectively in ASI. To get the values at c onstant prices of 1993-94, we have used corresponding sectors’

    Output_ Growth_ NFCS_ Growth_ NFCS_ Growth_ Technological Wage_Cost_per_unit_ Total_Input_Cost_per_ unit_ Labour_Market_
    Rate_Lag1 Rate_Lag1 Rate_Lag2 Change Output_ Growth_Rate Output_ Growth_Rate Flexibility_Growth_Rate
    Output_ Growth_Rate_Lag1 1.0000
    NFCS_ Growth_Rate_Lag1 0.2058 1.0000
    (0.0001)
    NFCS_ Growth_Rate_Lag2 0.0004 0.0509 1.0000
    (0.9944) (0.3414)
    Technological change 0.0077 0.0595 0.0305 1.0000
    (0.8849 ) (0.2653) (0.5686)
    Wage_Cost_per_unit_Output_ Growth_Rate 0.2502 0.0223 0.0695 -0.0139 1.0000
    (0.0000) (0.6767) (0.1930) (0.7946)
    Total_Input_Cost_per_ unit_Output_ 0.2149 -0.1420 -0.0411 -0.0334 0.0491 1.0000
    Growth_Rate (0.0000) (0.0076) (0.4416) (0.5321) (0.3584)
    Labour_Market_Flexibility_Growth_Rate -0.0583 -0.0571 -0.0184 0.0136 -0.1065 0.1462 1.0000
    (0.2757) (0.2857) (0.7310) (0.8002) (0.0462) (0.0061)
    Values in parenthesis give the level of significance.
    Economic & Political Weekly may 9, 2009 vol xliv no 19 47
    EPW
    Table 3: Correlation Coefficient of Explanatory Variables in Regression Equation 2

    Labour_Market_ Technological Wage Growth Output_Growth_ Flexibility_Growth_ Change Rate Rate Rate

    Labour_Market_Flexibility_ Growth_Rate 1.0000
    Technological change 0.0136 (0.8002) 1.0000
    Wage growth Rate -0.0217 (0.6847) 0.1383 (0.0094) 1.0000
    Output_ Growth_Rate 0.1142 (0.0325) 0.0961 (0.0718) 0.2268 (0.0000) 1.0000

    Values in parenthesis give the level of significance.

    WPI series with the base 1993-94=100. Instead of the consumers’ price index (CPI) number for industrial workers, the wage bill of workers is being deflated by WPI. It is because we are interested here to know the cost of labour to entrepreneurs. For a similar reason we have used WPI to get the value of total input costs at 1993-94 price.

    The EPWRF database does not include data for the workers e mployed through contractors. We have taken this data from ASI of CSO and by using concordance table provided by EPWRF prepared the time series data for the 1995-96 to 2003-04.

    The data series for stock of fixed capital has been constructed by using both ASI and National Accounts Statistics (NAS) database of CSO. The ASI has provided data on net fixed capital stock at book value. Our analysis requires current net fixed capital stock, which is available for present production. The data series provided by NAS satisfies this definition of fixed capital stock. NAS has provided it at the aggregated level of organised manufacturing sector. Since ASI provides the data at three-digit NIC level, following Chaudhuri (2002) by using perpetual inventory accumulation method, we have combined the ASI and NAS data to estimate net fixed capital stock at 1993-94 prices

    between employment growth rate and growth in labour market flexibility. The dataset available to us for estimating these two models are balanced panel data. It consists of eight years’ data on 44 sectors. Since, it is a panel data, to begin with, we have assumed that the constant term of both the equations consists of both sector specific and time specific effect. And then we have done some statistical test to find the existence of sector and time specific effect.

    Relation between Output Growth and Labour Market Flexibility

    Each sector is the aggregation of firms which produce that particular sectoral product. A firm’s (representative of a sector) decision to set a production target for the current year depends upon four factors – first, its expectation about the demand for its product in the current year, second, the production capacity, third, the choice of technology, and fourth, the variable cost of production.

    We assume that firm’s expectation about the demand for its product in the current year depends upon its previous year’s experience. So output growth of the current year depends upon the output growth of the previous year. If the manufacturing economy as a whole is booming then it is most likely that the production target for the current year will be revised upward. And during slowing growth, the opposite will most likely happen. In the study, our period of econometric analysis is 1996-97 to 2003-04. The Indian manufacturing sector experienced a slow growth in the substantial part of this period. Hence, it is most likely that the output growth of the current year is negatively related with the output growth of the previous year.

    The production capacity of a firm depends upon the capital stock it has. Since the gestation period of investment is generally

    for the three-digit NIC 1998 level manu-Table 4: Equation1: Pooled Panel Regression with Robust Standard Error

    facturing sectors.

    To trace the relation of the growth in l abour market flexibility with output growth rate at 1993-94 prices and employment growth, the correlation coefficients are calculated. We found that both the values of the

    Independent Variable Value of Std Err t P>|t| Statistical Coefficient Significance at 5% level

    Growth rate of output lag1 -.1215029 .0610402 -1.99 0.047 Significant
    Technological change .001715 .0005385 3.18 0.002 Significant
    Growth rate of wage cost per unit of output -.5846804 .1037982 -5.63 0.000 significant
    Growth rate of total input cost per unit of output -.0866525 .6405072 -0.14 0.892 Insignificant
    Growth rate of net fixed capital stock lag1 -.0789331 .1768525 -0.45 0.656 Insignificant

    correlation coefficient are close to zero Growth rate of net fixed capital stock lag2 -.0047898 .1260702 -0.04 0.970 Insignificant

    ( Table 1, p 47). So, there is no linear relation-Growth rate of labour market flexibility .0254202 .0755917 0.34 0.737 Insignificant

    ship of growth of labour market flexibility Constant term .083756 .0255637 3.28 0.001 Significant with output growth and employment growth.

    Dependent variable: Number of F(7,344) = R-squared Root <SE = Growth rate of output observation: 11.53 = 0.2569 .24532

    The scatter diagram of growth in labour mar

    352 Prop>F = ket flexibility and output growth shows no 0.000

    pattern of relation. This is true for the scatter d iagram of growth in labour market flexibi-Table 5: Equation 2: Pooled Panel Regression with Robust Standard Error

    lity and output growth also.

    In order to further investigate the causal link between the variables we undertake a simple econometric exercise. We shall estimate two regression equations for the period 1996-97 to 2003-04. The equations are meant to (1) estimate the relationship between output growth rate and growth in labour market flexibility, and (2) estimate the relationship

    Independent Variable Value of Std Err t P>|t| Statistical Coefficient Significance at 5% level

    Growth rate of output .3312669 .055072 6.02 0.000 Significant

    Technological change .0001137 .0002802 0.41 0.685 Insignificant

    Growth of wage rate -.3689253 .0812437 -4.54 0.000 Significant

    Growth rate of labour market flexibility -.0084198 .0290925 -0.29 0.772 Insignificant

    Constant term -.0090189 .0081839 -1.10 0.271 Insignificant

    Dependent variable: Number of F(7,344) = 13.80 R-squared = Root MSE = .1453 Growth rate of workers observation: Prob > F = 0.000 0.2997 352

    may 9, 2009 vol xliv no 19

    SPECIAL ARTICLE

    Graph 1: Net Fixed Capital Stock-Labour Ratio (in Rs lakh/worker, at 1993-94 prices)

    16

    12

    8

    4

    0 1986 1988 1990 1992 1994 1996 1998 2000 2002 2003

    Source: National Accounts Statistics of various years for net fixed capital stock. Annual Survey of Industries for workers.

    Graph 2: Net Fixed Capital Stock-Output Ratio (at 1993-94 prices)

    1.4

    1.3

    1.2

    1.1

    1.0

    0.9

    0.8 1986-87 1988-89 1990-91 1992-93 1994-95 1996-96 1998-99 2000-01 2002-03 2003-04 Source: National Accounts Statistics of various years for net fixed capital stock.

    Annual Survey of Industries for output. RBI Handbook of Indian Economy for WPI of manufacturing sector.

    one to two years, the output growth of the current year should depend upon the investment decisions of previous years. So we assume that the output growth of the current year depends upon the growth of capital stock of the previous year and of the year before the previous year. As an increase in production capacity allows firms to increase the production, it is most likely that output growth is positively dependent upon growth in capital stock.

    The choice of technology decides the combination of capital and labour that is to be used in production. Technological change can alter the factors of production mix which, in effect, can change the output level. So we can say that the output growth in the current year depends upon the technological change in the current year, which is measured as the ratio between growth in capital stock in the current year and growth in workers employed in the current year.

    The firm’s ability to increase output also depend upon its variable costs of production. The major components of the variable costs are labour, raw materials and energy. So we can say the firm’s output growth also depends upon the growth in wage bill per unit of output. This relationship is expected to be negative. The other major constituent of variable cost for the firm is the raw material, energy, etc. We have clubbed all of them together as input cost. The growth rate of input cost per unit of output should also negatively influence the growth rate of output.

    Now the growth in labour market flexibilities should positively influence output growth through its influence on greater encouragement to invest, influencing the technological choices and reducing the labour cost. The inclusion of the growth of l abour market flexibilities along with the growth in capital stock, technological choice and labour cost as factors to influence output growth may create multicollinearity problems. We have to test it statistically.

    Economic & Political Weekly

    EPW
    may 9, 2009 vol xliv no 19

    Following the above discussion more formally we can write the model as,

    (Y – Yt-1)

    1/Yt-1t

    1 1 1 = α + β1 –Yt-2) + β2 – Kt-2)+ β3 – Kt-3)

    (Yt-1(Kt-1(Kt-2Yt-2 Kt-2 Kt-3

    (K – Kt-1)

    t

    1 W 1 INPINPt-1

    Kt-1 t Wt-1t

    + β4+β5+β6 (L – Lt-1) Wt-1 Y Yt-1 NPt-1 Y Yt-1

    ttt

    Lt-1 () ()

    Yt-1 Yt-1

    1 CL

    t CLt-1

    + u...(1)

    + β7 t CLt-1 L

    t Lt-1 Lt-1 ()

    where Y is output of a particular sector at the tth period, K is

    tt

    capital stock of a particular sector at the tth period, L is total

    t

    workers of a particular sector at the tth period, W is total wage bill

    t

    of workers of a particular sector at the tth period, INP is total input

    t

    cost (it includes raw materials, energy, etc) of workers of a particular sector at the tth period, CL is total workers employed

    t

    through contractor of a particular sector at the tth period.

    The expression in the left hand side of the equation is output growth of a particular sector at the tth period. The second term from the left in the right hand side of the equation is output growth of a particular sector with lag 1. The third and fourth term from the left in the right hand side of the equation are capital stock growths of a particular sector with lag 1 and lag 2. The fifth, sixth, seventh and the eighth terms are technological change, growth in wage cost per unit of output, growth in total input cost per unit of output and growth in labour market flexibility respectively for a particular sector.

    The expected sign of the β coefficients are β1 < 0, β2 > 0, β3 > 0, β4 > 0, β5 < 0, β6 < 0 and β7 > 0.

    We have estimated the correlation coefficient among all the explanatory variables to test the multicollinearity problem. The values of correlation coefficient are very low (Table 2, p 47). So there is no multicollinearity problem here. Though theoretically there is a possibility of multicollinearity between the growth in labour market flexibilities and the three other explanatory variables – growth in capital stock, technological change and growth in wage cost per unit output – the values of the correlation c oefficients between them are substantially low making multicollinearity unlikely. Only the value of the correlation coefficient of growth in labour market flexibilities with growth in wage cost per unit output is statistically significant. The values of the c orrelation coefficient of growth in labour market flexibilities with growth in capital stock and technological change are not statistically significant.

    Initially we have assumed that in this panel data series, both sector-specific and time specific effects exist. To capture the time specific effect, we have introduced seven dummy variables to

    Graph 3: Output and Workers Trends (at 1993-94 prices)

    70 60 50 40 8 6 4 2Number of workers (in lakhs)Output (in Rs lakh crore) Workers Output

    1986-87 1988-89 1990-91 1992-93 1994-95 1996-96 1998-99 2000-01 2002-03 2003-04 Source: Annual Survey of Industries for output and workers. RBI Handbook of Indian Economy for WPI of manufacturing sector.

    separate out the effect of each year. To start with we have estimated the random effect model. We found that for the sectors pecific effect ‘u’ (by using Breusch and Pagan Lagrangian multiplier test for random effects).

    H0: Var (u) = 0 cannot be rejected even at 10% level of s ignificance.

    So the pooled panel estimation may be more accurate model to estimate. Also we found that the hypothesis that the coefficients of the dummy variables are equal to zero cannot be rejected even at 10% level of significance. Even for the pooled panel model we cannot reject this hypothesis. So we have dropped the time s pecific dummy variables from the model. The estimation results are shown in Table 4 (p 48).

    The growth rate of wage cost per unit of output of the current year has statistically significant (at 5% level) negative influence on the current year’s output growth. By averaging over time out of the 44 sectors, 36 sectors have a negative average growth rate of cost per unit of output of the current year. And by averaging over the sectors, out of eight years, in six years this growth rate is negative. So lowering of wage cost per unit output has helped output to grow.

    The growth rate of output of the previous year has a statistically significant (at 5% level) negative impact on output growth of the current year. Since, in the substantial part of the period 1996-97 to 2003-04, Indian manufacturing industry experienced a slowdown, it is expected that the producers adjusted their output growth target for the current year lower from the previous year’s output growth. In fact, out of 44 sectors, in 30 sectors the previous year’s output growth on average (averaged over time) are higher than the current year’s output growth. This finding also implies that during this period, Indian organised manufacturing suffered from lack of demand. This is further reflected in the fact that, when all most all the sectors on average (over the time) have positive growth in capital stock, growth of capital stock with lag of one and two years has no statistically significant influence on output growth of the current period. This indicates the existence of excess capacity in the Indian organised manufacturing sector (this is similar to finding of Guha 2008).

    The technological change in the current year, which measures the extent of substitution of labour by capital, also has a statistically significant positive influence on output growth of the current year. But the value of coefficient of technological change is very small. A majority of the sectors, out of 44 sectors 26 have on average (average over time) increased capital intensity. So the increasing capital intensity has resulted in output growth by a very narrow extent.

    Lastly, we found that growth rate of labour market flexibility has no statistically significant influence on growth rate of output. It is expected that increasing labour market flexibilities should reduce the wage cost per unit of output. As growth of wage cost per unit of output negatively influences output growth, the growth in labour market flexibility should have a positive influence on output growth. But the value of the coefficient (in the regression equation) of the growth of wage cost per unit of output is less than 1, around -0.58. So the impact of the growth in labour market flexibilities on output growth through the reduction in growth of wage cost per unit of output will be low. The increasing labour market flexibility should result in adopting more labourintensive technologies and growth in net fixed capital stock. But more capital-intensive technologies are being adopted. And the growth of net fixed capital stock has no statistically significant influence on output growth. So, we are not getting any statistically significant influence of the growth rate of labour market flexibility on growth rate of output.

    Employment Growth and Labour Market Flexibility

    It is assumed that employment growth in a sector depends upon the demand for labour by the firms of that sector. The demand for labour in the current year is positively dependent upon the firm’s target of production in the current year. So the employment growth in the current year should be positively dependent on output growth rate in the current year.

    The demand for labour also depends upon the choice of technology. Choice of capital-intensive technologies will reduce l abour demand. So the technological change in the current year, which is measured as the ratio between growth in capital stock in the current year and growth in workers employed in the current year (here, the underlying assumption is that there is no labour constraint for the Indian manufacturing sectors) should negatively influence employment growth of the current year. Another factor influencing the demand for l abour negatively is its cost. So the growth in wage rate of the

    Graph 4: Average Wage Rate (Rs per worker, at 1982-83 prices) 11500

    11000

    10500

    1000

    9500

    9000

    8500 1986-87 1988-89 1990-91 1992-93 1994-95 1996-96 1998-99 2000-01 2002-03 2003-04

    Source: Annual Survey of Industries for wage bill and workers. RBI Handbook of Indian Economy for CPI of Industrial worker.

    current year should negatively influence employment growth of the current year.

    The growth in labour market flexibility should positively influence output growth, influence technological choices towards the adoption of labour-intensive technologies and reduce labour cost. All three should create a positive influence of growth in labour market flexibility on employment growth. The inclusion of the growth of labour market flexibility along with the change in t echnology and growth in wage rate as factors to influence

    may 9, 2009 vol xliv no 19

    SPECIAL ARTICLE

    e mployment growth may create multicollinearity problems. We have any statistically significant sign of the coefficient for technohave to test this statistically. logical change in the estimated equation. Following the above discussion more formally we can write Lastly, the growth in labour market flexibility does not have the model as any statistically significant influence on employment growth. (L – Lt-1)

    1/Lt-1t

    (K – Kt-1) 1 1 W Wt-1

    Kt-1t

    = α + β1 (Y) + β2 + β3

    t –Yt-1

    t()

    Yt-1 (L – Lt-1) Wt-1 L

    tt Lt-1

    Lt-1 Lt-1

    1 CL

    t CLt-1

    + u...(2)

    + β4 t

    CLt-1 L

    t Lt-1 Lt-1 ()

    The expected sign of the β coefficients are β1 > 0, β2 < 0, β3 < 0, > 0.

    β4

    The values of the correlation coefficient among the explanatory variables are very low (Table 3, p 48). So there is no multicollinearity problem here. Though theoretically there is a possibility of multicollinearity between the growth in labour market flexibilities and the two other explanatory variables – technological change and growth in wage rate – the values of the correlation coefficient between them are substantially too low to have any multicollinearity problem.

    To identify the more accurate regression model for this equation we have followed the same procedure of the previous equation, i e, to begin with we have assumed that this panel data s eries have both sector-specific and time specific effects. To capture the time specific effect, seven dummy variables are introduced to separate out the effect of each year. To start with we have estimated the random effect model. We found that for the sector specific effect ‘u’ (by using Breusch and Pagan Lagrangian multiplier test for random effects):

    H0: Var (u) = 0 cannot be rejected even at 10% level of s ignificance.

    So the pooled panel estimation may be a more accurate model to estimate. Also we found that the hypothesis that the coefficients of the dummy variables are equal to zero cannot be rejected even at 10% level of significance. Even for the pooled panel model we cannot reject this hypothesis. So we have dropped the time specific dummy variables from the model. The results are reported in Table 5 (p 48).

    The growth rate of output has a statistically significant (at 5% level of significance) positive influence on employment growth. The growth of wage rate also has a statistically significant (at 5% level of significance) negative influence on employment growth. It appears that growth rate of employment is more sensitive to growth in the wage rate than growth in output. Technological change does not have any statistically significant influence on employment growth. As we have mentioned that the use of technologies are becoming more capital-intensive, technological change should have a negative impact on employment growth. But this technological change has a positive impact on output growth and output growth has a positive influence on employment growth. So technological change has both a positive and negative impact on employment growth. As a result, we do not

    Economic & Political Weekly

    EPW
    may 9, 2009 vol xliv no 19

    Conclusions

    From the regression analysis that we have undertaken, we conclude that increasing labour market flexibility – defined as an i ncrease in the proportion of non-permanent/casual workers in total workers – has no positive impact on output and employment growth. The neoliberal proposition that an increase in labour flexibility would lead to higher output growth and greater labour absorption does not seem to be valid as far as Indian manufacturing industries are concerned. Therefore, the reasons for an amendment to the Contract Labour Act, 1970 and IDA, 1947, which are being proposed in order to facilitate greater labour market flexibilities, appear quite dubious.

    It will be rather interesting to see what the other consequences of this labour market flexibility are. If we start looking from the mid-1980s onwards, the Indian organised manufacturing has witnessed a monotonically increasing trend in capital-labour r atio1 (Graph 1, p 49). The growth rate of the ratio is lowest in the second half of the 1980s. In the first half of the 1990s, the growth was higher and thereafter, till 2001-02, it was the highest. It then declined marginally. This indicates that manufacturing companies are increasingly adopting capital-intensive technologies.

    Graph 5: Trends of Wage and Profit Share in Net Value Addition (in %)

    50

    40

    30

    20

    10

    Profits Wage

    1986-87 1988-89 1990-91 1992-93 1994-95 1996-96 1998-99 2000-01 2002-03 2003-04 Source: Annual Survey of Industries for wage bill, profit and net value addition.

    The capital-output ratio2 measured at 1993-94 prices shows a different trend (Graph 2, p 49). It has a very marginal upward trend and is actually dominated by cyclical fluctuation of a narrow 5% band spread on either side of the average trend. Now, if capital-intensive technologies are being adopted to replace l abour, who are difficult to retrench due to labour market rigidities, then there should have been a monotonic rising trend in the capital-output ratio. This is not the case with the Indian manufacturing sector in the 1990s. Rather it seems capital- intensive technologies are being adopted to produce commodities for two reasons – first, the cost of capital has been reduced substantially for large companies due to financial market liberalisation (Guha 2008) and second, the demand pattern for I ndian industries has changed as even the domestic demand pattern is becoming similar to the global. And this global p attern of demand is highly tilted towards capital-intensive products (Guha 2008; Papola 2008).

    With the exception of a few years in the mid-1990s, the total number of workers employed are more or less stagnant throughout the whole period 1986-87 to 2003-04, whereas the actual manufacturing output at constant prices of 1993-94 has a mono-of the output growth. This is being done through the tonically increasing trend with the exception of 1998-99 combination of adopting capital-intensive technologies and (Graph 3, p 50). The average wage rate3 at 1982-83 prices has fluc-greater labour market flexibility. Therefore, more labour tuated within the narrow band of 15 to 20% of the average wage market flexibility has no influence on output, employment rate. In the mid-1990s the average wage rate was highest. And, in growth, apart from making a redistribution of income in favour late 1990s and thereafter, it was lower than in the late 1980s and of the capitalist class. early 1990s (Graph 4, p 50). Lastly, the share of wages in net value If this is the situation, as during the period of the substantial a ddition has steadily declined from a little more than 30% in positive growth of the manufacturing sector, then it is obvious 1986-87 to 15% in 2003-04. The decline is at a much faster rate in that during a recession a substantial portion of the adjustment the last three years. On the other hand, the profit share in net burden of the production process will be shifted to the workers. value addition has increased from 18% in 1986-87 to 45% in In mainstream economic theory, capitalists earn profit in return 2003-04. Here we have also seen a marked jump in the increase for taking risk. The Indian capitalist class, in this process of takof the profit share from the mid-1990s onwards (Graph 5, p 51). ing risk in business, wants to grab the profits in the good times

    All these trends together indicate that even when there is a but do not want to take the burden of adjustment in the business, substantial growth in manufacturing output, the workers are when it goes through bad times. They want to shift the burden of not benefiting. The capitalist class is reaping the larger benefits adjustment on to the workers.

    Notes

    1 Capital is net fixed capital stock of organised manufacturing sector at 1993-94 prices. The data source is NAS. Labour is “workers” in organised manufacturing sector provided by EPWRF’s ASI database.

    2 The data of output for organised manufacturing sector has been taken from EPWRF’s ASI database and it is deflated by the WPI series of RBI to transform it in 1993-94 prices.

    3 Nominal wage bill for workers has been taken from EPWRF’s ASI database. Dividing the wage bill with the number of worker we got the Nominal average wage rate. Nominal average wage rate is deflated by consumer price index number of industrial workers which is reported in H anbook of Indian Economy (Reserve Bank of India).

    References

    Baker, D, A Glyn, David Howell and John Schmidt (2002): “Labour Market Institutions and Unemployment: A Critical Assessment of the Cross-Country Evidence”, CEPA Working paper 2002-17 (Center for Economic Policy Research, New School University, New York).

    – (2004): “Labour Market Institutions and Unemployment” in D Howell (ed.), Questioning Liberalisation: Unemployment, Labour Markets and the W elfare State (Oxford: Oxford University Press).

    Chaudhuri, S (2002): “Economic Reforms and Industrial Structure in India”, Economic & Political Weekly, pp 155-61.

    Debroy, B (2005): “Issues in Labour Law Reform” in Bibek Debroy and P D Kaushik (ed.), Reforming the Labour Market (New Delhi: Rajiv Gandhi Foundation).

    Deshpande, Lalit K et al (2004): Liberalisation and L abour: Labour Flexibility in Manufacturing (New Delhi: Institute for Human Development).

    Deshpande, Sudha et al (1998): Labour Flexibility in a Third World Metropolis (Indian Society of Labour Economics: New Delhi and Commonwealth P ublishers, New Delhi).

    Dutta Roy, S (1998): “Lags in Employment Ad- justment and Inter-Industry Differentials: An Analysis Using Dynamic Inter-related Factor D emand Functions”, Discussion Paper No 149, I ndira Gandhi Institute of Development Research, Bombay.

    Fallon, P and R Lucas (1991): “The Impact of Changes in Job Security Regulations in India and Zimbabwe”, World Bank Economic Review, 5.

    Government of India (2003): Report of the Second National Commission on Labour, Chapter 1, p 12.

    Guha, Atulan (2008): “Evolution of Indian Organised Manufacturing Industrial Structure” in S R Hashim, K S C Rao, K V K Ranganathan and M R Murthy (ed.), Indian Industrial Development and Globalisation (Delhi: Academic Foundation).

    International Labour Organisation (1996): ILO Jobs Report 1996-97, www.jobsletter.org.nz/jbl05210.htm.

    International Monetary Fund (2003): “Unemployment and Labour Market Institutions: Why Reforms Pay Off”, Chapter 4, World Economic Outlook, April, Washington DC.

    Nagaraj, R (2004): “Fall in Organised Manufacturing Employment: A Brief Note”, Economic & Political Weekly, Issue 30, Volume 39, p 3387.

    OECD (1994): OECD Jobs Study, Evidence and Explanations, Parts I and II, Paris.

    Papola, T S (2008): “Industry and Employment: Dissecting Recent Indian Experience” in S R Hashim, K S C Rao, K V K Ranganathan and M R Murthy (ed.), Indian Industrial Development and Globalisation (Delhi: Academic Foundation).

    Sen, Sunanda and B Dasgupta (2006): “Labour in I ndia’s Organised Manufacturing” (mimeo) presented in International Conference on Developmentin Open Economies: Industry and Labour organised by the Academy of Third World Studies, Jamia Milia Islamia in Delhi.

    – (forthcoming): Unfreedom and Wage Work: L abour in India’s Manufacturing Industry (Delhi: Sage-India).

    Sharma, Alakh and S K Sasikumar (1996): “Structural Adjustment and Labour”, VV Giri National Labour Institute, NOIDA (mimeo).

    World Bank, World Development Report, 1990 (New York: Oxford University Press).

    THE POSTNATIONAL CONDITION

    March 7, 2009

    The Postnational Condition –Malathi de Alwis, Satish Deshpande, Pradeep Jeganathan, Mary John, Nivedita Menon,

    M S S Pandian, Aditya Nigam, S Akbar Zaidi South Asia? West Asia? Pakistan: Location, Identity –S Akbar Zaidi The Practice of Social Theory and the Politics of Location –Satish Deshpande Reframing Globalisation: Perspectives from the Women’s Movement –Mary E John Postnational Location as Political Practice –Malathi de Alwis The Postnational, Inhabitation and the Work of Melancholia –Pradeep Jeganathan Empire, Nation and Minority Cultures: The Postnational Moment –Aditya Nigam Nation Impossible –M S S Pandian Thinking through the Postnation –Nivedita Menon

    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

    may 9, 2009 vol xliv no 19

    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