EDITORIAL

Manufacturing Industry
The pattern of human consumption has been changing over the decades and changing
quite fast in recent times. A shift from consumption of (hard) goods to services (soft goods
included) is clearly visible. We also notice a corresponding change in the pattern of
production. In fact, each influences the other. It is, however, true that Manufacturing
Industry has to continue supporting all production processes in a big way. Thus, the role
of ‘manufacturing’ remains quite significant. Of course, with ‘lean manufacturing’ concepts
and practices and with automation in its advanced form, size of manufacturing units in
terms of the number of workers or employees will tend to be smaller in general. This fact,
along with the traditional practice of treating units with at least 10 workers (20 in case
electricity is not used for production purposes) as constituting the “organized
manufacturing” sector, may imply an increasing segment of manufacturing industry units
will be marked as ‘unorganized’. Having said all this, ‘organized manufacturing’ sector
continues to be the anchor of economic activities in the country and this anchor must get
stronger and more accessible to provide a boost to the National Economy.
The Annual Survey of Industries (ASI) being conducted by the Industrial Statistics Wing
of the Central Statistical Office is possibly the only source of credible information about the
organised manufacturing industry in the country, with gradually increasing information
content. And, plans are afoot to cover the Service Industry also in course of time. Once this
expansion in coverage takes place, we will get a much more comprehensive picture about
the production set-up in the country and its contribution to National Development.
One thing, however, has to be realized as circumscribing the contribution of ASI data to a
portrayal of performance of Manufacturing Industry (for the present). This is the inseparable
connections between ‘organised’ and ‘unorganised’ manufacturing. The latter plays a big
role in supplying inputs to the former, while the former in some cases provides technical,
financial and other support to the latter. In fact, most unorganized manufacturing units are
engaged in producing accessories or spares or semi-processed materials or finished
components to large manufacturing units covered by the organised sector. The performance
of any registered factory generally depends on the performance of its vendors including
‘unorganised’ production establishments. At the same time, most small units outside the
purview of ASI depend heavily on demand from and support extended by the organized
sector.
In a somewhat different sense, the performance of any industry sector is affected by
sectors which provide inputs to or receive outputs from this sector. A study of
interdependence among different sectors which are linked in terms of customer-supplier
relations should be undertaken in respect of important dimensions of performance. In fact,
there have been situations where a customer industry had to wind up its operations in the
absence of required inputs coming from the domestic supplier industry. The converse also
has happened with dwindling demands from domestic customer industry sending out
signals or threat of extinction to the domestic supplier industry. Exploring foreign markets
for customers or suppliers may not be convenient or economical in all cases. It is also
interesting to note that the nature and extent of such inter-dependencies have also been
changing over time and deserve appropriate investigation.
ii

Statistics relating to manufacturing industry and meant to portray a reliable picture of this
very significant economic activity should be understood and interpreted in a holistic manner
that provides a balanced view about the intentions of and the contributions by different
players—both domestic and foreign—in the growth and development of manufacturing
industry. Thus, inflow of foreign investment which is often associated with some strings
that pull the recipient industry in a certain direction—not necessarily desired in our
country—is welcome in some sense, inflow of foreign production technologies –some of
which draw upon more energy or generate more pollution or lead to reduction in
employment—in the name of Research & Development initiative may not be always a
welcome move.
All this and many more related issues call for adequate attention of investigators who can
–in their turn—demand more information about different aspects of functioning of
Manufacturing Industry.

September 2016 S. P. Mukherjee

Kolkata Editor-in-Chief
 iii

SECTION I : ARTICLES

Page No.

• Ramaa Arun Kumar 99

Subcontracting and Industrial Agglomeration :
Related Phenomena in India’s Unorganised Manufacturing Sector

• Nilabja Ghosh, Roopal Jyoti Singh and M. Rajeshwor 122

Understanding Demand, Supply and Price Behavior
in the Dairy Sector: Using Official Indian Statistics

• A. K. Panigrahi 138
Contract Workers in India’s Organised Manufacturing Sector

• Sarmishtha Sen and Subrata Majumder 154

Employment-Productivity Profile and Labour Demand Elasticity :
A Preliminary Study of the Organized and Unorganized Indian
Textile and Garment Firms

• Sajal Jana and Maniklal Adhikary 179

On Spatial Concentration of Organized Manufacturing Industries :
A Look at Regional Perspective

• Shiney Chakraborty 192

The Price of Prejudice : Employment Trend and Wage Discrimination of
Women Workers in India

• Panchanan Das, Abhishek Halder and Rahul Dutt 226

Export Competitiveness and Intensity of Technology in
Indian Manufacturing Industries –Analysis with ASI Unit Level Data
 The Journal of Industrial Statistics (2016), 5 (2), 99 - 121 99

Subcontracting and Industrial Agglomeration: Related Phenomena in

India’s Unorganised Manufacturing Sector
Ramaa Arun Kumar1, Delhi School of Economics, New Delhi, India

Abstract

The unorganised manufacturing sector in India has been an integral part of the economy
and its inter-linkage with the organised part of the sector in the form of subcontracting is
well established. The benefits arising out of this inter-linkage may be enhanced due to
the presence of agglomeration economies generated at the regional level. The study tries
to estimate the effect of such economies on the prevailing subcontracting arrangements
in the unorganised manufacturing sector. Using the quinqennial enterprise surveys
conducted by the National Sample Survey Organisation (NSSO) on Unorganised
Manufacturing Sector Enterprises for the 56th (2000-01), 62nd (2005-06) and 67th (2010-
11) Rounds, the paper establishes that the effect of industrial agglomeration on the
probability of a firm being in a contractual arrangement in the unorganised manufacturing
sector varies across different industries at NIC 2-digit level. Overall, the localisation
economies at district level seem to have a stronger effect on subcontracting than industrial
diversity.

1. Introduction

1.1 Production subcontracting in manufacturing sector is one of the key characteristics

of the interaction between organised and unorganised manufacturing sector. A firm may
choose to undertake all activities in its manufacturing process or outsource/subcontract a
part of the manufacturing process to an outside firm, whether it is sourcing inputs (backward
linkage) or after production processes (forward linkage). In the Indian context, subcontracting
has seen a significant and persistent growth, which is considered to be an offshoot of the
economic reforms that were initiated in the late 1980s and early 1990s when industrial
licenses were abolished for a number of industries along with trade and investment
liberalisation reforms. There was an advent of increased competition due to these
developments for industries in India as it encouraged private sector to grow. The competitive
pressures thus, created on manufacturing sector were aggravated by the strict labour laws
pertaining to the formal sector enterprises which led to the proliferation of the unorganised
manufacturing firms (Moreno-Monroy et. al, 2012).

1.2 A healthy inter-firm relation between large and small firms entails a lot of benefits
for both the contracting parties. On the other hand, subcontracting that takes place between
firms of different sizes creates inequality of power which may also be detrimental to the
outcomes. It remains to be seen how subcontracting has impacted the smaller firms that
belong to the unorganized manufacturing sector in India.

1.3 For any subcontracting relation to persist in the long run, the access to a large
number of potential input suppliers is crucial as competition among them would reduce the
cost of subcontracting to the parent firm. Likewise, a number of larger firms concentrated in
the same region would enhance the probability of the small firms to get contracts. This
1
e-mail: [email protected]
10 0 The Journal of Industrial Statistics, Vol. 5, No. 2

mutual benefit arising out of industrial agglomeration in a region is the key factor that will
be the focus of this study. Economies arising out of industrial agglomeration have been
subjected to a number of studies and discussions. The earliest explanations date back to
Marshall (1890) when he explained three kinds of agglomeration economies that is, labour
market interactions, buyer-supplier linkages and knowledge spillovers through inter-industry
interactions which increase the tendency of firms to cluster in a region.

1.4 The presence of own industry firms in unorganised sector in a region may provide
a bigger market for the larger firms to choose smaller firms for giving out contracts which
raises the probability of such firms to be working under a contract. On the other hand,
concentration of different industries in one region leads to knowledge spillovers generated
by inter-industry linkages. Both these forces of agglomeration have an impact on the
probability of a firm in unorganised manufacturing sector to be operating under a contract.

1.5 This paper analyses this interrelationship between agglomeration and

subcontracting for the unorganised manufacturing sector in India. The National Sample
Survey Organisation (NSSO) conducts a quinqennial survey at the enterprise level for the
unorganised manufacturing sector. The data from three quinqennial surveys, that is, 56th
(2000-01), 62nd (2005-06) and 67th (2010-11) are used to empirically establish the interrelationship
described above.

1.6 The effect of industrial agglomeration is envisaged by the likely economies that it
generates for the firms at the industry and regional level. However, the effects on different
industries’ subcontracting status vary as there is an element of heterogeneity across
industries. The subsequent sections discuss and analyse this hypothesis empirically to
throw some light on this inter-relationship that is existing in the unorganised manufacturing
sector. Section 2 begins with an overview of the subcontracting status in India and the
literature that explains the nature of interactions taking place between unorganised and
organised manufacturing sector. The review moves on to discuss the economic geography
literature that throws light on the process of agglomeration and various economies it
generates to enable firms to thrive therein. The literature on the inter-relationship between
industrial agglomeration and subcontracting is then revoked to enable us to understand
how these two phenomena are linked.

1.7 Section 3 explains the measures of agglomeration that have been used in the
empirical estimations. The descriptive statistics based on the enterprise survey data are
presented in section 4 that tries to correlate industrial agglomeration and the existing
subcontracting status of the unorganised manufacturing sector firms at the state and
industry level. Section 5 gives a bird’s eye view of the data sources and the variables that
have been included in the study. This section also discusses the likely effect each variable
is expected to have on the dependent variable and the methodology used in the empirical
estimation. Section 6 presents the empirical results followed by the conclusion in section 7.

2. Subcontracting and Agglomeration-Related Phenomena: Review of Literature

The literature providing evidence of this inter-relatedness of subcontracting and

agglomeration dates back to 1890 when Marshall proposed that agglomeration allows for
sharing of inputs and facilitates the emergence of specialised intermediate input producers.
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 10 1

On a similar path, Stigler (1951) argued that increasing local market size leads to greater
vertical disintegration.

2.1. Subcontracting

Subcontracting in terms of manufacturing process takes the form of either assembly

subcontracting or product subcontracting, among others. Where sub contractors produce
the entire product and parent firm essentially performs the marketing activity for the sale of
the product, it is called product subcontracting. Sometimes assembling components to
form a finished good may be a labour intensive process, for example, in case of electronic
industries. Firms may farm out this activity to small firms, which is called assembly
subcontracting. There may be conditions where the large firms employ labour on a contract
basis, or they are temporary in nature. However, this would not amount to the process of
subcontracting, but would be in nature of informal employment. The study restricts itself to
contractualisation of a firm’s production process, and not that of labour.

2.1.1. Subcontracting in Indian Context

2.1.1.1 The unorganised manufacturing sector in India is highly labour intensive as it has
easy access to cheap labour, while it lacks access to capital. Owing to the small scale of
these firms, employing mostly cheap labour which works with limited capital, makes firms in
this sector low cost firms. Nagaraj (1984) explains that this comparative cost advantage is
the principal basis for the growth of sub-contracting. It is also believed that subcontracting
was fallout of the rigidity of regulations governing factory employment. Micro-economic
studies have documented the widespread use of market mediated work practices and
outsourcing of products [Bhushan (1984) and Ramaswamy (1988)].

2.1.1.2 In the Indian context, subcontracting dates back to the 1970s and 1980s much
before the liberalisation reforms were introduced in 1991. Ramaswamy (2013) notes that the
subcontracting intensity in Indian manufacturing sector which is characterised by the
phenomenon of ‘missing middle’ in terms of size distribution, was significantly high for the
employment size group with turnover below rupees 50 million, that is exempt from excise
duty under the General Excise Exemption Scheme. Further, within this group of small scale
firms, the subcontract intensity is found to be higher in the labour intensive firms which are
located in states that are inflexible in terms of labour regulations.

2.1.1.3 Evidence from Indian data on unorganised manufacturing sector indicates that
there has been significant subcontracting taking place in the past two decades. Bairagya
(2010) reports that the percentage of units operating on contracts has risen from 17 per cent
in 1999-2000 to 31.7 per cent in 2005-06. However, in the subsequent period of 5 years, the
share of subcontracting taking place in unorganised manufacturing sector has reduced to
20 per cent in 2010-11.

2.1.1.4 According to the NSS 62nd round for 2005-06, it is found that 87 per cent of the
OAMEs under a contract work solely for the contractors, while the similar figures for the
NDMEs and DMEs are 63 percent and 70 percent, respectively. However, in 2010-11, 91
percent of OAMEs worked solely for the contractors, while around 86 per cent of the
NDMEs and DMEs together were working under contract.
10 2 The Journal of Industrial Statistics, Vol. 5, No. 2

2.1.1.5 Subcontracting usually takes place between firms of different sizes and thus,
unequal power. The large or parent firms can exercise control over its subcontractors. Also,
they may provide the subcontracting firms with raw materials, capital, product designs etc.
which may help small firms to enhance production as well as their productivity with better
resources, compared to firms operating independently.

2.1.1.6 Based on a field survey of unorganised sector firms in India in 2005, Sahu (2007)
found that technology transfer has been happening where small firms have got assistance
from large ones in the form of technical drawings and specifications, tools and marketing,
etc. In addition, technical advice is provided in respect of machinery, process, and materials
to be used, and so on. In rural areas, a large proportion (57 per cent) of the subcontracting
units worked with the technology completely prescribed or handed over by the parent
company, and another 23 per cent operate with the technology that is partially or occasionally
prescribed by their parent companies; only 21 per cent of such units operated with their
own technology.

2.1.1.7 There is enough literature on the impact of subcontracting production overseas

on the parent firms. However, the studies on impact of domestic subcontracting on the
parent as well as subcontracting firms are very few in case of India. This may be because
the subcontracting relations between large and small firms in unorganized manufacturing
sector have not been that strong. In addition, data poses the greatest barrier to studying
these issues. This study fills this lacuna in the literature by using the enterprise surveys
conducted by the National Sample Survey (NSS) that reports the contractual status of each
firm in the unorganised manufacturing sector. The study very innovatively combines the
firm level factors with district and industry level factors that affect the contractual status of
the firm.

2.2. Subcontracting and Agglomeration

2.2.1 The factors influencing the decision of firms to subcontract or agglomerate in a

region may be numerous, but there are fallouts emerging from these processes that are
important to understand and analyse. The location of the firms, large as well as small, play
an important role in determining the possibility and pattern of subcontracting taking place
in unorganised sector. For example, locating firms nearer to the large industries may fetch
greater opportunities to get contracts as it enables larger firms to ensure the quality of
products supplied by the smaller firm. Location of firms, therefore, plays an important role
in affecting the performance of firms in unorganised manufacturing sector.

2.2.2 Industrial agglomeration is coming together of firms within the same industry or
different industries in the same region. There are a number of factors that may influence the
decision of a firm to locate itself in a particular area. The earliest literature on agglomeration
phenomenon dates back to Marshall (1890). According to him, the tendency of industries
to cluster arises out of the agglomeration economies from three sources: labour market
interactions, linkages between intermediate and final good suppliers and knowledge
spillovers. Firms within the same industry may come together to take advantage of the
localization economies such as specialized know-how, presence of buyer-supplier networks
and opportunities of efficient subcontracting. In addition, the inter-industry benefits also
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 10 3

exist such as complementary services, availability of labour, varied skills, inter-industry

information transfers and availability of low cost infrastructure, that also contribute to
firms locating in close proximity to other firms in the same industry or other industries. A
deeper and a more fundamental explanation on why industries choose a certain area and
not others is asked and attempted to answer in the New Economic Geography (NEG)
literature (Krugman, 1991, Fujita and Krugman, 2004; Venables, 1996 etc.).

2.2.3 A competitive market of input suppliers lowers the cost of outsourcing, thus,
reducing production costs of the firm which is outsourcing (Ono, 2001).

2.2.4 Ono (2007) estimated the effect of larger markets on the likelihood of outsourcing
business services. He develops a theoretical model to show that greater local demand for
an input induces more suppliers to set up shop in that locality. This causes greater
competition among the suppliers and thus, reduces the market price of the input. This
raises the probability for the final good producer to outsource. Using cross section data for
U.S manufacturing firms, the study found that plants in larger local markets have greater
opportunities to outsource services.

2.2.5 The clustering of firms in the electronic industry in the districts of Madrid facilitating
greater subcontracting has been shown in the studies by Suarez-Villa and Rama (1996) and
Rama et al. (2003). Clustering creates externalities for firms located therein that are internal
to the cluster but external to the production establishment. Proximity makes it easier to
build up inter-firm trust and reciprocity that leads to greater subcontracting in such clusters.

2.2.6 Taymaz and Kilicaslan (2005) have studied the subcontracting relations in Turkish
textiles and engineering industries. The study looks at the subcontract offering as well
receiving firms. It finds that geographical concentration of firms in the industry, that is,
clusters are important determinants for establishing subcontracting relationships.

2.2.7 Thus subcontracting decision of a firm is influenced by the regional factors.

However, a firm’s decision to locate in a particular region may not be an independent
decision. Holl (2008) argues that in estimating the effect of agglomeration on subcontracting
decision of a firm, there may be a bias in the estimates as the location decision of a firm may
be correlated with the error term that reflects management ability. According to him, the
firm’s management strategy may affect the decision of the firm to locate. For example, if a
firm is efficiently carrying out all production activities in-house, with no intention to engage
in subcontracting, the firm may locate to the peripheral regions to take advantages of lower
labour and land costs. However, firms that may not be able to carry all activities in-house
may need to subcontract and thus, decide to locate in agglomerations of economic activity.

2.2.8 In the Indian context, the effect of agglomeration on the subcontracting status of
a firm in the unorganised manufacturing sector has not been studied. This paper would try
to analyse what determines the firm’s subcontracting decision and the role of agglomeration
in influencing that decision.

3. Measuring Agglomeration

There are a number of studies that have dealt with the measurement of agglomeration.
Agglomeration effects are envisaged as the various economies arising out of the industrial
10 4 The Journal of Industrial Statistics, Vol. 5, No. 2

agglomeration process that are beneficial to the firms located therein. The different measures
used in this study depict two crucial linkages: intra-industry and inter-industry. The spatial
unit considered in this study is the district level.

A region co-located by firms in same industry enables access to specialised know-how,

adequate pool of labour and efficient subcontracting through intra-industry linkages. On
the other hand, the presence of diverse industries in the same region leads to greater inter-
industry linkages. These linkages are a breeding ground for innovation and knowledge
transfers (Chinitz, 1961 and Jacobs, 1969). In addition, regions with diverse economic
activities are endowed with financial and physical infrastructure such as access to better
banking services, advertising, improved transportation etc.

For the purpose of this study we consider three measures of agglomeration: location
quotient depicting the intra-industry linkages and Herfindahl Index to measure the extent
of diversity or concentration of economic activities in both organised as well as unorganised
manufacturing sector.

3.1. Localisation Economies

3.1.1 First, we consider the degree to which industries are concentrated in each location
in each industry, which is measured by location quotients. They are simple measures of
regional concentration used in regional studies (North, 1955; Lall and Chakravorty, 2005;
Pardo and Arauzo-Carod, 2011)

3.1.2 Measure of industrial concentration shows the degree of concentration of

economic activity in a particular location, region/spatial unit, while, industrial agglomeration
shows how this concentration is distributed spatially.

3.1.3 Our analysis relates to the agglomeration effects on firm characteristics, therefore,
the dimension of industries as well as location should be contained in one measure. Location
quotient (LQ) possesses both spatial and industrial dimension. It is the ratio of the share of
employment of industry “i” in a particular region to that in a bigger geographical area, say
state or the economy as a whole. We take the share of the industry in the economy as a
whole for comparison. Thus, the location quotient would represent the extent of
agglomeration of an industry in a particular district relative to its total presence in the
economy.

The formula for LQ is given by the following:

LQir = (Eir/Er ) / (Ei /E)

Where,

Eir is the employment of industry “i” in region “r” (district in our study),

Er is the employment in region “r”,

Ei is the total employment in industry “i” and

E is the total employment in the economy

 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 10 5

3.1.4 If the LQ is greater than one, the industry “i” would be considered to be
concentrated in the region “r”. LQ, being a proportion of two proportions, can range from
zero to very high values. For this study LQ would be calculated for each industry at NIC 3-
digit level located in different districts of the 18 Indian states.

3.1.5 There is no consensus on what the magnitude of LQ should be in order to consider

an industry being concentrated in a region or area. Various studies have used different cut-
offs for indentifying clustered areas based on LQs. Some studies have used a cut-off of
1.25 for identifying clusters, such as the study by Miller et al. (2001) that studies clusters
within a range of UK industries. Whereas, Isaksen (1996) and Malmberg and Maskell (2002)
have identified industrial agglomeration for industries within an area having an LQ larger
than 3. More recently, Chandrasekhar and Sharma (2014) have also undertaken a study on
spatial concentration of employment in India using a cut off of one using location quotients.

3.1.6 O’Donoghue and Gleave (2004) have modified the LQ to a standardised LQ measure
to incorporate the importance of statistical significance in using this measure. The motivation
was taken from Duranton and Overman (2002) which asserts that most cluster measures are
indices that do not report levels of significance. Therefore, O’Donoghue and Gleave (2004)
in their study identify agglomerations where the regions have statistically significant LQ
values, rather than an arbitrary cut-off.

3.1.7 The location quotient is a discrete measure of industrial concentration measured

at the spatial as well as industrial level. The objective of this study is to look at all areas
whether or not they are industrially agglomerated, in order to distinguish between the
effects of a firm located in either of the areas on its contractual status. Thus, the industrial
as well as the regional dimension are important in choosing the agglomeration variable,
which the location quotient possesses. In order to have a preliminary understanding of the
extent of concentration levels in different districts in the 18 states considered in the study,
we have used a simple cut-off of LQ greater than one to identify areas that have concentration
of a particular industry at NIC 3-digit level.

3.2. Industrial Diversity

3.2.1 Herfindahl Index has been used as a measure of the industrial diversity in a region
or the composition of local economic activity. This study includes two indices based on
Herfindahl index for the organised as well unorganised manufacturing sector separately.
The index is measured as given below:

2
E 
HI r   i  ir 
 Er 

Here, Eir is the employment of industry ‘i’ in region ‘r’ and Er is the total employment in
region ‘r’. The maximum value of the Herfindahl index can be one that indicates that the
region is dominated by only one industry, while lower values of the index indicate diversity
of economic activity. The presence of greater number of industries within a region in both
organised as well as unorganised manufacturing sector acts as a catalyst to the process of
10 6 The Journal of Industrial Statistics, Vol. 5, No. 2

subcontracting. For example, the firm in the wearing apparel firm belonging to NIC Division
14 may outsource its textile or fabric requirements to the small weaving firms that belong to
another industry (NIC Div. 13), or the pharmaceutical industry firms subcontracting their
chemical inputs to the firms in the chemical industry.

4. Some Descriptive Statistics

4.1 Before any empirical analysis based on robust regression method is undertaken,
it is useful to understand the trends in the different parameters or variables across time or
cross section. This is important to understand the framework in which our hypothesis is
based. Therefore, this section provides an overview of the trend in subcontracting and
agglomeration in different states and across industries at various points of time. The main
source of data at enterprise level for unorganised manufacturing sector is the NSS
Unorganised Manufacturing Sector Survey in 2000-1, 2005-6 and 2010-11. The survey
collects data on firm characteristics, different operating costs, wages and salaries of hired
employees, employment, capital, revenue, gross value added and its assets and liabilities.

4.2 The overall growth of unorganised manufacturing sector had slowed down over
the 11 year period from 1994 to 2005. Although, initially there was a significant growth in the
real value added in the unorganised manufacturing sector in 2000-01, vis-à-vis, 1994-95,
with 6.7 per cent compounded annual growth rate over the six year period as shown in
Table 1. However, in the period of next five years, the growth rate fell drastically to 3.8 per
cent in 2005-06. The trend was reversed with real gross value added rising at an annualized
growth rate of 6.8 percent between the five year period 2005-06 to 2010-11.

4.3 The real gross value added is calculated by deflating the nominal GVA in various
NSS rounds using the wholesale price index (WPI) for manufactured products with base
year 2004-05. The data on WPI which have been sourced from Handbook of Statistics,
Reserve Bank of India have the current base year as 2004-05. However, the various time
points considered in Table 1 had different base years, therefore, the base years of the WPI
for the two time points before 2004-05 were changed to 2004-05. The new indices were then
used for deflating the nominal GVA from the various NSS rounds.

4.4 Employment in the unorganised manufacturing sector, on the other hand, has
seen a falling trend since 2000-01. Total labour employed in the sector was 37 million in 2000
which has come down to 34 million in 2010-11.

4.5 The corresponding GVA per labour employed, or in other words, the labour
productivity has been on the rise at the aggregate level. The period 2005-06 to 2010-11
witnessed a dramatic increase in labour productivity, a growth of about 6.4 percent. However,
the commensurate high growth of GVA at 6.8 percent and a fall in total labour employed in
the sector in 2010-11 can be construed as the reason behind the high growth in labour
productivity in the five year period. This means that with relatively less labour employed,
the unorganised sector has been able to generate a high value added in 2010-11.

4.6 At the industry level, labour productivity has also been rising over the period of
11 years, spanning 2000-01 to 2010-11, based on the real value added and total labour
employed from the three NSS rounds (Table 2). However, some industries experienced a fall
in their labour productivity from 2005-6 to 2010-11.
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 10 7

4.7 These industries include basic metals industry (NIC Div. 24) where productivity
fell from Rs. 100,629 in 2005-06 to Rs. 60,329 in 2010-11. Likewise, computer and electronic
industry (NIC Div. 26), cotton ginning and cleaning (NIC Subdiv. 01632) and other
manufacturing industry (NIC Div. 32) also experienced a fall in productivity during the same
period.

4.8 Around 90 per cent of employment that is provided by 16 industries in the

unorganised sector also account for about 93 per cent of the total firms within each industry
operating under a contract from a larger firm. Within industries, the highest percentage of
subcontracting is being carried out by tobacco industries, followed by the textile sector.
Tobacco industry is also one of the largest job providers in the unorganised manufacturing
sector in India. According to the Ministry of Labour, Annual Report 1999-2000, the total
number of beedi-workers in India was around 4.5 million. The entry of foreign players in
tobacco industry as a result of economic liberalisation has forced the main beedi
manufacturers to shift to more backward and poverty-ridden areas in search of unorganised
and cheaper labour to subcontract their manufacturing activity (ILO, 1997). This may explain
the high and rising share of contractual work in this sector, which rose from 35 per cent in
2000-01 to 45.8 per cent in 2010-11 which is mostly (99.5 per cent) carried out by own
account manufacturing enterprises (OAME).

4.9 Textiles sector is another sector where substantial subcontracting is present.

Taking all industries under textiles and apparel sector (NIC 13 and 14 division), around 38
per cent of total subcontracting in unorganised manufacturing sector is accounted for by
this sector.

4.10 In the context of subcontracting of production activity by a firm, the location of an

industry plays an important role. The next section throws some light on the trends across
states and districts of the unorganised manufacturing sector.

4.11 Regional Scenario: District Level

4.11.1 To measure the extent of industrial concentration in various states at district level,
the location quotient has been calculated for all three NSS rounds. The location quotient,
as explained in the previous section, is a measure of the extent of industrial concentration
in a geographical unit compared to that at a bigger geographical unit. In this study, the
district level employment at 3-digit NIC level as compared to the national level employment
share has been considered.

4.11.2 A measure of location quotient greater than 1 is considered to be indicating

high concentration in a particular district or region. Based on the calculations of this
measure at district level for the set of 18 states chosen for this study, comparisons have
been made for the highly concentrated industries across time over the three time periods.
There were 858 district-industry pairs identified as highly concentrated areas (data for all
districts can be shared by author).

4.11.3 It is observed that the state of Andhra Pradesh has 16 out of 23 districts that are
agglomerated by textile and wearing apparel industry. These districts include East Godavari,
Vishakhapatnam and Hyderabad among others. The other industries that are concentrated
10 8 The Journal of Industrial Statistics, Vol. 5, No. 2

in different districts include paper industry in Vijayanagaram and Hyderabad, jewellery

industry in Hyderabad etc.

4.11.4 Towards the north of India, the district of Firozabad in Uttar Pradesh specialises
in all types of glass products. According to the district industrial profile given by the
Directorate Commissioner of the Ministry of Micro, Small and Medium Enterprises
(DCMSME), Government of India, there are 4222 units operating in the district alone with
an employment of 33,776 and a total investment of Rs. 12,666 lakh. Various glass products
like glass bangles, ware, tumblers, tubes form the major clusters in this district.

4.11.5 Likewise, sports goods industry in Jalandhar forms a major cluster. It is one of
the highest employment generating industries in the district. The footwear industry in the
district of Agra in Uttar Pradesh is another industry that has a dominant presence. There
are about 5,000 functional units of leather footwear in this district, including 150 exporting
units. Employment in this cluster is reported to be around two lakh in the district profile of
the DCMSME.

4.11.6 The district of Howrah in West Bengal also has a number of foundries producing
railway components and spares, apart from other casting metals for industrial purposes.
The Foundation for MSME Clusters (FMC) reports that the Howrah foundry cluster employs
around 8000 workers.

4.11.7 Location quotient is a simple measure of concentration that spans the regional as
well as the industrial dimension together. This measure has also been incorporated as an
explanatory variable in the empirical section.

5. Inter-relation between Subcontracting and Agglomeration: Data Sources and

Empirical Model

5.1. Data Sources

5.1.1 This section explains the factors influencing subcontracting decision of firms in
the unorganized sector. There are very few studies and probably none for India in particular
that looks at the subcontracting firms in different manufacturing industries in the unorganised
manufacturing sector. The factors considered, in accordance with the theory, that determine
whether a firm receives a contract or not are taken at firm level, industry and regional level.

5.1.2 The study uses the NSS 56th (2000-01), 62nd, (2005-06), and 67th (2010-11)
Unorganised Manufacturing Sector Surveys to analyse the effect of agglomeration on
subcontracting firms. The survey collects firm level data on the characteristics of the firm,
its operating costs, principle receipts, assets and liabilities etc. The firm level data has been
adjusted for annual levels, as the reference period in some of the characteristics of the firm
such as operating expenses, receipts and value added was one month.

5.1.3 In case of India, the country is divided broadly into states and each state is further
divided into districts. The study includes 18 states which contribute about 96 per cent of
total unorganised sector GVA. These states are: Maharashtra, Tamil Nadu, Gujarat, Uttar
Pradesh, West Bengal, Andhra Pradesh, Karnataka, Delhi, Rajasthan, Kerala, Punjab, Madhya
Pradesh, Haryana, Bihar, Orissa, Assam, Jammu & Kashmir and Himachal Pradesh. The
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 10 9

districts of Chhattisgarh, Jharkhand and Uttarakhand were not named in the NSS 56th
round, which would pose a problem of comparison across time, therefore, these three
states were not included, although they are important manufacturing states.

5.1.4 There are various factors at the firm, industry as well as district levels that determine
whether a firm in the unorganised manufacturing sector would undertake subcontracting
activity or not. Since the subcontract status of the firms is given by a yes or no answer in
the survey questionnaire, the dependent variable would be a binary variable.

5.1.5 Apart from the internal characteristics of the firms, there are several external factors
that play an important role in the unorganised sector firms to get contracts from them. The
concentration of firms within the same industry as well as industrial diversity in a region
generates numerous positive externalities that the firm can take advantage of. The study
has considered two effects that may enhance the probability of a small firm in the unorganised
manufacturing sector to obtain a contract from a larger firm. These are the location quotients
(localisation economies) and Herfindahl index (industrial diversity).

5.1.6 For the Herifndahl index organised manufacturing sector employment data is
collected by the Central Statistical Organisation under the Government of India, called the
Annual Survey of Industries (ASI). The ASI collects data on factories that are registered
under Sections 2m(i) and 2m(ii) of the Factories Act, 1948.

5.2. Empirical Model:

5.2.1 As mentioned before, the dependent variable is a binary variable, that takes a
value of 1 or 2 in the NSS round schedules to show whether a firm is having a contract with
a larger firm or not, where code 1 signifies that the firm has a contract. The empirical
estimation therefore, will be the estimation of probabilities based on maximum likelihood
procedure. The probit regression analysis is one of the techniques used in this case. Since
the dependent variable is a binary variable, using an ordinary least squares (OLS) would
not give efficient estimates as the changes in independent variable affecting the probability
that the dependent variable takes a value 1 cannot be bounded within the [0,1] range that
probability lies in. The probit regression coefficients give the change in the z-score or
probit index for a one unit change in the predictor.

5.2.2 The coefficients on independent variable are indicative of the direction of the
impact of a change in variable on the probability index. The magnitude of change in
probability owing to the change in an independent variable, however, is not a constant as
in a linear regression. The marginal effect of a change in any of the independent variables
is based not only on the value of that independent variable, but all other independent
variables too.

5.2.3 The estimation results in Table 3 are based on the coefficients obtained from the
estimation which indicate whether an explanatory variable increases or decreases the
probability of a firm to be working under a contract.

5.2.4 Variables and Empirical Equation:

5.2.4.1 The study attempts to test the inter-relation between the phenomenon of
subcontracting at the firm level and agglomeration at the industrial level. The factors affecting
11 0 The Journal of Industrial Statistics, Vol. 5, No. 2

the subcontracting of a firm will be looked at, at the firm level, in addition to the industrial
agglomeration, which is at the district-industry level.

5.2.5 Firm Level Independent Variables:

The firm level characteristics that are considered are listed below. Since they are firm
specific, they will be time varying.

a) Operational costs: Operational costs are one of the important factors that determine
whether the input suppliers would be subcontracted by a larger firm. A firm with lower
operation cost would be more likely to be working under a contract. Operational costs
include electricity charge, fuel, rent payable on machinery, transport expenses,
communication expenses etc. These expenses are an important part of the overall cost of
operation of a firm. Hence, they may be an important determinant for a firm to receive
contracts. The log of real operational cost per unit of value of output has been considered
in the estimation using wholesale price index (WPI) for fuel, power, light & lubricants
provided for industries together.

b) Size in terms of value added: Size of the firm is also a factor that may influence the
likelihood of a firm working under a contract. Bigger the size of the firm, larger is the scale
economies which enable them to undertake specialised operations. Log of real gross value
added has been used.

c) Capital Intensity: Capital intensity at the firm level is measured as the ratio between
real value of assets owned by the firm and total workers employed in the firm.

d) Labour Productivity: Labour productivity of a firm is an indicator of the efficiency

with which firms are operating. It may be an important determinant for the large firms
choosing to contract out a part of their production process or inputs to the smaller firms in
the unorganised manufacturing sector. The variable is measured as the log of labour
productivity calculated as the real gross value added per worker employed in the respective
firm.

e) Status of firms in last 3 years: Status of the firm in last 3 years in the NSS rounds
is enquired, that is, whether the sample firm has been expanding, remained stagnant or
contracted in the last 3 years. The variable is reported in codes: expanding – 1, stagnant –
2, contracting – 3, operated for less than 3 years – 9. Hence, three separate dummy variables
have been included in the estimation, viz., firms that are expanding, remained stagnant and
those which have contracted for the last three years. The firms which have operated less
than three years act as a baseline category against which the coefficients on the other three
dummies would be interpreted.

f) Whether accounts are maintained by firms: A firm that maintains records and
accounts of its operations would be more preferred to be given a contract than a firm that
does not maintain any accounts. The NSS schedule contains a binary code for firms who
do or do not maintain accounts. Therefore, a dummy variable has been used for this
characteristic of the firm, where the dummy takes a value 1 if the firm maintains accounts,
zero otherwise.
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 11 1

g) Location of firm’s premises: Two separate dummy variables for firms that have
premises within the household (Code 1) and those that have fixed and permanent structures
(Code 2) have been included. The rest of the categories that do not have any permanent
and/or fixed structures are taken as the baseline category.

5.2.6 District Level Agglomeration Variable:

5.2.6.1 This study considers the own industry concentration of the unorganised
manufacturing sector firms as well as concentration of all industries in the unorganised as
well as organised sector at the district level. The presence of own industry firms in
unorganised sector in a region may provide a bigger market for the larger firms to choose
smaller firms for giving out contracts. Therefore, the chances of the firm located in such a
region would have better probability to be working under a contract.

5.2.6.2 Secondly, it is the geographical proximity of the small and the large firms that
plays a bigger role in the small firms in the unorganised sector to get contracts. Although,
in the new age of advanced telecommunication and better transportation services, it may
be easy to have a long distance inter-firm relationship of this kind, the chances of a contractual
relationship between a larger firm and a smaller one that are in proximity would be in any
case higher.

5.2.6.3 The measures of agglomeration included in the study have already been
discussed in section 3 above. These are own-industry concentration measured by location
quotient, industrial diversity taken as the Herfindahl Index for unorganised and organised
manufacturing sector.

5.2.7 Industry Dummies:

5.2.7.1 The concentration of industries differs across states as well as industries, thus,
it is necessary to see to what extent the inter-relationship between agglomeration and
subcontracting varies across industries. For this, industry dummies at two digit NIC level
have been interacted with the corresponding district level agglomeration variables. The
coefficients on these dummies would indicate the extent of the effect that industry
concentration/diverstiy has on the subcontractual status of the unorganised manufacturing
firm across industries.

5.2.8 Time factor

5.2.8.1 In a pooled cross section, the effect of time on the dependent variable can be
gauged through time dummies. In order to look at the effect of any variable overtime, an
interaction variable between time dummy and the concerned variable would be included in
the estimation. Refer to equation 1 below that shows the empirical equation with all the
variables explained above estimated in the study.

SUBCitd  f ( LgOCit , RGVA, KIit , EXPit , STAGit , CONit , ACCit , HOUSEit , PERMit ,
LQdt , ORGdt ,UNORGdt , LOGLPit , Year05i , Year10i , Industryt , DISTT )      (1)

In equation 1, SUBC is the binary variable for contractual status of the firm, LgOC is log of
real operating costs, RGVA is the log of real GVA, KI is capital intensity, EXP, STAG and
11 2 The Journal of Industrial Statistics, Vol. 5, No. 2

CON refer to dummies for expanding, stagnant and contracting firms in the last three years,
ACC is the dummy for firms maintaining contracts, HOUSE and PERM are the respective
dummies for the location of the premises of the firm, LQ is location quotient for the
unorganised manufacturing sector at district- NIC 3-digit level and ORG and UNORG is the
industrial diversity measured as the Herfindahl index for organised and unorganised
manufacturing industries, respectively.

5.2.8.2 LOGLP is log of labour productivity at the firm level. The unorganized
manufacturing firms are mostly labour intensive, therefore, the productivity of labour is an
important factor. Including an endogenous variable in the probit estimation of equation 1
above may lead to biased coefficients. On the other hand, omitting a variable from the
equation may create a specification bias. Although, in literature, the omitted variable bias in
probit or logit regression is not observed to be affecting the coefficients of other independent
variables in the regression equation, our study includes this variable.

5.2.8.3 Year05 and Year10 are the respective dummies for years 2005 and 2010, with year
2000 as the baseline category. District dummies (DISTT) have been included to control for
the time invariant district fixed effects like infrastructure and government policies such as
the SEZs that may affect contracting. The industry fixed effects are controlled for by the
NIC-2 digit industry dummies (Industry).

6 Estimation Results

6.1 The probit regression estimation results are given in Table 3. Probit estimations
have been undertaken for all three years on a pooled cross section data. Model 1 includes
all firm specific variables, agglomeration variables and individual industry dummies. Model
2 includes individual industry dummies interacted with location quotient to capture industry
specific effects of agglomeration. Model 3 includes industry dummies interacted with the
organised sector diversity variable, while Model 4 includes interacted industry dummies
with unorganised manufacturing sector diversity. The different models have been chosen
to see how individually the three measures of agglomeration, namely, own-industry
agglomeration measured by location quotient and organised and unorganised manufacturing
sector industrial diversity have an effect on subcontracting probability for different
industries.

6.2 In pooled time series cross section where the sample in the cross section is not the
same across the years, the effect of time is estimated by including time dummies in the
regression equation. The year dummies for 2005 and 2010 have been included in the regression
equations, with year 2000 as the base year.

6.3 From the empirical estimations, we can see that own industry concentration
measured by location quotient (LQ) of the unorganised manufacturing sector has an overall
negative coefficient, except models 3 and 4 where the coefficient is positive but statistically
insignificant. The weak overall effect of LQ can be explained by the fact that the effect of LQ
across industries varies significantly, although the overall effect is negative. This shows
that the likelihood of a firm getting a contract in an industrially agglomerated region is
different across different industries.
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 11 3

6.4 However, an overall negative coefficient indicates that irrespective of the inter-
industry differences there is less likelihood of manufacturing firms to be operating under a
contract that are located in areas with greater agglomeration of own industry unorganised
sector industries.

6.5 Agglomeration Effect: Across Industries

6.5.1 Effect of Location Quotient across Industries

6.5.1.1 The effect of own industry agglomeration of unorganised sector firms has a positive
and significant coefficient for most of the industries. These include, in order of statistical
significance, tobacco industry (NIC Div. 12), textiles (NIC Div. 13), wearing apparel (NIC
Div. 14), motor vehicles (NIC Div. 29), furniture industry (NIC Div. 31), among others. This
implies that the firms in these industries that are situated in an agglomerated area have a
significant likelihood that they may be working under a contract.

6.5.1.2 On the other hand, negative coefficients are found for relatively few industries
such as coke and refined petroleum (NIC Div. 19), food products industry (NIC Div. 10),
beverage industry (NIC Div. 11), paper and printing (NIC Div. 17 and 18) and repair and
installation of machinery (NIC Div. 33), among others. The likelihood of the firms in these
industries is not affected by the economies arising out of the industrially agglomerated
area they are located in.

6.5.1.3 This is an important result as agglomeration cannot be construed to be creating

a similar advantage to unorganised manufacturing firms across various industries in terms
of acquiring. There is significant heterogeneity across industries and therefore, the effects
of agglomeration may not be uniform for all.

6.5.2 Effect of Industrial Diversity in Organised Sector across Industries

6.5.2.1 Industries with significant negative coefficient include food products industry
(NIC Div. 10), beverage industry (NIC 11), wearing apparel industry (NIC 14), rubber and
plastic industry (NIC 16), coke and petroleum industry (NIC Div. 19), chemicals (NIC 20)
and pharmaceuticals (NIC 21). A negative coefficient is indicative of a positive effect of
industrial diversity in organised manufacturing sector on the probability of smaller firms in
unorganised manufacturing in getting contracts. This is because a rise in the value of
Herfindahl index implies greater concentration than diversity of a particular industry in the
region. Thus, a negative coefficient implies that as the region becomes more industrially
diverse, the greater is the probability of unorganised sector firms to be under a subcontract.

6.5.2.2 However, most industries have a positive coefficient such as tobacco industry
(NIC Div. 12), textiles industry (NIC 13) and leather industry (NIC 15), paper industry (NIC
17), printing industry (NIC 18), transport industry (NIC 29), among others. For these
industries subcontracting probability is higher when they are located in a region that is less
diverse industrially.

6.5.3 Effect of Unorganised Sector Concentration across Industries

6.5.3.1 Apart from own industry concentration measured by location quotient, the
industrial diversity within the unorganised manufacturing sector in a region is also an
11 4 The Journal of Industrial Statistics, Vol. 5, No. 2

important factor that may be affecting the probability of smaller firms to be working under
a contract.

6.5.3.2 The empirical results of model 4 in Table 3 pertain to the interaction of industry
dummies with the unorganised manufacturing sector diversity measured by Herfindahl
index. The overall effect is seen to be negative and statistically significant. However, the
coefficient across many industries is positive implying that industrial concentration, rather
than diversity is affecting the subcontracting probability more in some industries. These
industries are tobacco, textiles, leather, paper, printing, rubber and plastic, basic metals,
fabricated metal products, electrical and electronic, transport, furniture and other
manufacturing. The knowledge spillovers enabled by inter-industry linkages that is
associated with industrial diversity does not have a discernible effect on the probability of
small firms to be operating under a contract.

6.5.3.3 The weak effects of industrial diversity on firms’ likelihood of working under a
contract, compared with the results of a positive relation of location quotient on the same
are indicative of strong localisation economies at work in the unorganised manufacturing
sector. A contradictory result on the localisation economy effects prevailing in India’s
organised manufacturing sector has been estimated by Lall et al. (2003). They found a
significant cost saving effects of industrial diversity using plant level data for 1998-99 from
the Annual Survey of Industries (ASI).

6.6 Firm Specific Variables

6.6.1 A relatively more capital intensive firm has a higher likelihood of operating under
a contract as the coefficient of capital intensity is positive and significant. It is known fact
that the small firms in the unorganised manufacturing sector are plagued by technological
backwardness. Therefore, larger firms’ decision to contract out a part of their production
process or input requirement would prefer to approach firms within the region that are
technologically more advanced. Also, a firm with its own premises, whether operating in a
house or having a fixed and permanent location outside the house, has greater chances of
working under a contract, compared to a firm that does not have any fixed or permanent
premise, the baseline category of firms.

6.6.2 The size of the firm in terms of the real gross value added (GVA) has a significant
positive coefficient in all models. A larger firm has a higher likelihood of getting a contract.
Also, maintaining accounts increases the probability of fetching a contract for an
unorganised manufacturing sector firm.

6.6.3 Productivity of labour employed in the firm has a negative but statistically
insignificant effect on the probability of getting a contract. The operating cost of the firm,
on the other hand, has a positive and significant coefficient. This result does not tally with
the expected outcome. It was explained that a higher cost firm may not be very likely to be
working under a contract because it would be costly for the larger firm to subcontract to a
firm that has a higher cost. However, the variable included in the analysis contains charges
for electricity, fuel and lubricant, raw material for own construction etc. It does not contain
wages and salaries to employees or raw material expense. A small firm using electricity may
have a high operating cost, however, it may be beneficial for the productivity of the firm,
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 11 5

that may partially explain the positive and significant coefficient for the operating cost
variable.

6.6.4 A firm which in the owner’s opinion has expanded, remained stagnant or contracted
in the last 3 years of its operation does not have a likelihood of being under a contract. The
coefficients for all three dummy variables are negative and in some cases, statistically
significant.

6.7 Time Effect

6.7.1 The overall probability of a firm getting a contract has been going down over the
years, as shown in the negative and significant coefficients of time dummies in all models.
One explanation can be that the number of firms that have been working under a subcontract
has declined over the years.

7. Conclusion

7.1 This study has empirically established that the inter-relation between the firm
level phenomena of subcontracting is affected by the spatial phenomenon of industrial
agglomeration. Using the NSS unorganised manufacturing enterprise surveys for the years
2000-01, 2005-06 and 2010-11, the study has estimated the effect of various agglomeration
economies on the probability of a firm working under a contract. Since the variable denoting
the status of the firm is a binary one, the estimation strategy used in the study is a probit
regression.

7.2 As discussed in theory, the proximity of input suppliers to the larger firms enhances
the likelihood of more subcontracting taking place. However, the empirical analysis gives
mixed results. Districts with high own industry agglomeration (as captured by the location
quotients) has a positive effect on the likelihood of a firm receiving a contract for a large
number of industries. The agglomeration economies arising out of industrial diversity of
organised and unorganised manufacturing sector, however, seem to be less dominant in
getting the smaller firms to work under a contract.

7.3 The overall likelihood of firms in getting contracts has fallen overtime, everything
else being the same. This may be indicative of the fact that in the recent times the percentage
of small firms working under a contract has come down since 2000. Consequently, for a
number of industries such as tobacco industry, textiles industry, leather industry, paper
industry, printing industry, transport industry, coke and refined petroleum industry there is
a negative effect of organised manufacturing industry diversity on the likelihood of a
subcontract. That is, the presence of larger firms in different industries does not seem to be
enhancing the linkage between large and small industries. This result does not imply that
lack of presence of larger firms will benefit smaller firms in terms of getting contracts. It only
means that localisation economies are stronger than those generated from industrial
diversity.

7.4 One of the reasons behind the fading linkage between large and small firms is the
inability of industries in unorganised manufacturing sector to upgrade their technology to
match the quality requirements of the buyers in the organised part of the sector (Ramaswamy,
2013). Our study finds that the firms that are capital intensive by nature are more likely to be
11 6 The Journal of Industrial Statistics, Vol. 5, No. 2

operating under a contract from larger firms. This indicates the potential of the unorganised
manufacturing sector which is capable of having a strong linkage with larger firms through
subcontracting. However, the inability of the sector to cope up with technology and other
requirements limits this phenomenon.

7.5 Subcontracting in unorganised manufacturing sector in India is an important

phenomenon and has a lot of potential to enhance the performance of small firms in the
unorganised manufacturing sector in India. Given that the sector employs about 85 per
cent of total employment in manufacturing sector, it is all the more important to ensure that
this sector should be made as viable as possible.

7.6 Subcontracting has been taking place mainly in the highly unskilled part of the
sector presumably due to the lack of an appropriate channel for larger firms to outsource
their activities to the better performing firms in the unorganised manufacturing.
Subcontracting exchanges that are in operation have not performed efficiently to help
larger firms to get access to smaller firms through them (Sahu, 2007). This calls for
maintenance of a database of a large number of small firms operating in the unorganised
manufacturing sector, with information on their financial health and other relevant information
made available. This information can help larger firms to forge access to the small firms
through these exchanges.

7.7 In order to encourage greater subcontracting and also allow smaller firms to gain
access to institutional funds the financial institutions can provide incentives to small firms
which are linked to larger firms through a contract to get cheaper credit from them. This
would enable small firms to gain access to formal credit and also bind them to perform well.
Secondly, being linked to a larger firm, they also benefit from knowledge spillover effects in
terms of technology upgradation and efficient production techniques, which in turn would
ensure that quality standards are maintained.

7.8 The recent initiative of launching the Micro Units Development and Refinance
Agency (MUDRA) Bank by the Government of India is a step in the direction of including
the micro, small and medium firms (MSMEs) in the gamut of institutional borrowing. The
bank would re-finance the loans up to Rs. 10 lakhs made available to MSMEs through the
scheduled commercial banks, regional rural banks, non-bank financing institutions,
cooperative banks and micro finance institutions. First time and young entrepreneurs and
women entrepreneurs would be encouraged through special schemes designed to suit
their needs.

7.9 The government should be cognisant to inter-relatedness of the phenomenon of

subcontracting and industrial agglomeration while formulating policies affecting industrial
agglomeration in order to maintain and/or enhance the linkage between organised and
unorganised manufacturing sector.
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 11 7

Reference:

Bairagya, Indrajit 2010, ‘Liberalization, Informal Sector and Formal-Informal Sectors’

Relationship: A Study of India’, Paper Prepared for the 31st General Conference of The
International Association for Research in Income and Wealth St. Gallen, Switzerland, August
22-28, 2010.

Bhushan, Bharat 1984, ‘Ancillarisation: A New Production Strategy’, Business India, New
Delhi, January 16-29, 1984.

Chinitz, B. 1961, ‘Contrasts in Agglomeration: New York and Pittsburgh’, American Economic
Review, 51, pp. 279-289.

Duranton, Gilles, and H. G. Overman 2005, ‘Testing for Localization Using Micro-geographic
Data’, The Review of Economic Studies 72(4), pp. 1077–1106.

Fujita, Masahisa and Paul Krugman 2004, ‘The New Economic Geography: Past, Present
and the Future’, Papers in Regional Science, Springer, 83(1), pp. 139-164.

Holl, Adelheid 2008, ‘Production Subcontracting and Location’, Regional Science and
Urban Economics, 38, pp. 299–309.

International Labour Organisation. ILO’s Pilot Action Project for Beedi Women Workers in
India. [INT/02/M57/NET]. Available: http://www.ilo.org/wcmsp5/groups/public/@asia/@ro-
bangkok/@sro-new_delhi/documents/projectdocumentation/wcms_125467.pdf (accessed
on 6th February, 2016).

Isaksen A 1996, ‘Towards Increased Regional Specialisation? The Quantitative Importance

of New Industrial Spaces in Norway, 1970-1990’, Norsk Geografisk Tidsskrift- Norwegian
Journal of Geography, 50, pp. 113-123.

Jacobs, J. 1969, The Economy of Cities. New York: Vintage.

Krugman, Paul 1991, ‘Increasing Returns and Economic Geography’, Journal of Political
Economy, 99, pp. 483–499.

Lall, Somik. V., Jun Koo and Sanjoy Chakravorty 2003, Diversity Matters: The Economic
Geography of Industry Location in India (Vol. 3072). World Bank Publications.

Lall, Somik V. and Sanjoy Chakravorty 2005, ‘Industrial Location and Spatial Inequality:
Theory and Evidence from India’, Review of Development Economics, 9(1), pp. 47-68.

Malmberg A, Maskell P. 2002, ‘The Elusive Concept of Localization Economies: Towards a

Knowledge-Based Theory of Spatial Clustering’, Environment and Planning A, 34(3), pp.
429 – 449.
11 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Marshall, A., 1890, Principles of Economics, 8th ed. Macmillan, London. Published in 1920.

Miller, P., R Botham, G Gibson and B Moore 2001, ‘Business Clusters in the UK- A First
Assessment’, Report for the Department of Trade and Industry by a consortium led by
Trends Business Research. London: TSO, 238pp. 2001.

Moreno-Monroy, Ana I., Janneke Pieters, and Abdul A. Erumban 2012, ‘Subcontracting
and the Size and Composition of the Informal Sector: Evidence from Indian Manufacturing’,
IZA Working Paper No. 6785.

Nagaraj, R 1984, ‘Subcontracting in Indian Manufacturing Industries: Analysis, Evidence

and Issues’, Economic and Political Weekly, August 1984.

North, Douglass C. 1955, ‘Location Theory and Regional Economic Growth’, Journal of
Political Economy, 63(3), pp. 243-258.

NSSO, Unorganised Manufacturing Sector Survey, 56th Round. 2000-01.

NSSO, Unorganised Manufacturing Sector Survey, 62nd Round. 2005-06.

NSSO, Unorganised Manufacturing Sector Survey, 67th Round. 2010-11.

O’Donoghue, D., and B. Gleave 2004, ‘A Note on Methods for Measuring Industrial
Agglomeration’, Regional Studies, 38(4), pp. 419–427.

Ono, Yukako 2001, ‘Outsourcing business service and the scope of local markets’, Working
Paper Series, WP-01-09, Federal Reserve Bank of Chicago.

Ono, Yukako 2007, ‘Outsourcing Business Services and the Scope of Local Markets’,
Regional Science and Urban Economics, 37(2), pp. 220-238.

Pardo, Ángel Alañón- & Josep-Maria Arauzo-Carod 2011, ‘Agglomeration, Accessibility

and Industrial Location: Evidence from Spanish Municipalities’, Documentos de trabajo
de la Facultad de Ciencias Económicas y Empresariales 11-02, Universidad Complutense
de Madrid, Facultad de Ciencias Económicas y Empresariales.

Rama, R., Ferguson, D., Melero, A. 2003, ‘Subcontracting Networks in Industrial Districts:
the Electronics Industries of Madrid’, Regional Studies, 37(1), pp. 71–88.

Ramaswamy, K.V. 2013, ‘Understanding the ‘Missing Middle’ in Indian Manufacturing:

The Role of Size-dependent Labour Regulations and Fiscal Incentives’, Issue 480 of V.R.F.
Series, Institute of Development Economics, Japan External Trade Organisation.

Ramaswamy. E A. 1988, Worker Consciousness and Trade Union Response, Oxford

University Press. New Delhi.
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 11 9

Sahu, Partha Pratim. 2007, ‘Subcontracting in India’s Small Manufacturing Enterprises:

Problems and Prospects’, Institute for Studies in Industrial Development, Working Paper
2007/01.

Stigler, George J. 1951, ‘The Division of Labour Is Limited by the Extent of the Market’,
Journal of Political Economy, 59(2).

Suarez-Villa, L., Rama, R. 1996, ‘Outsourcing, R&D and the Pattern of Intra-Metropolitan-
Location: the electronics industries of Madrid’, Urban Studies 33(7), pp. 1155–1197.

Taymaz, E., Kilicaslan, Y. 2005, ‘Determinants of Subcontracting and Regional Development:

An Empirical Study on Turkish Textile and Engineering Industries’, Regional Studies,
39(5), pp. 633–645.

Venables, A. J. 1996, ‘Equilibrium Locations of Vertically Linked Industries’, International

Economic Review, 37(2), pp. 341-359.
12 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 1: Trend in Real Gross Value Added and Labour Productivity in Unorganised
Manufacturing Sector
Real GVA Compounded Labour Productivity
Employment
NSS Rounds Year Annual Growth (Compounded Annual
(in Million)
(Rs. in Crores) Rate (in per cent) GR in per cent)
st
51 1994-95 47,795 -- 33.2 14396
th 19051
56 2000-01 70,642 6.7 37.08
(4.8 per cent)
23470
62nd 2005-06 85,533 3.8 36.4
(3.5 per cent)
34087
67th 2010-11 118,924 6.8 34.8
(6.4 per cent)
Source: NSS Reports on Enterprise Surveys in Unorganised Manufacturing Sector

Table 2: Labour Productivity in Unorganised Sector at NIC 2-Digit Level

NIC 2 Digit Industry name 2000-01 2005-06 2010-11
10 Food products 18,388 23,235 36,510
11 Beverages 15,508 16,164 26,017
12 Tobacco Products 8,222 6,650 10,091
13 Textiles 16,469 18,911 29,182
14 Wearing Apparel 19,494 18,575 29,681
15 Leather and Leather Products 21,848 26,946 37,362
16 Wood an d Products 12,443 13,148 26,930
17 Paper and Products 24,396 20,095 45,202
Printing and reproduction of Recorded
18 Media 37,950 42,408 67,442
19 Coke and Petroleum Products 28,469 35,628 48,739
20 Chemicals 20,264 17,294 32,803
21 Pharmaceuticals 14,156 53,542 71,949
22 Rubber and Plastics 44,466 49,498 49,456
23 Other Non-metallic Products 18,881 28,150 34,040
24 Basic Metals 54,231 100,629 60,329
25 Fabricated Metal Products 28,409 40,154 61,064
26 Computer, electronic and optical prd 50,071 82,251 73,843
27 Electrical equipment 47,588 62,098 63,205
28 Machinery & Equipment nec 39,827 70,513 84,476
29 Motor vehicles, trailers and semi-trailers 51,015 57,648 92,459
30 Other transport equipment 55,303 47,314 75,913
31 Furniture 28,229 32,928 49,001
32 Other Manufacturing 27,680 42,198 39,769
Repair and installation of machinery and
33 equipment 27,555 31,660 50,416
01632 Cotton ginning, cleaning and bailing 16,009 45,015 35,301
 Subcontracting and Industrial Agglomeration: Related Phenomena in India’s ..... 12 1

Table 3: Estimation Results for Probit Regression

Dep. Variable: Contract Dummy
Independent Variable
Model 1 Model 2 Model 3 Model 4
-0.0001 -0.008 0.0009* 0.0007*
Location Quotient
(-0.09) (-1.26) (3.04) (2.4)
-0.49 -0.36* -0.08 -2.11 #
Unorganised Sector Diversity
(-1.41) (-3.34) (0.77) (-1.67)
-0.47* 0.04 -1.91 0.04
Organised Sector Diversity
(-2.99) (0.66) (-1.92) (0.56)
-0.03 0.04* -0.004 -0.01
Labour Productivity
(-0.69) (3.68) (-0.32) (-0.96)
0.15* 0.10* 0.12* 0.12*
Operating Cost
(7.11) (15.4) (18.3) (18.2)
0.06* 0.03* 0.03* 0.03*
Capital Intensity
(4.25) (6.8) (6.3) (6.3)
0.04 -0.04 # -0.04 # -0.03
Expanding
(0.62) (-1.86) (-1.87) (-1.23)
-0.13* -0.16* -0.16* -0.15*
Stagnant
(-2.15) (-7.7) (-7.8) (-7.06)
-0.036 -0.14* -0.15* -0.14*
Contracting
(-0.44) (-5.5) (-6.2) (-5.4)
0.225* 0.26* 0.21* 0.16*
Location: House
(3.05) (7.7) (6.17) (4.8)
0.16* 0.34* 0.23* 0.19*
Location: Fixed & Permanent
(2.05) (10.3) (6.9) (5.6)
0.16* 0.05* 0.09* 0.09*
Log Real GVA
(5.51) (5.4) (9.7) (9.9)
-0.081 0.03 0.025 0.008
Account Dummy
(-1.38) (1.38) (1.16) (0.38)
-2.9 -2.5 -2.3 -2.2
Constant
(-4.5) (-20.1) (-18.3) (-17.1)

-0.05 -0.04* -0.05* -0.04*

Year2005
(-0.95) (-2.3) (-2.9) (-2.2)
-0.91* -0.85* -0.85* -0.86*
Year2010
(-15.28) (-35.8) (-35.8) (-34.8)

District Dummies YES YES NO YES

Individual Industry Dummies YES -- -- --
LQ Interacted Dummies -- YES -- --
Organised Sector Interacted Dummies -- -- YES
Unorganised Sector Interacted
Dummies -- -- -- YES

No. of Observations 120765 117355 125492 117823

Pseudo R-Squared 0.28 0.19 0.12 0.198
Note: Symbols *, **, # denote levels of significance at 1%, 5% and 10%, respectively .
Parentheses include the corresponding z-statistic.
12 2 The Journal of Industrial Statistics (2016), 5 (2), 122 - 137

Understanding Demand, Supply and Price Behavior in the Dairy

Sector: Using Official Indian Statistics
Nilabja Ghosh1, Institute of Economic Growth, New Delhi, India
Roopal Jyoti Singh, Institute of Economic Growth, New Delhi, India
M. Rajeshwor, Institute of Economic Growth, New Delhi, India

Abstract

Balancing the pressures of WTO compliance with direct and derived consumer demands
and the interests of the producers is a looming challenge in India’s Milk economy. After
being driven by a successful cooperative movement for decades, the sector needs a review.
In an attempt towards understanding the milk market this paper identifies different data
sources along with data gaps to trace the market movements methodically in the recent
two decades.

1. Introduction

1.1 Milk today is not just a consumer item. It is an exportable item in a competitive
world and, with the rise of the organized food processing industry it is also becoming an
industrial input. A very large number of agents such as farmers, vendors, traders, processors,
cooperative workers, manufacturing companies, exporters and consumers depend on the
dynamics of the sector. The challenges on the demand and supply sides are compounded
by the pressure mounted on India to open up the dairy market as part of India’s obligation
towards WTO.

1.2 To monitor the dairy sector high quality and reliable data will be needed for broader
public view. Attributable perhaps to this inadequacy a lack of statistics-based holistic
focus on the structure of the dairy economy is observed. We filter out relevant data from
different secondary official data sources to examine the directions of milk supply, of the
different components of its disposition and to assess the movement of price and the
supply-demand gap. Conducted at the all India level for the post liberalization period 1990-
91 to 2010-11, the study keeps validation and consistency of estimates in context throughout
the analysis.

2. Background to Indian dairy sector

2.1 Milk accounts for 9% of India’s GDP but an estimated 70 million households many
of whom are small farmers (Birthal et al., 2008) earn from livestock ownership, 69% of
workforce being women (GOI, 2013). Despite being nutritive and a part of the regular diet of
Indian people, milk did not get the recognition in policies for India’s food security in the
way cereals did. Little effort was made to ensure that the production met the demand of the
consumers, let alone understand the possibilities of expanding the market2. Seen commonly

1
e-mail: [email protected]
2
Public initiatives such as the Milk Control Board and the Operation Flood (OF) can however be
compared with the public distribution system and the green revolution respectively,
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 12 3

as a useful tool for poverty alleviation, dairy remained largely an individualistic, unorganized
and subsistence activity.
2.2 After independence India had a large number of dairy animals but produced a
meagre 17 million tonnes of milk owing to poor animal health, low quality forage and humid
and hot conditions in many parts of the country (Raj, 1969). The National Dairy Development
Board (NDDB) was founded at Anand in 1965 by a parliamentary Act, focusing on the
grass-root level producers with the principles of cooperation. By 2001 India became the
leading milk producer in the world but even this was the beginning of yet another journey.
2.3 Today, visible shifts in consumer’s expenditure pattern towards milk products3,
the appalling status of malnutrition among Indian children (Arnold et. al, 2009) and a close
nexus between milk price and food price inflation (milk has a weight of 3.24% in wholesale
price index in the 2004-05 series) are empirical findings that link milk closely with consumer
welfare. On the other hand, farmers with little or no land look for an expanding market for
livelihood and incentive. In view of the significance of welfare of the diverse stake-holders
of the dairy sector, the continual rise in the price of milk in recent times interspersed with
intimidating slumps in the international market makes understanding the market necessary.
Cooperative functioning
2.4 The World Bank-financed Operation Flood (OF) launched by NDDB in the 1960s
was a landmark achievement in cooperation that transpired on Indian territory. Although
many milk cooperative4 predated OF the government’s active participation in the process
was a turning point. Every cooperative in Asia has tried to emulate this success story of
Amul. Though other cooperatives too made a mark5, by and large, the cooperative movement
in totality actually failed in most regions of India and large sections of households and
food providers even in the urban sector remain to be supplied by informal traders and milk
vendors.
2.5 NDDB addressed the milk producers’ access to technology, feed and facilities at
door step or proximity. Technological progress was mainly manifested in planned breeding
through artificial insemination (AI) with selected genetic features acquired from other
countries by cross-breeding so that the share of indigenous cattle diminished in India over
time. NDDB also made access to veterinary facilities easier.
2.6 Grazing animals on pastures was a traditional practice but grazing on crop lands
after harvest serves the dual purpose of clearing and manuring land for the next sowing
while also nourishing them. Dry crop residues from cereal and sugarcane fields are less
nutritious but they are common as fodder while cultivation of green fodder or forage crops6
is a superior means, rare but promoted by NDDB.A switch from extensive to an ‘intensive
feeding regime’ in which animals are fed in their stalls with commercially prepared nutritious
concentrates made from grains and oilseed meals is however emerging (World Bank, 2008).
3
According to NDDB chairman Amrita Patel, rising incomes have led to a shift away from cereals to
vegetables, milk and meat (Srivastava, 2011).
4
The Kaira District Cooperative in Kheda district of Gujarat in 1946 is an example of early cooperative
venture.
5
Brands like Vijaya (Andhra Pradesh), Verka (Punjab) and Saras (Rajasthan) are examples.
6
Guinea grass, paragrass, lucerne, berseem, cowpea, velvet bean are nutritious forage crops.
12 4 The Journal of Industrial Statistics, Vol. 5, No. 2

Liberalization and the emerging Market channels

2.7 While the OF was certainly a mile stone, the liberalization of the Indian economy
in the 1990s introduced India’s dairy sector to another new era. De-licensing in 1991 and
1992 and signing of the WTO agreement in 1995 gradually opened up the dairy sector to
private capital and foreign markets7. Exports of dairy products began to grow from a miniscule
level in the 2000s, creating pressure on the domestic capacity even while imports of cheaper
milk persisted as a threat to the viability of small milk producers. The government, yet
unprepared for the conflicting pressures responded with trade controls (occasional levy,
invocation of SPS measures8, sometimes outright bans) that created discontent among
exporters and processors and even perhaps deprivation to farmers. While trade partners
especially EU clamor for access at the multi-lateral level, Free trade Agreements (FTA) with
individual partner countries will not make managing the economy easier (Sengupta and
Gautam, 2011).
2.8 The National Dairy Plan (NDP) made in 2012 and the National Livestock Policy in
2013 in consistency with the National Policy of Farmers of 2007 were responses of the
union government to create a framework for a sustainable livestock economy underlining a
departure in approach from intense cross-breeding. The new government, in particular is
reported to be placing considerable emphasis on the upgradation of indigenous ‘desi’
breeds through a national Gokul Mission.
2.9 With the launch of economic liberalization, besides the state-promoted
cooperatives, profit oriented, private, organized and more resourceful firms of both domestic
and multinational ownership have also shown interest in milk. Even contracting, despite all
the ideological and historical contention it encounters is already seen as a promising
channel (Kolekar et al., 2012). Not surprisingly, even though cooperation, seen as an ‘ethical
social behaviour’ (Abbott, 2014) proved itself to be a successful model worldwide (Chahal
and Gupta, 1986, Ibitoye 2012), reappraisal of the NDDB’s role has been proposed (Kapoor,
2014) at different levels.
Sources of Demand for Liquid Milk
2.10 Tastes vary within India but with globalization, dairy products have gained greater
significance in food habits9 across the spectrum. The response to the changing domestic
and international demand was also manifested in the rise of food processing in the organized
sector which grew by a massive 378% between the years 2001-02 to 2010-11 against a rise
of wholesale price index of all commodities by 66.4% over the same span.
2.11 Milk is highly perishable and so the net available amount falls short of the
production. A large part of the fresh milk produced is used for self-consumption of producers,
for informal sale in local markets and delivery at door-steps in rural and urban neighborhoods
7
The government was limited by a commitment of zero tariff binding made in the WTO agreement. The
tariff commitments on milk powder were re-negotiated in 2001-02 when Quantitative Restrictions on
imports were removed. The Milk and Milk Product order (MMPO) imposed in 1992 was amended in
1999 to allow flexibility to the processors to adjust their scales of production.
8
Agreement on the Application of Sanitary and Phytosanitary Measures (SPS Agreement) subjects WTO
members to base their methodologies on three recognized standards (Wikipedia)
9
Milk is seen as a ‘high value’ (HV) product whose expenditure elasticity food is shown to be relatively
high (Ravi and Roy 2006)
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 12 5

in unprocessed form besides being marketed through the cooperatives and dairy firms.
The price received by the farmers (farmer price or producer price) differentiates for the
quality and fat content and the data even at the cooperative level are not systematically
reported on public domain. It is indicated by The National Commission of farmers (NCF)
that the milk producers in the Anand Model of milk production get, net of intermediation,
about 60% of the final price but given the spatial and qualitative variations and with the
organized market and the unorganized sector both involved in milk marketing, this figure is
only indicative. A second estimate comes from Kulkarni (Kulnarni, 2014) which puts the
same figure at somewhere between 75% and 85% or at a median figure of 80%. Given the
retail price as officially reported the Kulkarni estimate indicates a higher producer price
than NCF estimate does. The matter remains unresolved.
2.12 Milk can be purchased in raw and liquid form (to be boiled at the domestic level)
or in pasteurized liquid form or it can also be purchased skimmed of all or part of the fat
content. A first stage processing for preservation of milk is the evaporation, spray drying
and similar techniques to produce milk powder or condensed milk more stable than the
liquid form. Further downstream, both milk powder and liquid milk can be processed into
dairy items like curd, yoghurt, cream, butter, cheese which in turn can be intermediate
inputs for other food processing activities such as bakery, confectionery. The different
constituents of milk such as the fat, the butter milk, the milk protein and casein also have
important uses in pharmaceutical and other sectors.
3. Data on the Milk market
3.1 Production of fresh milk is historically reported by Ministry of Agriculture,
Department of Animal Husbandry, Dairying & Fisheries (DAHDF)10.In this reporting
protocol, milk is treated as of uniform quality. Data on wastage in the supply chain are not
reported on a regular basis but quantitative post-harvest losses (PHL) were assessed by
CIPHET (2010) by a onetime survey. Also the PHL for milk was estimated only through
verbal enquiry and field observations and only at select stages11.While this source is by far
the most reliable one owing to the methodological rigour, implications of underestimation
cannot be ruled out.
3.2 Unlike crops that flow from land as the ‘capital’ stock’, output of milk flows from
live animals constituting livestock. Animals are not only valuable sources of farm household
nutrition but they also serve in land preparation and organic manuring and provide fuel,
haulage, transport and farm energy while also serving to stabilize incomes. The milch
animals are comprised of cattle and buffalo as also other smaller animals. Buffalo, considered
more productive than cattle in feed utilization (Khajarern and Khajarern, 1998), makes up
over 40% of the milch animals in India but around the world they are a minor source of milk.
The data on livestock by type of animal, sex, milch status and also by indigenous or cross-
bred classification are reported through the Livestock Census only at five years interval.

10
Collected from annual sample surveys conducted by the state Animal Husbandry Departments under the
‘Integrated Sample scheme’ in three seasons namely summer, monsoon and winter.
11
Not covered in the survey are operations of sorting and grading either because such operations are not
relevant for milk or required facilities and suitable methodologies were not at hand. Storage at godown,
warehouse and cold stores, storage at wholesale level are also not covered because the units were not
available in the specific districts surveyed.
12 6 The Journal of Industrial Statistics, Vol. 5, No. 2

3.3 Feed is the major constraint for milk production. Although animal products and
crop products are alternative human food, because the availability of dry fodder is associated
positively with the access to land and its cultivation with certain crops and that of higher
grade feed competes with crop land, indirectly milk production too depends on the
performance of the crop sector and the weather. Area under different cereal crops, their
production and wholesale prices are readily available from Ministry of Agriculture sources
(website) and so are corresponding statistics on sugarcane and oilseeds. The same Ministry
also provides Land use statistics with data on pastures. Aggregate estimate of the area
under fodder crops is also reported by MOA but detailed information is scarce.
3.4 Household consumption data are reported as value of milk and milk products by
National Sample Survey Office (NSSO, various) on per capita per month basis in its annual
(thin) and the quinquennial (large) surveys of consumer expenditure. Data reported by
NSSO are known to be reliable. Using population data from Census (1991,2001, 2011)
undertaken at decennial intervals and interpolated for the interim years using compound
annual average growth rate, the total consumption value of milk can be calculated for all
Indian households and for households of different types. Producers’ demand can be seen
as the consumption in agricultural households presumably comprising of NSSO reported
categories ‘self-employed in agriculture’ (SEA) and the ‘agricultural labour’ (AL) households
as they are both likely also to be milk producers. The NSSO however reports annually for
the composite milk consumed both in fresh and primary processed forms (dairy)12.
3.5 To bring the value of milk consumption to physical units it is necessary also to
have milk prices at which consumers buy. The wholesale price (WP) of milk is reported for
major markets or mandis by Agricultural Prices in India (MOAa, various)and these prices
are averaged to get all India prices. Price index of milk and dairy products at the all India
level is reported in the website maintained by Office of the Economic Adviser, Government
of India. However, the consumers buy milk at retail prices (RP) the data for which are not
continuous. Rural consumption raises further complications because milk is either farm
produced or purchased locally from producers13 and the rural market functions differently
from the urban market. The published retail prices can be collected from the same sources
as WP and averaged across months and centres to arrive at all India annual retail prices
(RP) but these prices possibly do not represent rural consumer prices.
3.6 Under such constraints consumer prices for different sections can be conjectured
only using their empirical relation to the secondary data on WP. In the years 2001-02 to
2011-12 for which both sets of data (RP and WP) are accessible monthly, we expressed the
WP as a proportion of the corresponding RP. While the proportion varies between 0.90 and
1.08, the average ratio is 0.967, which is used for projecting the retail prices for all the sample
years using the yearly wholesale prices as benchmarks. The retail price so derived is
assumed to be the price paid by urban consumers.
3.7 If representative estimates of prices paid by producers and other rural people
were available, the rural price can be obtained as an weighted average, the weights being
their shares in total rural households as available from NSSO data. The producer price can
12
NSSO includes consumption of milk and milk products: baby food, condensed/ powder milk, curd, ghee,
butter, ice-cream, other milk products.
13
The NSSO values the consumption at the actual paid cost or at producer prices.
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 12 7

be used as a proxy for the consumer price of farm households with the intuitive rationale
that these households are likely to forego this price when they do not sell the milk that they
retain for home consumption. However, a consistent data-base of the producer prices even
for the cooperative channel, is not in public domain. We have assumed the ratio to be 60%
of the corresponding RP going by the estimate given by NCF as stated earlier. We also take
into account the alternate estimate of the ratio at 80% based on Kulkarni’s estimated range
and produce a second set of estimates of the producer price.
3.8 What the non-farm rural consumers pay is a matter of conjecture. To the extent
that sales to these households are residual to cooperative sales and the cost of marketing
is lower, this consumer price will be lower than the producer price. On the other hand, faith
placed on the quality, the visibility or production and door-step services may translate to
higher value but, the fact that such sales are made locally on foot or by cycle and mostly
not in graded forms14 makes this less likely. In any case obviously there is a competition
between the cooperative channel that provides assured sales and prices and the less
demanding informal sales process in local markets so that the two prices are likely to
converge. We therefore assume that all rural households pay the same price equivalent to
the producer price.
3.9 The use of secondary available price data to derive the consumer prices is justified
and validated by matching the derived producer price and the retail price with imputed
consumer price actually obtained using NSSO’s quinqennial survey data published in
Agricultural Statistics at a Glance (MOAb, various). Consumer prices of milk imputed using
value and quantity of milk and dairy products consumed (NSSO data) in 2004-05 are Rs
12.22 and 16.30 for the rural and urban sectors respectively compared to the derived producer
price and retail price of fresh milk at Rs 10.42 and Rs 17.37 that we estimate using wholesale
price as the anchor (Table A1). Interestingly, the rural consumer price derived using Kulkarni
ratio is found to be closer to the rural price imputed using NSS data in both years 2004 and
2009 compared to NCF. The total value of milk consumption is then divided by the
corresponding consumer price to obtain the quantity consumed. To obtain per capita
consumption of different rural sections, population ratios calculated from NSS reports are
used as weights.
BOX 1: Specifications for Milk and Milk Product with Harmonized System Code (HSC)
Forms Specification as reported in DGCIS
Liquid Milk Milk & Cream Not Concentrated Nor Containing Added Sugar or Other
(HSC-0401) Sweetening Matter
Milk Powder Milk & Cream Concentrated/Containing Sugar/Sweetening Matter
(HSC-0402)
Dairy Products
Butter Milk Butter Milk, Curdled Milk & Cream, Yogurt, Khir& Other Fermented
(HSC-0403) Acidified Milk & Cream
Whey and Products Whey & Products Consisting of Natural Milk Constituent Not
(HSC-0404) Containing Added Sugar or Sweetening Matter.
Butter and Fat Butter and Other Fats & Oils Derived from Milk; Dairy Spreads
(HSC-0405)
Cheese And Curd Cheese and Curd

14
In organized milk marketing technical methods are employed (such as the 2-axis formula and the
use of the lactometer) to maintain compliance.
12 8 The Journal of Industrial Statistics, Vol. 5, No. 2

3.10 Liquid milk occupies a marginal place in international trade. Powdered milk is
exported both as whole milk powder (WMP) and as low fat skimmed milk powder (SMP).
Trade data is reported by Directorate General of Commercial Intelligence (DGCIS) in terms
of components by HS codes15.FAO also in its website FAOSTAT reports India’s exports
and imports of liquid and powdered milk further disaggregated as skimmed and whole milk.
We use DGCIS data on both milk and dairy food products as follows.
3.11 We make a crude estimate of the trade in pure milk interpreted to be either liquid or
powdered milk specified in Box 1 in liquid milk equivalence, though these component forms
vary in fat content. Using their proportions in trade as reported by FAO, the respective
components of skimmed milk powder and whole milk powder are expressed as liquid milk
equivalent with no distinction employing conversion rates16 obtained from industry sources
and added up. The dairy products also use milk as an input but need to be treated differently.
To obtain movements of total milk import or export in real values, their total value is expressed
as index and deflated by a composite milk price index which again is the weighted average
of wholesale price indices milk and dairy products with the same base. Admittedly,
generalizations are compulsions given the data reporting system and the heterogeneity of
milk and its products.
3.12 The Annual Survey of Industries (ASI) conducted by the CSO generates data on
the amount of fresh milk processed annually covering only the factory sector. The informal
sector (using less than 10 employees or no power) is covered by the NSSO in its Survey but
only at fairly long intervals17. The value of fresh milk consumed as industrial inputs for
dairy products (NIC 152) as well as all products together obtained from ASI is deflated by
the wholesale price of milk to obtain estimates of the quantities of milk processed in the
organized sector for different purposes. We use unit level data on ‘input consumed’ provided
by ASI for the organized sector in the decade starting 2001 and identify the item fresh milk
using the commodity code ASICC.
4. Results
4.1 In table 1 the production of milk is seen to have risen impressively in the recent
years from 54 to 128 million tonnes from 1990 to 2011. It is further expected to cross 140
million tonnes in 2014 (World Bank, 2008). The compound growth rate of total production
and net production accounting for wastage works out to 4.2%. The performance is reflected
in the productivity measured by production per hectare of crop area, per hectare of fodder
area and per hectare of area under all potential feed materials much of which however are
only by-products. Dedicated fodder area constitutes only 4% of India’s cropped area.
Animals are increasingly fed in the stalls and are also known to stray about for feeding.
Stable cereals and sugarcane acreages, declining pastures and highly volatile fodder area
taking a special downturn in drought years like 2002 are other major features of the change
in the feed scenario.
4.2 Production per capita and productivity per animal grew at lower rates. The
composition rather than the number of animals or the amount of feed produced seems to
stand out as a factor behind the production breakthrough as seen in Table 2 also. The
15
Harmonized System
16
100 liters of whole milk is equivalent to 9kg of SMP or 13 kg of WMP.
17
The data has also serious problems with the codes for identifying the items.
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 12 9

animal stock remained nearly stable in the study period especially after 1997 census. The
decline in the number of indigenous cattle after 1992 which is offset by the sharp growth of
the numerically minor group of cross bred animals and a slow growth of the number of
buffalos shifted the composition of animals steeply in favour of cross-bred cattle suggesting
that NDDB’s emphasis on improved breeding was successful (see also IUF, 2011).

4.3 The average Wholesale price, Rs7.72 per litre in 1990-91 and Rs 10.30 in 1993-94
crossed Rs 30 in 2011-12 recording a growth rate of 6.04% in the period 1993-94 to 2011-12
which is comparable, but a little less than the overall price rise (inflation in WPI) of 6.14%
(Table 3).Thus milk prices did not increase significantly relative to general price level
especially in the mid-1990s and in most part of 2000s until it caught up in 2009-10 (Figure 1).

4.4 At 3.9%, total consumption grew nearly at par with production (Table 4). The rural
consumer’s spending on milk and dairy products grew at a slightly higher rate than the
urban consumer, but the latter’s total physical consumption increased considerably faster
reflecting the urbanization of population. Between 1993-94 and 2011-12 the urban share in
India’s domestic milk and dairy products consumption went up from 26% to 30% but the
rural sector remained an important contributor (70%) to milk consumption.

4.5 Farm households make up 45% of total population in the country and about 60%
of the rural population. Using the assumptions given in earlier section we calculate that
about 42% of milk consumption in India is accounted for by farm households down from
45% in 2004-05 (appendix Table A2) and the SEA type households are larger milk consumers
than AL18.We have assumed uniform consumer price equal to the retail price in the urban
sector. Since not all the urban households are served by the organized sector (Birthal et al
2008)19 both the rural and the urban price may arguably be an over estimate. Based on NCF
estimate, probably considerably less than 30% of milk consumption (Figure 2) attributed to
urban sector passes the formal route. If the estimate from Kulkarni is accepted then rural
price is higher (Table 3) and rural physical consumption is considerably less in 2011-12 at 47
million tonnes compared to 63 million by NCF estimate and the rural share comes down to
63% from 69%. NDDB estimated 20% of production (128 Million Tonnes) i.e., 26 million
tonnes in 2011-12 passed the organized channel (NCF, website). This is comparable to the
less than 27 million tones of urban consumption for that year calculated using NCF-ratio.
NCF states that only 20% of the produced milk reaches the organized sector and of this
only a half passes the cooperatives. We estimate urban consumption to be 19% of
production. The IUF (2011) reports that 65% of milk is consumed in the unorganized sector,
including producers and vendors, compared to over 70% estimated using the NCF ratio
and 63% using Kulkarni ratio.

4.6 Milk is traded mostly as milk powder and baby food. Items like butter, cheese and
whey are other traded items. Since the import tariffs were renegotiated in 2001-02, India’s
trade in dairy goods went through a transition. Table 5 summarises trade movements
between the years 2001-02 and 2011-12. After rising to a peak in 2005-06, exports of milk
18
The AL and SEA households spend 5.3% and 9.9% of their expenditures respectively on milk and
milk products. They spend 58.4% on food against 51.8% spent by all households.
19
Direct transaction is common in rural and peri-urban areas. Restaurants, hotels, sweet shops and many
households are known to buy from informal vendors. The informal sector is the one that is not recognized
by the tax administration. Mutual trust and not rules and regulations guide their transactions.
13 0 The Journal of Industrial Statistics, Vol. 5, No. 2

powder and dairy product declined significantly. Betterment of technology and managerial
practices in competing countries and increased subsidies in EU are some of the reasons
cited in literature. Exports of most dairy products declined after a lag only while powder milk
took the major impact. Meanwhile imports of milk powder as also of all the dairy goods like
cheese and butter rose sharply. Some dairy products are both exported from and imported
in India at all times. Liquid milk export sustained increases but its net export actually
became negative.

4.7 Domestic industries process liquid fresh milk mostly to produce powder milk and
dairy products. Milk however enters as input in industries in various forms. To avoid over-
counting only ‘fresh’ milk, identified by commodity code ASICC, is considered to obtain
the extent of processing. All primary processed versions such as milk powder, pasteurized
milk and condense milk are excluded as inputs. Besides, bakery, confectionery and other
industries also use milk as an input but the dairy industry has a predominant role (Figure 3).
Both components of food processing rose in the middle of the decade and came down
thereafter. The extent of processing of fresh milk for producing milk powder and dairy
products has been about 11% and less than 1% for other products, but came down in 2010-
11. NDDB (2009) provides another set of estimates of milk processing as ‘butter, milk
powder and western type manufacture products’ and in the form of ‘traditional products’
without further elaboration, and their measures indicate that between 8% and 25% of fresh
milk is processed. The processed milk is both consumed in the domestic economy and
exported.

The Demand Supply Gap and Price behaviour

4.8 Data drawn from the different sources are explored to track demand-supply gap in
the country and the prices in the milk market. Given the issues we face in dealing with
composite data on milk and milk products and with complexities of conversion to milk
equivalence, the results are indicative at best. A reasonable assumption about the producer
price or consumption price of farmers is also made difficult by the absence of and variation
between alternate estimates of the retail price’s equation with the producer price. Moreover,
there is a large informal market both in rural and urban area for milk on which data is sparse
and can be collected only from primary survey. The prices in this market are likely to be
lower than the formal prices in the respective areas but the assertion cannot really be made
with conviction in the absence of evidences. Comparing our estimated total consumption
and net production a large gap is found. The gap would be larger if Kulkarni estimate is
used though it would narrow if possibly lower prices in informal markets are taken account
of.

4.9 Figures 4 and 5 show that the real price of milk (WP of milk deflated by WP all
commodities) increased only from 2008-09 bearing little if any relation to the estimated
surplus while the price rise is accompanied by falling exports, rising imports and decline of
milk processing. Indices plotted with a common base however suggest a fatigue in domestic
production that may not keep pace with domestic demand in future (Figure 6). In practice,
the surplus measured as derived consumption quantity less net production is attributable
to processing in formal manufacturing industries, informal enterprises and in the service
sector, or used for other purposes such as religion and storage as basic products or arising
from methodological flaws or even remain unexplained.
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 13 1

5. Conclusion

5.1 In context of the current changes in food demand within the country as well as the
opportunities and threats in the global markets and to harness the opportunity of value
addition and employment generation the milk economy needs to be monitored, encouraged
and guided with understanding.

5.2 To understand the milk market, data available in the public domain is explored. The
analysis of the price rise does not indicate that producers have gained substantially
subsequent to liberalization. The interests of milk producers and employment generation
may be compromised by lower demand from processors. Since processors may be holding
stocks in the form of milk powder, data on which are not available, their long run incentive
is also undermined creating a vicious cycle. On the contrary price rise creates discontent
among urban consumers and exporters too. The challenge of matching sporadic declines of
international milk price as happening currently creates policy challenges with little
preparation.

5.3 With globalization the milk market will have to be monitored continuously requiring
improved intelligence and analysis. Although much of the information is available publicly,
inadequacy and fusion of multiple information necessitate assumptions. Besides after
counting the varied sources of milk disposition reported there remains a gap between
supply and demand that cannot be explained without data on its use in the service sector
and the informal sector. More consistent estimates of producer prices of milk are required
for which primary surveys in different parts of the country have to be encouraged. It is
possible that the NCF estimated relation between retail and wholesale prices is an
underestimate while Kulkarni’s ratio seem more consistent. This shows the importance of
data reconciliation, coordination and strengthening of data collection.

References

Abbott, S. and Panwar, R. (2014). A Study into Marketing-cum-Processing Co-operative

societies in Punjab with Special Reference to Mullanpur Society. Indian Journal of
Economics and Development, Vol. 10 No. 3, 291-297.

Arnold, F., Parasuraman, S., Arokiasamy, P., & Kothari, M. (2009). Nutrition in India. National
Family Health Survey (NFHS-3) India 2005-06.

Birthal, P. S. (2008). Making Contract Farming Work in Smallholder Agriculture. New Delhi:
National Centre for Agricultural Economics and Policy Research.

Birthal, P. S., Jha, A. K., Tiongco, M. M., &Narrod, C. (2008). Improving farm-to-market
linkages through contract farming: A case Study of Smallholder Dairying in India. Discussion
Paper 00814. IFPRI.

Census of India (1991, 2001, 2011). Office of the Registrar General & Census Commissioner,
Ministry of Home Affairs, Government of India.
13 2 The Journal of Industrial Statistics, Vol. 5, No. 2

Central Institute of Post-harvest Engineering and Technology (CIPHET), (2010). Project on

Post Harvest Technology, Ludhiana Punjab, Report submitted to Ministry of Agriculture,
Government of India.

Central Statistical Organisation (CSO), (1998, 2004, 2008). National Industrial Classification
(All economic activities). Ministry of Statistics and Programme Implementation, Government
of India.

Chahal, S.S. and Gupta, J.R. (1986). A study into growth of milk cooperatives in Punjab.
Dairy Guide. 8 (9): 11-16.

Department of Animal Husbandry, Dairying and Fisheries, Ministry of Agriculture.

Directorate General of Commercial Intelligence, DGCIS (various). Ministry of Commerce

and Industry, Government of India.

FAOSTAT (website). Food and Agriculture Organization of United States. http://

faostat.fao.org/

Government of India (GOI) (2013). National Livestock Policy. Department of Animal

Husbandry, Dairying and Fisheries, Ministry of Agriculture.

Government of India (GOI) (website). Office of the Economic Advisor, Ministry of Commerce
& Industry, Department of Industrial Policy & Promotion. http://www.eaindustry.nic.in/

Ibitoye, S.J. (2012). Survey of the performance of Agricultural cooperative societies in

Kogi state, Nigeria. European Scientific Journal. 8 (24): 98-114.

International Union of Food (IUF) (2011). Indian Dairy Industry. IUF’s Dairy Division
Country and Company Reports. http://cms.iuf.org/

Kapoor, Rana (2014). In search of a second white revolution. Business Line, February 10.

Khajarern, S. and J.M. Khajarern (1998). Feeding Swamp Buffalo for Milk Production. Book
(Eds.) by Andrew Speedy, Feeding Dairy Cows in the tropics. Food and Agriculture
Organisation of the United Nations. Daya Publishing House, New Delhi.

Kolekar, D. V., Kokate, L. S., Bangar, Y. C., & Khillare, G. S.(2012). Review on contract dairy
farming: to boost Indian dairying.

Kulkarni Vishwanath, (2014) Amul, mother dairy have not form the cartel. July 1, Business
Line.

Ministry of Agriculture (MAO) (Various), Agricultural Statistics at a Glance, Directorate of

Economic and Statistics, Department of Agriculture and Cooperation, Government of India.

Ministry of Agriculture (MOA) (Various), Agricultural Prices in India, Directorate of

Economic and Statistics, Department of Agriculture and Cooperation, Government of India.
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 13 3

Ministry of Agriculture (MOA), Directorate of Economics and Statistics, Department of

Agriculture and Cooperation (DES), Government of India. http://eands.dacnet.nic.in/.

National Commission on Farmers, NCF (website). http://krishakayog.gov.in/NDDB.pdf

National Dairy Development Board NDDB, (2009). Base working paper on Strategy and
Action Plan for Ensuring Safety of Milk and Milk Products. www.fsssai.gov.in/portals/0/
baseworkingpaper_june2009.pdf. Expert group on milk and milk poducts.

National Sample Survey Office NSSO (various). Household Consumer Expenditure. Ministry
of Statistics & Programme Implementation, Government of India.

Raj, K.N. (1969). Investment in Livestock in agrarian Economies. Centre for Advanced
Studies, Department of Economies, Delhi School of Economics, University of Delhi.

Ravi, C. and D. Roy, (2006) Consumption patterns and food demand projection: A regional
analysis. Paper presented at the Workshop ‘From Plate to Plough- Agricultural Diversification
and Its Implications for Smallholders, organized by the International Food Policy Research
Institute, New Delhi Office, and Institute of Economic Growth, New Delhi, India, 20-21
September.

Sengupta, R. and Gautam, K. (2011). A Case Study of the Food Processing Industry. India’s
Free Trade Agreements and Micro, Small and Medium Enterprises. InditeGlobal, New Delhi.

Srivastava, Samar (2011). When the Milk Runs Dry: India’s New Dairy Policy. India Forbes,
April 4.

Wikipedia.http://en.wikipedia.org/wiki/
Agreement_on_the_Application_of_Sanitary_and_Phytosanitary_Measures

Wikipedia (website). www.wikipedia.com

World Bank (2008). Agriculture for Development. World Development Report.

13 4 The Journal of Industrial Statistics, Vol. 5, No. 2

Figure 1: Price (wholesale) rise in Milk and all Commodities

Figure 2: Share (%) of physical Milk consumption in Rural households based on

alternative estimated ratios

Figure 3: Quantity of liquid milk processed in industry (Vertical axis for Milk powder
and Dairy is in the left and for the other industries in the right)
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 13 5

Figure 4: Movements of Real (wholesale) price of Milk and domestic surplus (net
production less derived consumption demand) in the 2000s

Figure 5: Imports, Exports, Processing and Real (wholesale) price of Milk

Figure 6: Net Production and derived Domestic Consumption of liquid milk

13 6 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 1: Availability and productivity indicators of Milk economy

Year Availability of milk Productivity of milk in relation to
Production Net Per capita net animals crop area fodder area feed area
production production
Million Million Kg Kg/head Kg/hectar Kg/hectare Kg/hectar
tonnes tonnes e e
1990 53.9 53.5 64.4 555.9 411.4 6505.7 357.7
2005 97.1 96.3 87.3 896.9 739.7 12038.2 648.3
2012 132.4 131.3 106.6 1100.2 1031.9 17236.6 905.3
CAGR% 4.17 4.17 2.35 3.22 4.18 4.46 4.23
Note: C
on. Post-harvest loss is 0.8%. Animal: Cattle and buffalo, Crop area: Area under all cereals, oilseeds, sugarcane; Feed:

from data from Ministry of Agriculture.

Table 2: Mi lch animal stocks and their composition

Number of animals ('000) Percentage share in total
Calendar year Cross-bred Indigeno Buffalos Total Cross-bred Buffalos
us
1992 5792 52001 40275 98068 5.91 41.07
1997 8355 49875 42730 100960 8.28 42.32
2007 14407 48042 48641 111090 12.97 43.79
2012 19420 48120 51050 118590 16.38 43.05
Growth rate % 6.24 -0.39 1.19 0.95
Note: CAGR is compound annual growth rate between 1992 and 2012. Source: MOA a,(various),

Table 3: Price behaviour of milk over two decades by alternative estimates (Rs. /100 litre)
Year Wholesale Retail Producer price WPI
price price All commodities
NCF Kulkarni 2004-05=100
1993-94 1029.1 1063.8 638.3 851.1 53.4
2000-01 1537.8 1589.7 953.8 1271.8 83.1
2005-06 1711.3 1769.0 1061.4 1415.2 104.4
2011-12 2959.4 3059.3 1835.6 2447.4 156.1
CAGR% 6.04 6.04 6.04 6.04 6.14
Source: MOA a,(various), GOI (website).
2011-12. Wholesale price index (WPI) of all commodities at base 2004-05

Table 4: Derivation of Milk consumption in India

Consumption Value Estimated Quantity Annually Consumed
(Rs) (Million Tonnes)
Per capita monthly NCF ratio Kulkarni ratio
Year Rural Urban Total Rural Urban Total Rural Urban Total
1993-94 27 45 31.74 33.38 11.94 45.31 25.03 11.94 36.97
2000-01 43 76 52.21 40.07 16.46 56.53 30.05 16.46 46.51
2005-06 51 85 61.06 45.22 19.00 64.22 33.92 19.00 52.91
2011-12 115 184 136.64 63.04 27.20 90.24 47.28 27.20 74.48
CAGR 8.38 8.14 8.45 3.60 4.68 3.90 3.60 4.68 3.97
Note: CAGR is compound annual growth rate between 1993-94 and 2011-12. Source: NSSO and calculated by our method.
 Understanding Demand, Supply and Price behavior in the Dairy sector:... 13 7

Table 5 : Years of maximum and minimum of Exports and Imports

Exports Imports
Commodity Highest Lowest Highest Lowest
Liquid Milk 2011-12 2001-02 2008-09 2002-03
Milk Powder 2005-06 2011-12 2011-12 2001-02
Total Liquid Milk Equivalent 2005-06 2011-12 2011-12 2001-02
Butter Milk, Etc. 2007-08 2002-03 2009-10 2001-02
Whey etc. 2007-09 2011-12 2011-12 2002-03
Butter, Etc. 2008-09 2001-02 2009-10 2005-06
Cheese And Curd 2008-09 2001-02 2010-11 2004-05

Appendix
Table A1: Validation of Prices and per capita consumption derived from NSS quinquennial consumer survey
and Wholesale price data
Rural Urban Urban
Imputed Price Price used Rural
Consumption Consumption
Whole
Expen Qua Expen Qua Rural Urban Estimated Retail
sale Consumption Estimated
se ntity se ntity (Rs/ (Rs/ Producer's (Rs/kg
(Rs/ Kg/Capita
Year (Rs.) (kg) (Rs.) (kg) kg) kg) (Rs/ kg) )
kg)
Estim Estim Kulk Estim Kulk Kulk
NSS NSS NSS NSS MOA NCF NCF NCF
ated ated arni ated arni arni
2004 47.31 3.87 83.30 5.11 12.22 16.30 16.8 13.89 10.42 17.37 3.41 4.54 4.80 4.80
2009 80.55 4.12 137.01 5.36 19.55 25.56 24.17 19.99 14.99 24.99 4.03 5.37 5.48 5.48
Source: Monthly per capita expenditure and population taken from NSSO (various), MoA ,
Census Data.

TableA2 : Consumption Expenditure and Share of Agricultural Households in Consumption of

Milk (2000s)
Monthly per capita Share in total Physical consumption (%)
Consumption Rs NCF estimate Kulkarni Estimate
Type of Household 2004 2009 2004 2009 2004 2009
Agricultural Labour 20.9 43.7 7.7 9.3 7.0 8.5
Self-employed in 64.2 109.5 37.3 32.6 33.9 29.5
Agriculture
Agricultural Household 47.5 82.1 44.9 42.0 40.8 38.0
Rural Household 47.3 80.6 69.7 68.4 63.3 61.8
Source: Monthly per capita expenditure and population taken from NSSO (various).
13 8 The Journal of Industrial Statistics (2016), 5 (2), 138 - 153

Contract Workers in India’s Organised Manufacturing Sector

A.K. Panigrahi1, Central Statistics Office, Kolkata, India
Abstract
The trend of contract workers in organized manufacturing sector in recent years has
been increased significantly. In this scenario an attempt has been made to examine the
contract workers participation rate and wage difference between direct workers and
contract workers in the organized manufacturing sector in India. The unit level data of
Annual Survey of Industries (ASI) of 2000-01, 2005-06, 2010-11 and 2012-13 have been
used for the analysis. From the analysis it has been observed that the proportions of
contract workers have increased significantly and reached at 34 percent in 2012-13. It
has been observed that industrial activities, namely, tobacco products, where highest
(73.29 percent) proportions of contract workers are engaged in the production process
followed by other non-metallic mineral products (57.98 percent) and manufacture of
coke, refined petroleum products (49.56 percent). Contract workers participation rate
with respect to different States/UTs in India has also varies significantly. States, namely,
Bihar (70.05 percent), Odisha (58.47 percent), Uttarakhand (51.98 percent), Andhra
Pradesh (47.87 percent) and Haryana (47.06 percent) where, significantly higher
proportions of contract workers have been engaged in the organized manufacturing
sector. There is significant wage difference between contract workers and direct workers
has also been observed and this is also true with respect to industrial activities and major
States/UTs in India. From the analysis it has been observed that contract workers average
wage per day is Rs. 156 during 2012-13, whereas, direct workers average wage per day
is Rs. 404. From this analysis it is clearly observed that contract workers are getting 60
percent less wage than that of the direct workers.
1. Introduction
1.1 In the present era, outsourcing, contractualisation, contract workers etc. are the
predominant issues. Most of the organizations are in favour of contractual employment
rather direct employment for smooth functioning of the day today activities. These
contractual workers/labours are available from the market on the prevailing market price
through certain agencies/contractors. There is no direct relationship between the contract
workers and organization where they are contributing their labour. The agencies have to
supply the contract workers and received the commission charges under certain terms and
conditions. The suppliers have to manage all the issue relating to these contract workers.
In this scenario the entrepreneurs are little bit free from issues relating to the contract
workers and prefer to maximize their profit. If the entrepreneurs are not satisfied with the
work done by these contract workers can discontinue their service as and when they like.
These factors are tending industries to hire more and more numbers of contract workers to
have greater flexibility to adjust the number of workforce based on economic efficiency,
better utilization of resources, optimization of profit and bringing cost effectiveness, despite
the risk of lower worker loyalties and lousy pay.
1.2 What are the differences between direct workers and contract workers? Therefore,
it is better to understand the difference between direct workers and contract workers. The
1
e-mail: [email protected]
 Contract Workers in India’s Organised Manufacturing Sector 13 9

major difference between direct and contract workers are discussed here. Direct workers
are directly recruited by the employer whereas contract workers are taken from the
contractors. Recruitment rules are applied for the direct workers whereas no such recruitment
rules are applied among the contract workers. On the aspect of job security, direct workers
are highly secured whereas contract workers are not secured at all. On the basis of working
hours direct workers are benefitted through certain rules and regulation whereas contract
workers do not have any regulation of working hours. Direct workers wages & salaries are
based on certain rules and regulation whereas contract workers wages & salaries are not
regulated properly. On the social security aspects, direct workers are in advantages which
includes medical allowance and sick leave whereas contract workers are not get such
protection. Among direct workers leave rules are applicable whereas among contract workers,
leave rules are not applicable. Finally, direct workers are protected by labour laws including
right to freedom of association and collective bargaining whereas no such rules for the
protection of contract workers.
1.3 Some of the studies have observed that employers in a globalised economic
environment favour flexible labour strategies where they ask for the freedom to hire workers
for a fixed term even for perennial activities and discontinue their services when not needed
(Sood, Nath and Ghosh, 2014). A study by Neethi (2008) has observed that contractualisation
prevails in almost all industry groups and it is highly region-specific and industry-specific
factors have their influence in determining contract work intensity.
1.4 All India Organisation of Employers’ has studied on the issues relating to the
industrial relations & contract labour and observed that during the recent years, employment
of contract labour has become a contentious issue and a key reason for the increasing
labour unrest in the form of strikes and protests. They have cited that the major reasons for
the rise in industrial unrest could be increasing dependence of industries on contract
labour for requirement of flexibility. This segment of worker due to anxiety of job security,
lack of social security, exploitation in the hands of contractors, low wages, unequal treatment
by Trade Unions and even abusive behavior of the permanent workers and supervisors
develop rebellion feelings. This study has also cited some of the instances of industrial
unrest during recent past and conclude that the surge in violence disturbing industrial
relations has become a concerning situation for all. On September 22, 2008 the CEO of
Graziano Transmissioni India, the Indian unit of an Italian auto component maker, was
clubbed to death by a group of 200 workers. In another incidents, in March 2011, a Deputy
General Manager (Operations) of Powmex Steel, a unit of Graphite India Ltd. was killed after
his vehicle was set afire by irate workers, in November 2010 an Assistant General Manager
of Allied Nippon, an auto parts maker, was stoned to death by angry workers, in September
2009 the Vice-President (HR) of Pricol was beaten to death by agitating workers, and many
more. The most recent worst form of industrial unrest was witnessed in the Maruti Suzuki
India Ltd., Manesar plant, where workers went into riotous, leaving its General Manager
(HR) dead and 100 other officials laid up in hospital with serious injuries (http://www.aioe.in/
htm/IndustrialRelations.pdf).
1.5 In this context an attempt has been made to examine the contract workers
participation and wage differentials in organized manufacturing sector in India with the
following objectives.
14 0 The Journal of Industrial Statistics, Vol. 5, No. 2

2. Objectives of the Study

(i) To examine the participation rate by contract workers in organized manufacturing
sector in India.
(ii) To study the differentials in participation of contractual workers with respect
to major industrial activities and major states in India.
(iii) To examine the wage differentials among contract workers with respect to
direct workers.
3. Data and Methodology
3.1 Annual Survey of Industries (ASI) data are available for the organized
manufacturing sector and those industries are registered under the Factories Act. 1948 are
covered in the survey. ASI schedule is the basic tool to collect required data for the factories
registered under Sections 2(m)(i) and 2(m)(ii) of the Factories Act, 1948. Block E of the ASI
schedule collects the information with respect to employment and labour cost including
contract workers in the organized manufacturing sector. In this block, information with
respect to direct workers, contract workers, supervisor & managerial staff, unpaid family
members, other employees, man-days worked, average number of persons worked, wages/
salaries etc. are available. Therefore, in this paper an attempt has been made to examine the
contract workers participation rate and wage differentials in the organized manufacturing
sector. Before analyzing the contract workers participation, we should have better idea
about who are the contract workers? Who are the direct workers? How are they different
from each others?
3.2 Contract Worker: All persons who are not employed directly by the factory
owner/employer but engaged through a third party i.e. agency/ contractor, are termed as
contract workers. Such agency charges from the factory for this job. In ASI schedule,
Block E: item 4 collects the information on workers employed through contractors. In more
specific terms those workers employed purely on contract basis are reported in item 4 of
Block E (Govt. of India, 2014).
3.3 Direct Worker: It includes those workers employed directly by the factory. In
ASI schedule, Block E: items 1 & 2 collects the information with respect to male and female
workers directly employed which include all persons employed directly on payment of
wages or salaries and engaged in any manufacturing process or its ancillary activities like
cleaning any part of the machinery or any premises used for manufacturing or storing
materials or any kind of work incidental to or connected with the manufacturing process
(Govt. of India, 2014).
3.4 In this paper, the contract workers participation rate is defined as the proportion
of contract workers employed in the total workers and contributing in the organized
manufacturing process.
Thus,

Total Contract Workers

Contract Workers Part icipation Rate = × 100
Total Workers
 Contract Workers in India’s Organised Manufacturing Sector 14 1

3.5 ASI unit level data from 2000-01 to 2012-13 have been used for the analysis of
employment composition in the organized manufacturing sector. However for depth analysis
of contract workers participation and their wage rate calculation, the unit level data of ASI
of 2000-01, 2005-06, 2010-11 and 2012-13 have been used.
4. Results and Finding

4.1 Employment Scenario in Organised Manufacturing Sector

4.1.1 It is interesting to analyses the composition of employment scenario in the

organized manufacturing sector. Figure 1 presents the employment scenario in the organized
manufacturing sector during 2000-01 to 2012-13. From the figure it is clearly shown that in
2000-01 the total employment in the organized manufacturing sector was around 8 million.
However, over the period it has increased drastically and reached 13 million in 2012-13. It is
interesting to see the proportions of workers in the total employment. From the figure it is
evident that workers constitute around 80 percent of the total employment in the organized
manufacturing sector and the remaining 20 percent are other than workers, which includes,
supervisor & managerial staff, unpaid family members, other employees those are not
directly involved in the manufacturing process. Over the period both workers as well as
total employees have increased drastically. From this analysis it is clearly understood that
workers in the manufacturing sector are the back bone for the organization for manufacturing
process.

4.2 Contract Workers in Organised Manufacturing Sector

4.2.1 Now it will be interesting to analyse the composition of the workers with respect
to direct workers and contract workers. Figure 2 presents the proportions of direct and
contract workers in the organized manufacturing sector during 2000-01 to 2012-13. From the
figure it is evident that in 2000-01 direct workers constitute around 80 percent of the total
workers and the remaining 20 percent are contract workers. However, during last decade it
has been observed that proportions of direct workers declined significantly and reached at
66 percent in 2012-13. Moreover, the proportions of contract workers increased significantly
and reached at 34 percent in 2012-13. In beginning of the decade there was a big gap
between direct workers and contract workers participation in the organized manufacturing
sector. However, over the period the gap has become narrowed.

4.3 Workers Size vs Contract Workers

4.3.1 An attempt has been made to examine the percentage of contract workers with
respect to the workers size. Figure 3 presents the proportions of contract workers in total
workers by workers size of factories during 2000-01, 2005-06, 2010-11 and 2012-13. From the
figure, it is clearly evident that higher proportions of contract workers are engaged with
respect to higher workers size. It is observed that those factories having 5000 and above
workers, more than 50 percent workers are contract workers. Those factories, where workers
size is less than 50 significantly lower percentage of contract workers are engaged in the
organized manufacturing sector. There is an upward trend of contract workers participation
with respect to workers size in organized manufacturing sector.
14 2 The Journal of Industrial Statistics, Vol. 5, No. 2

4.4 Contract Workers vs Direct Workers

4.4.1 An attempt has also been made to examine the number of factories in operation
with respect to the contract workers size. Table 1 presents the contract workers size class
with respect to factories in operation during 2000-01, 2005-06, 2010-11 and 2012-13. From
the analysis of ASI unit level data, it has been observed that the contract workers are
predominant among the 25 percent of the total factories in the manufacturing sector. It has
also been revealed that around 75 percent of the factories are functioning without any
contract workers. From the table it has been revealed that around 7 percent (11891 factories)
of factories are in operation with contract workers 1-9 followed by 20-49 contractor workers
(6 percent) and 10-19 contract workers (4 percent).

4.4.2 It is surprising to see the number of factories where numbers of contract workers
are more than that of the direct workers. Table 2 presents the numbers of factories in
operation with respect to the contract workers size among those factories where numbers
of contract workers are more than that of the direct workers. It has been observed that in
2000-01 around 9 percent of factories are in operational where numbers of contract workers
are more than that of the numbers of direct workers. Over the period in 2012-13, it is also
interesting to see the huge number of factories (31908 factories during ASI 2012-13), where
numbers of contract workers are more than that of the direct workers which constitute
around 18 percent of the total factories in the organsied manufacturing sector in India. By
examining contract workers size class among these factories, it has been observed that
higher proportions of factories (around 25 percent) are in the size class of 20-49 contract
workers. It has also been observed that there are significant proportions (around 38 percent)
of factories where contract workers are above fifty. It has also been observed that 12
percent (3744 factories) of factories are functioning where more than 200 contract workers
are in the production process.

4.4.3 Table 3 presents the number of contract workers and number of direct workers per
factory among those factories where numbers of contract workers are more than that of the
number of direct workers with respect to contract workers size class. From the table it has
been revealed that in contract workers size class 1-9, where around 6000 factories are in
operational, on an average six contract workers are engaged in the production process and
in the same industry on an average around one direct worker is engaged. Similarly, in size
class 10-19, the average numbers of contract workers per factory are around 14 whereas the
average numbers of direct workers are only 3. Similarly, in size class above 200, the average
numbers of contract workers per factory are around 473 whereas the average numbers of
direct workers per factory are around 117 in 2012-13. From this analysis it has been revealed
that the numbers of contract workers are proportionately increasing with respect to the
contract workers size class. Similarly, irrespective of contract workers size class the average
number of contract workers per factory has significantly more than three times that of the
number of direct workers.

4.5 Industry-wise Variations in Contract Workers Participation

4.5.1 It is interesting to see the industry wise variation in contract workers participation.
Figure 4 presents the industry wise variation in contract workers participation. From Figure
4, it is evident that there is significant variation with respect to contract workers participation
 Contract Workers in India’s Organised Manufacturing Sector 14 3

in industrial activities. Industrial activities, namely, tobacco products, where the highest
(73.29 percent of all workers) proportion of contract workers are engaged in the production
process followed by other non-metallic mineral products (57.98 percent) and manufacture
of coke, refined petroleum products (49.56 percent). Industrial activities, namely,
manufacturing of wearing apparel (12.35 percent), textiles (14.07 percent), and printing &
reproduction of recorded media (16.85 percent), where, significantly low percentage of
contract workers are engaged.
4.6 Interstate Variation in Contract Workers Participation
4.6.1 Figure 5 presents the percentage of contract workers with respect to major States2/
UTs in India. From the figure, it is evident that there is significant variations have been
observed with respect to the participation of contract workers in major States/UTs in India.
Top five States, namely, Bihar (70.05 percent), Odisha (58.47 percent), Uttarakhand (51.98
percent), Andhra Pradesh (47.87 percent) and Haryana (47.06 percent) where, significantly
higher proportions of contract workers have been engaged in the organized manufacturing
sector. However, bottom five States/UTs, namely, Delhi (10.83 percent), Kerala (14.17
percent), Tamil Nadu (19.54 percent), Assam (19.86 percent) and Karnataka (20.09 percent),
where significantly low proportions of contract workers have been engaged in the organized
manufacturing process. It has also been observed that the most industrialized States/UTs,
namely, Maharashtra (40.31 percent), Gujarat (36.55 percent), Uttar Pradesh (35.93 percent)
and Andhra Pradesh (47.87 percent) significantly higher proportions of contract workers
have been engaged in the organized manufacturing sector which is higher than the national
average. However, one industrialized States, namely, Tamil Nadu (19.54 percent) has
significantly low proportions of contract workers have been engaged in the organized
manufacturing sector.
5. Why Contract Workers?
5.1 From the above analysis it has been observed that percentage of contract workers
are increasing over the study periods. But exact reason is not revealed from the above
analysis. There are many reasons behind increasing contract employment in organized
manufacturing sector. However, some of the possible reasons are given below:
i. Contract workers are the substitution against the direct workers:It is argued
that employers facing stringent labour laws do not want to employ more people in
the production process. Employers in a globalised economic environment favour
flexible labour strategies where they ask for the freedom to hire workers for a fixed
term even for perennial activities and discontinue their services when not needed.
ii. For short run/seasonal production: The enterprise may interest for short run
production for some specific kind of job. Therefore, the enterprise may hire contract
workers to meet the short term demand. It may happen that for seasonal items/
production, enterprise may hire contract workers when required and fire them in
completion of the project.
2
Major States/UTs are Andhra Pradesh, Assam, Bihar, Chattisgarh, Delhi, Gujarat, Haryana, Himachal
Pradesh, Jammu & Kashmir, Jharkhand, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Odisha,
Punjab, Rajasthan, Tamil Nadu, Uttar Pradesh, Uttarakhand and West Bengal. Other States/UTs includes
A & N. Island, Chandigarh, Dadra & N Haveli, Daman & Diu, Goa, Manipur, Meghalaya, Nagaland,
Puducherry, Sikkim and Tripura
14 4 The Journal of Industrial Statistics, Vol. 5, No. 2

iii. Special kind of skilled workers for technical work: It may happen that for
fulfillment of special kind of technical work which may not require regularly. In this
context, enterprises are very much interested to outsource the work through contract
workers.
iv. Minimizing monitoring cost/regulatory cost: Employing contract workers
through contractors minimizes the monitoring cost. It is the responsibility of the
contractor to follow the terms and conditions for specific job. Therefore, contractor
will be very much responsible for the completion of specific job as per the terms
and condition otherwise forfeit the payment.
v. Avoid fringe benefits like annual leave with wages, gratuity, bonus, etc.:
Contract workers are not in the pay-role of the factories. Therefore, leave rules and
other welfare measures such as gratuity, bonus etc. are not applicable for the
contract workers.
6. Wage Differentials
6.1 In the following section an attempt has been made to look into the wage differential
in the registered organized manufacturing sector with respect to contract workers and
direct workers in India.
We define,

Average contract workers wage

Wage Differential Ratio =
Average direct workers wage

6.2 The average wage has been calculated on the basis of total annual wages3 to total
man-days4 worked. So, the average contract workers wage has been calculated on the basis
of total contract workers annual wage to total man-days worked for contract workers.
Similarly, the average direct workers wage has been calculated on the basis of total annual
wage to total man-days worked for direct workers.
6.3 Table – 4 presents the average wage per day, wage difference and wage differential
ratio with respect to contract workers and direct workers. From the table it has been revealed
that the average wage rate per day for direct workers is Rs. 164 in 2000-01. However, the
wage rate has increased significantly during last decade and reached at Rs. 404 in 2012-13.
Whereas the average wage rate per day for contract workers is Rs. 90 in 2000-01 and
increased to Rs. 221 in 2010-11 and Rs. 156 in 2012-13. It has also been revealed that there

3
Wages: Wages are defined to include all remuneration capable of being expressed in monetary terms and
also paid more or less regularly in each pay period to workers (defined above) as compensation for work
done during the accounting year. It includes:
(i) Direct wages and salary (i.e. basic wages/salaries, payment of overtime, dearness, compensatory, house
rent and other allowances;
(ii) Remuneration for period not worked (i.e. basic wages), salaries and allowances payable for leave
period, paid holidays, lay-off payments and compensation for unemployment (if not paid from source
other than employers);
(iii) Bonus and ex-gratia payment paid more or less regularly (i.e., incentive bonuses and good attendance
bonuses, production bonuses etc.).
 Contract Workers in India’s Organised Manufacturing Sector 14 5

is significant wage difference with respect to direct workers and contract workers. Over the
period, the wage rate difference between contract workers and direct workers has increased
significantly. From wage differential ratio, it has been revealed that the contract workers
wage rate is forty five (45%) percent less than that of direct workers wage in 2000-01. There
is an indication of declining wage differential ratio between direct workers and contract
workers over the last decades.
6.4 From the Figure 6 it is clearly understood the wage differentials between contract
workers and direct workers during 2000-01 to 2012-13.
6.5 Wage differential Ratio in Industrial Activities
6.5.1 It will be interesting to examine the wage differential ratio with respect to industrial
activities. Table A3 presents the average wage rate of direct workers, contract workers and
wage differential ratio. From the table it has been revealed that there is significant wage
difference has been observed in coke & refined petroleum products (0.29) followed by
motor vehicles & trailers (0.42) and basic metals (0.46). But there are industrial activities,
namely, recycling (0.93), cotton ginning (0.88), leather & related products (0.86) and products
of wood (0.85) wage differential ratio is comparatively low. It has also been observed that
there are industrial activities, namely; wearing apparel, leather & related products, cotton
ginning contract workers are getting higher wages than direct workers. Activities of wearing
apparel in 2000-01, leather and related products in 2000-01 and 2010-11, cotton ginning in
2005-06 and 2010-11 contract workers are getting higher wages than director workers. It has
also been observed that in 2012-13 the wage rate of contract workers was almost four times
(3.85) of direct workers in the industry of tobacco products. This is in sharp contrast with
the earlier years of 2000-01, 2005-06 and 2010-11 in that activity. It may happen that for
special kind of technical work enterprises are very much interested to outsource the work
through contract workers by paying higher wages.
6.6 Wage differential Ratio in Major States/UTs
6.6.1 Table A4 presents the wage difference in organized manufacturing sector in India
with respect to major States/UTs between contract and direct workers. From the table it has
been revealed that contract workers are always getting lower wage than that of direct
workers. There is significant variation in wage difference has been observed among major
States/UTs between contract and direct workers. States, namely; Odisha, Bihar, Andhra
Pradesh, Uttarakhand, and Maharashtra wage difference is higher than that of other States.

It excludes layoff payments and compensation for employment except where such payments are for this
purpose, i.e., payments not made by the employer. It excludes employer’s contribution to old age
benefits and other social security charges, direct expenditure on maternity benefits and crèches and other
group benefit in kind and travelling and other expenditure incurred for business purposes and reimbursed
by the employer. The wages are expressed in terms of gross value, i.e., before deductions for fines,
damages, taxes, provident fund, employee’s state insurance contribution etc. Benefits in kind (perquisites)
of individual nature are only included(Govt. of India, 2014).
4
Man-days Worked: These are obtained by summing up the number of man-days worked by persons
working in each shift over all the shifts on all days, i.e. both manufacturing and non-manufacturing days.
This figure excludes persons who are paid but remain on leave, strike, etc. Manufacturing days will mean
and include number of days on which actual manufacturing process was carried out by the unit where as
Non-manufacturing days will mean and include number of days on which only repair/maintenance and
construction work were undertaken (Govt. of India, 2014).
14 6 The Journal of Industrial Statistics, Vol. 5, No. 2

It has also been revealed that the eastern region significantly wage difference is higher
followed by central and western region.
7. Conclusion
7.1 From the above analysis it has been revealed that contract workers participation
rate has significantly increased during the last decade. This is also true with respect to
industrial activities and major States/UTs. Because of flexibility in labour laws, contract
workers are engaged as the substitution against the direct workers. There is significant
wage difference has been observed between contract and direct workers.
References
Govt. of India (2014), Instruction Manual: Annual Survey of Industries (Concepts,
Definitions and Procedures). Central Statistics Office (IS Wing) and Field Operation
Division (NSSO), Govt. of India, Ministry of Statistics and PI.
Industrial Relations & Contract Labour in India by All India Organisation of Employers’,
Federation House, Tansen Marg, New Delhi. http://www.aioe.in/htm/
IndustrialRelations.pdf
Neethi P. (2008) Contract Work in the Organised Manufacturing Sector: A Disaggregated
Analysis of Trends and their Implications. The Indian Journal of Labour Economics, Vol.
51, No. 4, Pp. 559-573.
Sood, Atul, Paaritosh Nath, and Sangeeta Ghosh (2014) Deregulating Capital, Regulating
Labour: The Dynamics in the Manufacturing Sector in India, Economic and Political
Weekly, Vol. XLIX, No. 26 & 27, Pp. 58-68.
 Contract Workers in India’s Organised Manufacturing Sector 14 7

Figure 1: Employment Composition in Organised Manufacturing Sector during 2000-01

to 2012-13 in India

Figure 2: Percentage of Direct and Contract Workers in Organised Manufacturing

Sector during 2000-01 to 2012-13 in India
14 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Figure 4: Contract Workers in Organised Manufacturing Sector with respect to

Industrial Activities during 2000-01 and 2012-13.
 Contract Workers in India’s Organised Manufacturing Sector 14 9

Figure 5: Contract Workers in Organised Manufacturing Sector with respect to Major

States/UTs during 2000-01 and 2012-13.

Figure-6: Wage Rate Differentials in Organised Manufacturing Sector

15 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 1: Number of Factories in Operation with respect to Contract Workers Size

during 2000-01, 2005-06, 2010-11 and 2012-13.
Contract Percentage of factories in operation
Workers Number of factories in operation (%)
Size Class
2000-01 2005-06 2010-11 2012-13 2000-01 2005-06 2010-11 2012-13
No contract
workers 147531 138312 124449 133486 85.63 80.48 72.28 74.53
1-9 8387 9551 13022 11891 4.87 5.56 7.56 6.64
10-19 4765 6066 8506 7668 2.77 3.53 4.94 4.28
20-49 6138 8213 10928 10831 3.56 4.78 6.35 6.05
50-99 3064 5040 7157 6813 1.78 2.93 4.16 3.80
100-199 1303 2505 4123 3974 0.76 1.46 2.39 2.22
Above 200 1102 2178 3991 4439 0.64 1.27 2.32 2.48
All India 172290 171865 172176 179102 100.00 100.00 100.00 100.00
Table 2: Number of Factories in Operation with respect to Contract Workers Size during
2000-01, 2005-06, 2010-11 and 2012-13 among those Factories where Contract Workers
are more than the Direct Workers .
Contract Number of factories in operation Percentage of factories in operation
Workers
Size Class 2000-01 2005-06 2010-11 2012-13 2000-01 2005-06 2010-11 2012-13
1-9 3278 4079 6865 5894 21.70 17.82 20.15 18.47
10-19 3155 4393 6026 5344 20.88 19.20 17.69 16.75
20-49 4644 6533 8661 8255 30.74 28.55 25.42 25.87
50-99 2322 4135 5932 5477 15.37 18.07 17.41 17.16
100-199 916 2002 3383 3194 6.06 8.75 9.93 10.01
Above 200 793 1743 3202 3744 5.25 7.62 9.40 11.73
Total 15108 22885 34069 31908 100.00 100.00 100.00 100.00
Percentage 8.77 13.32 19.79 17.82
All India 172290 171865 172176 179102
Table 3: Number of Contract Workers and Direct Workers per Factory in Operation with
respect to Contract Workers Size during 2000-01, 2005-06, 2010-11 and 2012-13.
Contract Contract workers per factory Direct workers per factory
Workers
Size Class 2000-01 2005-06 2010-11 2012-13 2000-01 2005-06 2010-11 2012-13
1-9 6 6 6 5 1 2 1 1
10-19 14 14 14 14 4 4 4 3
20-49 31 32 32 32 8 8 7 7
50-99 67 68 70 70 14 14 16 15
100-199 131 130 131 131 33 30 33 33
Above 200 635 504 502 473 103 98 116 117

Table 4: Wage Rate and Wage Differential Ratio in Organised Manufacturing Sector
in India
Wage Rate (INR) per day Wage
Years Contract differential
Direct workers ratio
workers
2000-01 164 90 0.55
2005-06 198 116 0.59
2010-11 313 221 0.70
2012-13 404 156 0.39
 Contract Workers in India’s Organised Manufacturing Sector 15 1

Appendix Tables
Table A1: Percentage of Contract Workers in Organised Manufacturing Sector in India
during 2000-01, 2005-06, 2010-11 and 2012-13 with respect to Industrial Activities.
Activity Description Percentage of Contract W orkers (%)
2000-01 2005-06 2010-11 2012-13
MANUFACTURE OF TOBACCO PRODUCTS 63.39 68.33 67.59 73.29
MANUFACTURE OF OTHER NON -METALLIC MINERAL
PRODUCTS 33.07 49.26 56.05 57.98

MANUFACTURE OF COKE, REFINED PET ROLEUM PRODUCTS 19.25 43.82 49.83 49.56
MANUFACTURE OF OTHER TRANSPORT EQUIPMENT 12.56 31.93 45.23 48.20
MANUFACTURE OF FABRICATED METAL PRODUCTS, EXCEPT
MACHINERY AND EQUIPMENTS 27.70 39.73 46.54 43.89
MANUFACTURE OF BASIC METALS 23.56 33.73 41.40 43.44

MANUFACTURE OF CHEMICALS AND CHEMICAL PRODUCTS 20.12 31.22 38.24 41.79

MANUFACTURE OF MOTOR VEHICLES, TRAILERS AND SEMI -
TRAILERS 11.55 30.79 41.65 39.79
RECYCLING 36.15 46.24 20.84 39.04
MANUFACTURE OF ELECTRICAL EQUIPMENT 12.64 28.04 36.02 38.24
COTTON GINNING, CLEANING AND BAILING 27.57 33.46 49.75 35.48
MANUFACTURE OF COMPUTER, ELECTRONIC & OPTICAL
PRODUCTS 11.46 20.81 33.84 33.86
MANUFACTURE OF MACHINERY AND EQUIPMENT N.E.C. 10.77 22.70 33.66 32.73
MANUFACTURE OF RUBBER AND PLASTIC PRODUCTS 13.28 24.14 30.64 32.61
MANUFACTURE OF FOOD PRODUCTS AND BEVERAGES 20.53 26.15 30.58 30.63
MANUFACTURE OF PAPER AND PAPER PRODUCTS 21.94 27.30 28.59 26.38
MANUFACTURE OF WOOD AND OF PRODUCTS OF WOOD 9.39 24.37 26.48 26.33
MANUFACTURE OF FURNITURE 14.90 17.63 22.77 23.10

MANUFACTURE OF LEATHER AND RELATED PRODUCTS 18.85 19.91 16.01 19.66

PRINTING AND REPRODUCTION OF RECORDED MEDIA 5.71 10.50 19.02 16.85
MANUFACTURE OF TEXTILES 9.17 12.52 14.94 14.07
MANUFACTURE OF WEARING APPAREL 5.79 13.26 14.46 12.35
OTHER ACTIVITES 21.74 24.03 28.09 26.67
All India 20.42 28.54 33.94 34.26

Table A2: Percentage of Contract Workers in Organised Manufacturing Sector in

India during 2000-01, 2005-06, 2010-11 and 2012-13 with respect to Major States/UTs.
Percentage of Contract Workers (%)
State Name 2000-01 2005-06 2010-11 2012-13
Andhra Pradesh 44.88 53.43 48.07 47.87
Assam 7.22 16.33 19.75 19.86
Bihar 38.24 55.24 64.28 70.05
Chattisgarh 24.77 36.08 42.91 40.81
Delhi 6.31 9.21 12.57 10.83
Gujarat 26.91 34.12 36.06 36.55
Haryana 30.26 44.66 46.65 47.06
Himachal Pradesh 15.74 20.19 26.35 29.99
Jammu & Kashmir 25.01 31.09 48.34 44.98
Jharkhand 12.40 12.33 23.24 35.45
Karnataka 11.30 13.48 21.13 20.09
Kerala 4.16 9.22 16.32 14.17
15 2 The Journal of Industrial Statistics, Vol. 5, No. 2

Table A2: Percentage of Contract Workers in Organised Manufacturing Sector in India during
2000-01, 2005-06, 2010-11 and 2012-13 with respect to Major States/UTs. (Contd.)
Percentage of Contract Workers (%)
State Name 2000-01 2005-06 2010-11 2012-13
Madhya Pradesh 23.62 27.49 33.15 33.50
Maharashtra 18.84 31.07 40.51 40.31
Odisha 28.74 42.01 47.97 58.47
Punjab 16.46 27.90 28.52 27.89
Rajasthan 22.73 33.47 36.29 37.22
Tamil Nadu 8.03 14.57 19.95 19.54
Uttar Pradesh 25.21 30.38 36.44 35.93
Uttarkhand 21.22 43.02 50.18 51.98
West Bengal 10.50 18.86 30.41 33.27
Others 21.20 31.16 41.45 39.16
All India 20.42 28.54 33.94 34.26

Table A3: Wage Differenc in Organised Manufacturing Sector in India during 2000-
01, 2005-06, 2010-11 and 2012-13 with respect to Industrial Activities between
Contract and Direct Workers.
Direct workers average wage Contract workers average
per day (Rs.) wage per day (Rs.) Wage Differential Ratio Ave
rage
2000- 2005- 2010- 2012- 2000 2005 2010 2012 2000 2005 2010 2012
Ratio
Activity Descriptions 01 06 11 13 -01 -06 -11 -13 -01 -06 -11 -13
COKE, REFINED
PETROLEUM 521 586 1181 1407 132 197 406 331 0.25 0.34 0.34 0.23 0.29
MOTOR VEHICLES,
TRAILERS 277 349 503 634 116 149 261 200 0.42 0.43 0.52 0.32 0.42
BASIC METALS 274 334 477 637 143 137 236 254 0.52 0.41 0.49 0.40 0.46
246 274 401 555 103 134 240 188 0.42 0.49 0.60 0.34 0.46
MACHINERY AND
EQUIPMENT 228 281 422 572 111 143 267 148 0.49 0.51 0.63 0.26 0.47
CHEMICAL PRODUCTS 200 251 390 504 95 129 235 229 0.47 0.51 0.60 0.45 0.51
COMPUTER &
ELECTRONIC PRODUCTS 194 250 430 555 119 145 254 154 0.61 0.58 0.59 0.28 0.51
PRINTING &
REPRODUCTION 180 214 326 462 98 144 236 65 0.55 0.68 0.72 0.14 0.52
PAPER PRODUCTS 160 200 290 358 96 124 211 116 0.60 0.62 0.73 0.32 0.57
OTHER TRANSPORT
EQUIPMENT 202 284 458 525 129 173 276 281 0.64 0.61 0.60 0.54 0.60
RUBBER & PLASTIC
PRODUCTS 138 180 278 358 95 115 209 151 0.69 0.64 0.75 0.42 0.63
FABRICATED
METAL PRODUCTS 173 202 341 434 94 135 282 236 0.54 0.67 0.83 0.54 0.65
FURNITURE;
MANUFACTURING 161 219 312 392 131 161 298 90 0.82 0.73 0.96 0.23 0.68
FOOD PRODUCTS
AND BEVERAGES 118 139 228 298 85 109 200 127 0.73 0.78 0.88 0.43 0.70
TEXTILES 129 145 222 278 109 127 203 52 0.84 0.87 0.92 0.19 0.70
OTHER NON-METALLIC
MINERAL PRODUCTS 141 172 270 355 99 110 186 347 0.70 0.64 0.69 0.98 0.75
WEARING APPAREL 90 129 215 275 106 130 210 42 1.17 1.01 0.98 0.15 0.83
 Contract Workers in India’s Organised Manufacturing Sector 15 3

Table A3: Wage Differenc in Organised Manufacturing Sector in India during 2000-
01, 2005-06, 2010-11 and 2012-13 with respect to Industrial Activities between
Contract and Direct Workers. Contd.
Direct workers average wage Contract workers average
per day (Rs.) wage per day (Rs.) Wage Differential Ratio Ave
rage
2000- 2005- 2010- 2012- 2000 2005 2010 2012 2000 2005 2010 2012
Ratio
Activity Descriptions 01 06 11 13 -01 -06 -11 -13 -01 -06 -11 -13

PRODUCTS OF WOOD 87 117 194 276 82 123 200 107 0.94 1.05 1.03 0.39 0.85
LEATHER & RELATED
PRODUCTS 101 129 201 258 109 121 231 73 1.08 0.94 1.15 0.28 0.86
COTTON GINNING 73 90 163 248 70 94 169 120 0.96 1.05 1.03 0.49 0.88
RECYCLING 112 166 195 321 95 98 322 198 0.85 0.59 1.65 0.62 0.93
TOBACCO PRODUCTS 72 93 197 211 46 47 68 811 0.64 0.51 0.35 3.85 1.34
OTHERS 129 172 320 431 85 123 268 112 0.66 0.71 0.84 0.26 0.62
All India 164 198 313 404 90 116 221 156 0.55 0.59 0.70 0.39 0.56

Table A4: Wage Difference in Organised Manufacturing Sector in India during 2000-01,
2005-06, 2010-11 and 2012-13 with respect to Major States/UTs between Contract and
Direct Workers.
Direct workers average wage Contract workers average Ave
per day (Rs.) wage per day (Rs.) Wage Differential Ratio rage
State/UTs Rati
2000- 2005- 2010- 2012- 2000- 2005- 2010- 2012 2000 2005- 2010 2012-
Name o
01 06 11 13 01 06 11 -13 -01 06 -11 13
Jammu &
Kashmir 135 142 224 297 89 105 183 240 0.66 0.74 0.82 0.81 0.76
Himachal
Pradesh 120 154 260 330 77 110 234 280 0.64 0.71 0.90 0.85 0.78
Punjab 125 166 245 314 94 112 174 258 0.75 0.67 0.71 0.82 0.74
Uttarkhand 317 303 329 411 92 115 214 271 0.29 0.38 0.65 0.66 0.50
Haryana 186 223 353 435 97 135 252 294 0.52 0.60 0.71 0.67 0.63
Delhi 138 171 295 434 124 140 271 342 0.90 0.82 0.92 0.79 0.86
Rajasthan 137 169 273 388 111 120 211 299 0.81 0.71 0.77 0.77 0.77
Uttar
Pradesh 151 183 294 355 94 118 191 242 0.62 0.64 0.65 0.68 0.65
North 164 189 284 371 97 119 216 278 0.59 0.63 0.76 0.75 0.68
Bihar 174 202 477 306 58 83 138 206 0.33 0.41 0.29 0.67 0.43
Assam 93 124 250 264 79 89 144 244 0.85 0.72 0.57 0.92 0.77
West Bengal 189 221 310 381 121 123 210 289 0.64 0.56 0.68 0.76 0.66
Jharkhand 296 360 517 801 293 118 229 245 0.99 0.33 0.44 0.31 0.52
Odisha 214 305 435 579 74 89 231 293 0.34 0.29 0.53 0.51 0.42
East 193 242 398 466 125 100 190 255 0.65 0.41 0.48 0.55 0.52
Chattisgarh 228 229 370 678 138 138 196 280 0.61 0.60 0.53 0.41 0.54
Madhya
Pradesh 167 209 322 413 122 103 208 268 0.73 0.49 0.65 0.65 0.63
Central 197 219 346 545 130 120 202 274 0.66 0.55 0.58 0.50 0.57
Gujarat 158 205 304 376 102 136 232 299 0.64 0.66 0.76 0.79 0.72
Maharashtra 229 288 446 547 108 144 264 350 0.47 0.50 0.59 0.64 0.55
West 193 246 375 462 105 140 248 325 0.54 0.57 0.66 0.70 0.62
Andhra
Pradesh 137 182 310 423 49 77 165 193 0.35 0.42 0.53 0.46 0.44
Karnataka 153 185 333 454 103 140 258 390 0.67 0.75 0.78 0.86 0.77
Kerala 143 164 250 330 128 127 190 265 0.90 0.77 0.76 0.80 0.81
Tamil Nadu 119 143 248 327 88 131 268 316 0.74 0.91 1.08 0.97 0.92
South 138 169 285 383 92 119 220 291 0.67 0.70 0.77 0.76 0.73
Other States 127 169 305 361 89 122 210 292 0.70 0.72 0.69 0.81 0.73
All India 164 198 313 404 90 116 221 286 0.55 0.59 0.70 0.71 0.64
15 4 The Journal of Industrial Statistics (2016), 5 (2), 154 - 178

Employment-Productivity Profile and Labour Demand Elasticity: A

Preliminary Study of the Organized and Unorganized Indian Textile and
Garment Firms
Sarmishtha Sen1, Syamsundar College, Burdwan, India
Subrata Majumder, Sundarban Mahavidyalaya, South 24 Parganas, India

Abstract

The present work seeks to trace the pattern of change in labour demand by both organized
and unorganized T&G enterprises in India. Changes in average labour demanded across
different size-groups of firms classified by firm-specific average labour productivity as
well as intermediate input costs have been analysed. The changing share of these groups
in estimated aggregate labour demand is also noted and the presence of some kind of
stickiness in adjustment of labour demand with respect to productivity performances
comes out. Estimation of labour demand elasticity all-India level and for selected textile-
major states suggested the limits of price signals in guiding such optimal adjustments. It
is necessary to trace the possible sources of this stickiness in the structural rigidity
inherent especially in the vast unorganized segment. Such an analysis has to treat also
the endogeneity arising from two-way causality between efficiency-enhancing policy
efforts and the two segments’ existing employment and growth behaviour.

1. Introduction

1.1 Recent interventions in the Indian textile and garment (T&G) industry and
especially the National Textile Policy 2000 changed the pre-1985 policy approach radically
by gradually removing the major protective measures for the small-scale decentralized
section while withdrawing restrictions on their large-scale counterpart (Srinivasulu 1996,
Roy 1996, 1998, Niranjana & Vinayan 2001, Galab et al 2009, Kathuria and Mamta 2012). This
process signified a shift of two types: first, from a policy era with greater emphasis on
employment objective to one – explicitly focused on efficiency enhancement based on
equalization principle of the market, and secondly from policy-guided inter-firm linkage to
market-driven connectivity between organized and unorganized segments. Deregulation
of the industry including de-reservation of the garment sector, the latter initiated by the
Textile Policy 2000 and mostly completed by 2005, is expected to have removed major
blocks to profitable restructuring by the large units - including those with 100% FDI - while
allowing them entry practically in every area of T&G production (Singh and Sapra 2007,
Chaudhary 2011). With efficiency-enhancing optimal re-allocation of resources by organized
units becoming easier and possible now, we can expect a significant change in the
employment generating potential of these two constituent segments of the Indian T&G
industry especially in the post-MFA period. Increased scope for firm-level optimal
adjustment of labour demand suggests that greater part of this employment be concentrated
in the relatively high-productivity segments of the industry.2

1
e-mail: [email protected]
2
Assuming a perfectly elastic labour supply, it is plausible to argue that observed changes in employment
are guided mainly by the hiring behaviour of the firms.
 Employment-Productivity Profile and Labour Demand Elasticity .... 15 5

1.2 Increased trade-liberalization is expected to have offered significant opportunities

for expanding production activities especially in the organized segment. There are evidences,
on the other side, of unorganized units becoming a part of this technical and structural
reorganization through the large-small linkage like subcontracting (Maiti 2008, Sahu 2011).
Labour demand in the unorganized segment may fall due to substitution effect arising from
use of more sophisticated production techniques associated with such reorganizations. If
the organized segment however, contracts out the relatively labour-intensive operations to
these enterprises, there is a high possibility of increase in labour demand in the latter due
to scale effect. This potential however may not be realized if there is little increase in
outsourcing by large units interested in more intense supervision of production and quality
control through in-house production.3 The unorganized units in such a situation may
experience very slow expansion of production and labour demand eve in absence of any
significant substitution effect.

1.3 Thus, we can trace this process of adjustment at two levels: through following
changes in aggregate labour demand at the sectoral level, and by studying the micro-level
process of making optimal hiring decisions. Through an exploratory analysis, this paper
seeks to examine the pattern of change in aggregate as well as sector-specific demand for
labour across the two constituent sections of the Indian T&G firms in the post-2000 years.
Further, it attempts to evaluate how individual firms in different locational set-ups adjusted
their demand for labour during this period, in response to changes in wage. The estimation
of elasticity of labour demand (LD) for selected Indian states with considerable presence of
organized and unorganized T&G units may help us locate the significance among others of
various policy parameters in inducing this demand. The broad research questions are the
following:

 How did the distribution of LD in the T&G sector and in its organized and
unorganized parts change in the period following Textile Policy 2000? Is there any evidence
of direct association between productivity & change in LD in the period following the
announcement of National Textile Policy 2000?

 How does firm-level labour demand adjust with wage & how do relevant locational
parameters & policy factors influence labor demand elasticity? Is there any variation across
the states?

1.4 The first part of the analysis assumes importance in view of a stagnant productivity
performance in the organized T&G segment and an actually falling relative performance of
even their bigger unorganized counterparts during the first decade of the present century
as noted by Sen and Majumder (2015). In this context, can we expect a direct association
between productivity growth and higher employment due to movement of employment
from low to relatively high-productivity sub-groups expected in optimal re-allocation of
resources made possible by reforms? Here, we treat the observed employment as the
amount of LD assuming a perfectly elastic labour supply especially in the vast unorganized
segment with approximately fixed wage. More rapid rate of increase in LD among the low-
productivity groups however, would indicate the presence of some kind of stickiness in

3
Reduced traditional marketing support from government is expected to have increased dependence of
the unorganized units for access to market on the large and organized ones.
15 6 The Journal of Industrial Statistics, Vol. 5, No. 2

adjustment of labour demand despite poor productivity performances. Next, we examine

the firm-specific adjustment of labour demand by Indian T&G units - both organized and
unorganized - in response to wage while controlling other factors potentially influencing
labour demand in thirteen textile-major states.

1.5 Rest of the paper is organized as follows: Section 2 delineates, based on the
available empirical findings, the possible link between employment, labour productivity
and labour demand elasticity in the organized and unorganized T&G segments. The next
section entitled ‘Methodology and the Database’ first discusses the methodological
requirements of this kind of an issue. Variables and the source of data used for this purpose
are described in two following subsections in it. Section 4 presents the findings of both the
exploratory analysis and the regression exercise respectively in two subsections while
section 5 contains the concluding observations.

2. Links between Employment and Productivity: An Outline

2.1 Using the Flow Chart 1, we attempt to explain the possible channels through
which segment-specific labour demand and its average labour productivity may be related
in the context of heightened market-based competition in the domestic economy along with
increased integration with the global economy. Trade liberalization and complete abolition
of MFA require that the T&G units, both organized and unorganized, enhance productivity
and competitiveness. Indian textile production base is known to have suffered from
accumulated structural weaknesses resulting from fragmentation of productive capacity,
lack of technological upgrading especially modernization, and absence of advantages in
terms of scale economies. Restrictive textile policies of the pre-1985 era did not allow
capacity expansion or changing input-mix in the organized segment according to the
requirements of production efficiency. Thus such a policy strategy was viewed as an
obstacle to the achievement of export-competitiveness. The restrictions on the large
organized units for optimal restructuring - through capacity expansion or capital imports
and subsequent capital deepening process - were lowered in the post-1985 years especially
after the National Textile Policy 2000 had allowed entry of large units in garment production
including 100% automatic approval of FDI. Organized sector in most of the Indian
manufacturing industries have resorted to increased use of labour-displacing technologies
despite the remaining labour laws as has been noted by a number of empirical studies in the
recent years (Nagaraj 2004, Rani and Unni 2004, Banga 2005, Sen and Dasgupta 2006, Das
and Kalita 2009, among others). This may have exerted a negative pressure on employment
through substitution effect.4

2.2 Trade liberalization on the other hand, especially in the post-MFA years, offer
greater scope for venturing in markets hitherto restricted by MFA quota. Traditional trade
theories would suggest an expansion of the labour-intensive exporting segment. In fact,
Indian T&G exports are known to have created a niche market of their own while producing
innovation- or design-intensive high value-adding customized items. It has achieved export
success particularly in the high-end fashion segment on the basis of economies of scope.
The T&G enterprises are also catering to the expanding domestic, especially urban market

4
Increased use of capital may have been prompted by absence of skilled labour while increased capital
intensity may also be a reflection of increased labour productivity.
 Employment-Productivity Profile and Labour Demand Elasticity .... 15 7

of consumers with rising purchasing power and demand for tailor-made and highly
differentiated T&G products (Roy 1998, Teewari 2005, Singh and Sapra op cit. for example).
The expanding market base presents opportunities for greater specialization, efficiency-
enhancing restructuring and improvement in labour productivity at least in certain segments
of the organized T&G units.5

2.3 Organized segment however subcontract a part of its operations, especially the
labour-intensive ones, to the unorganized segment as a means to achieve cost-
competitiveness. There are empirical evidences indicating use of increased share of low-
cost labour in the Indian T&G sector as a cost-cutting and efficiency-enhancing strategy
especially in the years after the implementation of Agreement on Textiles and Clothing
(Abraham & Sasikumar 2010, for instance). Maiti and Marjit (2008) on the other hand
argued that with increased international price of T&G product, it becomes profitable for the
organized section of the T&G industry to outsource increasing proportion of production
activities to the unorganized units while retaining and specializing in the marketing-related
activities. Ability to serve the increasingly globalized domestic and international markets -
of such a linkage between organized and unorganized firms however hinges on developing
technical and organizational capability of the latter under the supervision of the parent
firms. In such a scenario it is plausible to argue that export success of the organized parent
unit may be accompanied by an improvement of productivity performance of the unorganized
subcontracting firm. Downward pressure on employment arising from any possible increase
in capital intensity may be outweighed by expansionary tendency of employment in the
segment via scale-effect.
Flow Chart 1
Productivity-Employment Link in Organized T&G and Induced Effect in the
Unorganized T&G Segments

Search for
ORGANIZED T&G EMPLOYMENT
ORGANIZED T&G EMPLOYMENT Greater
Increased Trade openness
Competitiveness (-) (+)

Use of Labor -saving Increased Scale of

Relaxed Optimal Technique Operation Expanding
Restructuring by Domestic Market
ORGANIZED
Organized Specializing Labor-intensive
in Marketing & Operations Outsourced
Monitoring

UNORGANIZED T&G EMPLOYMENT

Survivalist UNORGANIZED: Modern UNORGANIZED:

Productivity-Employment Disconnect & Relatively Rapid Employment Increase in
High-Productivity Deciles

Source: Authors’ Understanding

5
Organized firms are expected to respond more effectively than their unorganized counterpart to these
opportunities.The phenomenon of relatively high increase of employment in labour-intensive exporting
industries has found support in empirical studies (Kakarlapudi 2010, for instance).
15 8 The Journal of Industrial Statistics, Vol. 5, No. 2

2.4 To accommodate the phenomenon of intrinsically heterogeneity and diversity of

the unorganized T&G segment in our analysis, we have divided it in two distinct groups:
one mostly survivalist6 and the other dynamic7 section (following Sen 2016). Enhanced
market-based competition of the recent years and changing policy regime are expected to
influence these two parts differentially. For instance, market incentives are expected to
work better with the latter group, which is better-placed to reap benefits from policy support
too. We can expect a direct association here between productivity improvement and
employment generation through competition as well as collaboration with the organized
segment. The traditional group, on the contrary, is not supposed to display any such
definite pattern as the urge to survive, and not production efficiency, is the driving force
behind its operations.

2.5 Trade liberalization has been found to be associated with increasing labour demand
elasticity in the organized manufacturing industry. Unorganized segment however falls
outside the scope of labour legislation, wage variation may be low and competitive labour
supply does not allow effective assessment of the effect of wage-variation on demand for
labour.8

3. Methodology and the Database

3.1 Methodology:

3.1.1 In the first exercise, we examine the change in segment-specific labour demanded
(LD) across size-groups based on level of unit-specific average labour productivity as well
as on the firm-level expenses on intermediate inputs. We undertake further an exploratory
analysis of the inter-relationship of change in average labour demand with change in
productivity and changing share of different size-groups in estimated aggregate labour
demand. This part of the analysis makes use mainly of summary statistics on average
labour productivity (APL) and LD to demonstrate whether or not any increase in LD is
located in the relatively growing section of T&G firms during the post-2000 years.
3.1.2 The issues involved in the estimation of labour demand elasticity are mainly
twofold: specification of demand function and choosing the proper estimator, and selecting

6
This group refers to typically low-productivity units either operating independently in the local market
or is connected with distant markets through traditional and often exploitative chain of intermediaries.
These firms generally lack the preparedness to adapt with efficiency-improving market signals or to
respond to policy support.
7
This group consists of units, generally mechanized and modern, linked with globalized market through
collaborations with large-scale sector resulting in technological spillover, increased capital intensity and
enhanced production performance.
8
If increased wage is a reflection of use of labour with higher productivity (evident from fall in the share
of low-productivity self-employment, (Goldar and Sadhukhan 2015)) in the unorganized section and
scale effect dominates substitution effect we may not observe the effect of wage rise as expected on
labour demand. Increased exposure to trade may give rise to asset-specificity and the export obligations
may call for long term relationship with the skilled-worker. In this case considerations other than wage
may dominate employment decisions of the unit concerned). In fact, elasticity of wages with respect to
productivity was higher than unity in most of the organized manufacturing industries including textiles
between 1990 and 2006 (Mitra 2013). Nagaraj (2004) observed that relative cost of labour did not seem
to have any influence on employment decisions even in organized manufacturing at the turn of the
present century.
 Employment-Productivity Profile and Labour Demand Elasticity .... 15 9

the variables for estimation and consideration of their characteristics. Assuming an unlimited
labour supply and identical technology with varying factor price ratio (assuming comparable
unit cost of capital and different labour remuneration rates) across the organized and
unorganized segments we attempt to estimate segment-specific labour demand elasticity
from the following labour demand function (following Hasan et al 2007 and Goldar 2009):

... (1)

where, (with competitive labour market), (total emoluments per labor-

day), is the logarithmic value of the wage series; (-), =organized segment dummy
(+), = proportion of subcontracting big establishments in unorganized T&G
segment in a state (+), G=garment-dummy (‘0’ if textile) (-), =effort toward productivity
improvement through capital structure (simple mean of unit’s ranks in terms of log Capital-
labor ratio, capital in land & non-land capital-intensity) (+), =effort on productivity
improvement through input structure (simple mean of the units’ ranks by & log
material costs) (-), = simple mean of % spending on electricity in total output (exceeding
1%) and state-level proportion of such power-driven establishments in the unorganized
&G segment, Y=output value-added of a firm (+); =random error..

3.1.3 As individual firm’s labour demand and its output (a specific measure for T&G
value-added of a firm) are simultaneously determined within the model, both are likely to be
correlated with the random error term 9. This may result in biased and inconsistent estimates.
Use of instrumental variable (IV) regression method can give us consistent estimators by
reducing unobserved or hidden bias especially in the observational studies like this. An
instrumental variable has to satisfy two criteria: it has to be highly correlated with the
problematic endogenous explanatory variable but should not affect the outcome
independently.

3.1.4 We use instrumental variable regression method with the help of two-stage least
squares (2SLS) estimator where the estimation is done at two stages. Suppose, the population
regression relating the dependent variable and the output-regressor is as follows:

... ... (2)

Where represents other covariates, and is the error term representing omitted factors
that determine . A non-zero correlation between and , due to endogeneity for example,
will yield inconsistent estimates. At the first stage, the variable under consideration, firm’s
output is decomposed into two parts: a part that is not correlated with the random error and
therefore is non-problematic, and the other that is correlated with the error term. Here
individual firm’s output is regressed on the chosen instrument, which is correlated with
9
This happens due to possible reverse causation of labour demand to output giving rise to the problem of
endogeneity. Endogeneity of labour price or wage is taken care of by the assumption of perfectly elastic
labour supply typical of a developing economy.
16 0 The Journal of Industrial Statistics, Vol. 5, No. 2

but not with . The second stage estimates the concerned regression coefficient with the
help of this exogenous variation component i.e. the problem-free part of variation in firm’s
output. Thus, the first stage of the regression links firm’s output with the chosen instrument,
.

... ... (3)

Here, is the intercept, the slope and the error term. From the regression, we have the
predicted variation in firm’s output, as - the part uncorrelated with the error term
in the original regression. In the absence of actual regression coefficients and the
second stage regresses on this predicted variation in i.e. . As is not
correlated with a simple OLS will give us consistent estimate of i’s.

3.1.5 In this research exercise, we attempt to estimate the labour demand function by
the 2SLS IV estimator. At the all-India level we tried two alternative instruments: Y1_ins=per
capita Net State Domestic Product at constant prices10 as has been used in similar studies
of labour demand estimation & Y2_ins=each firm’s share of NSDP apportioned according
to its share in total GVA of these establishments for each state).11 As we are working on two
repeated cross-section database we estimate the elasticity with the instrumental variable
estimator separately for the years 2001 and 2011. We carried out this regression analysis for
sample units in India as a whole and separately for thirteen selected states with major
presence of T&G establishments in both organized and the unorganized segments (details
given in Table 11) to examine state-specific variations. Variables e.g. LINKAGE (proportion
of unorganized units linked with bigger firms through subcontracting), TECH (the average
degree of mechanization of unorganized establishments in a state), and per capita NSDP
have single values for each state. So these cannot be used in the state-level regression
exercise.

3.2 Choice of Variables:

3.2.1 Studies explaining the pattern of Indian manufacturing employment growth and
those estimating labour demand have concentrated on the following factors: the indicators
marking labour market rigidity, increased trade-openness encouraging adoption of capital-
intensive and often imported production techniques as well as cheaper inputs, presence of
major supply-side bottlenecks in different states (Nagaraj 2004, Das and Kalita 2009, ICRIER
2008, Kannan & Raveendran 2009, Nataraj 2010, Goldar 2011, Mitra op. cit.). Unlike these
works (generally dealing with the organized manufacturing sector, Nataraj 2010 being an
exception), any analysis of changing labour demand condition of one particular industry
has to take in explicit consideration the specific interaction of that industry’s structure with

10
This measure, representing the state’s level of prosperity, influences the overall T&G production of the
state but is expected to be minimally affected by the T&G industry-specific developments, which are
likely to influence the sector’s labour demand also.
11
Pairwise correlation between firms’ output measure and the proposed second instrument was quite high
at least at 5% level of significance.
 Employment-Productivity Profile and Labour Demand Elasticity .... 16 1

changes in market conditions on the one hand, and the same with relevant policy parameters
there12 on the other.

3.2.2 The analysis has to be sensitive to potential variation in effect of same structural
or policy-level change in an estimation that covers both the organized and unorganized
segments. Change in policies instilling greater flexibility in the labour market may have
differential impact on the two segments. Similarly, increased trade-openness is also expected
to influence the two segments’ response in terms of production and employment differentially,
especially depending on the concerned unit’s structural location with respect to new market
opportunities. The impacts may vary less if production in the organized and unorganized
sections is linked through subcontracting.

3.2.3 These interventions - through influencing productivity - may affect the labour
demand of the organized and unorganized sections of the industry. As noted in the
introductory section, recent policy efforts targeting the Indian T&G sector have attempted
to establish a strong connectivity of large and small firms through subcontracting. Thus,
the extent of variation in subcontracting percentage would indicate the extent of organized-
unorganized linkage as well as the pro-competitive stance of the policy regime. The process
of mechanization (alongwith other infrastructural factors) adds to the speed of production
and increases the scope of these linkages by ensuring adhering to the delivery schedule as
well as the quality-specifications of crucial and big national/global buyers. Indicator of
mechanization - a proxy of productivity enhancing technology-centered effort of the firm,
based on the proportion of electricity consumption in T&G units especially in the
unorganized segment will be an important factor to be controlled. Individual firm’s
organizational location as well as the product-group catered by it (textile or garment, for
instance) serves to incorporate segment-specific variation in the Indian T&G industry at
the unit level and across states. Following Biswas et al (2014), we have controlled various
aspects of firm’s productivity-enhancing effort through the channel of capital (by
constructing a composite index of different measures of capital i.e., different sub-categories
of capital) and channel of other non-capital inputs in the LD elasticity estimation.

3.2.4 Due to non-accessibility of systematic and direct information on many such factors
over time, we have tried to control or reduce the extent of heterogeneity by controlling
various structural and location parameters of the sample units.13 Thus, we incorporate
structural location of the firm: the segment it operates in (ORG), the product-group it

12
We have not included policy variables, such as amendments to the Industrial Disputes Act, the reforms
introduced to scale up the social security contribution by the firms or to dilute them effectively as a direct
measure of rigidity in the structure of the labour market due to our limited access to such data. These
changes, labour being a state subject also, can vary across the states considered here. Neither had we had
access to data required for constructing measures e.g. index of state-business relation or investment
climate (as in Caliet al 2009, for example). We could only devise indirect measures against those potential
influences. Similarly, information on industry-specific and productivity-enhancing interventions (such as
Textile Upgradation Funds Scheme) targeting conduct of the firm operating in the market and marketing
helps to aid optimal response to market-driven forces in presence of pre-market constraints, might be
relevant. Firm-level data used here however do not provide such information.
13
Use of direct information is also constrained by availability of comparable statistics on the same
variable at the two ends of the study period. For example, we could not construct any such comprehensive
index of state-specific policy efforts and structural features using indiastats.com.
16 2 The Journal of Industrial Statistics, Vol. 5, No. 2

produces and markets (G), degree of labour-market rigidity (RGDT)14, extent of connectivity
between unorganized units with bigger concerns (LINKAGE).15 The last two factors also
reflect the current policy inclination in respective state’s textile and garment sector. On the
other side, we have controlled firm’s efficiency-enhancing effort in the production sphere
like KP16, INP and extent of mechanization. By concentrating only on the relatively big T&G
establishments and excluding unorganized T&G firms with no hired worker from this
estimation exercise, we can somewhat reasonably treat the effect of trade-liberalization
policies symmetrical across the two segments.17 This exclusion allows us to focus on
changes in domestic structure of the industry especially in terms of variation in the effect
of policy on the labour demand at the state level where the relevant policies include existing
labour laws and pro-competitive interventions meant for the T&G sector.

3.3 Database:

3.3.1 Unit level data on the organized factories from the Annual Survey of Industries
(ASI) database were pulled together with the information on unorganized manufacturing
units provided by the successive rounds of NSSO (using the framework of Sen and Majumder
2015, summed up in Table A1). The exact choice of time points was guided by our access
and availability of unit level data and the comparability of the two sets of industrial statistics.
The data are in the form of repeated cross-sections and our study period refers to the years
between 2001 and 2011.

3.3.2 Each observation in our data set collates information on a number of variables
including those on input and output bundles for different individual industrial enterprises.
Necessary adjustments were made to ensure comparability of categories representing the
same variables. We have considered only the perennial enterprises in the unorganized
sector, which operate regularly throughout the year to make the comparison more compatible.
For the same reason, we have only included unorganized firms hiring worker on a regular
basis. Two full-time hired workers were taken as the relevant cut-off to get the final set of
observations as in Sen and Majumder 2015. Similarly, both the data sets were converted at
2004-05 prices - for the exploratory part - by deflating with the help of wholesale price index
(WPI) for ‘Manufacturing Products’18. The pre-reform analysis has been carried out on the

14
Scope of profitable adjustment in labour demand in response to changing wages is expected to be limited
by the extent of rigidity prevailing in the state’s labour market thereby putting a downward pressure on
labour elasticity of demand
15
LINKAGE acts here roughly as a proxy of state-specific textile policy effort and labour market
flexibility (i.e., indicator of the general state of development in the T&G industry in each state - as that
would indicate the scope for restructuring the entire value chain spanning both the segments);
16
KP - (capital pillar) indicates the nature of technology development through capital structure. The
interactive terms of wage with capital-based effort and other inputs-based effort are likely to capture the
effect of these channels generally on employment generation and specifically on labour demand elasticity
17
The one- or two-worker units (known as own account manufacturing enterprises, the OAMEs) were
found to face significant participation constraints leading to lower incidence of large-small linkage there
(Sen and Majumder 2014).Estimation of labour demand elasticity also loses relevance somewhat in units
with preponderance of unpaid worker and with difficulty in distinguishing between the worker’s and
entrepreneurial roles.
18
To simplify we have not resorted to use of multiple indices although use of WPI for textile products and
WPI for textiles machinery or that for general machinery to deflate output and fixed capital stock
respectively would be more appropriate. However, this is a limitation of the measures used in this work.
 Employment-Productivity Profile and Labour Demand Elasticity .... 16 3

basis of 9056 unit level observations and the analysis of the later period uses 6522
observations19. Composition of the firms however changed in the estimation of LD elasticity
as the latter required fulfilling different criteria of analysis. Per capita NSDP value was
downloaded from the RBI Handbook of Statistics.

4. Findings:

4.1. Change in LD across Productivity-groups and Firm-size: An Exploration:

4.1.1 In this section, we attempt to examine whether or not any increase registered in
labour demand is concentrated in the relatively high productivity firms or in the higher size-
groups (based on firm-level expenses on intermediate inputs) in the industry. We start with
checking if average labour demand - measured in labour-days generated - has increased in
the relatively high productivity sectors or in the low-productivity ones. For this, we have
classified the sample units in different size groups based on the values of average labour
productivity (APL). In Table 2, we have divided all the T&G firms into four quartile groups
of APL distribution. Summary statistics of LD - obtained after applying population weights
- shows that greater mean LD values were recorded in the upper two productivity quartiles
by T&G units of these two sections taken together in both the years. Just the opposite was
true for the unorganized segment at both time-points while organized section displayed the
same pattern in the last year. Median LD values followed the same pattern of recording
greater values in the lowemost quartiles both for all the T&G units of the industry and for
its unorganized part (the same table). For the organized segment, median value was relatively
high in the top quartiles, particularly in the second and third quartiles.

4.1.2 Table 3 shows that increase in mean LD over time seems to have been concentrated
in the rather higher productivity groups. Disaggregation into the organized and unorganized
segments however indicates that the increase was greater in the lower productivity segments
of the organized part of the industry20. On the other hand, firms at both the lower and higher
productivity ends of the unorganized segment seem to have experienced this rise21. Reliability
of these findings however seems to be compromised as the coefficient of variation of the
extreme two productivity quartiles is quite high (see Table 4, for the selected summary
statistics) i.e., the firms in those quartiles are not homogenous in terms of the criterion of
grouping viz. average labour productivity.

4.1.3 Thus, we disaggregated the units further into productivity-deciles and tried to
examine the pattern of change in LD across these deciles. It is clear from the Table 5 that the
firms constituting different deciles (except the topmost one22) of the T&G enterprises and
their organized and unorganized sectors are relatively homogenous groups in terms of their
respective productivity performances.

19
The final data set was obtained after excluding firms with non-positive values of output (the relevant
measures-gross output or GVA) and/or inputs (capital, labour and material-fuel) to make it more reliable.
20
This was also true when we considered the rise in median LD values across the productivity quartiles of
the organized segment (as Table A1 in the appendix shows).
21
Median LD also followed a similar pattern for the unorganized units and for all the sample units taken
together.
22
The concerned CV has however decreased significantly at the last time-point.
16 4 The Journal of Industrial Statistics, Vol. 5, No. 2

4.1.4 Table 6 presents the change in mean values of estimated LD (obtained again by
applying population weights) across productivity deciles at the two time-points concerned.
It does not throw any clearly discernible pattern of relationship between the productivity
performances of the firm and change in estimated labour demanded by it. There was rise in
mean estimated LD in almost all the productivity deciles (except the third decile in the entire
set of T&G firms as well as fourth and eighth productivity deciles of sample organized
units) during the study period. The increase of the mean value for all the T&G units was
concentrated towards the medium deciles especially sixth through ninth deciles. Estimated
mean LD for the unorganized establishments increased more at both lower and upper ends
than around the middle of the productivity distribution, as is expected from our
conceptualization of the segment in Section 2. The pattern remains the same even if we
consider the median LD (see Table 7 presenting selected descriptive statistics for the
distribution of estimated LD).23 The pattern is not so clear for organized segment: the rise
was greater in lowest three deciles, generally low in the topmost three deciles and fluctuated
around the medium deciles. Thus, changes in both mean and median LD values do not
indicate any definite relation between size-groups of productivity and change in average
level of LD for the organized units during the study period.

4.1.5 The other possible way to get an idea of the distribution of LD across productivity
deciles is to check if greater part of labour demand generated now is concentrated in the
higher or lower productivity deciles than before. Thus, we proceed to examine whether
firm-level change in LD has translated in any improvement in the share of higher productivity
deciles in total labour demand. Table 8 presents the changing share of productivity deciles
in total LD over time, taking all T&G units together and the organized and unorganized
segments separately. The sample units in the sector as a whole definitely show an increased
concentration of share in total LD around the middle section of the productivity distribution.
Once we locate the changes in the two constituent segments, we find a clear deterioration
in the distribution of estimated LD as the share has increased mainly in the lower productivity
deciles of organized (except for the 7th decile) and especially unorganized T&G firms. The
striking observation here is that of a significantly high increase in share of the topmost two
productivity deciles in the unorganized part indicating the possibility of a sustainable
growth of firm in these productivity size-groups.

4.1.6 Next, we carry out a similar exercise by examining distribution of estimated LD in

both the years across size-groups of firms classified in terms of firm-level expenses on
intermediate inputs.25 Firms were grouped initially in quartiles as well as deciles based on
this criterion. However, the coefficient of variation in intermediate input costs is very high
across both the quartiles and the decile-groups. To make the comparative analysis more
meaningful we have resorted to a classification by trial whereby all the sample units were
categorized in 17 clusters. Our aim in this categorization was to retain as many observations
as possible in the analysis for both the time-points without allowing a very high coefficient
of variation of intermediate input costs in individual clusters. Table 9 presents the details of
this classification and the selected summary statistics for expenses on intermediate inputs
23
This was true also for all the sample units.
25
This seems better than the possible alternative schemes of size-classifications - based on size of
employment or by stock of fixed capital. This is so because size-classification based on intermediate input
costs is expected to minimize the problem of slack or underutilization.
 Employment-Productivity Profile and Labour Demand Elasticity .... 16 5

in these groups. Findings from the exploratory analysis of change in estimated mean LD and
share of the groups so obtained in total LD are presented in Table 10.26

4.1.7 Almost all the size groups experienced fall in mean APL but quite a few of them
registered positive change in mean LD. Considering all the T&G firms in the sample together
we find clearly that the increase in mean LD is higher around the middle-sized firms. In fact
the medium-sized firms had recorded higher share of LD in the aggregate demand for labour
in both the years. Notwithstanding the declining productivity performance in most of the
size-groups, a few also exhibited increase in their respective shares in total labour demand.
It is clear, however from the above discussion that there is no definite and positive
correspondence between the change in LD and level as well as over time growth in
productivity in organized sphere of the industry. Employment growth in the unorganized
part however seemed to be concentrated towards the two ends of productivity distribution.
This probably indicates at some kind of stickiness in optimal adjustment of LD according to
productivity performances in different segments of T&G firms.

4.2 Estimation of Labour Demand Elasticity:

4.2.1 With the help of two-stage least square estimator we have carried out an
instrumental variable regression analysis for estimating the labour demand elasticity. We
could incorporate all the variables enlisted in the section 3.1 only in the LD estimation at the
all-India level. Here we have reported only the results of the second stage of instrumental
variable regression (presented in Table 12)27. As mentioned earlier, we tried two possible
instruments against firm output for the all-India regression: one used in similar studies on
estimating labour demand elasticity (Y1_ins), the per capita net state domestic product
(PCNSDP); and the alternative instrument devised in our study (Y2_ins).28 What we find
here is that the labour demand elasticity had the expected negative sign at the first time
point while it was statistically insignificant at the end of the study period. The responsiveness
of LD with respect to wage increases significantly if the concerned firm was in the organized
sector while it decreased when the firm belonged to a state with greater degree of connectivity
between organized and unorganized segments. The direction of influence on the same
responsiveness of LD however changed during the study period for factors such as labour
market rigidity, individual firm’s effort in terms of improving inputs, degree of mechanization
of the state in which the firm was located. Firms had responded more through adjusting LD
to wage changes at the first time-point while it responded less at the end of the study
period in states with higher labour market rigidity or in states with greater extent of
mechanization in the unorganized segment.29

4.2.2 Labour demand elasticity became significantly lower with firm’s effort toward
improvement in the basic non-capital inputs it used but did not differ significantly with its
effort in terms of improvement in the capital stock and capital intensity in the first period. It
26
Similar results for the mentioned quartiles and deciles are not reported due to the reason specified here.
27
Results significant at least at 5% level are discussed only.
28
It was explained earlier in section 4.1. The first instrument however, never proved to be statistically
significant with alternative measures of labour market rigidity or even after excluding the regressors with
t-values less than one. Thus, we had reported only the results of the regression using the second instrument.
29
Labour market rigidity does not allow optimal adjustment in labour demand despite wage rise. Higher
extent of mechanization in the unorganized segment probably facilitates greater subcontracting by
organized units.
16 6 The Journal of Industrial Statistics, Vol. 5, No. 2

is noteworthy however that the elasticity increased significantly at the last time-point with
enhanced effort toward improvement on both input and capital fronts. This seems
encouraging as very little lowering of wage can bring about a substantially high and
statistically significant increase in labour demand through complementary efforts in terms
of physical capital and other inputs e.g. raw materials. Individual firm’s location in the
garment sector lowered the effect of one percent change in wage on the LD adjustment in
the first year possibly indicating the relatively greater dominance of non-wage factors
(especially stronger scale effect) in such adjustments compared to that among the textile-
producing firms. The product-segment wise location of the firm however had no statistically
significant effect on the labour demand elasticity at the latest year under consideration.

4.2.3 Results of estimation of labour demand through the 2SLS-IV regression for the
selected thirteen states with major presence of both organized and unorganized T&G units
are reported in Table 13 (second state dealing with the instrumented variable). Only the
second instrument Y2_ins is used in this regression exercise. Like in the all India-level
regression the instrument concerned was found to be positively significant in all states and
at both the time-points.30

4.2.4 Adjustment of labour demand with changing wage was of varied degree and in
different directions in the selected states. Estimated labour demand elasticity had the
expected negative sign in seven out of the 13 states in 2001(Andhra Pradesh, Gujarat,
Karnataka, Kerala, Maharashtra, Uttar Pradesh and West Bengal). But in no state it was
negative significant in 2011. The value was positive and statistically significant only in
Tamil Nadu in the first year, while it became positive significant also in Delhi, Karnataka,
Maharashtra, Rajasthan and West Bengal at the end of the study period. Among the latter
set of states, Delhi and Rajasthan displayed statistically insignificant elasticity value in
2001. The estimated elasticity was insignificant at both time-points in Haryana, Madhya
Pradesh and Punjab. In the states viz., Andhra Pradesh, Gujarat, Kerala and Uttar Pradesh,
labour demand elasticity was negative significant in 2001 and it became statistically
insignificant in the latter year. This indicates again the possibility of non-wage factors
dominating individual firm’s optimal adjustment of LD.

4.2.5 There was a few uniform patterns also, similar to the one exhibited in the estimation
for all-India T&G labour demand function. For example, labour demand elasticity always
increased when the firm in question operated in the organized segment rather than in the
unorganized part of the industry.31 This increase was statistically significant in all the
states at both ends of the study period. Labour market rigidity generally exerted an expected
negative significant effect on the elasticity values (except registering a positive significant
effect in Punjab, Rajasthan and West Bengal and without any significant effect Gujarat and
Kerala in 2001). It was negative and statistically significant at least at 5% level in all the
states in the last year.

4.2.6 Similarly the responsiveness of LD to changing wages was lowered by the efforts
toward improvement of non-capital inputs (except an insignificant effect in Andhra Pradesh
30
The coefficient value increased for Karnataka, Rajasthan and Tamil Nadu, remained almost the same
in Andhra Pradesh, Delhi,Gujarat, and Kerala, and declined somewhat in the others.
31
As is expected from our discussion in Section 2, labour demand elasticity has actually increased at
the end of the study period for a typical organized unit in all the states and at the all-India level.
 Employment-Productivity Profile and Labour Demand Elasticity .... 16 7

in 2011).32 This might indicate that further possibility of enhancing labour demand by
reducing wages may be limited if the effort in terms of improving non-capital inputs is
already at a higher level. Operating in the garment sector also seems to have exerted a
statistically significant negative influence on estimated labour demand elasticity. The effect
was negative significant at both the time points in the states of Maharashtra, Rajasthan,
Uttar Pradesh and West Bengal. It was negative significant in Andhra Pradesh, Gujarat and
Kerala in 2001 but later became statistically insignificant. On the contrary, it was insignificant
in the first year and became statistically significant (-) at the end of the study period in
Madhya Pradesh and Tamil Nadu. The elasticity value did not differ significantly depending
on the product-markets catered by the firms in both the years in the remaining states. Why
labour demand adjusted in particular ways with changing wages and why the estimated
labour demand elasticity was influenced by relevant locational and other factors in the
above-discussed manner in the selected states however, needs a thorough mapping of
more concrete state-specific characteristics and their effects on the labour demand condition
in a comparative framework. This however lies beyond the scope of this preliminary analysis.

5. Conclusion

5.1 Indian textile policies in the recent years have gradually removed most important
protective measures for decentralized and significant restrictions on the organized sectors
giving way to the free interplay of market forces. In the near absence of growth-restricting
regulations, optimal re-allocation of productive resources including labour through greater
connectivity between the organized and unorganized segments seems to be an ongoing
process. It is then logical to expect that the segments with lower-productivity and efficiency
will be outcompeted and labour demand will mainly concentrate in the relatively growing
section of T&G firms. On this backdrop, the present paper seeks to trace the pattern of
change in employment in both the organized and unorganized firms in the Indian T&G
industry. For this, we attempted to examine the inter-relationship between change in labour
demand and the same in average labour productivity. Steeper rise in mean LD as well as
share in aggregate LD was concentrated in groups of all T&G units, classified by intermediate
input costs, which also registered substantial fall in APL. Disaggregated exploration for the
organized and unorganized segments indicated no definite pattern in the organized part
while both mean LD and share in aggregate rose in the extreme size-groups of the unorganized
T&G establishments. Thus the exploratory analysis pointed at the presence of some sort of
stickiness in the behaviour of certain segments e.g. the survivalist unorganized set, with
respect to productivity performances necessitating the study of firm-specific employment
behaviour.

5.2 This prompted us to inquire the nature of LD adjustment with respect to changes
in market structure and in policy shifts influencing the structure or having a direct bearing
on the firm’s conduct. The particular approach of estimating labour demand function and of
deriving estimate of LD elasticity with respect to wage was adopted while controlling certain
structural parameters. We sought to carry out the estimation using the two-stage least
squares IV estimator at the all India level as well as for selected states with major presence

32
Like in the regression for all India units, we used the effort in terms of only non-labour inputs among
the non-capital components such as raw-materials and fuels as a measure of input-based improvement
effort and found similar results.
16 8 The Journal of Industrial Statistics, Vol. 5, No. 2

of the T&G industry. Certain difficulties were encountered in the estimation across the
textile-major states especially regarding the choice of instrumental variable against firm-
level output and instrument devised for conducting this analysis turned out to be statistically
significant with an expected sign in all cases. What we learnt from this exercise, however, is
that the responsiveness of labour demand to wage in particular and market incentives in
general may be weak (the estimated coefficient of wage i.e., the labour demand elasticity
term turned out to be insignificant at the all-India level and for most of the states). This
stickiness possibly owes its origin to (i) inherent structural rigidities such as labour market
regulations facing the organized group and pre-market constraints faced by the unorganized
enterprises, (ii) lag in adjustment arising from inadequate preparedness of firms to respond
gainfully to market signals. An efficient re-allocation of employment through flexible and
optimal adjustment of labour demand in response to signals from price mechanism and
other market incentives will require, as a precondition, an in-depth inquiry into the source
of the observed stickiness.

5.3 The above findings are subject to limitations imposed by the quality of available
information and variables used in the analysis. Estimation of labour demand elasticity in a
specific sector with the presence of a huge unorganized section also has to introduce finer
distinction between different market forces/policies affecting this phenomenon of stickiness.
Methods of analysis have to be modified further as the estimation covers both the organized
and unorganized segments. The former – being registered under the Factories Act - is a
relatively homogenous group in terms of structural location while the unorganized segment
is characterized by inherent structural heterogeneity. Nature of industry-specific policy
interventions is also likely to vary across the two segments resulting in the error term being
correlated with the policy variable (since segment-specific policy formulation may itself be
motivated by changing employment composition of the two segments). Thus, any future
probe based on this preliminary analysis, needs addressing the possible endogeneity bias
arising from the two-way relationship between any state’s labour-market regulations and
industry-specific efficiency enhancing policies with the performance of the two segments
concerned.

Reference
Abraham, V. and Sasikumar, S.K. (2010): Labour Cost and Export Behaviourof Firms in
Indian Textile and Clothing Industry, MPRA Paper No. 22784, [Online] http://mpra.ub.uni-
muenchen.de/22784 (Accessed 6th January, 2015);
Banerjee S., Chattopadhyay, S. and Lahiri, K. (2016). Export Performance in Textile and
Garments with China as a Competitor: An Analysis of India’s Situation from the Perspective
of Structure-Conduct-Performance Paradigm in Roy M. & S. Sinha Roy (eds) International
Trade and International Finance Explorations of Contemporary Issues, Springer;
Banga, R. (2005), Impact of Liberalization on Wages and Employment in Indian Manufacturing
Industries, Working Paper No. 153, Indian Council for Research in International Economic
Relations;
Biswas, P.K., Banerjee, P., Choudhury, A. and Kukreja, P. (2014): Industry Structure and the
Pattern of Innovation: Pharmaceutical Industry of India 2000-2010, CSIR-NISTADS, New
Delhi;
 Employment-Productivity Profile and Labour Demand Elasticity .... 16 9

Cali, M., Mitra, S. and Purohit, P. (2009): State-Business Relations in Indian States: Paper
towards constructing an effective SBR index for Indian states, IPPG Discussion Paper
Series, No.25, Manchester;
Das, D.K. and Kalita, Gunajit (2009): Are Labour-intensive Industries Generating
Employment in India? Evidence from Firm Level Survey, The Indian Journal of Labour
Economics, Vol. 52 No.3, pp. 411-432;
Galab, S. and Revathy, E. (2009): Understanding Powerloom Weavers’ Suicides in Sircilla,
Economic and Political Weekly, Vol. 44 No. 8, pp. 12-15;
Goldar, B. (2011): Growth in Organised Manufacturing Employment in Recent Years,
Economic and Political Weekly, Vol. 46, No. 7, pp. 20–23.
Goldar, B. and Sadhukhan, A. (2015): Employment and Wages in Indian Manufacturing:
Post-Reform Performance, Working Paper No. 185, Employment Policy Department,
International Labour Office, Geneva;
Hasan, R., Mitra, D. and Ramaswamy, K.V. (2007): Trade Reforms, Labor Regulations, and
Labor-Demand Elasticities: Empirical Evidence from India, The Review of Economic and
Statistics, Vol. 89 No.3, pp. 466-481;
ICRIER (2008): A Study on Labour Intensity and Employment Potential of Indian
Manufacturing, a report sponsored by National Manufactuting Competitiveness Council;
Kathuria, L.M. and Mamta (2012): Understanding Financing Practices and Constraints of
Small and Medium Enterprises in Textile and Clothing Sector: An Empirical Study, Asia-
Pacific Journal of Management Research and Innovation [Online] http://abr.sagepub.com/
content/8/3/303 (Accessed 6th November, 2014);
Kakarlapudi, K. K. (2010): The Impact of Trade Liberalization on Employment: Evidence
from India’s Manufacturing Sector, MPRA Paper No. 35872, [Online] http://mpra.ub.uni-
muenchen.de/35872 (Accessed 6th January, 2015);
Kathuria, V., Raj, R.S.N. and Sen, K. (2010): Human Capital and Manufacturing Productivity
Growth in India, Paper submitted for International Conference on: Science, Technology
and Economy: Human Capital and Development (Annual Conference of IASSI & Knowledge
Forum Hosted by IIT Bombay during Nov. 11-12, 2010);
Kathuria, V.R., Natarajan, R.R.S. and Sen, K. (2013): Efficiency Comparison between Formal
and Informal Firms: Evidence from Indian Manufacturing, Journal of Industrial Statistics,
Vol. 2 No.1, pp.1-23;
Maiti, D. (2008): Organizational Morphology of Rural Industries and its Dynamics in
Liberalized India, Cambridge Journal of Economics, Vol. 32 No. 4, pp. 577-591;
Maiti, D. and Marjit, S. (2008): Trade Liberalization, Production and Informal Sector of the
Developing Countries, The Journal of International Trade & Economic Development: An
International and Comparative Review, Vol. 17, No. 3, pp. 453-461;
Mitra, A. (2013): Can Industry be the Key to Pro-Poor Growth? An Exploratory Analysis for
India, ILO Asia-Pacific Working Paper Series, DWT for Sout Asia and Country Office for
India, International Labour Organization
Nagaraj, R (2004): Fall in Organised Manufacturing Employment-A Brief Note, Economic
and Political Weekly,Vol. 39 No. 30, pp. 3387-90;
17 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Nagaraj, R (2007): Labour market in India, paper presented at a seminar on “Labour

Markets in Brazil, China and India”, Princeton University, New Jersey, [Online] http://
www.oecd.org/employment/emp/38355655.pdf (Accessed 29th July, 2015)
Nagaraj, R (2011):.Growth in organised manufacturing employment in recent years,
Economic and Political Weekly, Vol.46, No.39, pp.20–23.
Nataraj, Shanthi (2010): Productivity and Firm Size Distribution: Evidence from India’s
Organized and Unorganized Manufacturing Sectors, Unpublished Doctoral Thesis,
University of California, Berkeley;
Niranjana, S. and Vinayan, S. (2001): Report on Growth and Prospects of the Handloom
Industry, Study Commissioned by the Planning Commission, New Delhi: Government of
India;
Raveendran, G. and Kannan, K.P. (2009): Growth sans Employment: AQuarter Century of
Jobless Growth in India’s Organised Manufacturing, Economic and Political Weekly, Vol.
44 No. 10, pp. 80-91;
Reserve Bank of India (2014): RBI Handbook of Statistics, downloaded on 01-08-2015 from
http:///www.rbi.org.in/scripts/publicationsview.aspx?id=15795
Roy, T. (1996): Market Resurgence, Deregulation, and Industrial Response: Indian Cotton
Textiles in 1990s, Economic and Political Weekly, Vol. 31 No. 21, pp. M31-M41;
Roy, T. (1998): Economic Reforms and Textile Industry in India, Economic and Political
Weekly, Vol. 33 No. 32, pp. 2173-2182;
Sahu, P.P. (2011): Subcontracting in India’s Micro and Small Enterprises: An Exploratory
Analysis, Das, K. (ed), Micro and Small Enterprises in India:The Era of Reforms, pp. 185-
207, Routledge, New Delhi;
Sen, S. and Dasgupta, B. (2006): Labour in India’s Organized Manufacturing Sector, Indian
Journal of Labour Economics, Vol. 49, No. 1, pp. 79-102;
Sen, Sarmishtha. and Majumder, Subrata. (2014): Relative Efficiency in Adaptation to Pro-
Competitive Environment: An Exploration of Unorganized Indian Textile and Garments
Enterprises, Journal of Rural and Industrial Development, Volume 2(2), pp. 9-21
Sen, S. and Majumder, S. (2015): Relative Efficiency of Organized and Unorganized Segments
of Indian Textile and Garments Industry: A Preliminary Exploration, The Journal of
Industrial Statistics, Vol. 4 No. 1, pp. 37-57;
Sen, S. (2016): Adaptability to Pro-competitive Environment: A Study of Unorganized Indian
Textile and Garment Enterprises, Unpublished Doctoral Thesis, University of Calcutta;
Sharma, A.N. (2006): Flexibility, Employment and Labour Market Reforms in India,
Economic and Political Weekly, Vol. 41 No. 32, pp. 2078-2085;
Srinivasulu, K. (1996): 1985 Textile Policy and Handloom Industry Policy, Promises and
Performance, Economic and Political Weekly, Vol. 31 No. 49, pp. 3198-3206;
Teewari, M. (2005): Post-MFA Adjustments in India’s Textile and Apparel Industry: Emerging
Issues and Trends, Working Paper No. 167, Indian Council for Research on International
Economic Relations, pp. 1-55;
 Employment-Productivity Profile and Labour Demand Elasticity .... 17 1

Table 1
Variables used in the combined dataset and their definitional compatibility
Variable ASI data NSS data
INTER: Expenses on materials and
Intermediate fuels consumed = Annual value of
INTER: E xpenses on materials
input (total expenses – other operating
and fuels consumed
(Rs.) expenses + costs of electricity &
fuel consumed)33
K: Market value of fixed assets
K: Book value of fixed asset on (owned & hired) as on the closing
Capital (Rs.) the opening date of the date of the reference year – net
reference year addition to fixed assets during the
reference year

L = Total man -days worked L = (Total number of FTE workers34

Labour
(both manufacturing and * 30 * number of months operated *
(number)
non-manufacturing) average hours worked35)/8

GVA alt/L GVA alt/L

Firm-specific
(GVA alt=Ex-factory value of (GVA alt=Annual 37 value of total
APL
output – Costs of materials receipt38 – Expenses on materials
(Rs./manday)
and fuels consumed36) and fuels)
Source: Sen and Majumder, 2015

33
There are a very large number of missing entries in the series on ‘expenses on raw materials’ thus
making its direct use unreliable.
34
To make the labour-data comparable to the ASI data, we convert all workers (both full-time and part-
time) into full-time equivalents (FTEs), by treating 1 full-time worker = 2 part-time workers.
35
In the 56th round data, average number of hours worked daily is not given. We assume here that each day
consists of an 8-hour block.
36
Information on Gross Value Added (GVA) is not directly available in the ASI database but it can be
calculated by deducting expenditure on materials and fuels and other operating expenses from total
output (ex-factory value of output and other receipts e.g. value of electricity produced and sold).
However, to be consistent with the production framework used in this study an alternative series of GVA
values (GVAalt) was calculated by subtracting only costs of materials and energy from ex-factory value of
output. Similarly, a new series of value-added was computed from the given GVA data provided by the
NSSO to arrive at a similar measure. Observations with negative GVAalt values were dropped to make the
exercise meaningful.
. Annual values are obtained directly from NSSO 2000-01 data. But annual values of receipts (and other
37

categories e.g. expenses) for the year 2010-11are computed by multiplying the given series with number
of months operated.
38
It is necessary to note: (a) this category includes receipts from other activities e.g. ‘receipt from trade’
as well as ‘other receipts’ – in addition to ‘receipts from manufacturing activities’. More accurate values
of production could be obtained by deducting the non-manufacturing component from total receipts. Due
to the presence of considerable number of missing values (especially for NSS 67 th round) this operation
would have made the data series unreliable or a substantial number of data points would have been lost.
17 2 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 2
Summary Statistics for Estimated LD across Productivity Quartiles, 2001 & 2011
APL Quartiles/ 2001 2010
Summary
Statistics Obs. Mean SD Median Obs. Mean SD Median

All T&G
1 89656 2810.83 5499.34 2160 143748 3506.18 4979.37 2700
2 92919 3027.54 11613.65 2250 153098 3322.07 10659.55 2520
3 71979 3683.42 30176.82 2160 47792 8087.05 72505.80 2520
4 18300 11634.75 56786.22 1800 12254 18705.07 74868.17 2520
Organized T&G
1 2342 31345.48 92550.11 6820 2531 55608.33 244148.60 9150
2 2124 52930.66 184055.21 7518 2448 57295.57 200046.90 10496
3 1940 42890.29 105120.09 12671 2434 53657.81 142558.00 16174
4 1917 31805.08 88569.22 10300 2618 36977.30 89907.97 13505
Unorganized T&G
1 67074 2609.44 2078.12 1980 75093 3548.47 3103.39 2835
2 63970 2810.38 1792.95 2310 75310 3268.24 2648.13 2520
3 76644 2719.50 1573.16 2160 87223 2988.70 1981.51 2520
4 56842 2077.73 1682.03 1800 109235 3173.59 2653.55 2430
Note: Used population weights
Source: Authors’ Calculation
Table 3
Change in Mean LD across Quartiles of APL
APL Unorganized
All T&G Units Δ (%) Organized T&G Δ (%) Δ (%)
Quartiles/ T&G
(ALL) (ORG) (UN)
Mean LD 2001 2011 2001 2011 2001 2011

I 2810.83 3506.18 24.74 31345.48 55608.33 77.40 2609.44 3548.47 35.99

II 3027.54 3322.07 9.73 52930.66 57295.57 8.25 2810.38 3268.24 16.29

III 3683.42 8087.05 119.55 42890.29 53657.81 25.10 2719.50 2988.70 9.90

IV 11634.75 18705.07 60.77 31805.08 36977.30 16.26 2077.73 3173.59 52.74

Note: (i) ALL-All T&G Units; ORG-Organized Segment; UN-Unorganized Segment;

(ii) Used population weights
Source: Authors’ Calculation
 Employment-Productivity Profile and Labour Demand Elasticity .... 17 3

Table 4
Summary Statistics of APL within Productivity Quartiles of Sample T&G Units in India,
2001 and 2011
APL ALL T&G Organized T&G Unorganized T&G

Quartiles Obs. Mean Median SD CV Obs. Mean Median SD CV Obs. Mean Median SD CV

2001
I 2264 54.35 57.35 17.74 0.33 706 133.12 137.64 67.02 0.50 1558 48.79 50.67 14.36 0.29
II 2265 108.11 107.25 17.31 0.16 704 353.50 350.76 67.09 0.19 1559 86.67 86.06 10.65 0.12
III 2263 214.16 202.35 54.39 0.25 706 698.90 671.49 155.13 0.22 1559 133 130.31 17.9 0.13
IV 2264 1397.42 725.35 8790.51 6.29 705 3016.47 1789.02 15606.61 5.17 1559 407.29 248.63 706.86 1.74
2011
I 1631 69.57 72.34 22.86 0.33 844 202.70 220.20 86.22 0.43 788 54.11 56.14 14.64 0.27
II 1630 154.13 148.24 33.83 0.22 842 454.94 450.45 75.38 0.17 786 91.80 91.02 10.02 0.11
III 1631 391.64 378.30 116.34 0.30 843 781.50 766.39 127.13 0.16 788 132.27 131.40 14.07 0.11
IV 1630 1608.01 1065.53 2328.48 1.45 844 2327.64 1578.04 3051.91 1.31 787 289.73 212.82 367.56 1.27

Source: Authors’ Calculation

Table 5
Distribution of APL - Productivity Deciles in India, 2001 and 2011

Decile Summary APL_ALL T&G APL_Organized APL_Unorganized

Groups/Variables Statistics
2001 2011 2001 2011 2001 2011
Obs. 905 652 282 337 623 315
Mean 36.11 46.11 63.64 113.18 34.12 39.16
1 Median 38.74 48.49 66.32 120.16 36.51 41.85
SD 11.51 13.99 33.71 56.85 9.68 9.65
CV 0.32 0.30 0.53 0.50 0.28 0.25
Obs. 905 652 282 337 623 315
Mean 61.84 78.58 157.69 240.56 54.92 60.24
2 Median 61.99 78.84 157.90 242.45 55.26 60.68
SD 5.64 7.39 23.89 23.64 4.62 4.87
CV 0.09 0.09 0.15 0.10 0.08 0.08
Obs. 905 652 282 337 623 315
Mean 80.51 105.82 243.24 328.62 69.15 75.08
3 Median 80.25 105.71 243.94 327.08 69.24 75.16
SD 5.44 8.61 24.90 27.69 3.85 3.98
CV 0.07 0.08 0.10 0.08 0.06 0.05
Obs. 905 652 282 337 623 315
Mean 101.37 138.76 327.56 426.56 82.56 87.93
4 Median 100.98 138.23 326.26 424.89 82.46 87.86
SD 6.57 11.25 25.78 28.91 4.01 3.77
CV 0.06 0.08 0.08 0.07 0.05 0.04
Obs. 905 652 282 337 623 315
Mean 126.10 189.72 423.63 533.65 97.81 102.38
5 Median 125.64 187.22 423.67 536.29 97.64 102.27
SD 8.23 19.58 30.97 31.28 4.81 4.39
CV 0.07 0.10 0.07 0.06 0.05 0.04
17 4 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 5
Distribution of APL - Productivity Deciles in India, 2001 and 2011 (Contd.)
Decile Summary APL_ALL T&G APL_Organized APL_Unorganized
Groups/Variables Statistics
2001 2011 2001 2011 2001 2011
No. of Observation 905 652 282 337 623 315
Mean 162.09 274.84 546.38 654.07 115.20 117.95
6 Median 161.34 272.48 542.20 652.90 115.22 117.95
SD 13.20 31.78 43.25 40.04 5.31 4.58
CV 0.08 0.12 0.08 0.06 0.05 0.04
Obs. 905 652 282 337 623 315
Mean 222.46 420.34 732.45 812.14 137.00 136.49
7 Median 220.34 416.83 728.24 810.09 136.86 136.09
SD 23.18 51.40 67.68 52.13 7.31 5.95
CV 0.10 0.12 0.09 0.06 0.05 0.04
Obs. 905 652 282 337 623 315
Mean 341.70 625.96 1028.76 1036.13 170.25 159.60
8 Median 335.97 621.33 1020.01 1034.44 169.82 158.95
SD 48.85 69.70 110.47 75.94 12.00 8.12
CV 0.14 0.11 0.11 0.07 0.07 0.05
Obs. 905 652 282 337 623 315
Mean 621.44 960.59 1599.84 1432.42 229.31 198.66
9 Median 596.64 945.70 1572.59 1406.63 225.16 195.90
SD 131.08 132.50 241.86 167.91 24.84 16.32
CV 0.21 0.14 0.15 0.12 0.11 0.08
Obs. 911 654 283 340 628 314
Mean 2666.63 2710.03 5364.45 3817.38 695.02 443.10
10 Median 1582.82 1825.78 3304.60 2541.53 428.26 296.43
SD 13763.33 3383.92 24467.80 4401.06 1049.43 547.15
CV 5.16 1.25 4.56 1.15 1.51 1.23
Source: Authors’ Calculation
Table 6
Change in Mean LD across Decile-Groups of APL
APL All T&G Units Organized T&G Unorganized T&G
Change Change Change
Deciles/
2001 2011 (%) 2001 2011 (%) 2001 2011 (%)
Mean LD

1 2891.78 3881.49 34.22 23829.71 40429.08 69.66 2512.16 3591.28 42.96

2 2664.59 3524.81 32.28 40307.73 68659.32 70.34 2708.02 3718.59 37.32

3 2967.57 2964.90 -0.09 45339.76 53861.32 18.79 2658.42 3353.88 26.16

4 2925.65 3169.22 8.33 59124.31 58123.16 -1.69 2920.34 3508.97 20.16

5 3234.78 3621.61 11.96 39680.28 61848.23 55.87 2762.64 2906.28 5.20

6 2802.50 5628.13 100.83 45701.31 52334.29 14.51 2972.01 2982.55 0.35

7 3657.62 14567.99 298.29 35762.74 56583.89 58.22 2511.18 3187.26 26.92

8 8160.41 23640.63 189.70 44879.70 44842.64 -0.08 2441.69 2754.12 12.80

9 12051.51 22101.82 83.39 34866.95 47202.84 35.38 2283.56 3164.99 38.60

10 13972.81 15418.24 10.34 25194.24 26867.42 6.64 1753.63 3266.34 86.26

Note: Used population weights

Source: Authors’ Calculation
 Employment-Productivity Profile and Labour Demand Elasticity .... 17 5

Table 7
Distribution of Estimated LD across Productivity Deciles in India, 2001 and 2011
Productivity APL_ Unorganize
Summary
Deciles/Variable APL_ALL T&G APL_Organized d
Statistics
s 2001 2011 2001 2011 2001 2011
Obs. 905 652 282 337 623 315
Mean 36.11 46.11 63.64 113.18 34.12 39.16
1
Median 38.74 48.49 66.32 120.16 36.51 41.85
SD 11.51 13.99 33.71 56.85 9.68 9.65
Obs. 905 652 282 337 623 315
Mean 61.84 78.58 157.69 240.56 54.92 60.24
2
Median 61.99 78.84 157.90 242.45 55.26 60.68
SD 5.64 7.39 23.89 23.64 4.62 4.87
Obs. 905 652 282 337 623 315
Mean 80.51 105.82 243.24 328.62 69.15 75.08
3
Median 80.25 105.71 243.94 327.08 69.24 75.16
SD 5.44 8.61 24.90 27.69 3.85 3.98
Obs. 905 652 282 337 623 315
Mean 101.37 138.76 327.56 426.56 82.56 87.93
4
Median 100.98 138.23 326.26 424.89 82.46 87.86
SD 6.57 11.25 25.78 28.91 4.01 3.77
Obs. 905 652 282 337 623 315
Mean 126.10 189.72 423.63 533.65 97.81 102.38
5
Median 125.64 187.22 423.67 536.29 97.64 102.27
SD 8.23 19.58 30.97 31.28 4.81 4.39
No. of Observation 905 652 282 337 623 315
Mean 162.09 274.84 546.38 654.07 115.20 117.95
6
Median 161.34 272.48 542.20 652.90 115.22 117.95
SD 13.20 31.78 43.25 40.04 5.31 4.58
Obs. 905 652 282 337 623 315
Mean 222.46 420.34 732.45 812.14 137.00 136.49
7
Median 220.34 416.83 728.24 810.09 136.86 136.09
SD 23.18 51.40 67.68 52.13 7.31 5.95
Obs. 905 652 282 337 623 315
8 Mean 341.70 625.96 1028.76 1036.13 170.25 159.60
Median 335.97 621.33 1020.01 1034.44 169.82 158.95
SD 48.85 69.70 110.47 75.94 12.00 8.12
Obs. 905 652 282 337 623 315
Mean 621.44 960.59 1599.84 1432.42 229.31 198.66
9
Median 596.64 945.70 1572.59 1406.63 225.16 195.90
SD 131.08 132.50 241.86 167.91 24.84 16.32
Obs. 911 654 283 340 628 314
Mean 2666.63 2710.03 5364.45 3817.38 695.02 443.10
10
Median 1582.82 1825.78 3304.60 2541.53 428.26 296.43
SD 13763.33 3383.92 24467.80 4401.06 1049.43 547.15
Note: Used population weights
Source: Authors’ Calculation
Table 8
Change in Share of Productivity-Deciles in Total LD
Productivity All T&G ORG_T&G UNORG_T&G
Change (%) Change (%) Change (%)
Groups Share_01 Share_10 Share_01 Share_10 Share_01 Share_10
Decile 1 9.37 12.64 34.93 6.52 8.17 25.39 9.24 11.16 20.83
Decile 2 10.40 12.58 20.98 10.82 13.62 25.92 10.85 8.47 -21.92
Decile 3 9.26 11.60 25.25 13.58 10.47 -22.86 11.03 7.75 -29.72
Decile 4 11.29 12.33 9.25 15.45 12.05 -22.05 9.56 10.00 4.61
Decile 5 12.28 12.94 5.44 9.94 10.96 10.31 11.41 8.47 -25.74
Decile 6 11.09 13.13 18.39 10.70 10.18 -4.86 11.78 8.87 -24.72
Decile 7 9.46 7.36 -22.14 8.94 10.73 20.01 13.10 10.54 -19.56
Decile 8 10.50 6.75 -35.70 10.09 8.82 -12.61 11.15 8.33 -25.26
Decile 9 8.85 5.95 -32.77 8.64 9.15 5.99 7.15 12.05 68.55
Decile 10 7.67 4.64 -39.56 5.39 5.93 9.85 4.80 14.38 199.37

Note: Share_01 and Share_10 are respective shares of productivity deciles in total LD in 2001 & 2010;
LD estimates obtained using population weights
Source: Authors’ Calculation
17 6 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 9
Descriptive Statistics for Size-Groups based on the Values of Intermediate Inputs, 2001 & 2011
2001 2011
Size-Group Value of Intermediate Inputs
Obs. Mean Median SD CV Obs. Mean Median SD CV
1 <=500 162 272 300 133.41 0.49 22 328.18 360 143.15 0.44
2 >500 & <=1500 330 1003.13 1020 284.52 0.28 113 1031.87 980 296.76 0.29
3 >1500 & <=4000 410 2597.12 2520 712.96 0.27 284 2722.80 2712 675.01 0.25
4 >4000 & <= 11000 569 7372.01 7320 2121.63 0.29 462 6992.98 6960 2022.41 0.29
5 >11000 & <=30000 1174 19814.78 19452 5476.57 0.28 575 19571.06 19200 5493.85 0.28
6 >30000 & <=80000 1129 50242.27 48681 13848.71 0.28 586 50923.04 48920 13603.96 0.27
7 >80000 & <=200000 828 128295.20 122400 34279.43 0.27 345 125068.80 118200 33677.55 0.27
8 >200000 & <=500000 559 319902.80 307200 84617.38 0.26 250 325144.00 315600 82326.21 0.25
9 >500000 & <=1250000 519 836336 808200 224169.10 0.27 278 826172.60 812370 212584.10 0.26
10 >1250000 & <=3000000 615 1946529 1825200 504439.60 0.26 271 2062392.00 2026764 491482 0.24
11 >3000000 & <=7500000 443 4790441 4606810 1263527 0.26 331 5110867 4992290 1367846 0.27
12 >7500000 & <=20000000 503 1.28E+07 1.24E+07 3610075 0.28 456 1.29E+07 1.26E+07 3596787 0.28
13 >20000000 & <=55000000 599 3.47E+07 3.41E+07 9696519 0.28 631 3.58E+07 3.51E+07 1.01E+07 0.28
14 >55000000 & <=150000000 621 9.32E+07 8.98E+07 2.67E+07 0.29 754 9.48E+07 9.25E+07 2.71E+07 0.29
15 >150000000 & <=375000000 355 2.34E+08 2.16E+08 6.36E+07 0.27 602 2.42E+08 2.34E+08 6.41E+07 0.27
16 >375000000 & <=1000000000 191 5.56E+08 5.17E+08 1.52E+08 0.27 376 5.90E+08 5.53E+08 1.64E+08 0.28
17 >1000000000 49 1.83E+09 1.35E+09 1.57E+09 0.86 186 2.30E+09 1.56E+09 2.35E+09 1.02

Source: Authors’ Calculation

Table 10
Change in Mean APL, Estimated LD & Share in Estimated LD by Groups based on
Intermediate Input-Costs
Intermediate Input Mean AP L Mean LD Share in Total LD
Cost (Rs.) Groups 2001 2011 Δ (%) 2001 2011 Δ (%) 2011 2011 Δ (%)
1 75.55 64.85 -14.16 2229.33 1493.35 -33.01 1.56 0.16 -89.87
2 86.66 74.87 -13.60 2275.97 1895.08 -16.74 3.96 1.93 -51.20
3 100.17 89.31 -10.84 2502.07 2157.64 -13.77 4.02 3.78 -5.99
4 118.17 107.25 -9.24 2190.41 2527.53 15.39 5.63 7.48 32.86
5 120.78 125.43 3.85 2123.22 2605.91 22.73 11.58 9.94 -14.20
6 138.51 124.85 -9.86 2541.16 3308.67 30.20 15.45 13.19 -14.60
7 154.62 129.76 -16.08 2937.92 3874.89 31.89 8.92 10.49 17.60
8 150.75 186.44 23.67 3333.20 4476.69 34.31 6.09 8.16 34.00
9 213.15 183.56 -13.88 2920.01 3709.68 27.04 4.06 4.21 3.83
10 301.32 321.15 6.58 3100.73 3828.56 23.47 4.43 4.78 7.99
11 484.78 402.75 -16.92 5421.48 6957.86 28.34 2.97 3.88 30.54
12 784.78 649.25 -17.27 10733.97 6562.04 -38.87 2.70 2.53 -6.26
13 1796.18 976.06 -45.66 25386.82 20248.27 -20.24 4.34 3.34 -23.04
14 1094.29 1061.60 -2.99 61714.41 42772.72 -30.69 6.17 5.48 -11.08
15 1175.77 965.72 -17.86 141039.13 101069.46 -28.34 6.92 6.03 -12.86
16 1038.30 1206.43 16.19 334531.97 232869.06 -30.39 7.74 6.81 -12.05
17 1634.34 1528.50 -6.48 708048.89 605618.24 -14.47 3.52 7.73 119.31
Note: LD estimates obtained using population weights; Coefficient of Variation is very high for the two
extreme size-groups here and the estimates are not reliable;
Source: Authors’ Calculation
 Employment-Productivity Profile and Labour Demand Elasticity .... 17 7

Table 11
Distribution of Sample Units across Selected Textile-Major States, India
2001 2010
States
Organized Unorganized Total Organized Unorganized Total
Andhra Pradesh 124 341 465 156 270 426
Delhi 173 235 408 120 290 410
Gujarat 225 452 677 262 237 499
Haryana 169 98 267 255 48 303
Karnataka 230 284 514 168 163 331
Kerala 130 354 484 84 173 257
Madhya Pradesh 65 144 209 42 80 122
Maharashtra 245 1040 1285 332 222 554
Punjab 172 299 471 304 64 368
Rajasthan 161 181 342 202 110 312
Tamil Nadu 878 1181 2059 1114 560 1674
Uttar Pradesh 162 990 1152 255 368 623
West Bengal 113 645 758 139 564 703
Source: Authors’ Calculation

Table 12
Results of Labour Demand Estimation at All-India Level, 2001 and 2011
Variables/Year 2001 2011
0.73 *** 0.81 ***
L_Y2_ins
(116.04) (126.78)
-0.16 *** 0.02
L_W
(-9.53) (0.98)
0.18 *** 0.56 ***
L_W_ORG
(34.84) (43.79)
0.00 ** -0.00 ***
L_W_RGDT
(4.06) (-10.82)
-0.03 ** -0.03 ***
L_W_LINKAGE
(-2.58) (-8.87)
-0.05 *** 0.02
L_W_G
(-14.05) (1.39)
0.00 * 0.03 ***
L_W_KP
(1.96) (26.41)
-0.07 *** 0.00 ***
L_W_INP
(-38.85) (6.56)
0.00 *** -0.12 ***
L_W_TECH
(2.86) (-44.68)
Observation 9056 6521
F-value F(9, 9046) =6339.00 F(9, 6511) =6414.32

Prob.>F 0.0000 0.0000

Adj. R2 0.8674 0.9001

d.f. 9055 6520

Source: Authors’ Calculation

17 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 13
Results of Labour Demand Estimation for Textile-major States, 2001 and 2011
State/Year/Variables L_Y2_ins L_W L_W_ORG L_W_RGDT L_W_G L_W_KP L_W_INP Obs. F-Value Prob.>F Adj. R2 d. f.
0.78 -0.17 0.14 -0.05 -0.06
F(5, 458)
2001 (34.92) (-2.63) (6.44) (-3.08) (-7.57) 464 0.0000 0.9025 463
=858.48
*** *** *** *** ***
AP
0.77 -0.14 0.59 -0.00
F(6, 417)
2011 (38.66) (-1.87) (13.63) (-2.38) 424 0.0000 0.7937 423
=272.28
*** * *** **
0.67 -0.16 0.14 -0.00 -0.01 0.00 -0.07
F(7, 400)
2001 (30.02) (-1.80) (5.44) (-1.31) (-1.09) (3.00) (-8.93) 408 0.0000 0.8529 407
=338.03
*** * *** ** *** ***
DEL
0.71 0.64 0.31 -0.00 -0.00 -0.10
F(6, 403)
2011 (33.83) (8.08) (8.98) (-6.80) (-3.08) (-11.67) 410 0.0000 0.8966 409
=592.35
*** *** *** *** *** ***
0.70 -0.26 0.13 0.00 -0.05 0.00 -0.05
F(7, 665)
2001 (37.58) (-4.07) (9.12) (1.13) (-3.50) (1.75) (-8.18) 673 0.0000 0.9104 672
=976.40
*** *** *** *** * ***
GUJ
0.74 0.60 -0.00 0.03 0.00 -0.11
F(6, 491)
2011 (31.86) (13.35) (-6.65) (-1.83) (3.92) (-11.89) 498 0.0000 0.8821 497
=620.60
*** *** *** * *** ***
0.76 0.09 -0.07
F(3, 263)
2001 (28.04) (2.96) (-8.49) 267 0.0000 0.8552 266
=524.58
*** ** ***
HAR
0.60 0.22 0.49 -0.00 -0.02 0.00 -0.09
F(7, 294)
2011 (21.87) (1.42) (6.06) (-9.73) (-1.84) (2.61) (-8.54) 302 0.0000 0.9178 301
=481.14
*** *** *** * ** ***
0.56 -0.17 0.33 -0.05
F(4, 504)
2001 (10.04) (-2.32) (9.52) (-5.74) 509 0.0000 0.8858 508
=834.53
*** ** *** ***
KAR

0.69 -0.10 0.27 -0.07 -0.07

F(5,
(-
2001 (47.53) (19.66) (-9.77) (-18.43) 1281 1275) 0.0000 0.8641 1280
2.63)
*** *** *** *** =1628.37
MAH ***
0.59 0.36 0.65 - 0.00 -0.04 0.00 -0.14
F(7, 542)
2011 (25.83) (4.00) (13.88) (-8.24) (-3.75) (6.17) (-16.99) 550 0.0000 0.9077 549
=772.48
*** *** *** *** *** *** ***
0.77 0.13 0.00 -0.08
F(4, 463)
2001 40.14) (6.75) (2.13) (-14.99) 468 0.0000 0.8709 467
=796.61
*** *** ** ***
PNJ
0.56 0.68 - 0.00 -0.01 0.00 -0.07
F(6, 359)
2011 (20.20) (10.81) (-12.76) (-1.44) (2.74) (-6.51) 366 0.0000 0.9352 365
=879.66
*** *** *** *** ***
0.64 0.11 0.00 -0.11 0.00 -0.09
F(6, 335)
2001 (19.22) (3.62) (3.77) (-5.18) (3.63) (-12.26) 342 0.0000 0.8409 341
=301.34
*** *** *** *** *** ***
RAJ
0.73 0.35 0.51 -0.00 -0.06 0.00 -0.13
F(7, 304)
2011 (23.68) (2.83) (7.74) (-6.61) (-3.38) (4.08) (-10.19) 312 0.0000 0.9084 311
=441.45
*** *** *** *** *** *** ***
0.62 0.15 0.27 0.00 -0.10 F(5, 2051)
2001 (46.95) (3.68) (21.77) (6.43) (-21.04) 2057 =2579.41 0.0000 0.8625 2056
*** *** *** *** ***
TN
0.72 0.27 0.53 -0.00 -0.02 -0.11 F(6, 1630)
2011 (63.86) (5.40) (19.09) (-16.05) (-3.54) (-18.60) 1637 =2405.48 0.0000 0.8982 1636
*** *** *** *** *** ***
0.77 -0.35 0.10 -0.10 0.00 -0.05 F(6,1135)
2001 (46.79) (-11.82) (7.86) (-4.24) (2.32) (-14.92) 1142 =1000.43 0.0000 0.8410 1141
UP *** *** *** *** ** ***

Source: Authors’ Calculation

 The Journal of Industrial Statistics (2016), 5 (2), 179 - 191 17 9

On Spatial Concentration of Organized Manufacturing Industries: A

Look at Regional Perspective
Sajal Jana1, Garhbeta College, Midnapore, India
Maniklal Adhikary, Burdwan University, Burdwan, India

Abstract

The Present paper has been designed to explain and analyze the regional pattern of
organized manufacturing sector in India at three time points: 1987-88, 1997-98, and
2010-11. We intend to examine the extent of spatial concentration of the same across the
states at selected disaggregated two digit industry levels. The specific objectives have
been examined by looking at employment and output figures of organized sectors in
India. The results suggest that concentration has increased during the years 1997/98 &
2010/11 respectively and the concentration across the states is not uniform at the time
points being considered to the present study.

1. Introduction

1.1 The question of regional spread of industries is very important to understand the
development potential of the sub national regions in India, since traditionally industrialization
is considered as sine qua non for economic growth (Kaldor, 1967; Hirschman, 1958). Empirical
studies point out that owing to increasing dominance of the Private sector in industrialization
location concentration of industries has increased during the reform period & this may be
considered as one of the major causes of regional inequality in India. The literature of
regional economies suggest that industries tend to concentrate in order to realize tangible
benefits from being close to other firms & to consumers, market access, thick labour markets,
better infrastructure, transportation, raw materials & resources, agglomeration benefits,
knowledge & technology spillover, externalities etc ( Saikia, 2011).

1.2 The location decisions of state owned industries are influenced by consideration
of balanced regional development. The role of state as industrial owner and industrial
location regulator has been substantially curtailed under the regime of liberalization &
structural reforms. Therefore, with the increasing dominance of private sector
industrialization, it is expected that industries will be more spatially concentrated in leading
industrial regions having developed socio-economic infrastructure, which will further lead
to higher levels of regional inequality. For balanced growth, the concentration of industrial
activities must decline over time and industrially backward states must also attract good
share in total output of the state thus in turn creating good employment opportunities.

1.3 A brief review of literature highlights that numerous studies on regional

specialization and concentration have been undertaken both in national and international
context. To cite a few, Glenn Elisson (1997) developed a model to show that localized
industry specific spillovers, natural advantages, and pure random chance all contribute to
geographical concentration of manufacturing industries in United States. The study by M.
Brulhant and J. Torestensson (1998) showed that industrial specialization among European

1
e-mail: [email protected]
18 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Union (EU) countries has increased in the 1980s and increasing returns industries tend to
be highly localized, concentrated in central EU countries and subject to relative low intra
industry trade. A. Hildebrandt and J. Worz (2004) applied regression analysis on individual
industries to investigate the determinants of the patterns of regional concentration and
specialization in Central and Eastern European countries (CEECs) over the years 1993 to
2000. The study concluded that a massive reallocation of production and labour force
strongly affects the pattern of concentration both in terms of production and employment
generally increased in the CEECs. In 2006 the research studies on Economic Transition and
Industrial Concentration in China conducted by Canfei He, et.al. focus on the result that
country’s employment in the manufacturing sector has been increasingly concentrated
since the early 1980’s while industrial output experienced decentralization in the 1980’s
followed by a centralization process in 1990’s. The study undertaken by Z.Goschin, et al.
(2009) have used the measures Herfindhal Index, Krugment dissimilarity Index and Lilien
Index to explore the main characteristics and the interaction of industries in Romania on the
basis of GVA & employment figure. In contrast the present studies considers two key
factors viz. number of total persons engaged & gross output as a proxy of employment &
output figures respectively and applies location Herfindahl Index and location Gini Index to
analyze spatial concentration of industries across the states.

1.4 In the Indian context, empirical research carried out by Alagh et. al.(1971) concluded
that the least and moderately diversified states like Maharastra, Tamil Nadu, West Bengal
specialized in resource based industries, while the less diversified states namely, Bihar,
Rajasthan, Orissa and Kerala specialized in capital and demand oriented consumer goods
industries. Ghosh (1975) computed Gini’s coefficient and Herfindhal Index to show that a
declining trend exists in concentration of twenty-two industries over the period 1948 to
1968. Awasthi (2000) in a district level study in Gujrat has also observed that the flow of
investment has occurred to the districts that have proximity to some major industrial
concentration with the advantage of forward and backward linkages. Another study by Lal
& Chakraborty (2005) concluded that new private sector industrial investments in India are
biased toward existing industrial and coastal districts and that the structural reforms increase
spatial inequality in industrialization.

1.5 In 2006 S. Athraye et.al studied the impact of economic liberalization on industrial
concentration by using a dynamic model based on time series data on twelve industries
over the period 1970-99. Singh (2012) has computed industrial concentration levels for the
states based on Gini’s coefficient and Herfindahl index for each year between1979-80 and
2006-07 using ASI data. In 2011 D. Saikia examined with the help of Gini’s coefficient the
spatial concentration of the unorganized manufacturing sector at the state level. The findings
revealed that there is a decline in industrial share of the leading states in post reform period.
Empirical studies in the recent past have shown evidences of a wide range of factors that
influence industrial location. The present study adds value to the existing body of literature
by considering the following specific objectives.
a) To analyze the spatial distribution of organized manufacturing sector by looking
at the share of states in terms of total employment, value of output in all India.
b) To examine the spatial concentration of organized manufacturing sector and its
dimension of change across the states at disaggregate two digit industry levels.
 On Spatial Concentration of Organized Manufacturing Industries .... 18 1

1.6 These two objectives have been addressed by selecting eight major industries
groups across the 15 states of India. Another notable feature is the cross-section time
points chosen to the present analysis are very recent compared to that of earlier studies.

1.7 The scheme of the paper is organized in the following sections. The data base,
Coverage and methodology are explained in section 2. Section 3 highlights regional
distribution of industries in India. Section 4 concentrates on the results & discussions.
Section 5 outlines scope of further work. Section 6 draws conclusions & policy implications.

2. Database, Coverage & Methodology

A) Database & Coverage of the Study

2.1 Database & Aggregation

2.1.1 The Present study uses the data from Annual Survey of Industries (ASI) compiled
by Central Statistical Organization (CSO), Government of India. The ASI organizes the data
on the basis of National Industrial Classification (NIC). Until 1997-98 the ASI data was
organized according to the NIC-1987 classification and then the NIC-1998 classification
has followed until 2003-04 and since then the NIC 2004 Classification has been followed
until 2007-08. Annexure 1 provides a brief explanation of selected industries corresponding
to their National Industrial Codes. A concordance between NIC 1970, NIC 1987 & NIC 2008
two digit levels has been made to build a comparable dataset at disaggregated two digit
levels. The variables to be used are gross output & total number of Persons engaged as a
measure of output & input respectively. The industry groups chosen for investigation can
be mentioned as: a) Food & Food Products b) Beverage & Tobacco Products c) Leather &
Leather products d) Metal & Metal products e) Machine & Machine tools f) Textile &
Textile Products v) Transport equipment. These industries have been selected on the basis
of their contribution to Gross output in organized manufacturing sector of India. We have
included leather industry despite its low share in gross output because of its higher
contribution to export earnings.

2.1.2 For the purpose of a comparative analysis we have considered three time points:
1987-88, 1997-98 & 2010-11. The Present analysis has been carried out across 15 states. It is
to be noted that the contribution of manufacturing sector in total employment for each
state exceeds one percent.

2.2 Coverage of the Study

2.2.1 The present analysis confines to only the organized / registered manufacturing
sector & excludes the unorganized manufacturing sector along with electricity, Water &
Gas supply undertakings & repair services units, all of which count as industry.

2.2.2 We have used employment and gross output data of the organized manufacturing
sector to represent industrial activities. By employment, we use the concept of “total
persons engaged” as used in the ASI frame. On the other hand, Gross output comprises
total ex-factory value of products & by-products manufactured as well as other receipts.
Suitable price deflators have been constructed with the help of the official series on
wholesale price indices (Index numbers of Wholesale Prices in India, prepared by the office
18 2 The Journal of Industrial Statistics, Vol. 5, No. 2

of the Economic advisor, Ministry of Industry) to deflate gross output at constant 1993-94
prices.

2.2.3 The study encompasses 18 major states of India (three of which were bifurcated in
November 2000) which are listed as 15 states. The bifurcated states are: Bihar, Madhya
Pradesh and Uttar Pradesh. Three new states were carved out of Bihar, Madhya Pradesh
and Uttar Pradesh respectively. To make comparability the data for the newly created states
have been added to the respective states from which it was created. We have marked the
bifurcated states with “*”. Thus, the states included in this study, arranged in alphabetical
order are: Andhra Pradesh (AP), Bihar (BIH*), Delhi (DEL), Gujarat (GUJ), Haryana(HAR),
Karnataka(KAR), Kerala( KER), Maharashtra (MAH), Madhya Pradesh (MP*), Orissa (
ORI), Punjab(PUN), Rajasthan ( RAJ), Tamil Nadu(TN), Uttar Pradesh(UP*), and West
Bengal(WB).

B) Methodological Issues

2.3 Measures of Spatial Concentration

2.3.1 The term spatial concentration refers to the extent to which a given industry is
concentrated in a few geographical units. Geographic concentration of a specific industry
reflects the distribution of its regional shares. A specific industry i is considered spatially
concentrated if a great part of the production is carried out in a few regions only. Several
statistical indices of spatial concentration have been proposed in the literature over the
years, which vary from traditional measures like coefficient of variation, location
concentration ratio, location Herfindahl Index, location Gini index, location entropy index
and location quotient etc. to the more recent measures like Ellison-Glaeser index and Moran’s
I etc. However, each of the measures has its own limitations.

2.3.2 The Present study uses a set of traditional measures namely, Herfindahl index,
Location Gini & Concentration ratio to measure spatial concentration of organized
manufacturing sector across the states in India since any single index is inadequate to
arrive at fairly reliable conclusion.

2.4 Indicators of Spatial Concentration of Industries: Location Herfindhal Index

2.4.1 The location Herfindahl Index of an industry i is defined as the sum squares of
employment (or output) shares of all states in the industry. If Eik is the employment (or
output) of Kth state in ith industry and Ei is employment / output of all the states in the ith
industry.

Symbolically:

The Herfindahl index is increasing with the degree of concentration reaching its upper limit
of 1 when the industry i is concentrated in one region/ state. The lowest value of
concentration is 1/n i.e. all regions have equal shares in industry i (i=1/n).
 On Spatial Concentration of Organized Manufacturing Industries .... 18 3

2.4.2 Concentration ratio is defined as the percentage share of employment / output of

an industry located in the largest four states, ranked in descending order of shares of the
states. Higher the ratio implies higher concentration.

2.4.3 Location Gini Index

Following CeaPraz (2008) we have applied the measure location Gini as the sum of the
differences of the concentration rates by the addition of the differences of the weights of
each industry and the weights of the arithmetic mean are obtained after the decreasing
classification of each state’s concentration rates. The index values between zero and one.
An index which approaches zero value indicates that the distribution of the concentration
in kth state corresponds to the national distribution. A value of the index equal to one means
that a specific state presents a strong concentration in a specific industry.

Symbolically:

Where n is the number of states. Ck = Sik /Sk for every state in the ith industry. Sik is the share
of employment in ith industry from kth state in total employment/output of ith industry. Sk is
the share of the employment of the kth state in the total employment/output of all the states.
Àk is rank of kth state in the ranking of Ck in descending order and is the mean value of
Ck after arranging in descending order.

Finally the empirical results based on these measures are reported & analyzed to focus on
spatial concentration in section 4.

3. Regional Distribution of Industries in India

We begin with analyzing the regional distribution of manufacturing industries across 15
states and four major regions.
3.1 Table 1 reports the employment share of regions and states in organized
manufacturing sector in 1987-88, 1997-98 and 2010-11 respectively.
3.1.1 It is obvious that the southern region has gained employment share from about 29
percent in 1987/88 to 34.8 percent in 1997/98, whereas the eastern region has considerably
lost their share from 17.7 to 14.4 during the same period. The employment share of the
Central region remains more or less same during the period of study and that of the north-
west region fluctuates over the years.
3.1.2 A state level analysis of the employment share corresponding to Eastern region
shows that the sharp decline in the share of employment share is largely due to sharp
decline in the share of West Bengal & to some extent Bihar. While the share of Assam &
Orissa remains more or less the same.
18 4 The Journal of Industrial Statistics, Vol. 5, No. 2

3.1.3 On the other hand, the insignificant fall in the employment share of the Central
region between 87/88 & 97/98 is mainly due to decline in the share of Maharashtra while the
share of Gujarat & Madhya Pradesh decline marginally and that of Rajasthan remains more
or less same.

3.2 Empirical result showing the share of states in terms of Gross output is reported in
Table 2.

3.2.1 The share of eastern region has significantly declined from about 16 percent in
1987/88 to 8 percent in 2010/11 whereas that of southern region has increased marginally
between the same time points. The share of north-West region has consistently declined in
the respective years 1997/98 and 2010/11.

3.2.2 Viewed in terms of Gross output, the eastern regions’ declining share is mainly
due to the decline in West Bengal & Bihar. The increasing share of Southern region is
attributed to the share of Andhra Pradesh & Karnataka in contrast to Tamil Nadu that has
consistently showed lower share during the time points as considered in the present analysis.

3.2.3 Individually, West Bengal, Maharashtra, Bihar and Tamil Nadu have significantly
lost their share in gross output, while Gujarat, Andhra Pradesh, Haryana and Karnataka
have improved their position in the years 1997/98 & 2010/11 respectively. It is remarkable to
note that though Tamil Nadu has improved its share in terms of employment over the years,
it has a stable position in terms of gross output throughout the years.

4. Results & Discussions

The tabular data presented in the preceding section (Table 1 & Table 2) cannot provide
information on the extent of spatial concentration of these industries. In this section we
examine the extent of spatial concentration of organized manufacturing industries across
the states at two digit disaggregated industry levels. The results derived by employing the
statistical measures as mentioned in Section 2.5 & 2.6 are summarized in the following
section.

4.1 Analysis of Results based on Herfindahl Index of Concentration

The Herfindhal Index computed in terms of Gross output & employment of organized
manufacturing sector disaggregated by two digit level industries is reported in Table 3.

4.1.1 A close look at Table 3 provides a better understanding of the degree of

concentration and the variation in the direction of change in concentration across the
industries. Concerning the evolution of disaggregated two digit level industries in terms of
employment, Herfindahl Index shows a dominant regional concentration of Beverage &
Tobacco Products, Leather & Leather Products, and Chemical & Chemical Products in the
year 2010-11. It is noteworthy to mention that same ranks have been preserved by these
three industries compared to the previous years, 1987/88 &1997/98 respectively.

4.1.2 In the year 2010-11, Textile, Chemical, Metal, Machine & Transport industries
have shown an increasing pattern of concentration in terms of employment compared to
both the previous years, 1987/88 &1997/98. However, Concentration has drastically declined
 On Spatial Concentration of Organized Manufacturing Industries .... 18 5

for Food & Food Products industry in the year 2010-11. Out of the eight industries
concentration has increased in as many as five industries in terms of employment and in
four industries in terms of gross output in the year 2010.

4.2 Analysis of Results based on Gini index of concentration

Table 3 also reports Location Gini measure of concentration for the selected disaggregated
two digit level industries.

4.2.1 The most concentrated sectors in the year 2010/11 are Leather & Leather Products
followed by Transport and Beverage & Tobacco products industries in terms of employment.
It is noteworthy to mention that Beverage & Tobacco product industry is the most
concentrated industry in the year 1997 with a spectacular drop in 2010. In the consequent
year, Leather & Leather product occupies second rank in concentration with an insignificant
drop in 2010.

4.2.2 A significant increase in concentration in terms of gross output has been found in
five industries, namely, Textile, Chemical, Leather & Leather Products, Beverage & Tobacco
products, machine& machine tools in the year 2010 compared to the year 1997-98.

4. 3 Results based on Concentration Ratio

Table 4 presents the results based on the computation of concentration ratio.

4.3.1 States like Maharashtra, Tamilnadu, Uttar Pradesh, Gujarat, and West Bengal
appeared more frequently in the list of four leading states in many of the industry groups.
Other states like Andhra Pradesh, Karnataka, Punjab, Rajasthan, Haryana, and Bihar appeared
single or couple of times in the list of four leading states in few industries.

4.3.2 Concentration of chemical industry has shifted from Uttar Pradesh & Tamilnadu to
West Bengal & Andhra Pradesh. Machine & Machine tools Industry has shifted from Uttar
Pradesh to Gujarat. Leather has shifted to Karnataka to Haryana in the year 2010 compared
to the year 1997-98.

5. Scope for further Work

The present study uses a rather broad classification of industries (2 digit NIC level). This
however leaves a scope to the researchers to study regional industrial scenario at more
disaggregated level in Indian context. Though the present study tries to examine the spatial
concentration of organized sector based on two key variables, viz. number of Persons
engaged & gross output but the trend of concentration over the years needs to be examined
based on the available time series data. Other important variables may be included to get
better insights.

6. Conclusions & Policy Implications

6.1 Differences in the results derived by applying several measures of concentration

are clearly noticed. This may be due to the size & importance attributed to the industry/
state. Despite the fact that the analysis presented in this paper are data explanatory, the
18 6 The Journal of Industrial Statistics, Vol. 5, No. 2

findings are important in understanding the pattern of spatial concentration in organized

manufacturing sector. It also helps to understand regional development in India.

6.2 The findings of the paper suggest that concentration is not uniform across the
different industry groups. Spatial concentration has increased during the years 1997/98 &
2010/11. The less diversification of Industries widens the regional inequality which is a
major bottleneck to achieve the balanced regional development.

6.3 In the year 2010 according to Gini index we observe an increase in spatial
concentration for the six industries, Food Products, Textile, Leather, Chemical, Machine &
Transport in terms of employment & a decline in concentration is noticed for the industries,
Beverage& Tobacco and Metal industries in the same year compared to the year 1997-98.

6.4 In the same direction, according to Herfindahl index we notice an increase in

concentration in terms of employment for five industries, namely, Textile, Chemical, Metal
& Transport industries and a constant trend for the food industry during the time point
2010 compared to the year 1997-98.

6.5 The probable reasons of industrial concentration throughout the years could be
some region specific factors such as cost structure, characteristics of labor force,
geographical characteristics, investment climate, political condition etc. The factors that
are likely to be more important to welcome new investment in industrialization are availability
of transport and communications, water & power, services & social amenities.

6.6 Therefore, the backward states should emphasize more on providing appropriate
physical infrastructure, legal & financial infrastructure (corporate law, accountancy norms,
banks, capital market), and social infrastructure to attract new industrial investment so that
inter-state variation in terms of industrial development can be reduced. Hence the states
are required to reconsider their development strategies, alter necessary policy decisions,
and change institutional structure to attract more industrial investment.

References

Alagh,Y. K., K. K. Subrahmanian and S. P. Kashyap(1971), “Regional Industrial Diversification

in India” Economic & Political Weekly, 6(15),795-802

Awasthi D. N (2000): Recent changes in Gujarat Industry: Issues and Evidence” Economic
and Political Weekly, 35(35/36), 3183-92

Athraye, S. and Kapur. S (2006)”Industrial concentration in a Liberalizing Economy: A

Study of Indian Manufacturing” Journal of Development Studies, 42(6), 981-999.

Brulhart, M., and Torstensson, J ( 1998), “ Regional Integration, Scale Economies and
Industry location in the European Union” School of Economic Studies, University of
Manchaster, research paper under the stimulation Plan for Economic Sciences in the
European Union (SPES-CT91-0058) and by the Swedish council for Research in the
Humanities and Social Sciences.
 On Spatial Concentration of Organized Manufacturing Industries .... 18 7

Bhattacharya BB, Sakthivel S. (2004): Regional Growth and Disparity in India: comparison
of pre- and post-reform Decades. Economic and Political Weekly, 39(10), 1071-77

Canfei He, Yehua Dennis Wei, Xiuzhen Xie (2008), “Globalization, Institutional Change and
Industrial location: Economic Transition and Industrial Concentration in China”, Regional
Studies, Vol.427, August 2008, pp. 923-945

CeaPraz IL (2008): The concepts of specialization and spatial concentration and the process
of economic integration: Theoretical relevance and Statistical measures. Journal of Romanian
Regional Science Association, 2(1), 68-93

Ellison, G. and Glaeser, E. L. (1997) “Geographic concentration in U.S. Manufacturing

industries: A Dartboard Approach” Journal of Political Economy, 105(5), 889-997

Ghosh, A. (1975): “Concentration and Growth of Indian Industries 1948-68,”, Journal of

Industrial Economics, 23(3), 203-222.

Goschin, Z. Constatin D. L., Roman,M. and Illeanu, B. (2009). “Regional Specialisation and
Geographic Concentration of Industries in Romania”, South-Eastern Europe Journal of
Economics, Vol.1, pp.99-113.

Hirschman (1958): A Strategy for Economic Development, Yale University Press, New Havan.

Hildebrandt, A. and Worz, J., (2004), “Determinants of Industrial Location Patterns in CEECs,
Working Paper N0. 32,The Vienna Institute for International Economic Studies.

Kar, S. Sakthivel, S. (2007): Reforms and Regional inequality in India, Economic and Political
Weekly, 42(47), 69-77

Lall, S.V., Chakraborty, S. (2005): Industrial Location and Spatial inequality: Theory and
Evidence from India. Review of Development Economics, 9(1), 47-68

Chakraborty S. (2003) Industrial location in Post-reform India: patterns of inter-regional

divergence & Intra-regional Convergence, Journal of Development Studies, 40(2), 120-
52

Saikia, D. (2011): “Unorganized Manufacturing Industries in India: A Regional Perspective”

African Journal of Marketing Management, 3(8), 195-206.

Singh, F. P. (2012). “Economic Reforms and Industrial Concentration in Indian Manufacturing

Sector: An Inter-temporal Analysis” International Journal of Marketing, Financial Services
& Management Research, 1(7), 36-53.
18 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Annexure 1: Selected industries with respective NIC Code

Industry description NIC 1970 NIC 1987 NIC 2004 NIC
2008
Food & Food Products 20-21 20-21 15 10
Beverage & Tobacco Products 22 22 16 11
Textile & Textile Products 23+24+25+26 23+24+25+26 17+18 13+14

Leather & Leather Products 29 29 19 15

Chemical & Chemical Products 31 30 24 20

Metal & Metal Products 33+34 33+34 27+28 24+25

Machinery & Equipments other 35+36 35+36 29+30+31+ 28

than Transport 32+33+34

Transport Equipments 37 37 35 30

Annexure 2: List of Industries with Abbreviations.

Abbreviation Industry Name

FOP Food & Food Products

BTP Beverage & Tobacco Products
TEXT Textile & Textile Products
LEATH Leather & Leather Products
CHEM Chemical & Chemical Products
METAL Metal & Metal Products
Machinery & Equipments other than
MACH Transport

TRANS Transport Equipments

Annexure 3: List of States with Abbreviation.

AP ANDHRA PRADESH

ASSAM ASSAM

BIH BIHAR

GUJ GUJARAT

HAR HARYANA

KAR KARNATAKA

MAH MAHARASHTRA

MP MADHYA PRADESH

ORI ORISSA

PUN PUNJAB

RAJ RAJASTHAN

TAMIL TAMIL NADU

UP UTTAR PRADESH

WB WEST BENGAL
 On Spatial Concentration of Organized Manufacturing Industries .... 18 9

Table 1: Share of States in Total Employment of Organized Manufacturing Sector.

Share of States/ Regions Absolute change
States/ Regions
1987-88 1997-98 2010-11 1987-88 to 1997-98 1997-98 to 2010-11

Assam 1.31 1.55 1.31 0.24 -0.24

Bihar 4.93 2.75 0.84 -2.18 -1.91

Orissa 2.06 1.80 2.23 -0.26 0.43

West Bengal 9.36 8.34 5.01 -1.02 -3.33

Eastern Region 17.66 14.44 9.39 -3.22 -5.05

Delhi 1.75 1.37 0.96 -0.38 -0.41

Haryana 3.04 3.16 4.30 0.12 1.14

Punjab 4.96 4.59 4.84 -0.37 0.25

Uttar Pradesh 9.66 9.73 8.64 0.07 -1.09

North- West Region 19.41 16.85 18.74 -2.56 1.89

Gujarat 8.67 8.80 10.20 0.13 1.4

Madhya Pradesh 4.82 4.67 4.86 -0.15 0.19

Maharashtra 15.66 14.76 13.38 -0.9 -1.38

Rajasthan 3.01 2.94 3.40 -0.07 0.46

Central Region 32.16 31.17 31.84 -0.99 0.67

Andhra Pradesh 9.17 12.09 10.25 2.92 -1.84

Karnataka 4.98 6.28 6.16 1.3 -0.12

Kerala 3.14 3.60 3.00 0.46 -0.6

Tamilnadu 11.45 12.85 15.31 1.4 2.46

Southern Region 28.74 34.82 34.72 6.08 -0.1

Total of 16 States 97.97 97.28 94.69 -0.69 -2.59

Other States 2.03 2.72 5.31 0.69 2.59

All India 100 100 100

Source: Based on ASI data, provided by CSO

19 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 2: Share of States in Gross output of Organized Manufacturing Sector.

Share of States/ Regions Absolute change
States/ Regions
1987-88 1997-98 2010-11 1987-88 to 1997-98 1997-98 to 2010-11

Assam 1.25 0.91 0.91 -0.34 0

Bihar 5.26 3.51 0.77 -1.75 -2.74

Orissa 1.90 1.81 1.97 -0.09 0.16

West Bengal 7.20 5.08 4.41 -2.12 -0.67

Eastern Region 15.61 11.31 8.06 -4.30 -3.25

Delhi 2.10 1.86 1.04 -0.24 -0.82

Haryana 3.50 3.94 4.60 0.44 0.66

Punjab 5.00 3.88 3.18 -1.12 -0.7

Uttar Pradesh 8.88 8.70 8.51 -0.18 -0.19

North- West Region 19.48 18.38 17.33 -1.10 -1.05

Gujarat 10.65 12.87 17.25 2.22 4.38

Madhya Pradesh 5.20 5.36 4.56 0.16 -0.8

Maharashtra 21.32 21.00 16.79 -0.32 -4.21

Rajasthan 2.83 3.45 3.21 0.62 -0.24

Central Region 40.00 42.68 41.81 2.68 -0.87

Andhra Pradesh 5.93 6.88 7.35 0.95 0.47

Karnataka 4.26 5.17 6.11 0.91 0.94

Kerala 2.71 2.42 1.75 -0.29 -0.67

Tamilnadu 10.62 10.01 10.10 -0.61 0.09

Southern Region 23.52 24.48 25.31 0.96 0.83

Total of 16 States 98.61 96.85 92.51 -1.76 -4.34

Other States 1.39 3.15 7.49 1.76 4.34

All India 100 100 100

Source: Based on ASI data, provided by CSO

 On Spatial Concentration of Organized Manufacturing Industries .... 19 1

Table 3: Spatial Concentration of Disaggregate Two-digit industries: 1987-88, 1997-

98, 2010-11
Location Herfindahl Index Location Gini

NIC Industry Employment Gross Output Employment Gross Output

2004 Descripti
Code on
1987- 1997- 2010- 1987- 1997- 2010- 1987- 1997- 2010- 1987- 1997- 2010-
88 98 11 88 98 11 88 98 11 88 98 11

15 FOP 0.101 0.093 0.093 0.098 0.965 0.095 0.319 0.458 0.476 0.504 0.306 0.447

16 BTP 0.371 0.414 0.313 0.108 0.111 0.102 0.753 0.760 0.603 0.564 0.540 0.657

17&1
TEXT 0.115 0.113 0.136 0.124 0.120 0.132 0.448 0.436 0.504 0.512 0.530 0.579
8

19 LEATH 0.277 0.272 0.259 0.284 0.242 0.190 0.693 0.690 0.786 0.764 0.768 0.849

24 CHEM 0.135 0.149 0.168 0.162 0.207 0.185 0.405 0.384 0.538 0.439 0.484 0.456

27&2
METAL 0.101 0.089 0.108 0.099 0.098 0.110 0.528 0.554 0.428 0.528 0.542 0.368
8

29-34 MACH 0.116 0.109 0.140 0.113 0.136 0.161 0.423 0.444 0.585 0.407 0.454 0.580

35 TRANS 0.117 0.105 0.137 0.163 0.153 0.184 0.583 0.525 0.647 0.667 0.642 0.757

Source: Authors’ Own estimation based on ASI data.

Note: See Annexure 2 for abbreviations used to denote the Industries.

Table 4: Location Concentration Ratio of Organized Manufacturing Sector by two digit

Industry.
Concentration Ratio Four Leading States(in terms of Gross output)
NIC Industry
Employment Gross Output
2004 Descripti
Code on 1987- 1997- 2010- 1987- 1997- 2010-
1987-88 1997-98 2010-11
88 98 11 88 98 11

UP,TAMIL,AP, MAH,AP,GUJ,
15 FOP 52.64 49.00 49.32 52.43 51.92 49.32 MAH,AP,UP, GUJ
PUN UP

AP,UP,MAH, AP,UP,KAR, MAH

16 BTP 82.03 84.89 51.29 57.16 58.11 51.29 AP,MAH,MP, UP
KAR

MAH,GUJ,PUN, TAMIL,MAH, TAMIL,GUJ, MAH,

17 & 18 TEXT 62.36 59.18 57.87 63.09 5844 57.87
TAMIL GUJ,RAJ PUN

TAMIL,UP,WB, TAMIL,UP,WB,
19 LEATH 82.48 80.58 71.94 81.42 77.37 71.94
MP KAR

MAH,TAMIL,GUJ, GUJ,MH,UP, GUJ,MAH,WB, AP

24 CHEM 66.21 69.95 59.11 67.68 75.29 59.11
UP TAMIL

MAH,BIH,WB,MP MAH,BIH,MP, AP MAH,GUJ, TAMIL,

27&28 METAL 54.66 48.89 43.43 54.41 53.11 43.43
WB

MAH,TAMIL, MAH,KAR,UP, MAH,GUJ,

29-34 MACH 55.62 56.00 62.56 54.26 60.80 62.57
GUJ,UP TAMIL TAMIL,KAR

MAH,HAR, HAR,MAH,PUN,
35 TRANS 60.50 55.26 66.54 67.21 66.87 66.54
TAMIL,UP TAMIL

Source: Authors’ Own estimation based on ASI data.

Note: See Annexure 3 for abbreviations used to denote the states.
19 2 The Journal of Industrial Statistics (2016), 5 (2), 192 - 226

The Price of Prejudice: Employment Trend and Wage Discrimination of

Women Workers in India
Shiney Chakraborty1, Jawaharlal Nehru University, New Delhi, India

Abstract
The present study reveals a persistent gap in labour force participation rate between
rural and urban areas and between males and females in manufacturing industry with a
substantial gender wage differential. The unequal distribution of women workers
belonging to different religious groups is also prominent in manufacturing industries.
Over the time period, a further disaggregated division of manufacturing industries at the
two digit level portrays that women workers are mainly concentrated in food, textile,
wearing apparel and tobacco industries. The Duncan dissimilarity index (ID) for
manufacturing industry reveals significant variations in the distribution of women and
men across major Indian states. The most important finding is the deceleration of women’s
participation rate in the job market. Wage estimation result also suggests a substantial
wage differential between men and women, but the difference has been reducing during
the post-reform period. This study also indicates that gender based wage difference cannot
be explained only by differences in education, experience and skills etc.
1. Introduction
1.1 Labour market discrimination has been a serious issue in India as for other
developing countries, particularly during the post reform period. Discrimination is normally
faced while entering into the labour market and in securing fair wages on participation in
the market, or in both. The International Labour Organization (hereafter ILO) defines
discrimination as “any distinction, exclusion or preference made on the basis of race,
colour, sex, religion, political opinion, national extraction or social origin, which has the
effect of nullifying or impairing equality of opportunity treatment in employment or
occupation”.2 Wage discrimination is generally expressed as an unequal treatment of workers
with roughly equal productivity in terms of their pay either in the form of allocative
discrimination, or other way (Peterson et al, 1997). Gender disparity in wages has been a
common feature in the labour market, particularly in the developing world, but the degree of
disparity varies widely across different ethnic and religious groups. The study provides
special emphasis on gender inequalities simply because it has serious implication on pro-
poor or inclusive growth in an economy (Birdsall and Londono, 1997; Deininger and Olinto,
2000). All over the world women face discrimination at work, with the global average gender
wage gap in paid employment at around 16.5 percent and over 21 percent in Asian countries
in 2008 (The report of the International Trade Union Confederation, 2008). In the Indian
subcontinent, also women have been suppressed and subdued by the hegemony of socio-
cultural patriarchy for several centuries (Ghosh, 2008).
1.2 Against this background, the present study focuses on some aspects of labour
market discrimination in Indian manufacturing industry by looking at employment pattern
and wage differentials across ethnic and religious groups during the post-reform period
1
e-mail: [email protected]
ILO defined discrimination in the Article 1(1a) during 1958 of the Discrimination (Employment and
2

Occupation) Convention. For details see ILO report 2003 (pp 16, Box 2.1).
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 19 3

(1993-2011).The data used in this analysis is from the National Sample Survey Organisation
(NSSO), which covers both organized and unorganized industries3. Disaggregated industry
level data at the two digit level for the time period 1993-1994 and 2011-12 is used for this
study. It is important to mention that four different industrial classifications are used over
this time period. For the surveys between 1993-94 and 1998-99, NIC-87 industrial
classification was used, between 1998-99 to 2004-05, the industrial classification used was
NIC-1998. Between 2004-05 and 2009-10, NIC -2004 and 2011-12 onwards NIC-2008 was
used. Therefore, a concordance exercise across these different classifications has been
undertaken to make the dataset comparable as per the NIC-2008 classification and finally 24
industries within the manufacturing industry is considered for the present purpose.
1.3 During the post reform period, the industry wise distribution of workers at the
disaggregated level indicates that women workers (about 63 per cent) remain concentrated
in the low paying agricultural sector, whereas the proportion of male workers migrating
from agriculture to other secondary and tertiary sector is increasing. The unequal distribution
of women workers of different religious groups is also prominent across industries. There
is also a considerable variation in terms of wage payments for male and female workers
irrespective of socio-religious groups. Although the observed differences in wages between
men and women provide a broad idea about the gender pay gap, this needs to be further
disaggregated across the industries. In this connection an attempt has been made to look
into the female work force participation and wage difference in manufacturing industry in
India as this industry is largely dominated by male. However literature on female participation
in manufacturing industry is very limited except few pioneering ones (Goldar, 2000;
Chaudhuri and Panigrahi, 2013; Ramaswamy, 2015). This study is one of the first few
comprehensive studies on gender wage gap in Indian manufacturing industry which has
focused on gender and socio religious groups. The labour economics literature is replete
with numerous studies on wages and earnings in India by legion of eminent scholars by
specially focusing on gender wage gap but there are very few studies on wages and
earnings in manufacturing industry across socio-religious groups. The present study has
tried to bridge the evidence gap in this regard.
1.4 As labour market participation is not likely to be random the study uses Heckman’s
selection model with two-step estimation techniques to find out the extent of wage gap
explained by education, work experience and other social factors. It analyses how gender
pay gap changed during the two decades of economic reforms in India with pooled data of
two independent samples taken from the 50th and 68throunds of the NSSO on employment
and unemployment in India.The rest of the paper has been structured as follows. Section II
presents an abridged overview of aggregate trends in employment and wages in
manufacturing industry. This section also outlines an assessment of the religion and social
groups in determining the employment pattern and wage levels in manufacturing industry
and indicates the presence of gender segregation in certain industries across the major
states. Section III contains a review of some relevant literature on gender wage gap. Section
IV describes the methodology adopted for addressing the questions and briefly describes
the data sources used for this study. Section V reports regression results that explain
variations in earnings for men and women, over the years and discusses them in the light of
previous findings. Section VI concludes the paper.

3
Annual Survey of Industries (ASI), also collects data annually on organised manufacturing in India.
19 4 The Journal of Industrial Statistics, Vol. 5, No. 2

2. Employment and Wage Trends

2.1 There has been much focus and discussion on the evidence of significant decline
in women’s labour force participation rate4, particularly, as the country is achieving a
sufficiently high rate of economic growth. The latest NSSO round data on employment and
unemployment (NSSO, 2011-12) shows a decline in labour force participation rate both for
males and females in rural area. But in the urban area for female workers there is an increase
of LFPR by about 1 percentage point and a constant LFPR for urban males. It can be seen
from Figure 1 that LFPR is the lowest for rural women in 2011-12 and the lowest for urban
women in 2009-10. It is surprising that rural employment for women fell by almost 8
percentage point inspite of the launch Mahatma Gandhi National Rural Employment
Guarantee Act (MGNREGA), in 2006-07.

2.2 Labour force is basically the ‘economically active’ population and therefore
includes both ‘employed’ and ‘unemployed’ persons and so a lower female labour force
participation rate is a result both of a smaller percentage of women than of men actually
working and a higher unemployment rate5 for females than those for males. Figure 2, indicates
that among those looking for work, a smaller percentage of women than men are getting
jobs which are also supported by a higher rate of unemployment among women than
among men, especially in urban areas.

2.3 However with an overall low and declining female workforce participation rate6,
WPR of women workers in manufacturing industry belonging to different socio-religious
groups indicates that after more than sixty years of independence, Muslim and Christian
women’s participation in the manufacturing industry is significantly lower in both the
sectors (See Table 1). This denotes that social norms confine women’s mobility and entry
into the labour force and keep more women tied to hearth and home. Overtime the WPR of
Hindu male and female has decreased in rural manufacturing industry. Among the Muslim’s
still, urban Muslim women are participating more in the manufacturing industry in
comparison to rural. It has also been observed that overtime the share of Muslim women
participation has increased in manufacturing industry. Among the social groups, WPR of
SC, ST is significantly lower in comparison to the others. One of the essential cruxes that
have emerged is the low probability of socially disadvantageous workers to get jobs in the
manufacturing industry but for all the industries from agriculture to services WPR of socially
backward group is higher. This implies the further disadvantageous position of these
backward groups in manufacturing industry which is considered to be the main engine of
economic growth.

4
The labour force participation rate (LFPR) is defined as the number of persons/ person-days in the
labour force per 1000 persons/ person –days.
5
Unemployment rate (UR) is defined as the number of persons/ person-days unemployed per 1000
persons/ person –days in the labour force.
6
The decline in women’s recorded/recognised workforce participation rate (WPR) is quite different from
actual work done by them as a lot of work done by women is simply not captured by the data. By using
more inclusive definition of work, and by including women who are involved in both paid and unpaid work
the overall workforce participation rate is consistently higher for women than for men (For details see
Ghosh, 2014).
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 19 5

2.4 With the onset of globalization it is believed that with trade openness, the
employment opportunities will increase. But in reality liberalisation does not encourage
female employment at least in organised manufacturing industry rather the use of new
capital intensive technology goes against the women workers (Banerjee and Veeramani,
2015). During the post reform period the Indian labour market witnessed an increasing
feminisation of agriculture and tertiary sector employment with a decrease in secondary
sector employment (Neetha, 2014). In this connection it will be interesting to compare the
change in the share of male and female workers in manufacturing industry and the trend
and pattern of ‘female workforce participation across different socio-economic background
and characteristics. The manufacturing sector currently employs 12.6% of India’s labour
force and is of dualistic structure i.e. the prevalence of a formal/organized sector which
coexists with a large “unorganized sector”. Over a long time period this dualism has persisted
in the manufacturing sector and this situation is not going to change over the next decade.
Goldar (2013) has perceived that over four-fifths of new jobs will be created in the unorganised
manufacturing industry. Even the organised manufacturing industry also witnessed a rise
in contractual employment at the expense of regular employment, which reflects deterioration
in the quality of jobs. This uninspiring performance of the manufacturing industry further
reveals a skewed distribution and concentration of men and women workers in a few
industries. This heterogeneity in absorbing the growing population in manufacturing
industry is mainly because economic growth has benefited capital intensive manufacturing
industries which depends more on skilled workers as opposed to unskilled/low skilled
workers.

2.5 Table 2, indicates that in 1993-94, more than 50 per cent of rural male workers were
mainly concentrated in manufacturing of food product, wood and products of wood and
cork, other non-metallic mineral products and textile industry. Rural female workers were
mainly working in manufacturing of textiles, tobacco, food product and wood and production
of wood and cork products. Major employment providing industries for urban male in terms
of percent share in overall employment is textile, food product and other manufacturing
industry. During the study period textile, tobacco and wearing apparel industries is the
major employment provider for urban female. Over the time period, there is a massive
increase in terms of percentage share of employment in wearing apparel industry both for
male and female in rural and urban areas though the share of female workers in this industry
is higher. Table 2, also reveals that tobacco, textiles and wearing apparel are the female
dominated manufacturing industries.

2.6 There exists an extensive literature on the organized manufacturing sector and its
lacklustre performance in employment generation despite extensive reforms (Goldar(2000),
Aghion et al. (2006), Besley& Burgess (2004), Gupta, et al. (2008), Mehrotra et al. (2014).
With this heterogeneity and limited contribution of the manufacturing sector to employment
generation an attempt has been made to look into the gender wise wage pattern in the
manufacturing industry. With the prevalence of wage differences in the formal manufacturing
sector women are forced to take part in the informal sector because of their economic
requirements. These requirements have pushed them accepting low-paying jobs such as
manufacturing of bidi, cigarette, masala and other related products for their livelihood. Next
to find out gender wage gap in manufacturing industry wage ratio (female to male wage
rate) is reported in Table 3. The value close to 1 indicates less gender wage gap whereas the
19 6 The Journal of Industrial Statistics, Vol. 5, No. 2

value close to 0 implies higher gender wage inequality. Table 3 indicates that gender wage
gap has reduced within the manufacturing industries which have a higher share of female
employment and also for manufacturing industry as a whole.

2.7 After discussing the employment distribution and wage gap in manufacturing
industry for both the gender further disaggregation across the socio-religious groups is
presented in Table 4 and 5. Out of 24 manufacturing industries in terms of National Industrial
Classification (NIC) 2-digit codes, tobacco, textiles and wearing apparel industries together
constitute more than 80 per cent of employment for Muslim female and over 60 per cent of
employment for Christian female in rural area during the entire time period. In urban area
also Muslim female workers share is higher in these industries. Table 4, discloses that over
the time period among the female workers Muslim female’s percentage share is double in
tobacco industries in comparison to Hindu females which echoes their pathos. During
1993-94 to 2011-12, major employment providing industries for Scheduled Caste (hereafter
SC) women workers in terms of percent share in overall employment are manufacturing of
food product, tobacco, textiles and wearing apparel (see Table 5). Over the same time,
among the socially disadvantageous groups more proportion of SC female workers are
involved in tobacco industries whereas the share of Scheduled Tribe (hereafter ST) female
workers are higher in textile industry in rural areas.In 1993-94, the percentage share of urban
ST female worker is higher in tobacco and textile industry in comparison to SC female.
However in 2011-12, employment distribution among the socially disadvantageous groups
in urban area has reversed in tobacco industry implying more percentage of SC women in
tobacco industry. The observations based on statistical findings that are noteworthy here,
is the nature or quality of work available for women workers belonging to disadvantaged
socio religious groups. Women workers are mainly involved in the low paying unskilled
manual jobs.

2.8 Next, to identify the wage gap across the socio-religious groups a disaggregated
analysis is carried out for usual status workers across genders. In Table 6 and 7, a simple
indicator of gender pay gap- ratio of female to male wage rates are reported which suggests
the presence of caste and other forms of social discrimination in India. Table 6, unveils that
in manufacturing industries like Tobacco, Textile and Wearing Apparel gender wage gap
exists in spite of higher representation of women workers. Women workers irrespective of
religion are getting less payment than their men counterparts in the manufacturing industry
which reveals the miserable situation of women workers in the industry. However during
the entire time period in the manufacturing industry as a whole, there was also a reduction
in gender wage inequality among all the religious groups. Within the sub divisions of
manufacturing industry where female’s share is higher gender wage gap reduced for Hindu
and Muslim women workers in rural area (see Table 6 (a)).Table 6 (b) indicates that over the
time period in manufacturing industry gender wage gap persists across all the religious
groups still for Hindu and Muslim female workers there is also a decrease in wage gap in
urban area. Within the sub divisions of manufacturing industry in food, tobacco, textiles
and wearing apparel industries wage gap reduced for Hindu female workers and for Muslim
female workers also wage gap reduced in these industries except food product and tobacco.
The gloomy situation of the Muslim female workers in comparison to the general Hindu
may be because of their low level of human capital endowment and less access to good
quality jobs.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 19 7

2.9 Table 7 depicts the wage gap among the social groups in manufacturing industry
in rural and urban area. Gender wage gap among the social groups has reduced in 2011-12
in both rural and urban area. Within the sub divisions of manufacturing industry, industries
with higher share of SC women workers have experienced a decline in wage gap for them in
rural area. Similarly for ST women workers also wage gap declined in textile industry though
in food and tobacco industries gender wage gap increased over time in rural area. Wearing
apparel industry also has witnessed a massive decline in gender wage gap in the recent
time period. Urban women workers’ belonging to socially disadvantage groups also has
witnessed a decline in gender wage gap in textile and wearing apparel industry. So from the
above analysis it is clear that gender is predominant over other forms of social discrimination
in terms of employment and wages in the manufacturing industry.

2.10 Finally the Duncan dissimilarity index (ID) has been constructed to measure the
industrial sex segregation in manufacturing industry. The Duncan index of dissimilarity is
defined as

(1)

Where denotes the number of female workers in the ith industry and denotes the
number of male workers in the ith industry. The value of the index lies between zero (no
segregation) to 1(full segregation). However the index is interpreted as the share of the
workers that would have to change industry in order to get the same relative distribution of
male and female across industries. Gender segregation across manufacturing industries
upto 2 digits, is computed for two time periods 1993-94 to 2011-12. The notable finding is
that gender segregation has increased sharply for causal wage labourers in both rural and
urban area. State-wise analysis of industrial segregation is done for the seventeen major
states and all around the Indian states there is a wide variation in labour regulations
between both the genders. The Duncan dissimilarity index (ID) for industries indicates that
there is a considerable variation in the distribution of women and men workers across major
Indian states, which is very alarming (see Table 9). Over the time period among the seventeen
major states Assam, Haryana, Jammu and Kashmir and Madhya Pradesh have witnessed a
deceleration in segregation in both rural and urban area. However in comparison to national
average low gender segregation in manufacturing industries is observed only in Madhya
Pradesh and Tamilnadu whereas in urban area all the states have higher gender segregation
in comparison to national average. The low index value of industrial segregation in few
states reflects the better performance of these states in comparison to the other states. At
the sub-national level the wide dissimilarity implies that a larger number of employed people
would need to change their existence industries to restore the distributional equality between
male and female across industries.

3. Review of Literature

3.1 The investigation of disparities in returns to work for same occupation or the
gender gap in wage and earnings has been a significant area of concern for theoretical and
empirical research in economics. In their pioneering papers, Blinder (1973) and Oaxaca
(1973) first formulated a quantitative measure of wage gap. The Blinder-Oaxaca
decomposition technique (hereafter, B-O) that distinguishes between explained variations
19 8 The Journal of Industrial Statistics, Vol. 5, No. 2

(such as level of education) and unexplained variations (which are said to include
discrimination) has been a key tool in the study of wage discrimination. The B-O
decomposition technique is frequently used even today as a summary measure of average
wage discrimination. Oaxaca (1973) analysed the average level of discrimination against the
female workers in the United States and found that women were mainly concentrated in the
low-paid jobs and the gender wage difference was mainly owing to the discrimination
component. Duraisamy and Duraisamy (1996) found a larger discrimination component (67-
77 per cent) for the year 1961-81 for persons with post-secondary schooling. They have
also observed that labour market experience tend to favour males whereas education favoured
females more. Kingdon (1997) analyzed the gender wage gap for urban Lucknow by using
1995 data and found that 45 per cent of the wage gap was due to discrimination and 55 per
cent was due to endowment. In an another study Kingdon and Unni (1998) investigated
determinants of wages on 1987-88 data for urban districts of Madhya Pradesh and Tamil
Nadu and found an average of 75-78 per cent discrimination. They concluded that women
suffer more wage discrimination in the urban labour market owing to their low educational
attainment. Das (2006) estimated the determinants of wages in 1999-2000 in the casual
labour market and found that 27.5 per cent of the difference in male-female wages was due
to the endowment effect whereas the rest (72.5 per cent) was due to discrimination. Bhaumik
and Chakraborty (2008) indicated a decline in gender earnings gap in India from 0.46 to 0.12
from 1987 to 1999.

3.2 In the Indian context, there are a few studies of wage discrimination based on
caste. Bhaumik and Chakraborty (2006) analyzed the wage gap between upper castes and
SC/ST during the period 1987-1999 and found a decline in wage differences across castes
but an increase in wage differences between Muslims and Non-Muslims. They found that
the inter-caste and inter-religion wage gaps were mainly due to endowment differences in
education and experience and not due to discrimination. Madheswaran and Attewell (2007)
examined the wage gap between higher castes and the SCs/STs in the urban regular salaried
job market and their study indicated a lower return of education for workers belonging to
lower castes than that to the upper castes and that 85 per cent of the wage gap across
castes was due to endowment effect whereas 15 per cent was due to discrimination. They
found occupational discrimination to be more pronounced than wage discrimination.

3.3 A review of these studies indicates that gender differentials in wages still exist,
but with varying trends. This is an issue which needs further examination and elaboration
at a disaggregated level with recent data to analyze the trend in wage rates over the period
and to capture the change in the post reform period. Wage equations are estimated to
analyse the role of various explanatory variables in wage determination and a decomposition
analysis of the wage gap is performed which is discussed in the following section.

4. Methodology and data sources

4.1 Most studies on gender wage gap use simple OLS regression techniques to estimate
the wage equations incorporating one cross section. Das (2013) and Sengupta and Das
(2014), on the other hand, estimated Mincerian wage regression in looking into different
aspects of wage discrimination in Indian labour market by taking 50th and 66th round NSS
data. This study has followed the methodology as used in Das (2013) and an independently
pooled cross section from 50th and 68th round unit level data, based on schedule 10 of the
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 19 9

survey is prepared. As the data are collected independently, it causes no problem in pooling
these data over time. In the estimation model the intercept is allowed to change over time
and a year dummy variable is generated to interact with other key explanatory variables to
see if the effect of that variable has changed over the time period. Information on wage and
salary earnings is collected separately for each of the wage/salaried work which is daily
recorded for a person. Here, earnings refer to the wage or salary income which is received
during the reference week by a worker on the basis of usual principal activity status7.

4.2 Following Mincer (1974)8, the wage equation in the frame of pooled data from
two independent random samples is specified as follows:

(2)

4.3 There is no universally accepted set of explanatory variables to estimate the

gender wage gap. The general consensus is to include factors, like education, marital
status, and number of children present in the household to explain the wage gap (Nordman
and Rouband, 2005).The natural logarithm of wages is used as the dependent variable. In
the regression equation two quantitative variables are age and age-squared. Here age is a
good proxy for experience. But the use of age instead of experience may lead to erroneous
result for the individuals with interrupted labour force participation. The quadratic term in
the variable age is to reflect the decreasing wage rate beyond the peak of the career, which
also reflects the possible diminishing return to human capital accumulation through
schooling. The qualitative variables used in the estimation method are as follows: is
a year dummy variable equal to 1 if the person belongs to the 68th round (2011-12) and 0 if
belongs to the 50th round (1993-94). The variable is another dummy variable, the value
of which is equal to 1 if the person is a female and 0 for male. is a general education
dummy variable, with four levels of education, among them illiterate is the reference category
and is the technical education dummy variable the value of which is equal to 1 for
having technical skill and 0 otherwise, is a random error term distributed N (0, e).Various
interaction dummies are also included in the equation (for details see appendix). The intercept

7
In NSS the persons surveyed are classified into various activity categories on the basis of the activities
pursued by them during certain specified reference periods. The usual activity status relates to the activity
status of a person during the reference period of 365 days preceding the date of survey. The activity status
on which a person spent relatively longer time (i.e. major time criterion) during the 365 days preceding
the date of survey is considered as the usual principal activity status, activity status determined with a
one-week reference period is the current weekly status (CWS) and current daily status (CDS) is based on
the daily activity pursued by individuals on each day of the reference week.
8
The standard Mincerian semi-logarithmic earning function is generally used to investigate the determinants
of earnings. In Mincerian earning equation the wage of an individual is assumed to depend upon level of
schooling and job experience.
20 0 The Journal of Industrial Statistics, Vol. 5, No. 2

for 1993-94 is  for male and for female it is (+ ) and in 2011-12 the intercept for male is
(+ )and for female is (+ + ). The coefficients of i and i represents the effect
religion and caste on the wage level for men and (i +i) and (i +i) respectively reflects
the effects of these on women in 1993-94. The similar effects of religion and caste on
women’s wage rate in 2011-12 is measured by (i +i + i) and (i +i + i). The coefficients
i act as the effects of general level of education at different levels for men and (i +i) for
women in 1993-94 and the effects for women in 2011-12 are measured by (i +i +i). Similar
to the general level of education, the effect of technical education on male wage rate can be
measured by  in 1993-94 and for women of the same period it can be measured by ( +).
In 2011-12 the effect of technical education on female wage rate can be measured by ( +
+). Here one more crucial assumption is the same effect of experience on the wages for
male and female workers in both the time periods.
4.4 Most of the studies on gender wage gap estimate the wage equations incorporating
wage earners only. But the presence of a large number of unemployed people can lead to
selectivity bias. Though the problem of selectivity bias arises at two stages of employment
process- first at the stage of joining of labour force and second when a specific occupation
is chosen, this is called occupational selectivity bias. Owing to occupational selectivity
bias wage differential takes place and due to entry barrier of the subordinate group another
types of discrimination takes place. In general, the first problem is taken into consideration
and correction is done. But correction for the second type of selection bias is not usually
done (Neuman and Oaxaca, 2003). In the sample, wages are observed for the working
individuals and simply by ignoring the individuals with no wage earnings will make the
sample non-random or incidentally truncated and here the problem of sample selection bias
will arise (Das, 2013). To improve the efficiency of the estimated coefficients of the wage
equation this study uses the two-step estimation procedure. By following Heckman (1979),
the equation for entering the labour market as follows:

(3)

where is the difference between the market wage and the reservation wage. The
reservation wage is the minimum wage at which the ith individual will be willing to work. If
the wage is below the reservation wage nobody will choose to work. is actually not
observed instead a dichotomous variable with value 1 when the person is participating
in the labour market and 0 otherwise is observed:

=1 if ;

= 0 if

The wage equation specified in (2) is appropriate only if is positive.

4.5 A decomposition technique is also used to divide the wage gap into explained
component and unexplained, or discrimination component. The decomposition of the wage
gap following Blinder – Oaxaca (1973) is widely used to measure labour market discrimination
against women. Decomposition method separates the wage gap into two factors. One can
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 20 1

be explained by different return to individual characteristics (endowment component),

another portion of the differential cannot be explained and is attributable to discrimination
(discrimination component). The gross wage differential can be written as

(4)

where and are the male and female wages, respectively. Becker (1994) extended the
model to include the influence of gender and other personal characteristics and propounded
that in the absence of discrimination wage difference between male and female is the pure
productivity difference (Q) which is defined as

(5)

where Wom/ Wof is the competitive wage ratio in the absence of discrimination. Blinder –
Oaxaca stated it in a different way

(6)

where Wm/ Wf is the observed wage ratio. Expressing it in logarithmic form this can be
written as:
(7)
4.6 In equation 7 the first term on the right hand side is for discrimination while the
second component is for the difference in male–female productivity-related characteristics.
The Binder–Oaxaca (B-O) wage decomposition technique requires an estimation of two
separate wage regressions for male and female workers and can be estimated by using
Heckman selectivity model9. Thus, in order to investigate the sources of gender differentials
in detail, the wage functions of men and women are estimated separately in the frame of
pooled data from two independent random samples in the rural and urban sectors.

(8)

(9)

where W denotes the geometric mean earnings,  is the error term with zero mean and
constant variance, m represents men, f stands for women. Separate estimation is needed for

9
The equations estimated separately for male and female workers are not shown in this paper, as the
inclusion of female dummy regressor in equation 2, is self explanatory. Further the decomposition in
Table 12 has been done by estimating equations 8 and 9.
20 2 The Journal of Industrial Statistics, Vol. 5, No. 2

equation 8 and 9. Now a slight modification of equation 8 and 9 by denoting X as the vector
of mean values of all the regressors, and  as the vector of estimated coefficients and then
by taking a simple log mean wage difference between men and women, the equations can
be written as:

(10)

where

By adding and subtracting Xm f in Equation 10, the equation can be written as

(11)
Again the addition and subtraction of Xf m with Equation 10 yields Equation 12.

(12)
4.7 Equation 11 implies that in the absence of discrimination, female wage structure
would prevail in the market where as equation 12 indicates the prevalence of male wage
structure in a non-discriminatory market. Blinder (1973) and Oaxaca (1973) developed
decomposition approaches to partition the gender wage differential into components caused
by two factors. The first term of the right hand side of the equation (11 and 12) captures
how the male-female wage differential changes in response to changes in the men-women
gap in characteristics. The first term is sometimes called ‘observed X’s’ or ‘observed gender
gap in characteristics’. The second term measures the unexplained wage gap for differences
in coefficients or returns. This term is considered to measure the level of ‘gender
discrimination’.
4.8 The data from the two quinquennial surveys (the 50th and 68th Rounds) on
employment and unemployment (Schedule 10) of the NSSO have been used in this paper.
This allows for an analysis of determinants as well as trends over time. The major aims of
these surveys have been to measure the magnitude of ‘employment and unemployment’ in
quantitative terms disaggregated by various household and population characteristics at
the national level. In order to capture the multi-dimensional aspects of employment and
unemployment, data on several correlates were also gathered. Each quinquennial round is
further segregated into four sub-rounds10 and covers the whole of the Indian Union except
few regions11. A stratified multi-stage sampling design was adopted for the survey both in
rural and urban areas12. In the analysis of the NSS unit record data, usual status workers of
the age group 15-59 years are considered13.

10
The sub-rounds are from July-September, October to December, January to March, and April to June.
The number of sample villages and blocks are allotted for these surveys in each of these four sub-rounds
are equal.
11
i) Leh (Ladakh) and Kargil districts of Jammu & Kashmir ii)interior villages of Nagaland situated
beyond five kilometres of the bus route and iii) villages in Andaman and Nicobar Islands which remain
inaccessible throughout the year
12
The first stage units (FSUs) are villages for rural areas and NSS urban frame survey (UFS) blocks for
urban areas. The ultimate stage units (USU) are households.
Only the age group 15-59 years is considered for the analysis as it is the most productive age group for
13

wage based employment.

 The Price of Prejudice: Employment Trend and Wage Discrimination ... 20 3

5. Results and Discussion

5.1 The sample selection corrected wage equation is estimated with 892,170 data
points from personal level information after excluding children up to age 14 and old-age
above 59 years in the pooled sample in which 652311 data points are censored and the rest
are uncensored. As the wage for nonworking people is unobserved, a probit model for
labour force participation needs to be estimated to test and correct for sample selection
bias. The estimated results are shown in Table 10. The inverse Mill’s ratio, the estimated
value of , as shown in the lower panel of Table 10, is statistically significant. Thus there
was sample selection problem in the data set and wage equation need to be estimated after
correcting for sample selection bias. In addition to education household size, and religion
and caste dummy variables are also included to capture the effects of these groups on the
selection equation. It is also assumed that, the household size have no effect on wage.

5.2 The intercept for 1993-94 is positive and statistically significant. The coefficient
value of the household dummy implies that as the number of members of a household
increases their probability of participation in the labour market decreases as they may be
obliged to spend most of their time on non-economic activities. Again the labour market
participation rate is lower for the workers in rural areas, one of the possible reasons for
which is job opportunities may be lower in the rural area. The estimation result also establishes
that one gender is lagging behind another gender in participating in the labour market over
the years in Indian economy and this is further intensified during 2011-12. Here only wage
earners (that is the paid work category) are considered. Hindu women had a higher chance
of entering into the labour market in1993-94, but the rate declined in 2011-12. On the other
hand Muslim women’s probability of participation in the labour market was lower than that
for women in other religious groups, and further deteriorated in the recent time period. The
positive coefficients of female dummies for backward social groups in 1993-94 imply that
the labour market participation for them was higher as compared with women in higher
castes, and women workers in scheduled caste had larger marginal effect than the tribal
women or women in other backward castes in the country. But the participation rate of
these lower caste women declined in the recent time period. In a nutshell during the post
liberalisation period gender dominates over all other forms of social discrimination as it is
evident from the estimation result.

5.3 The coefficients of dummy variables for females with various level of education
are negative, implying that education did not affect the decision of women’s labour market
participation in 1993-94. However in the recent time period the rate of entering into the job
market improved for women upto primary and middle level of education but deteriorated for
secondary and higher secondary level of education but comparing it with 1993-94, it has
increased marginally. Women without technical knowledge had less access to enter into
the job market in 1993-94, but their chance of getting employment although increasing over
time, at least in probabilistic sense as compared with their men counterparts. As the
hypothesis of sample selection bias is accepted, the use of the censored sample model
(OLS) would lead to incorrect estimates for the valuation of wage equation. This is also
indicated by the Wald chi2 test14, which indicates a significant correlation between error
terms in the selection equation and the wage equation. Hence, Heckman’s technique will
14
Wald chi2 (26) = 442579, Prob> chi2 = 0
20 4 The Journal of Industrial Statistics, Vol. 5, No. 2

provide better result. The estimated results of wage equation, specified in (2), by OLS using
participants in the labour market only and the normal hazard (the inverse Mill’s ratio)
estimated from the first step as an additional regressor are used in the earning function
estimation equation.
5.4 In the wage equation the constant term for 1993-94 is 9.56 and the constant for
2011-12 is 6.89. The weekly wage in logarithm form is used here in nominal rupees. The
negative coefficient value of the year dummy indicates deflationary factor for nominal
wage in 2011-12 as nominal wages grow simply due to inflation and the paper will concentrate
to see the effect of each explanatory variable on real wages. If wages are measured in 1993-
94 rupees experimentally, then deflating 2011-12 wages to 1993-94 rupees is required which
can be done by using the Consumer Price Index. But it turns out that this is not necessary,
provided a 1993-94 year dummy is included in the regression and log (wage) (as opposed to
wage) is used as the dependent variable15. The bottom line is that, for studying how the
return to education or the gender gap has changed, one need to turn nominal wages into
real wages in equation.On average, women workers got significantly lower wage than their
male counterparts in 1993-94,but the gender wage gap decreased notably in 2011-12. This
implies that the situation of women in terms of overall gender wage-gap improved over the
time period.Rural urban wage gap was significant irrespective of the gender dimension of
workers. In 1993-94, the average wage in rural areas was 52 per cent lower than urban wage.
This rural urban wage gap may in turn induce worker to migrate to the urban sector in
search of job.
5.5 It can be inferred that just being literate, or literate with primary and middle
schooling are not enough for men to earn better labour market rewards yet male workers will
get better labour market rewards with an increase in their level of education. However for
women workers different results are coming during the post reform period. With the onset
of liberalisation human capital accumulation plays a great role in deciding the wages of the
women worker though this trend does not hold in the recent time period. Alternatively there
is clear indication of a fall in rate of return for women workers for all levels of education
during the post-reform period in India. This means that education does not help women
workers to get higher level of wages in 2011-12 than in 1993-94. Skill premium, on the other
hand, was 68 per cent for men and 10 per cent for women in 1993-94. Here the negative
coefficient value implies that men workers without having technical education earn 68 per
cent lower wages than the male workers having the technical education similarly the technical
education of women workers help them to earn 10 per cent more wage than without these.
But the skill premium for women declined by 20 per cent in 2011-12 which implies that
during the liberalisation period women workers without having technical education earn 20
per cent lower wages than the pre reform period.
5.6 It can be suggested that Hindu women workers earn 26 per cent lower wages than
their male counterpart in 1993-94 whereas Muslim women earn 41 per cent less wages than

Using real or nominal wage in a logarithmic functional form only affects the coefficient on the year
15

dummy, Y11. To see this, let P11 denote the deflation factor for 2011-12 wages.Then, the log of the real
wage for each person i in the 2011-12 sample is

Now, while wage i differs across people, P11 does not. Therefore, log (P11) will be absorbed into the
intercept for 2011-12.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 20 5

the Muslim male workers during the same time period. In the post reform period there is a
fall in the rate of return for the Hindu women workers in comparison to 1993-94. On the other
hand for Muslim workers there is no change during the post reform period. Across the
castes, during 1993-94 ST and SC women workers earn 1 per cent and 19 per cent less wages
than their male counterparts in the same period. In 2011-12 there is no change for ST women
workers but for SC there is a decline by 6 per cent in comparison to the previous time period.
Comparing this it can be concluded that though gender gap has reduced from 1993-94 to
2011-12 but across the socio-religious group women from the marginalised section are
unable to reap the benefits of this and women from this group are becoming poorer.
5.7 Given the wide spread persistence of the gender wage gap, the final step is to
decompose the wage gap between men and women separately for the years 1993-94 and
2011-12 to bring out: (a) whether there has been a significant change in the gender wage
gap between these two years, and (b) whether the relative contributions of endowment
differential and gender discrimination have changed over time. The decomposition method
developed by Blinder (1973) Oaxaca (1973) is applied here. Here two separate Heckman
wage regression one for male another for female are performed as specified in equation 8
and 9 respectively and the coefficient estimates from the wage regressions are used to
decompose the wage gap between male and female. Decomposition closely follows Blinder’s
exposition and uses both his method and his terminology. Decomposition takes the average
endowment differences between male and female and weights them (multiplies them) by the
male wage workers estimated coefficients. The differences in the estimated coefficients are
weighted multiplied by) the average characteristics of the female-wage workers.
Conventionally, the high-wage groups (in this study high-wage group refers to male) are
regarded as the “non-discriminatory norm”, that is, the reference group.
5.8 Table 12 describes the results of the B-O decomposition technique. A positive
number indicates the percentage by which the gender gap would be reduced if male and
female are equal in respect to the characteristic assuming that the characteristic is rewarded
according to the estimated wage function for female/male. Negative number implies that if
women are more like men in this respect but the wage functions remains the same, then it
will lead to an increase in gender wage gap. The average gender wage gap was 89.4 per cent
in 1993 -94 which further reduced to 37.4 per cent in 2011-12. Table 12 indicates that the
discrimination component is larger than the endowment component in both the time periods.
6. Conclusion
6.1 The present study, based on data from the NSSO, seeks to analyze the structure
and trends of gender-specific wages and earnings in the manufacturing industry by focusing
on ethnic and religious groups in gender dimension during the post-reform period (1993-
2011). The study has considered only male and female workers who are in wage employment
and aged between 15 and 59 years. In the recent years with a minuscule growth in
employment generation Indian economy is experiencing an informalisation of the workforce
with a persistent gap in labour force participation rate between males and females. In the
urban areas, owing to better and diversified employment opportunities LFPR is higher. In
the rural areas, due to the agro-based subsistence economy and poor infrastructure such
employment opportunities are not only rare but also less remunerative.
20 6 The Journal of Industrial Statistics, Vol. 5, No. 2

6.2 Female labour force participation rate can be lower on account of the poor work
environment for them, which discourages their supply response. It can also be due to
limited employment opportunities. There could be other reasons too for low FLFPR: First,
women belonging to higher economic status may not enter into the labour market due to
family values, cultural or societal norms (demand side argument); second, shortage of well-
paying secure jobs for the educated women may reduce their work participation (supply
side argument). Over the years it has been observed that female workers are mainly
concentrated in tobacco, textiles and wearing apparel industry and gender wage gap (in
terms of female to male wage ratio) is decreasing in manufacturing industry.

6.3 Regression result from the estimation of participation equation portrays the
negative effect of the household size on the participation rate. It also outlines that labour
market participation rate are lower for the workers in rural areas than in comparison to urban
areas. One of the essential cruxes that have emerged during the post reform period is the
deceleration of women’s participation rate in the job market for all types of categories
alternatively gender trumps all other forms of social discrimination. With the primary level
of education women workers probability in finding a job increases in the after reform period
but higher level education is not going to help them to get a sophisticated job. Education
does not help women workers to get higher level of wages in 2011-12 than in 1993-94 this
may be an effect of a decline in Public sector jobs with social securities along with large
scale informalisation of various jobs. Wage estimation and Decomposition result suggest
that a substantial wage difference between men and women exists in the Indian labour
market, but the difference has been reducing during the post-reform period. Decomposition
result also highlights the higher incidence of discriminatory practices in India. The presence
of high discrimination component indicates that gender based wage difference is pervasive
and unless the stereotype behaviour of society changes or women’s position in the labour
market undergoes radical changes, the wage structure will continue to be imbalanced and
unequal in spite of the presence of the “equal remuneration act”. So any effort to reduce,
must address gender inequalities from a multi-dimensional perspective which accounts for
changing perceptions and notions regarding women’s role and contribution among different
agents of the labour markets, in addition to the enhancement of women’s employment. A
deliberate government policy and efforts are needed to reduce the wage difference and it
should be aimed at empowering the women who suffer from discrimination. Only an inclusive
growth strategy shall lead to lowering of wage differentials and removal of disparities in
living standards of people.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 20 7

References
Aghion, P., et al. (2006),”The unequal effects of liberalization: Evidence from dismantling
the License Raj in India.” CEP Discussion Paper no. 728.
Banerjee, P. and Veeramani, C. (2015), “Trade Liberalisation and Women’s Employment
Intensity: Analysis of India’s Manufacturing Industries”, Working Paper IGIDR, 2015,
Available at: http://www.igidr.ac.in/pdf/publication/WP-2015-018.pdf, Accessed on 8.7.2016.
Becker. G.S. (1985), “Human Capital, Effort, and the Sexual Division of labour”, Journal of
Labour Economics, Vol. 3, Issue 1, part 2: Trend’s in Women’s Work, education and Family
Building, S33-S58.
________ (1994), “Human Capital: A Theoretical and Empirical Analysis with Special
Reference to Education” (3rd Edition), The University of Chicago Press. Available at National
Bureau of Economic Research.
Besley, T. and R. Burgess.(2004)”Can regulation hinder economic performance? Evidence
from India.” Quarterly Journal of Economics, Vol.119, No. 1,pp. 91-134.
Bhaumik, S. K. and Chakraborty, M. (2006), “Earnings Inequality in India: Has the Rise of
Caste and Religion Based Politics in India had an Impact?”, IZA Discussion Papers, No.
2008, IZA, Bonn, Germany.
______ (2008), “Does Move to Market have an Impact on Earnings Gap across Gender?
Some Evidence from India”, Applied Economics Letters, No-15:8, pp. 601-605.
Birdsall, N. and Londono, J.L. (1997), “Asset Inequality Does Matter: Lessons from Latin
America”, OCE Working Paper, Inter American Development Bank, Washington, March,
pp. 1-27.
Blinder, A.S (1973), “Wage Discrimination: Reduced Form and Structural Estimates”, Journal
of Human Resources, VIII, 4,pp. 436-55.
Cotton, C. J. (1988), “On the Decomposition of Wage Differentials”, Review of Economics
and Statistics, 70, pp. 236-43.
Chandrasekhar. C.P and Ghosh. G. (2014), “Are Women Really Working Less in India?”
Available at http://www.macroscan.org/fet/aug14/pdf/Working_Less.pdfAccessed on 30
October 2014.
Chaudhuri, B and Panigrahi, A.K. (2013), “Gender Bias in Indian Industry”, Journal of
Industrial Statistics, Vol. 2, No. 1, pp.108-127.
Das, M.B. (2006), “Do Traditional Axes of Exclusion Affect Labour Market Outcomes in
India”, Social Development Paper, South Asia Series, June, No. 97.
Das, P. (2013), “Measuring Sample Selection Corrected Gender Wage Gaps in India: 1993-94
to 2009-10”, Bangladesh Development Studies, Vol. XXXVI, No. 14, pp. 81-97.
Deininger, K. and Olinto, P. (2000), “Asset distribution, Inequality, and Growth”, World
Bank Policy Research Working Paper, June, pp. 1-28.Available athttp://
elibrary.worldbank.org/doi/pdf/10.1596/1813-9450-2375,Accessed on17 April 2014.
20 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Duraisamy, M. and Duraisamy, P. (1996), “Sex Discrimination in Indian Labour Markets”,

Feminist Economics, Vol. 2, No. 2, pp. 41-61.
Ghosh, J. (2014), “What Exactly is Work?”, Available at http://www.macroscan.org/cur/
oct14/cur31102014Exactly_work.htm Accessed on 9 January 2015.
_______(2008), Never Done and Poorly Paid: Women’s Work in Globalising India, Women
Unlimited Press, New Delhi.
Goldar, B. 2000.”Employment Growth in Organised Manufacturing in India”. Economic and
Political Weekly, Vol 35, No 14, April 1-7.
_______(2013).”Sustaining a High Rate of Industrial Growth in India in the Next 10 Years”,
Paper presented at a Workshop on ‘Sustaining High Growth in India’, 25-26 July 2013,
Institute of Economic Growth, Delhi.
Gupta, P., et al. (2008),”Big Reforms but Small Payoffs: Explaining the Weak Record of
Growth in Indian Manufacturing”, In S. Bery, B. Bosworth, and A. Panagariya (eds.), India
Policy Forum, vol. 5, Sage, Delhi, pp.59-108.
Gupta, P., et al. (2008),”What constrains Indian manufacturing?” Working Paper no. 211.Indian
Council for Research on International Economic Relations.
Heckman, J. 1976, “The Common Structure of Statistical Models of Truncation, Sample
Selection, and Limited Dependent Variables and a Simple Estimator of Such Models.” Annals
of Economic and Social Measurement, 5, pp. 475-492.
________ (1979) “Sample Selection Bias as a Specification Error”, Econometrica, 47, pp.
153-63.
International Labour Organisation Report (2003), “Time for Equality at Work”, International
Labour Conference, Report I(B), International Labour Office, Geneva.
International Trade Union Confederation Report (2008), The Global Gender Pay Gap,
Belgium.
Kingdon, G. G. (1997), “Does the Labour Market Explain Lower Female Schooling in India”,
Available at http://www.econ.ox.ac.uk/Members/geeta.kingdon/Published papers/
JDevstudLabourlowerfemaleschooling.pdf.
Kingdon, G.G. &Unni, J. (1998), “Education and Women’s Labour Market Outcomes in
India: An Analysis Using NSS Household Data”, Economics Series Working Papers, 99201,
University of Oxford, Department of Economics.
Madheswaran, S. and Attewell, P. (2007), “Caste Discrimination in the Indian Urban Labour
Market: Evidence from the National Sample Survey”, Economic and Political Weekly,
October 13, pp. 4146-4154.
Mehrotra, S., Parida J., Sinha.S, and Gandhi, A.2014. “Explaining Employment Trends in the
Indian Economy: 1993-94 to 2011-12” Economic and Political Weekly, Vol XLIX No. 32.
Mincer, J. A. (1974), “Schooling, Experience, and Earnings”, NBER Books, National Bureau
of Economic Research.Available at http://www.nber.org/books/minc74-1, Accessed on 12
July, 2011.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 20 9

N, Neetha. (2014), “Crisis in Female Employment: Analysis across Social Groups”, Economic
and Political Weekly, November 22, pp. 50-59.
Neumark, D. (1988), “Employers Discriminatory Behaviour and the Estimation of Wage
Discrimination”, Journal of Human Resources, 23, pp. 279-95
Nordman, C. and Rouband, F. (2005), “Reassessing the Gender wage Gap: Does labour
Force Attachment Really Matter? Evidence for matched labour force and Biographical
Survey in Madagascar”, ECINEQ Working Paper-16 in Richard Sebaggala (2007), “Wage
Determination and Gender Discrimination in Uganda”, No. 50, Economic Policy and Research
Centre.
Neuman, S. and Oaxaca, R. (2003), “Estimating Labour Market Discrimination with Selectivity-
Corrected Wage Equations: Methodological Considerations and Illustration for Israel”,
Discussion Paper No. 2-2003.
Oaxaca, R. (1973), “Male-Female Wage Differentials in Urban Labour Markets”, International
Economic Review, Vol. 14, No. 3, October, pp. 693-709.
Oaxaca, L. R. and Ransom, R. M. (1994), “On Discrimination and Decomposition of Wage
Differentials, Journal of Econometrics, 61, pp. 5-21.
O’ Donnell, Owen et al, (2008), “Analyzing Health Equity Using Household Survey Data-
A Guide to Techniques and Their Implementation”, World Bank, Washington, DC.
Peterson, T., Snartland, V., Becken, L. E. Olsen, K. M. (1997), “Within Job Wage
Discrimination and the Gender Wage Gap: The case of Norway”, European Sociological
Review, Vol- 13, No-2, pp. 199-213.
Ramaswamy, K.V. (2015), “Labour, Employment and Economic Growth in India”, Cambridge
University Press, New Delhi.
Reimer, W.C. (1983), “labour Market Discrimination against Hispanic and Black men”, Review
of Economics and Statistics, 65, pp. 570-79.
Sengupta, A., Das, P. (2014), “Gender Wage Discrimination across Social and Religious
Groups in India Estimates with Unit Level Data”, Economic and Political Weekly, Vol.
XLIX, No. 21, pp. 71-76.
Unni, J. (2005), “Wages and Incomes in Formal and Informal Sectors in India”, Indian
Journal of Labour Economics, Vol.48, No.2, 311-317.
Wooldridge, J. M. (2009), Introductory Econometrics a Modern Approach (Indian Edition).
Cengage Learning India Private Limited, New Delhi.
21 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Figure 1: LFPRs in usual status (ps+ss) during 1993-94 to 2011-12 of all ages.

Source: NSS Report No. 554: Employment and Unemployment Situation in India, 2011-12
Figure 2: Unemployment rates (in percentage) according to usual status from 1993-94
to 2011-12

Source: NSS Report No. 554: Employment and Unemployment Situation in India, 2011-12
Table 1: WPRs (in percent) of various socio-religious groups in manufacturing
industry.
1993-94 2011-12
Rural Urban Rural Urban
Male Female Male Female Male Female Male Female
Hindu 85 83 78 73 79 71 73 70

Muslim 10 12 16 22 15 21 21 23
Religious Groups

Christian 2 4 3 4 4 5 2 4

Others 3 2 4 1 3 4 3 3

Total 100 100 100 100 100 100 100 100

SC 18 18 3 4 18 18 13 14

ST 7 10 9 10 10 11 3 4
Social Groups

Others 75 73 88 86 72 71 84 82

Total 100 100 100 100 100 100 100 100

Source:Estmated from Unit-level data, NSS Employment and Unemployment Survey, 1993-94 and
2011-12
Note: Total may not add up to 100 due to rounding off. Considered only the usual status workers of 15
to 59 years age group.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 21 1

Table 2: Distribution of usual status workers in manufacturing industry

1993-94 2011-12
Divisions of Manufacturing Industry
Rural Urban Rural Urban

Male Female Male Female Male Female Male Female

Food Product 18.06 13.91 9.43 9.62 14.68 9.06 10.37 7.56

Beverages 1.29 0.9 0.83 0.91 1.05 0.76 0.89 2.09

Tobacco 5.16 25.41 2.34 22.63 1.9 21.7 1.07 13.24

Textiles 11.46 23.36 17.1 24.08 7.69 19.55 15.08 26.84

Wearing Apparel 6.85 9.21 7.45 13.3 13.14 22.36 12.53 29.25

Leather and related products 1.38 0.53 2.57 1.92 1.09 0.66 2.88 1.17

Wood and products of wood and cork, except furniture 12.9 10.17 5.27 3.64 13.02 8.96 5.12 1.95

Paper and paper products 0.64 0.29 1.29 1.82 0.78 0.35 1.31 1.74

Printing and reproduction of recorded media 0.87 0.21 3.1 1.4 0.55 0.24 2.04 0.64

Coke and refined petroleum products 0.53 0.05 0.57 0.18 0.31 0.03 0.62 0

Chemicals and chemical products 2.11 1.94 4.19 5.43 1.44 1.88 2.34 1.46

Pharmaceuticals, medicinal chemical and botanical products 0.38 0.09 1.43 0.91 0.9 0.28 1.73 1.03

Rubber and plastics products 1.35 0.15 2.53 1.18 1.69 0.66 2.75 1.38

Other non-metallic mineral products 11.94 8.35 4.46 3.59 13.36 7.6 5.29 3.51

basic metals 1.95 0.26 5.36 0.93 2.59 0.42 4.17 0.21

Fabricated metal products, except machinery and equipment 4.46 0.47 6.81 1.08 6.21 1.67 7.43 0.85

Computer, electronic and optical products 0.57 0.12 1.65 1.31 0.33 0.17 1.42 0.5

Electrical equipment 0.75 0.12 2.27 0.8 1.35 0.21 2.28 0.67

Machinery and equipment n.e.c. 2.95 0.62 4.95 0.73 1.14 0.21 2.12 0.11

Motor vehicles, trailers and semi-trailers 0.23 0 0.88 0.03 1.32 0.07 2.07 0.25

Other transport equipment 0.49 0.06 2.4 0.32 0.34 0.03 1.59 0.11

Furniture 4.69 1.38 2.85 0.48 9.19 0.17 5.07 0.43

Other manufacturing 6.72 2.34 7.52 3.65 3.64 2.85 7.47 4.86

Repair and installation of machinery and equipment 2.26 0.08 2.76 0.04 2.25 0.1 2.37 0.14

Total 100 100 100 100 100 100 100 100

Source: Same as Table 1

21 2 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 3: Gender wage gap by usual status in manufacturing industry.

Divisions of Manufacturing Industry 1993-94 2011-12
Rural Urban Rural Urban
wage wage wage wage
Ratio Ratio Ratio Ratio
Food Product 0.52 0.28 0.65 0.40
Beverages 0.13 0.03 0.95 0.39
Tobacco 0.53 0.40 0.70 0.59
Textiles 0.31 0.29 0.65 0.58
Wearing Apparel 0.43 0.32 0.92 0.58
Leather and related products 1.98 0.43 0.78 1.19
Wood and products of wood and cork, except furniture 0.22 0.16 0.39 0.78
Paper and paper products 0.23 0.19 0.44 0.65
Printing and reproduction of recorded media 0.32 1.19 1.82 0.38
Coke and refined petroleum products 0.39 0.75 0.94 n.a
Chemicals and chemical products 0.23 0.31 0.46 0.37
Pharmaceuticals, medicinal chemical and botanical
products 0.74 0.77 0.63 0.81
Rubber and plastics products 0.39 0.52 0.80 0.30
Other non-metallic mineral products 0.42 0.31 0.73 0.50
Basic metals 0.35 0.90 0.38 0.27
Fabricated metal products, except machinery and
equipment 1.13 0.51 0.43 0.71
Computer, electronic and optical products 0.44 0.67 0.34 0.34
Electrical equipment 0.31 0.64 0.33 0.77
Machinery and equipment n.e.c. 0.79 0.65 1.39 0.82
Motor vehicles, trailers and semi-trailers n.a 0.37 0.25 0.89
Other transport equipment 0.76 0.64 0.41 0.61
Furniture 0.45 0.03 0.68 0.75
Other manufacturing 0.25 0.28 0.85 0.89
Repair and installation of machinery and equipment 0.00 1.17 n.a 1.37
Total Manufacturing 0.34 0.31 0.55 0.49

Source: Same as Table 1

Note: (n.a) implies no observations or unavailabilty of wage data.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 21 3

Table 4: Distribution of female usual workers by religious group in manufacturing

industry.
(a) Rural
1993-94 2011-12
Divisions of
Manufacturing Industry
Hindu Muslim Christian Others Hindu Muslim Christian Others

Food Product 15.06 7.89 12.24 3.96 10.42 6.58 4.93 2.7

Beverages 0.91 0.12 2.04 3.96 0.84 0 2.11 1.8

Tobacco 22.82 43.16 20.82 34.65 19.96 35.08 5.63 2.7

Textiles 22.31 26.14 38.37 21.78 14.9 32.38 28.17 25.23

Wearing Apparel 8.73 13.32 7.76 5.94 22.47 15.85 30.28 45.05

Leather and related

products 0.64 0 0 0 0.59 1.18 0 0

Wood and products of

wood and cork, except
furniture; 11.27 4.07 5.71 9.9 10.77 1.85 14.79 6.31

Paper and paper products 0.35 0 0 0 0.49 0 0 0

Printing and reproduction

of recorded media 0.18 0.25 0 1.98 0.29 0.17 0 0

Coke and refined

petroleum products 0.05 0 0 0 0 0 0 0.9

Chemicals and chemical

products 1.75 2.1 6.53 0 2.21 1.18 1.41 0

Pharmaceuticals, medicinal
chemical and botanical
products 0.11 0 0 0 0.2 0 2.11 0.9

Rubber and plastics

products 0.15 0 0.82 0 0.59 0.84 1.41 0

Other non -metallic mineral

products 9.69 1.73 0 8.91 9.34 1.35 4.93 12.61

Basic metals 0.31 0 0 0 0.59 0 0 0

Fabricated metal products,

except machinery and
equipment 0.56 0 0 0 2.16 0.34 1.41 0

Computer, electronic and

optical products 0.15 0 0 0 0.25 0 0 0

Electrical equipment 0.15 0 0 0 0.25 0.17 0 0

Machinery and equipment

n.e.c. 0.65 0 0 4.95 0.29 0 0 0

Motor vehicles, trailers and

semi-trailers 0.1 0 0 0

Other transport equipment 0.07 0 0 0 0.05 0 0 0

Furniture 1.58 0 0.41 3.96 0.2 0 0.7 0

Other manufacturing 2.42 1.23 5.31 0 2.95 2.87 2.11 1.8

Repair and installation of

machinery and equipment 0.09 0 0 0 0.1 0.17 0 0

Total 100 100 100 100 100 100 100 100

Source: Same as Table 1

21 4 The Journal of Industrial Statistics, Vol. 5, No. 2

(a) Urban
Divisions of Manufacturing 1993-94 2011-12
Industry Hindu Muslim Christian Others Hindu Muslim Christian Others

Food Product 10.66 5.85 12.55 8.43 8.3 4.13 17.09 3.7

Beverages 0.89 0.43 4.43 0 2.29 0.15 5.13 8.64

Tobacco 17.38 42.95 7.75 6.02 10.08 24.62 8.55 4.94

Textiles 26.32 20.06 11.07 3.61 22.85 39.76 17.95 32.1

Wearing Apparel 12.41 14.71 21.03 16.87 30.89 22.78 30.77 39.51

Leather and related products 1.96 1.85 2.21 0 1.42 0.15 3.42 0

Wood and products of wood and

cork, except furniture; 4.23 1.23 5.9 6.02 2.44 0.46 2.56 1.23

Paper and paper products 1.67 1.97 1.48 9.64 2.04 1.38 0 0

Printing and reproduction of

recorded media 1.48 0.31 5.54 3.61 0.87 0 0.85 0

Coke and refined petroleum

products 0.21 0.12 0 0

Chemicals and chemical

products 6.05 3.51 5.54 3.61 1.93 0.15 0.85 1.23

Pharmaceuticals, medicinal
chemical and botanical products 0.86 0.18 4.43 7.23 1.42 0.15 0 0

Rubber and plastics products 1.29 0.92 0.74 0 1.63 0.92 0.85 0

Other non -metallic mineral

products 4.3 1.11 2.21 12.05 4.48 1.53 0.85 0

Basic metals 1.2 0.12 0.74 0 0.25 0 0.85 0

Fabricated metal products,

except machinery and
equipment 1.22 0.25 2.21 4.82 0.66 1.07 1.71 2.47

Computer, electronic and optical

products 1.56 0.37 2.58 0 0.71 0 0 0

Electrical equipment 0.86 0.12 2.58 4.82 0.97 0 0 0

Machinery and equipment n.e.c. 0.67 0.18 3.69 6.02 0.15 0 0 0

Motor vehicles, trailers and

semi-trailers 0.04 0 0 0 0.36 0 0 0

Other transport equipment 0.36 0.12 0.74 0 0.15 0 0 0

Source: Same as Table 1

 The Price of Prejudice: Employment Trend and Wage Discrimination ... 21 5

Table 5: Distribution of female usual workers by social group in manufacturing

industry.
(a) Rural
1993-94 2011-12
Divisions of Manufacturing
Industry
SC ST Others SC ST Others

Food Product 15 29.23 11.77 7.58 5.56 9.97

Beverages 0.52 1.58 0.91 0.19 5.23 0.24

Tobacco 17.42 11.53 29.13 23.3 7.19 23.46

Textiles 10 18.01 27 12.5 16.01 21.9

Wearing Apparel 12.36 6.32 8.83 19.89 14.05 24.24

Leather and related products 1.63 0 0.33 1.33 0 0.59

Wood and products of wood and

cork, except furniture; 23 25.75 5.17 17.61 34.97 2.83

Paper and paper products 0.43 0 0.29 0.38 0.33 0.34

Printing and reproduction of

recorded media 0 0 0.29 0 0 0.34

Coke and refined petroleum

products 0.26 0 0 0 0 0.05

Chemicals and chemical products 2.06 0.32 2.12 1.52 0.98 2.1

Pharmaceuticals, medicinal
chemical and botanical products 0 0 0.12 0.19 0.65 0.24

Rubber and plastics products 0 0.32 0.16 0.19 0.33 0.83

Other non-metallic mineral

products 12 5 7.93 10.42 8 6.79

Basic metals 0 0 0 0.95 1 0.24

Fabricated metal products, except

machinery and equipment 0.69 0.63 0.39 0.95 3.59 1.56

Computer, electronic and optical

products 0 0 0.16 0.38 0 0.15

Electrical equipment 0 0 0.16 0 0.33 0.24

Machinery and equipment n.e.c. 0.26 0 0.78 0 0.33 0.24

Motor vehicles, trailers and semi -

trailers 0.19 0 0.05

Other transport equipment 0 0 0.08 0 0 0.05

Furniture 4.12 1.11 0.76 0 0.33 0.2

Other manufacturing 1.2 0 2.92 2.46 0.98 3.23

Repair and installation of

machinery and equipment 0.26 0 0.04 0 0 0.1

To tal 100 100 100 100 100 100

Source: Same as Table 1

21 6 The Journal of Industrial Statistics, Vol. 5, No. 2

(a) Urban
Divisions of Manufacturing Industry 1993-94 2011-12

SC ST Others SC ST Others

Food Product 10.45 5.8 10.05 6.72 3.54 7.9

Beverages 5.6 1.19 0.68 10.08 4.42 0.65

Tobacco 10.45 19.13 23.59 9.3 6.19 14.24

Textiles 16.79 22.69 24.57 17.31 27.43 28.4

Wearing Apparel 21 9.1 13.48 24.29 30.97 30

Leather and related products 0 10 1 3.1 0 0.91

Wood and products of wood and cork,

except furniture; 8.58 9.63 2.69 6.46 6.19 0.99

Paper and paper products 0.37 0.66 2.03 2.84 0 1.64

Printing and reproduction of recorded

media 1.49 1.85 1.34 2.07 0.88 0.39

Coke and refined petroleum products 0.75 0 0.18

Chemicals and chemical products 3.36 4.22 5.67 0.78 0.88 1.6

Pharmaceuticals, medicinal chemical

and botanical products 0 0.53 1 0.78 0 1.12

Rubber and plastics products 3.36 1.32 1.06 0.78 0 1.55

Other non-metallic mineral products 7.09 7.12 3.01 5.94 10.62 2.76

Basic metals 4.85 1.32 0.71 0.26 0 0.22

Fabricated metal products, except

machinery and equipment 0.75 0 1.22 0.26 0.88 0.95

Computer, electronic and optical

products 1.49 0.4 1.42 1.03 0 0.43

Electrical equipment 0 0.26 0.9 0.78 0 0.69

Machinery and equipment n.e.c. 2.61 0 0.74 0.26 0 0.09

Motor vehicles, trailers and semi-trailers 0 0 0.03 0.78 0 0.17

Other transport equipment 0 0.26 0.34 0 0 0.13

Furniture 1.12 1.32 0.35 0 5.31 0.26

Other manufacturing 0 2.9 3.9 5.94 2.65 4.79

Repair and installation of machinery and

equipment 0 0.4 0 0.26 0 0.13

Total 100 100 100 100 100 100

Source: Same as Table 1

 The Price of Prejudice: Employment Trend and Wage Discrimination ... 21 7

Table 6: Wage Ratio of religious groups by usual status in manufacturing industry.

(a) Rural
Divisions of Manufacturing 1993-94 2011-12
Industry Hindu Muslim Christian Others Hindu Muslim Christian Others
Food Product 0.52 0.44 0.57 n.a 0.66 0.73 0.25 0.55
Beverages 0.16 n.a n.a n.a 0.98 n.a 0.45 n.a

Tobacco 0.51 0.57 n.a 1.89 0.55 0.97 n.a n.a

Textiles 0.25 0.67 0.42 n.a 0.62 0.82 0.86 0.65

Wearing Apparel 0.40 0.57 n.a 0.84 0.84 0.61 n.a 0.85
Leather and related products 1.93 n.a n.a n.a 0.74 0.81 n.a n.a

Wood and products of wood and

cork, except furniture 0.19 0.80 0.07 n.a 0.39 0.65 0.28 n.a
Paper and paper products 0.24 n.a n.a n.a 0.42 n.a n.a n.a

Printing and reproduction of recorded

media 0.22 1.63 n.a n.a 2.02 n.a n.a n.a

Coke and refined petroleum products 0.36 n.a n.a n.a n.a n.a n.a 0.96

Chemicals and chemical products 0.25 n.a 0.12 n.a 0.44 0.58 0.62 n.a
Pharmaceuticals, medicinal chemical
and botanical products 0.71 n.a n.a n.a 0.57 n.a n.a n.a
Rubber and plastics products 0.40 n.a 0.30 n.a 0.76 1.26 0.45 n.a

Other non-metallic mineral products 0.43 0.63 n.a 0.36 0.74 0.51 0.88 0.68

Basic metals 0.36 n.a n.a n.a 0.37 n.a n.a n.a
Fabricated metal products, except
machinery and equipment 1.14 n.a n.a n.a 0.42 n.a 0.31 n.a
Computer, electronic and optical
products 0.45 n.a n.a n.a 0.38 n.a n.a n.a

Electrical equipment 0.31 n.a n.a n.a 0.33 n.a n.a n.a
Machinery and equipment n.e.c. 0.85 n.a n.a n.a 1.34 n.a n.a n.a

Motor vehicles, trailers and semi-

trailers 0.48 n.a n.a n.a 0.23 n.a n.a n.a

Other transport equipment 0.14 n.a n.a n.a 0.42 n.a n.a n.a

Furniture 0.42 0.38 0.92 n.a 0.68 n.a n.a n.a

Other manufacturing n.a n.a n.a n.a 0.79 n.a 5.54 n.a

Repair and installation of machinery

and equipment n.a n.a n.a n.a n.a n.a n.a n.a

Total Manufacturing 0.34 0.40 0.21 0.27 0.54 0.59 0.53 0.57

Source: Same as Table 1

Note: Same as Table 3 .
21 8 The Journal of Industrial Statistics, Vol. 5, No. 2

(b) Urban
Divisions of Manufacturing Industry 1993-94 2011-12
Hindu Muslim Christian Others Hindu Muslim Christian Others
Food Product 0.27 0.29 0.53 0.14 0.44 0.27 0.10 0.29

Beverages 0.03 n.a 0.09 n.a 0.38 n.a 1.00 n.a

Tobacco 0.38 0.48 n.a n.a 0.87 0.41 n.a n.a

Textiles 0.29 0.26 0.53 2.22 0.56 0.44 1.99 0.29

Wearing Apparel 0.33 0.14 0.60 1.99 0.56 0.54 0.65 0.26

Leather and related products 0.44 0.30 0.78 n.a 1.00 0.76 1.20 n.a

Wood and products of wood and cork,

except furniture 0.14 0.83 0.04 n.a 0.39 1.82 1.49 n.a

Paper and paper products 0.17 0.08 0.53 0.12 0.71 0.17 n.a n.a

Printing and reproduc tion of recorded

media 0.62 n.a 4.69 n.a 0.40 n.a 0.29 n.a

Coke and refined petroleum products 0.88 n.a n.a n.a 0.09 0.37 n.a n.a

Chemicals and chemical products 0.25 0.06 1.14 0.56 0.87 0.36 n.a n.a

Pharmaceuticals, medicinal chemical

and botanical products 0.48 0.24 3.77 2.18 0.27 0.13 0.06 n.a

Rubber and plastics products 0.53 0.90 0.29 n.a 0.32 0.42 0.56 n.a

Other non-metallic mineral products 0.29 0.52 0.27 0.33 0.52 n.a 1.61 n.a

Basic metals 0.94 n.a 0.34 n.a 0.58 0.29 0.28 n.a
Fabricated metal products, except
machinery and equipment 0.47 0.58 0.26 0.90 0.83 n.a n.a n.a

Computer, electronic and optical

products 0.62 0.29 1.38 n.a 0.56 n.a n.a n.a

Electrical equipment 0.67 1.07 0.99 n.a 0.64 n.a n.a n.a

Machinery and equipment n.e.c. 0.60 1.10 0.55 0.82 1.90 n.a n.a n.a

Motor vehicles, trailers and semi-trailers 0.36 n.a n.a n.a 0.50 n.a n.a n.a

Other transport equipment 0.59 1.18 0.82 n.a n.a n.a 0.17 5.86

Furniture 0.04 n.a n.a n.a 1.09 0.39 0.97 0.31

Other manufacturing 0.36 0.09 0.36 n.a 1.80 n.a n.a n.a

Repair and installation of machinery and

equipment 1.00 n.a n.a n.a n.a n.a n.a n.a

Total Manufacturing 0.30 0.23 0.85 0.70 0.49 0.46 0.38 0.26

Source: Same as Table 1

Note: Same as Table 3.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 21 9

Table 7: Wage Ratio (Female to Male) of social groups by usual status in

manufacturing industry.
(a) Rural
1993-94 2011-12
Divisions of Manufacturing Industry SC ST Others SC ST Others

Food Product 0.43 0.91 0.33 0.69 0.86 0.60

Beverages 2.60 n.a 0.05 1.16 1.01 0.95

Tobacco 0.57 0.63 0.52 0.86 0.58 0.74

Textiles 0.40 0.48 0.29 0.51 0.80 0.63

Wearing Apparel 0.42 0.32 0.42 1.44 0.77 0.94

Leather and related products n.a n.a 1.61 n.a 1.03 0.62

Wood and products of wood and cork,

except furniture 0.26 0.51 0.15 0.61 0.29 0.77

Paper and paper products n.a n.a 0.29 1.37 0.43 0.24

Printing and reproduction of recorded media n.a n.a 0.31 n.a n.a 1.82

Coke and refined petroleum products 0.47 n.a n.a n.a n.a 0.92

Chemicals and chemical products 0.37 0.39 0.21 0.57 0.57 0.42

Pharmaceuticals, medicinal chemical and

botanical products 0.00 0.00 0.74 1.03 0.65 0.85

Rubber and plastics products 0.00 1.02 0.38 0.57 0.98 0.78

Other non-metallic mineral products 0.72 0.59 0.26 0.71 0.78 0.72

Basic metals 0.00 0.00 0.35 0.76 0.42 0.47

Fabricated metal products, except n.a n.a

machinery and equipment 1.39 0.99 0.48 0.37

Computer, electronic and optical products n.a n.a 0.44 n.a 0.30 0.47

Electrical equipment n.a n.a 0.30 0.93 n.a 0.31

Machinery and equipment n.e.c. n.a n.a 0.83 n.a n.a 1.23

Motor vehicles, trailers and semi-trailers n.a n.a 0.57 n.a 1.01 0.22

Other transport equipment 0.02 n.a 0.25 n.a n.a 0.40

Furniture 1.87 n.a 0.42 n.a n.a 0.62

Other manufacturing n.a n.a n.a n.a 0.60 0.87

Repair and installation of machinery and n.a n.a n.a

equipment n.a n.a n.a

Total Manufacturing 0.41 0.53 0.30 0.68 0.64 0.52

Source: Same as Table 1

Note: Same as Table 3.
22 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 7: Wage Ratio (Female to Male) of social groups by usual status in

manufacturing industry. (Contd.)
(b) Urban
Divisions of Manufacturing Industry 1993-94 2011-12
SC ST Others SC ST Others

Food Product 0.54 0.56 0.26 0.38 0.58 0.37

Beverages n.a 0.13 0.03 n.a 0.22 0.38

Tobacco 1.30 0.57 0.37 n.a 1.98 0.55

Textiles 0.24 0.45 0.26 0.58 0.82 0.55

Wearing Apparel 0.09 0.19 0.33 0.92 0.71 0.56

Leather and related products n.a 0.58 0.42 n.a 0.55 1.35

Wood and products of wood and cork, n.a

except furniture 0.12 0.22 0.40 0.62 0.84

Paper and paper products n.a 0.04 0.19 n.a 0.85 0.63

Printing and reproduction of recorded

media 6.74 0.37 1.37 0.88 0.28 0.54

Coke and refined petroleum products 1.54 n.a 0.61 0.06 0.61 0.09

Chemicals and chemical products 0.11 0.37 0.31 n.a 0.97 0.79

Pharmaceuticals, medicinal chemical and n.a

botanical products n.a 0.45 0.83 0.37 0.21

Rubber and plastics products 0.54 0.43 0.58 0.23 0.35 0.35

Other non-metallic mineral products 0.20 0.42 0.30 n.a 4.84 0.45

Basic metals 0.74 0.20 1.32 0.37 0.22 0.39

Fabricated metal products, except

machinery and equipment 2.23 n.a 0.47 n.a 0.72 1.09

Computer, electronic and optical products 0.91 0.29 0.67 n.a 2.87 0.41

Electrical equipment n.a 0.80 0.64 n.a 0.63 0.71

Machinery and equipment n.e.c. 0.72 n.a 0.61 n.a 0.94 2.50

Motor vehicles, trailers and semi-trailers n.a n.a 0.39 n.a n.a 0.51

Other transport equipment n.a n.a 0.71 n.a n.a 0.28

Furniture n.a n.a 0.04 0.69 0.66 1.06

Other manufacturing n.a 0.18 0.29 n.a 0.87 1.94

Repair and installation of machinery and n.a n.a n.a n.a

equipment 1.04 n.a

Total Manufacturing 0.38 0.34 0.30 0.41 0.65 0.47

Source: Same as Table 1

Note: Same as Table 3.
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 22 1

Table 8: Industrial gender segregation in India 1993-94 and 2011-12

Regular salaried Casual wage
employees labourers

Rural Urban Rural Urban

Year
(ID) (ID) (ID) (ID)
1993-94 0.43 0.29 0.41 0.38

2011-12 0.38 0.27 0.45 0.40

So urce : Sa me a s Ta ble 1
Table 9: Gender segregation in Manufacturing Industry across major Indian states
from 1993-94 to 2011-12
1993-94 2011-12
Major States
Rural Urban Rural Urban

Andhra Pradesh 0.47 0.47 0.56 0.54

Assam 0.76 0.75 0.67 0.60

Bihar 0.25 0.51 0.43 0.65

Gujarat 0.63 0.50 0.49 0.55

Haryana 0.60 0.65 0.51 0.40

Jammu & Kashmir 0.72 0.58 0.53 0.54

Karnataka 0.49 0.50 0.64 0.58

Kerala 0.39 0.48 0.69 0.52

Madhya Pradesh 0.34 0.50 0.32 0.45

Maharashtra 0.47 0.34 0.52 0.52

Orissa 0.27 0.50 0.52 0.57

Punjab 0.62 0.46 0.71 0.46

Rajasthan 0.32 0.43 1.00 0.37

Tamilnadu 0.26 0.37 0.39 0.32

Tripura 0.28 0.38 0.46 0.68

Uttar Pradesh 0.30 0.40 0.47 0.46

West Bengal 0.29 0.37 0.62 0.36

India 0.42 0.33 0.42 0.33

Source: Same as Table 1

Table 10: Impact of Explanatory Variables on Female Work Participation: Results of

Probit Estimation
Variables Coefficients z-statistic P>z
_cons 0.25 58.57 0
hhsz -0.08 -132.7 0
DFemale -0.16 7.1 0
DYear 0.00 0.67 0.501
DYear2011_female -0.31 -8.43 0
DFemale_Hindu 0.11 10.13 0
22 2 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 10: Impact of Explanatory Variables on Female Work Participation: Results of

Probit Estimation. (Contd)
Variables Coefficients z-statistic P>z
DFemale_Muslim -0.11 -7.06 0
DFemale_ST 0.23 23.51 0
DFemale_SC 0.38 45.13 0
DFemale_belowprimary -0.27 -23.75 0
DFemale_uptomiddle -0.53 -59.69 0
DFemale_secandhs -0.31 -29.87 0
DFemale_techedu -0.82 -39.68 0
DRural -0.26 -82.67 0
DFemale_Hindu_11 -0.15 -7.95 0
DFemale_Muslim_11 -0.24 -8.96 0
DFemale_ST_11 -0.01 -0.69 0.493
DFemale_SC_11 -0.14 -9.49 0
DFemale_belowprimary_11 0.10 4.96 0
DFemale_uptomiddle_11 0.13 8.67 0
DFemale_secandhs_11 -0.22 -13.07 0
DFemale_techedu_11 0.24 7.09 0
Inverse Mill’s Ratio
-0.08 -58.23 0
rho -0.01
sigma 0.92
Source: Same as Table 1

Table 11: Estimated Coefficients for Men and Women Earnings Function
Explanatory variables Coefficients z-statistic P>z
_cons 9.56 386.93 0
age 0.06 54.11 0
age2 0.00 -36.52 0
DFemale -0.31 -10.97 0
DRural -0.52 -107.13 0
DYear -2.67 -565.54 0
DYear2011_female 0.46 10.14 0
Dlit_pri -0.25 -33.85 0
Dpri_middle -0.10 -18.29 0
Dsec_hs 0.26 45.02 0
DFemale_belowprimary 0.23 11.94 0
DFemale_uptomiddle 0.15 8.84 0
DFemale_secandhs 0.51 29.54 0
DFemale_gradandabove 0.00
DFemale_belowprimary_11 -0.35 -10.65 0
DFemale_uptomiddle_11 -0.35 -14.52 0
DFemale_secandhs_11 -0.52 -18.96 0
DTechedu -0.68 -74.68 0
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 22 3

Table 11: Estimated Coefficients for Men and Women Earnings Function (Contd.)
Explanatory variables Coefficients z-statistic P>z
DFemale_techedu -0.10 -3.93 0
DFemale_techedu_11 -0.20 -5.02 0
DFemale_Hindu -0.26 -16.5 0
DFemale_Muslim -0.41 -16.99 0
DFemale_ST -0.01 -0.92 0.355
DFemale_SC -0.19 -15.6 0
DFemale_Hindu_11 -0.07 -2.36 0.018
DFemale_Muslim_11 0.00 -0.09 0.926
DFemale_ST_11 0.00 0.11 0.909
DFemale_SC_11 -0.06 -2.7 0.007
Source: Same as Table 1

Table 12: Blinder and Oaxaca decomposition method

Components of Decomposition In %
1993-94 2011-12
Amount attributable: 161.1 74.6

- due to endowments (E): 38.4 13.9

- due to coefficients (C): 122.7 60.8

Shift coefficient (U): -71.7 -37.2

Raw differential (R) {E+C+U}: 89.4 37.4

Adjusted differential (D) {C+U}: 51 23.5

Endowments as % total (E/R): 42.9 37.1

Discrimination as % total (D/R): 57.1 62.9

Source: Same as Table 1

Note: a) Positive number indicates advantage to male group; negative number indicates advantage to
female group. b) The results from decomposition are presented using Blinder’s (1973) original formulation
of E, C, U, and D. The endowments (E) component of the decomposition is the sum of (the coefficient
vector of the regressors of the male-wage group)times (the difference in group means between the male-
wage and female-wage groups for the vector of regressors). The coefficients (C) component of the
decomposition is the sum of the(group means of the female-wage group for the vector of regressors)
times (the difference between the regression coefficients of the male-wage group and the female-wage
group). The unexplained portion of the differential (U) is the difference in constants between the male
wage group and the female-wage group. The portion of the differential due to discrimination is C+U. The
raw (or total) differential is E+C+U.
22 4 The Journal of Industrial Statistics, Vol. 5, No. 2

Appendix
A1: Explanations of the variables used in the regression method:
Female Gender dummy variable, equal to 1 for women and 0 for men

Y11 Time dummy variable, equal to 1 if the person comes from the 68th
round (2011 -12) and 0 if he/she comes from the 50th round (1993-94).

Y11_female Gender dummy interacted with time dummy, indicating the change of
the effect of gender dummy variable over the years

female_hindu Gender dummy variable for Hindus, equal to 1 for Hindu women and
0 otherwise

female_muslim Gender dummy variable for Muslims, equal to 1 for Muslim women
and 0 otherwise

female_st Gender dummy variable for STs, equal to 1 for ST women and 0
otherwise

female_sc Gender dummy variable for SCs, equal to 1 for SC women and 0
otherwise

female_Literatebelowprimary Gender dummy variable for literate below primary level of education,
equal to 1 for women with literate below primary level of education
and 0 otherwise
female_primary_ middle Gender dummy variable for primary level of education, equal to 1 for
women with primary level of education and 0 otherwise

female_Sec_HS Gender dummy variable for secondary and higher secondary level of
education, equal to 1 for women with secondary and higher secondary
level of education and 0 otherwise

female_graduate_ above Gender dummy variable for graduation and above level of education,
equal to 1 for women with graduation and above level of education
and 0 otherwise

female_tech Gender dummy variable for technical skill, equal to 1 for women with
technical skill and 0 otherwise.

female_hindu_11 Hindu gender dummy interacted with time dummy, indicating the
change of the effect of Hindu women over the years

female_muslim_11 Muslim gender dummy interacted with time dummy, indicating the
change of the effect of Muslim women over the years

female_st_11 ST gender dummy interacted with time dummy, indicating the change
of the effect of ST women over the years

female_sc_11 SC gender dummy interacted with time dummy, indicating the change
of the effect of SC women over the years
 The Price of Prejudice: Employment Trend and Wage Discrimination ... 22 5

Appendix
A1: Explanations of the variables used in the regression method: (Contd.)
female_Literatebelowprimary_11 Gender dummy with Literate below primary level of education
interacted with time dummy, indicating the change of the effect of
women with primary level of over the years.

female_primary_ middle_11 Gender dummy with primary and middle level of education interacted
with time dummy, indicating the change of the effect of women with
above primary level of over the years.

female_Sec_HS_11 Gender dummy with secondary and higher secondary level of

education interacted with time dummy, indicating the change of the
effect of women with secondary and higher secondary level of
education.

female_graduate_ above_11 Gender dummy with graduation and above level of education
interacted with time dummy, indicating the change of the effect of
women with university and higher degree,

female_tech_11 Gender dummy with technical skill interacted with time dummy,
indicating the change of the effect of women with technical skill over
the years.
22 6 The Journal of Industrial Statistics (2016), 5 (2), 226 - 239

Export Competitiveness and Intensity of Technology in Indian

Manufacturing Industries –Analysis with ASI Unit Level Data
Panchanan Das1, University of Calcutta, Kolkata, India
Abhishek Halder, University of Calcutta, Kolkata, India
Rahul Dutt, University of Calcutta, Kolkata, India

Abstract

This study looks into export competitiveness of the manufacturing firms in terms of firm
specific characteristics. The pooled unit level data compiled from the Annual Survey of
Industries (ASI) during the period 2008 -2012 are used for this purpose. A generalised
linear model with a binomial distribution and a logit link function is used in estimating
the regression relationship. In the pooled regression model, firm’s age has negative effect
on export competitiveness of a firm. Firm size has a significant positive effect on firm’s
export competitiveness. Larger the size of a firm, larger will be the export share to total
output. As expected, human capital favourably affects the export competitiveness of
manufacturing firms. The effect of human capital is, however, less than the effect of firm
size on export share of firm’s production. The capital intensity of a firm, on the other hand,
has no significant effect on its export competitiveness. The export performance of the firms
manufacturing apparels, leather and food products is better than that of the firms in other
manufacturing industries.

1. Introduction

1.1 The manufacturing sector has traditionally been treated as an engine of growth of
any economy (Kaldor, 1966). It enhances growth and productivity, generates employment
and it also strengthens agriculture and service sectors through forward and backward
linkages. The faster growth in India, as appeared since the late 1980s, however, is not
attributed significantly to the industrial sector growth. In the recent past, output from the
manufacturing sector grew at a higher rate and the annual growth rate reached a peak at 12
percent in 2006-07. Despite the high growth, the share of manufacturing sector in GDP has
remained at around 15 percent even in 2013-14. The manufacturing sector has not contributed
perceptibly to tackle the problem of unemployment and underemployment in the
unorganised sector (GOI, 2011). The exports from manufacturing sector, however, has
grown at a compound annual rate of more than 16 percent during 2007-08 to 2013-14
contributing over 60 percent to total merchandise exports2. The engineering goods have
emerged as the most contributors sharing roughly 40 percent of total manufacturing exports
followed by gems and jewellery (22 percent), and textiles (14 percent) in 2010-11.

1.2 Significant trade liberalisation together with the successive devaluation of domestic
currency were expected to improve the export competitiveness of Indian firms and lead to
increased contribution of exports to the Indian economy. The significance of exports from
the manufacturing sector has been increasing steadily in the Indian economy during the
period of trade liberalisation. The propagation in the use of nontariff measures mainly by
1
e-mail: [email protected]
2
Total merchandise trade in India was recorded at 37 percent of the GDP in 2010-11.
 Export Competitiveness and Intensity of Technology in Indian Manufacturing .... 22 7

the developed countries, in many cases, acts as one of the major constraints in export
competitiveness in a country like India. Sanitary and phytosanitary standards and technical
barriers to trade, for example, creates critical questions in increasing export performance by
the manufacturing firms in India (Kallummal 2006).

1.3 Against this observed facts, this study looks into export competitiveness of the
manufacturing firms in terms of firm specific characteristics with unit level data from Annual
Survey of Industries (ASI) during the period 2008 -2012. In analysing export competitiveness
of manufacturing firms in India, the rest of the study is organised in following manner.
Section 2 focuses on the determinants of export competitiveness as proposed in this study.
Section 3 describes the data, in short, and construction of variables for this study. Section
4 deals with econometric methodology applied in empirical exercise. Characteristic features
of the sample firms are summarised in section 5. Section 6 interprets the empirical findings.
Section 7 summarises and concludes.

2. Determinants of export competitiveness

2.1 The export competitiveness and the pattern of trade of a country are analysed
conventionally in terms of comparative cost advantage in a macroeconomic framework.
The comparative cost advantage of exporting a commodity arises from factor productivity
differentials and difference in factor endowments between two countries. The notion of
comparative advantage is extended further in the Heckscher-Ohlin model under the
assumption of same production technology, but with different factor endowments3. Later
on, human capital is treated as an additional important factor affecting trade in the neo
classical model. Scale economies and oligopolistic competition have become important
factors for determining trade patterns in the strategic trade models. Technology has also
become a crucial determinant of international trade.

2.2 Traditionally, export competitiveness of a country is measured by its share of

world markets for its products (Michael et al. 2008). It is the firm that decides ultimately if it
should trade or not. In this study we measure export competitiveness of a firm as its share
of export in total sales. While export competitiveness of a country is related to its comparative
advantage in producing and selling its products in international markets, trade
competitiveness of a firm is defined as the ability of the firm to export more in terms of its
total sales. The capability of a firm to perform in the world market depends largely on its
technological characteristics along with other micro and macroeconomic factors.
Technological progress improves technical efficiency that lowers comparative costs of the
exportable. Thus technologically advanced firms are more competitive than the other firms
(Posner 1961, Vernon 1966).

2.3 In this study, we have taken firm size, production age of the firm, human capital
employed, capital intensity and value added along with organisational type and

3
The Heckscher–Ohlin model is a general equilibrium mathematical model of international trade, developed
by Eli Heckscher and Bertil Ohlin. One of the theorems of this model states that a country will export
goods that use its abundant factors intensively, and import goods that use its scarce factors intensively.
Bertil Ohlin published the book which first explained the theory in 1933. Heckscher was credited as co-
developer of the model, because of his earlier work on the problem, and because many of the ideas in the
final model came from Ohlin’s doctoral thesis, supervised by Heckscher.
22 8 The Journal of Industrial Statistics, Vol. 5, No. 2

manufacturing type of a firm as the major determinant of export competitiveness of that

firm. Larger firms are generally regarded as more capable of coping with the large investments
and high risks associated with exporting (Aitken, Hanson and Harrison, 1997). The positive
relation between exports and firm size is expected because of efficiency advantage of larger
firms due to economies of scale. In some studies, however, firm size no longer plays a
significant role beyond a certain threshold level (Wagner, 2001). Firm’s production age is
also significant in analysing exports performance of a firm. Older firm may enjoy better
leverage to compete the world market. But, in some cases, younger firms may be more
flexible, aggressive and proactive when catering to world demand (Lefebvre and Lefebvre,
2001). Human capital, measured in this study by the share of engineers and managerial
staffs to total employment, of a firm is expected to have positive effect on export
competitiveness. Capital intensity is often treated as a determinant of a firm’s export
performance. A high capital-intensive firm may have high propensity to export because of
its past innovations. Value added is treated normally as an indicator of productive health of
a firm. More productive firm generates more value added. Thus, a firm with higher value
added is expected to be more export competitive.

3. Data and construction of variables

3.1 In this study we have used unit level data from the ASI, the main source of
information about registered industries, published each year by the Central Statistical
Office (CSO), Government of India. The CSO has published a separate set of unit level
information of ASI compatible for preparing unit level panel and the data are available from
1998 to 2012 for each year. However, one has to face a serious problem with this data set in
preparing a balanced panel. The major problem relates to the identification code as recorded
in the data set. Some identification code are not matching meaningfully for every rear. Thus
the same factory unit identified by the code does not appear in many years. For this reason
we have pooled the data by treating individual factories as cross section units. In this
study we have pooled the data for four years from 2008 to 2011 because export information
at the factory level are available since 2008. We have kept those factory units which are
currently operating (status of unit =1, as per the schedule). The number of currently operating
factory units in the pooled data set are 36020, 39109, 41211 and 41706 for 2008, 2009, 2010
and 2011 respectively. Thus, total number of sample observation in our data set is 158047.

3.2 The ex-factory value of output at constant prices is used as a measure of real
output. ASI reports output data in value terms (Rs. Lakh). The nominal values of output are
deflated by the wholesale price indices at 2004-05 base period for manufactured goods. We
have used two distinct types of labour inputs: manufacturing workers, and managerial and
technical employees. Unlike other factors of production, capital is used beyond a single
accounting period and measuring capital stock is rather problematic. Figures of fixed capital
shown in the ASI schedule include the values of plant and machinery along with other
types of assets used in production, transportation, living or recreational facilities, hospitals,
schools, etc., and are measured in terms of historical prices based on the book value of
fixed assets. We have used the net closing value of plant and machinery at the end of the
year as capital input. We define relative size of a firm as the total employment in the firm
relative to the total employment in the sector in which the firm lies. Thus, relative size of firm
i in textile industry, for example, is total employment in firm i divided by total employment in
the textile sector. Production age of a firm is calculated by taking the difference between the
 Export Competitiveness and Intensity of Technology in Indian Manufacturing .... 22 9

year of survey and year of initial production. The variable human capital is constructed by
taking the ratio of managerial and technical employees to total number of employees.
Capital intensity is defined here as the ratio of values of plant and machinery to total labour.
Value added is calculated at the unit level by taking the difference between total sales value
and total input costs. Export share of total sales value, a measure of export competitiveness
of a firm, is used as a dependent variable in our estimated model.

4. Econometric methodology

4.1 In exploring the relationship between export competitiveness and the proposed
determinants, we note that the dependent variable is bounded between zero and one. Thus
often used ordinary least square (OLS) method is not appropriate for assessing this
relationship. The relationship to be estimated in this study is specified in the following
form:

yit  0  1x1it  2 x2it  3 x3it  4 x4it  5 x5it

k l m n
  j D1 j  h D2h   p D3 p  q D4q   it (1)
j 1 h1 p 1 q1

Here, yit = export share to total output of firm i in time t as a proxy for export competitiveness

x1it = relative firm size defined as the number of employees of firm i in time t divided by total
number of employee in the sector where firm i belongs.

x2it = human capital in terms of share of engineers and managerial staff to total employee of
firm i in time t

x3it = capital intensity measured by capital to labour ratio of firm i in time t

x4it = value added, the difference between total sales value and total value of inputs used,
of firm i in time t

x5it = age of firm i in time t, the difference between the survey year and year of initial
production

D1j = Dummy variable related to year j

D2h = Dummy variable related to organisation type h ,

D3p = Dummy variable related to ownership type p

D4q = Dummy variable related to manufacturing group q

it is the random error
As the variable yit, 0  y it  1 , is to be explained by a set of explanatory variables, the
population regression model based on normality assumption as given in (2) rarely
provides the best description of E(y|x, D).
23 0 The Journal of Industrial Statistics, Vol. 5, No. 2

E y | x , D   0   1 x1   2 x 2   3 x 3   4 x 4   5 x 5
k l m n
 j 1
 j D1j  
h 1
 h D 2h  
p 1
 pD3p  
q 1
 q D 4q (2)

4.2 The primary reason is that y is bounded between 0 and 1, and so the effect of any
particular xj cannot be constant throughout the range of x (unless the range of xj is very
limited). The most common alternative to equation (2) has been to model the log-odds ratio
as a linear function. If y is strictly between zero and one then a linear model for the log-odds
ratio is

E log( y / 1  y ) | x, D    0  1 x1   2 x 2   3 x3   4 x 4   5 x5
k l m n
   j D1 j   h D2 h   p D3 p    q D4 q (3)
j 1 h 1 p 1 q 1

4.3 In equation (3), the log-odds ratio can take on any real value as y varies between
0 and 1, so it is natural to model its population regression as a linear function. Nevertheless,
there is a potential problem with equation (3). The expected value of log odds as shown in
equation (3) cannot be true if y takes on the values 0 or 1 with positive probability.
Consequently, given a set of data, if any observation, yi, equals 0 or 1 then an adjustment
must be made before computing the log-odds ratio. It is possible to estimate E(y|x, D) by
assuming beta distribution for y given x and D, and estimating the parameters of the
conditional distribution by maximum likelihood. One important limitation of the beta
distribution is that it implies that each value in [0,1 ] is taken on with probability zero. Thus,
the beta distribution is difficult to justify in applications where at least some portion of the
sample is at the extreme values of zero or one.
4.4 To overcome this problem, Papke and Wooldridge (1996) in their seminal paper
proposed a fractional response model that extends the generalised linear model with a
binomial distribution and a logit link function4 which may be appropriate in fractional
dependent variable. Papke and Wooldridge (2008) developed fractional response models
for panel data and use quasi-maximum likelihood estimator (QLME) to obtain a robust
method for estimating fractional response models without an ad hoc transformation of the
boundary values. Hausman and Leonard (1997) applied fractional logit to panel data on
television ratings of National Basketball Association games to estimate the effects of
superstars on telecast ratings. In using pooled QMLE with panel data, the only extra
complication is in ensuring that the standard errors are robust to arbitrary serial correlation
(in addition to misspecification of the conditional variance). Wagner (2003) analyses a large
panel data set of firms to explain the export-sales ratio as a function of firm size by
incorporating firm-specific intercepts in the fractional logit model. While including dummies
for each cross section observation allows unobserved heterogeneity to enter in a flexible
way, it suffers from an incidental parameters problem under random sampling when T (the
number of time periods) is small and N (the number of cross sectional observations) is
large.

4
The generalized linear model is a flexible generalization of ordinary linear regression that allows for
response variables that have error distribution models other than a normal distribution.It generalizes
linear regression by allowing the linear model to be related to the response variable via a link function and
by allowing the magnitude of the variance of each measurement to be a function of its predicted value.
 Export Competitiveness and Intensity of Technology in Indian Manufacturing .... 23 1

4.5 In this study we have estimated equation (1) by following Papke and Wooldridge
(2008) with pooled data constructed from four independent samples for four different time
periods (2008, 2009, 2010 and 2011) from the same population. The conditional expectation
of the frac-tional response variable is specified as

E  y | x, D   G[  0   1 x1   2 x 2   3 x3   4 x 4   5 x5
k l m n
   j D1 j    h D2h   p D3 p    q D4q ] (4)
j 1 h 1 p 1 q 1

In the non-linear fractional response regression, G(.) is specified as a logistic function. The
main advantage of the fractional response model is its ability to capture non-linearity.
Typically, G(.) is a distribution function similar to the logistic function -

exp z 
G z   (5)
1  exp z 

The condi-tional likelihood for observation i is

L   y i logG .  1  y i  log1  G . (6)

The estimated values of the coefficients as shown in equation (1) are obtained by
maximising equation (6).

5. Characteristics of sample firms by industry groups

5.1 Before analysing the export competitiveness of the firms in the registered
manufacturing sector we need to look at their structural characteristics. Table 1 presents
the summary view of the major factors likely to explain export competitiveness of the firms
in 2011-12, the terminal year in our sample. The mean export share of the firms in wearing
apparel industry was the highest followed by the manufacture of leather products. The
other manufacturing industries exhibiting notable export shares included manufacture of
pharmaceuticals, manufacture of computer, manufacture of textiles and manufacture of
fabricated metal. Export share was the lowest in beverages, tobacco products and the paper
industry in 2011-12. Thus, the export performance in terms of export shares varied widely
across the industry groups. The textiles and wearing apparel industries are traditionally
export led industry in India. In 2006-07 textiles and wearing apparel industries together
contributed 15 percent to total merchandise exports in India, and the share increased by
more than doubled in 2011-12 (Economic Survey, GOI, 2013). ASI data for 2011-12 also
indicate that the exports of manufacturing products are led primarily by the private limited,
public limited and partnership companies. Moreover, the manufacturing industries run
wholly by private ownership exhibited the highest share of exports to their total sales
value.
5.2 The mean production age of the sample firms varied widely across the different
manufacturing industry groups. As shown in Table 1, the mean production age in years
was the highest in tobacco industry exporting only 1 percent of their total sales in 2011-12.
23 2 The Journal of Industrial Statistics, Vol. 5, No. 2

The average production age of textiles and apparel industries, the leading manufacturing
exporters in India, was over 50 years. It is believed that older the manufacturing unit higher
will be its export share because of higher experience in production and innovation. However,
the relatively young industries in India, for example leather industry, export higher share of
their products than tobacco industry. Leather industry exported more than one fourth of
their products in 2011-12 (Table 1).
5.3 We have calculated relative size of a firm by taking the ratio of firm’s employment
to total employment in the sector in which the lies. The average firm size was the highest in
the repair and installation of machinery, followed by manufactures of furniture and petroleum
products. Firm size is one of the important factors affecting export performance of the firm.
In India, most of the manufacturing firms exporting larger share of their products are small
or moderate in size. Table 1 shows that average size of a firm in textiles, wearing apparels
and leathers was low as compared to many other manufacturing sectors which include
repairing and petroleum. Human capital, measured by the ratio of mangers and technical
persons to total employee, also varied widely across different manufacturing sectors. It
was the highest in manufacture of machinery and pharmaceuticals, and the lowest in tobacco
products.
5.4 Export intensive industries are normally viewed as technology intensive. Capital
intensity was different in different manufacturing sectors (Table 1). Manufacture of
computer, non-metallic products, basic metals, paper, petroleum products, pharmaceuticals
are high capital intensive sectors, while tobacco products, apparel were low capital intensive
sectors in registered manufacturing in India. Low or moderate capital intensive sectors
exported relatively larger share of their products as compared to high capital intensive
sectors. This kind of findings is consistent with the relative factor abundance in a country
like India. Labour abundant country enjoys comparative advantage in labour intensive
sectors. For this reason industries like wearing apparel and leather are exporting more as
their products are less capital intensive.
6. Empirical results
6.1 We have estimated equation (1) in a framework of the fractional response model as
described above. In this model we have considered export share to total output as the
dependent variable and firm’s age, firm size, human capital, capital intensity and some
dummies as explanatory variables. We have incorporated year dummy yr_11 and yr_12 for
years 2011 and 2012 to look into the time effect compared to the base year 2008 on export
competitiveness of manufacturing firms. Export competitiveness may vary across the
organisation type of the firms. In the ASI data, firm’s organisation is of 11 types. We have
regrouped the organisation type to divide them private and public sector organisations.
Government organisation is treated as a reference group. We include proprietorship,
partnership and limited companies as shown in the ASI schedule in the private sector. We
have used here three dummies (D_proprietary, D_partnership and D_limited) to capture the
differences in export performance among firms under private organisation as compared to
firms under government organisation. There are 24 manufacturing groups at the two digit
level in our data set. We have taken 23 sector dummies corresponding to different
manufacturing groups to avoid dummy variable trap by treating repairing and installation
as a reference industry group.
 Export Competitiveness and Intensity of Technology in Indian Manufacturing .... 23 3

6.2 The estimated values of the coefficients along with the corresponding odds ratios
of equation (1) are displayed in Table 2. From our regression analysis we can see that firm’s
production age has negative effect on export competitiveness of a firm. Thus, younger
firms export larger percentage of their total products as compared to the older firms in the
foreign market. This result supports the observed facts as shown in Table 1. Firm’s size has
a significant positive effect on export competitiveness. Larger the size of a firm larger is its
export share to total output. Large firms enjoy scale effect and thus they have more
comparative cost advantage than the smaller firm. This result supports the findings of
Wagner (2003). Human capital of a firm also has a positive effect on its export share as
expected. But, in Indian manufacturing industries, scale effect is higher than the effect of
human capital. Capital intensity, on the other hand, of the firm has no significant effect on
export competitiveness of the firm. India is a labour abundant country and the manufacturing
firms of the country exports largely labour intensive commodities by following the rule of
comparative advantage.
6.3 The year dummies included in the estimated equation capture the time effect on
export competitiveness. The sign of the coefficients for year dummies (yr_11, yr_12) are
significantly negative implying that the mean export share of the manufacturing firms, other
effects remain zero, declined in 2011 and further to 2012 from its value in 2008. Thus, export
competitiveness of the firms in registered manufacturing sector declined in the year of 2011
and 2012 as compared to 2008. We also observe that most of the coefficients corresponding
to sector dummies are statistically significant. Export performance of the firms in manufacture
of apparel, leather and food products performed better than the firms in other sectors in
manufacturing industries. However, the export performance of beverages, tobacco products
and paper was less than even the repairing sector. Textile, wearing apparel and other
manufacturing industries consisting jewellery, sports goods, medical instruments industry
performed better than the firms in repair and installation of machinery industry.
6.4 The estimated values of the odds ratio also provide the similar inferences on
export competitiveness. Odds ratios are greater than 1 for those industries whose coefficient
values are positive and industries having negative coefficient value, have less than 1 odds
ratio as well that seems probability of decreasing export share is higher over probability of
increasing exporting share.
6.5 The export competitiveness of the manufacturing industries is not affected similarly.
To find out the differential effects we have estimated the effects of firm specific factors as
mentioned above on export competitiveness for the leading exporting industries by using
the unit level information only for 2011-12. The selected major exporting industries include
manufacturing of food products, manufacturing of textiles, manufacturing of wearing apparel,
manufacturing of leather and related products, manufacturing of chemical and chemical
products, manufacturing of pharmaceuticals, manufacturing of fabricated products and
manufacturing of computer, electronic and optical products industry. The estimated results
are shown in Table 1. We observe that the effects of firms’ age, firm size, human capital and
capital intensity are not similar for all industries.
6.6 Firms’ production age has negative effect on export share in food products and
textile industry, but positive effect in pharmaceuticals. Thus, in the case of both food
products and textile industries the export share of older firms is less than the young firms.
23 4 The Journal of Industrial Statistics, Vol. 5, No. 2

But, for pharmaceutical industry the older firms export more than the new firms. However,
there is no significant effect of firms’ production age on export shares of the firms in the
manufacturing of leather, chemicals, metal products and computer. Size of the firm has
positive and significant impact on export share for all the industries as shown in Table 3.
Larger the size, larger is the share of export to firm’s output, but at different rates across the
industry groups. The size effect is the highest in food products industry followed by
textiles and metal products industries. The effect of human capital on firm’s export is
statistically significant in most of the industries. While the effect is favourable to firm’s
export share in textile, apparel and chemical industries, it is negative for leather and
pharmaceutical industries. Human capital has no significant role in export performance for
food products industry. Capital intensity has negative significant role in export share at the
firm level in food products and textile industry, but it has a positive role in pharmaceutical
industry. The firms under individual proprietorship, partnership and limited companies
exported more than the firms in the public sector for the major exporting manufacturing
groups. The share of export has increased more for individual proprietorship, followed by
partnership and limited companies for food products industry, textiles industry and leather
and related product industry.
7. Summary and conclusions
7.1 This study investigates how export competitiveness of the manufacturing firms is
affected by the firm specific characteristics with unit level ASI data during the period 2008
-2012. We have taken firm size, production age of the firm, human capital employed, capital
intensity and value added along with organisational type and manufacturing type of a firm
as the major determinant of export competitiveness of that firm. Export competitiveness is
measured by the fraction of total output exported by a firm. The generalised linear model
with a binomial distribution and a logit link function is used in estimating the regression
relationship with pooled data constructed from four independent samples for four different
time periods (2008, 2009, 2010 and 2011) from the same population.
7.2 The average export share of a firm in manufacture of wearing apparel was the
highest followed by the manufacture of leather products. The other sectors within
manufacturing showing notable export shares at the firm level included manufacture of
pharmaceuticals, manufacture of computer, manufacture of textiles and manufacture of
fabricated metal. Low or moderate capital intensive sectors exported relatively larger share
of their products as compared to high capital intensive sectors. This kind of findings are
consistent with the relative factor abundance in a country like India. Labour abundant
country enjoys comparative advantage in labour intensive sectors.
7.3 In the pooled regression model, firm’s production age has negative effect on
export competitiveness of a firm. Thus younger firms performed better than the older firms
in the foreign market. Firm size has a significant positive effect on firm’s export
competitiveness. Larger the size of a firm larger will be the export share to total output.
Larger firms enjoy scale effect and thus they have more comparative cost advantage than
the smaller firm. Human capital affected favourably the export competitiveness of
manufacturing firms as expected. But, the effect of human capital was less than the effect of
firm size on export share of firm’s production. Capital intensity, on the other hand, of the
firm has no significant effect on export competitiveness of the firm. Export performance of
 Export Competitiveness and Intensity of Technology in Indian Manufacturing .... 23 5

the firms in manufacture of apparel, leather and food products performed better than the
firms in other sectors in manufacturing industries.

7.4 This paper also analyses the effects of firm specific factors on the export performance
of firms in selected Indian manufacturing industries. Younger firms in food products and
textiles performed, while older firms in pharmaceuticals exported greater proportion of their
products. The size effect was the highest in food products followed by textiles and metal
products. The firms under individual proprietorship, partnership and limited companies
exported more than the firms in the public sector. The findings of this study are based on
output share to export by the manufacturing firms at the two digit level. The particular
products that dominate in merchandise trade need to be studied further in depth at the
disaggregated level to understand some other relevant issues relating to export
competitiveness in the context of trade liberalisation in India.

References

Aitken, B., G. Hanson and A. Harrison (1997), Spillovers, foreign investment, and export
behaviour, Journal of International Economics, 43(1-2): 103-132.

Hausman, J.A., Leonard, G.K., (1997). Superstars in the national basketball association:
Economic value and policy, Journal of Labor Economics, 15: 586–624.

Kaldor, N. (1966). Causes of the Slow Rate of Growth in the UK, Cambridge University
Press, London reprinted in Targetti and Thirlwall (1989)

Kallummal, Murali (2006), ٝ NonAgricultural Market Access Negotiations: Real Concernsٝ ,

Social Scientist, 34 (910), SeptemberOctober 2006.

Lefebvre, E. and L. Lefebrve (2001), “Innovative Capabilities as Determinants of Export

Performance and Behaviour: A longitudinal Study of Manufacturing SMEs” in Innovation
and Firm Performance: Econometric Explorations of Survey Data, Palgrave (MacMillan
Press) London at Basingstoke.

Michael E. Porter, Christian Ketels and Mercedes Delgado (2008), The Microeconomic
Foundations of Prosperity: Findings from the Business Competitiveness Index, The Global
Competitiveness Report 2007- 2008,

http://siteresources.worldbank.org/EXTEXPCOMNET/Resources/24635931213989126859/
03_Porter_GCI_ch1&2.pdf.

Papke, L. E. and J. Wooldridge (1996), Econometric Methods for Fractional Response Variable
with an Application to 401 (K) Plan Participation Rates, Journal of Applied Econometrics,
11(6): 619-632.

Papke, L. E., and Wooldridge, J. (2008). Panel data methods for fractional response variables
with an application to test pass rates. Journal of Economet-rics, 145(1-2): 121-133.

Posner, M. V. (1961), International Trade and Technical Change, Oxford Economic Paper,
13: 323-341.
23 6 The Journal of Industrial Statistics, Vol. 5, No. 2

Vernon, R. (1966), International Investment and International Trade in the Product Cycle,
Quarterly Journal of Economics, 80: 190-207.

Wagner, J. (2001), A Note on Firm Size—Export Relationship, Small Business Economics 17:
229-237.
Wagner, J., (2003). Unobserved firm heterogeneity and the size-exports nexus: Evidence
from German panel data, Review of World Economics, 139: 161–172.
 Export Competitiveness and Intensity of Technology in Indian Manufacturing .... 23 7

Table 1 Characteristics of sample firms by industry groups: 2011-12

Fraction Average
of age of Average Human Capital
Industry group
output factory size* capital** intensity***
exported units
Manufacture of food products 0.04 59.31 0.01 0.11 14.08
Manufacture of beverages 0.01 48.03 0.04 0.11 16.02
Manufacture of tobacco products 0.01 78.66 0.04 0.08 3.53
Manufacture of textiles 0.06 53.94 0.01 0.09 10.28
Manufacture of wearing apparel 0.32 50.66 0.01 0.09 3.72
Manufacture of leather products 0.26 39.39 0.03 0.10 6.28
Manufacture of wood 0.02 49.26 0.02 0.13 7.46
Manufacture of paper 0.01 48.92 0.02 0.12 21.63
Manufacture of printing 0.02 65.36 0.03 0.14 7.89
Manufacture of petroleum products 0.02 48.40 0.08 0.15 17.02
Manufacture of chemical products 0.04 51.09 0.01 0.14 15.76
Manufacture of pharmaceuticals 0.08 60.33 0.02 0.16 13.57
Manufacture of rubber 0.03 43.63 0.01 0.13 14.54
Manufacture of non-metallic products 0.02 44.98 0.01 0.11 22.31
Manufacture of basic metals 0.02 47.93 0.01 0.12 21.94
Manufacture of fabricated metal 0.06 50.34 0.01 0.13 9.15
Manufacture of computer 0.07 54.90 0.03 0.15 36.26
Manufacture of electrical equipment 0.03 52.28 0.02 0.14 14.12
Manufacture of machinery 0.04 46.30 0.01 0.16 5.68
Manufacture of motor vehicles 0.04 45.67 0.02 0.12 15.42
Manufacture of transport equipment 0.03 51.40 0.04 0.12 6.07
Manufacture of furniture 0.04 36.24 0.09 0.13 6.94
Other manufacturing 0.21 31.67 0.03 0.12 5.12
Repair and installation of machinery 0.01 34.87 0.12 0.15 3.76

Note: *Relative size of a firm by taking the ratio of firm’s employment to total employment in the
sector in which the lies; **Human capital is constructed by taking the ratio of managerial and technical
employees to total number of employees; ***Capital intensity is defined as the ratio of values of plant
and machinery to total labour employment.
Source: Authors’ estimation with ASI unit level data for 2011-12
23 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 2 QMLE of the fractional response model for all manufacturing units
Odds Z
variables Coefficients ratio statistic P>z

Intercept -4.50 0.01 -8.12 0.00

log (age) -0.11 0.89 -8.73 0.00
log (size) 0.42 1.52 36.14 0.00
log (human capital) 0.08 1.08 4.83 0.00
log (capital intensity) 0.01 1.01 1.17 0.244
yr_11 -0.15 0.86 -4.45 0.00
yr_12 -0.08 0.93 -2.29 0.022
D_proprietary 1.49 4.42 9.34 0.00
D_partnership 1.54 4.65 9.84 0.00
D_limited_company 1.15 3.16 7.47 0.00
D_owner_joint 0.91 2.47 2.51 0.012
D_owner_private 1.43 4.17 4.45 0.00
D_food 2.08 8.02 4.74 0.00
D_beverage -0.95 0.39 -1.8 0.071
D_tobacco -0.29 0.75 -0.56 0.576
D_textile 2.47 11.86 5.64 0.00
D_apparel 4.05 57.27 9.25 0.00
D_leather 3.38 29.35 7.71 0.00
D_wood 0.41 1.51 0.91 0.364
D_paper -0.08 0.92 -0.18 0.858
D_print 0.11 1.12 0.24 0.81
D_petro 0.15 1.16 0.29 0.775
D_chemical 1.68 5.36 3.82 0.00
D_pharma 1.83 6.24 4.16 0.00
D_ruber 1.10 3.02 2.49 0.013
D_nonmetal 0.93 2.54 2.11 0.035
D_basic_metal 1.15 3.17 2.61 0.009
D_fabri_metal 2.04 7.68 4.65 0.00
D_computer 1.67 5.33 3.78 0.00
D_elec_equipment 1.07 2.93 2.42 0.015
D_machinary 1.55 4.71 3.52 0.00
D_motor 1.24 3.47 2.81 0.005
D_transport 0.63 1.88 1.38 0.168
D_furniture 0.40 1.50 0.84 0.399
D_other_manu 3.05 21.03 6.94 0.00
Source: Authors’ estimation with ASI unit level data (2008-2011)
 Table 3 QMLE of the fractional response model for selected sectors: 2011-12

Note: ** significant at 1 percent level, * significant at 5 percent level, the rest are insignificant
Source: As for Table 1
Export Competitiveness and Intensity of Technology in Indian Manufacturing ....
23 9
 SECTION II 24 1

SECTION II

Page No.

• Selected Economic Indicators of 243

Manufacturing Sector of India : Table1

• Employment by Industry Division in 246

Manufacturing Sector : Table 2

• Employment by Industry Group in 247

Manufacturing Sector : Table 3

• All India ASI Data Based on Units with 249

100 and more Employees : Table 4

• 2-digit NIC Division and Description 252

• New Initiatives in Annual Survey of Industries (ASI) 254

 Selected Economic Indicators of Manufacturing Sector of India : Table 1 24 3

Table 1: Selected Economic Indicators by 2-digit Industry Div. based on ASI 2012-13
and 2013-14
All India
NIC- Value Added per Person Capital Productivity
2008 Description Engaged
(Rs. Lakh)
Div. 2012-13 2013-14 2012-13 2013-14
01* Crop and animal production, hunting and related
service activities 4.00 8.45 0.84 1.54
08** Other mining and quarrying 3.57 4.09 0.98 1.01
10 Manufacture of food products 3.63 3.79 0.42 0.42
11 Manufacture of beverages 8.66 7.47 0.49 0.44
12 Manufacture of tobacco products 2.87 2.70 2.26 2.16
13 Manufacture of textiles 3.32 3.06 0.39 0.23
14 Manufacture of wearing apparel 1.89 2.31 0.72 1.17
15 Manufacture of leather and related products 1.99 2.45 0.78 0.88
16 Manufacture of wood and of products of wood and
cork, except furniture; manufacture of articles of straw
and plaiting materials 2.87 3.05 0.47 0.50
17 Manufacture of paper and paper products 3.19 4.42 0.18 0.23
18 Printing and reproduction of recorded media 4.81 3.62 0.58 0.41
19 Manufacture of coke and refined petroleum products 113.11 92.91 0.70 0.53
20 Manufacture of chemicals and chemical products 12.10 11.20 0.52 0.50
21 Manufacture of basic pharmaceutical products and
pharmaceutical preparations 10.45 10.54 0.80 0.78
22 Manufacture of rubber and plastics products 4.71 6.22 0.42 0.51
23 Manufacture of other non-metallic mineral products 4.83 4.06 0.21 0.27
24 Manufacture of basic metals 7.81 11.90 0.16 0.22
25 Manufacture of fabricated metal products, except
machinery and equipment 5.10 4.40 0.69 0.56
26 Manufacture of computer, electronic and optical
products 8.39 9.92 0.83 0.59
27 Manufacture of electrical equipment 7.20 6.91 0.76 0.77
28 Manufacture of machinery and equipment n.e.c. 8.96 7.52 0.97 0.75
29 Manufacture of motor vehicles, trailers and semi-
trailers 7.46 5.70 0.46 0.31
30 Manufacture of other transport equipment 7.54 7.94 0.50 0.63
31 Manufacture of furniture 4.04 4.00 0.42 0.81
32 Other manufacturing 4.86 4.95 1.24 1.14
33 Repair and installation of machinery and equipment 5.76 6.43 0.57 0.60
38 Waste collection, treatment and disposal activities;
materials recovery 3.13 3.23 0.22 0.23
58 Publishing activities 11.60 11.24 0.74 0.88
Others Other Industries 9.27 9.60 0.13 0.14
Total 6.58 6.61 0.39 0.38

Value Added per Person Engaged: Net Value Added / Total Number of Persons Engaged
Capital Productivity: Net Value Added / Fixed Capital
*01: Includes only post harvest crop activities (0163) and seed processing for propagation (0164)
**08: Includes only the activities of extraction of salt (0893)
24 4 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 1 (cntd.): Selected Economic Indicators by 2-digit Industry Div. based on ASI
2012-13 and 2013-14
All India
NIC- Ratio of Total Output Output per Person
2008 Description to Total Input Engaged
(Rs. Lakh)
Div. 2012-13 2013-14 2012-13 2013-14
01* Crop and animal production, hunting and related
service activities 1.09 1.12 55.58 86.04
08** Other mining and quarrying 1.98 2.36 7.87 7.84
10 Manufacture of food product s 1.11 1.10 45.13 48.88
11 Manufacture of beverages 1.38 1.31 36.80 38.54
12 Manufacture of tobacco products 1.64 1.57 7.67 7.83
13 Manufacture of textiles 1.23 1.19 22.02 24.90
14 Manufacture of wearing apparel 1.27 1.24 9.80 12.77
15 Manufacture of leather and related products 1.19 1.22 14.46 15.29
16 Manufacture of wood and of products of wood and
cork, except furniture; manufacture of articles of straw
and plaiting materials 1.16 1.16 24.23 26.89
17 Manufacture of paper and paper products 1.18 1.21 29.12 33.89
18 Printing and reproduction of recorded media 1.39 1.30 20.34 19.66
19 Manufacture of coke and refined petroleum products 1.14 1.11 1056.73 1013.49
20 Manufacture of chemicals and chemical products 1.23 1.22 74.64 72.54
21 Manufacture of basic pharmaceutical products and
pharmaceutical preparations 1.48 1.51 35.77 34.69
22 Manufacture of rubber and plastics products 1.19 1.25 36.02 36.47
23 Manufacture of other non -metallic mineral products 1.38 1.36 21.90 19.89
24 Manufacture of basi c metals 1.15 1.20 75.54 87.89
25 Manufacture of fabricated metal products, except
machinery and equipment 1.29 1.25 25.53 25.39
26 Manufacture of computer, electronic and optical
products 1.25 1.24 48.72 56.16
27 Manufacture of electrical equipment 1.23 1.21 42.60 44.25
28 Manufacture of machinery and equipment n.e.c. 1.35 1.30 37.93 36.42
29 Manufacture of motor vehicles, trailers and semi -
trailers 1.22 1.19 52.07 49.99
30 Manufacture of other transport equipment 1.23 1.23 45.13 48.13
31 Manufacture of furniture 1.25 1.21 22.55 25.32
32 Other manufacturing 1.10 1.09 59.09 64.53
33 Repair and installation of machinery and equipment 1.32 1.53 27.35 21.60
38 Waste collection, treatment and disposal activities;
materials recovery 1.08 1.09 56.29 57.01
58 Publishing activities 1.82 1.84 28.96 27.34
Others Other Industries 1.24 1.24 67.18 66.83
Total 1.20 1.19 46.53 48.42
Ra tio of Tota l Ou tpu t to Tota l Inpu t: Tota l Ou tpu t / Tota l Inpu t
Ou tpu t per Person Enga ged: Tota l Ou tpu t / Tota l Nu mber of Persons Engaged
*01: Includes only post harvest crop activities (0163) and seed processing for propagation (0164)
**08: Includes only the activities of extraction of salt (0893)
 Selected Economic Indicators of Manufacturing Sector of India : Table 1 24 5

Table 1 (cntd.): Selected Economic Indicators by 2-digit Industry Div. based on ASI
2012-13 and 2013-14
All India
NIC- Wage Rate (Rs.)
2008 Description Direct Workers Contract Workers
Div. 2012-13 2013-14 2012-13 2013-14
01* Crop and animal production, hunting and related
service activities 60156 68488 54079 67195
08** Other mining and quarrying 67816 71085 52098 61694
10 Manufacture of food products 84221 100880 72414 87984
11 Manufacture of beverages 126262 137867 85886 96142
12 Manufacture of tobacco products 60771 35187 27143 33838
13 Manufacture of textiles 90264 102609 84814 92250
14 Manufacture of wearing apparel 82158 87576 74691 95402
15 Manufacture of leather and related products 77329 89654 80167 85055
16 Manufacture of wood and of products of wood and
cork, except furniture; manufac ture of articles of straw
and plaiting materials 81951 82894 81879 84114
17 Manufacture of paper and paper products 112952 126302 88591 96810
18 Printing and reproduction of recorded media 138319 144437 89665 109488
19 Manufacture of coke and refined pe troleum products 458864 575551 164672 132431
20 Manufacture of chemicals and chemical products 151922 168179 94856 101720
21 Manufacture of basic pharmaceutical products and
pharmaceutical preparations 164286 163585 92215 106042
22 Manufacture of rubber and plastics products 110322 127009 103320 97045
23 Manufacture of other non-metallic mineral products 101503 107958 65207 69291
24 Manufacture of basic metals 208229 247700 110898 119423
25 Manufacture of fabricated metal products, except
machinery and equipment 131121 137340 93002 98055
26 Manufacture of computer, electronic and optical
products 165576 217647 112676 116739
27 Manufacture of electrical equipment 167342 177993 94626 108464
28 Manufacture of machinery and equipment n.e.c. 172302 179267 115515 116383
29 Manufacture of motor vehicles, trailers and semi -
trailers 190880 207443 97859 110292
30 Manufacture of other transport equipment 160359 187909 109738 120293
31 Manufacture of furniture 122636 122032 94777 106126
32 Other manufacturin g 116781 134923 94109 137527
33 Repair and installation of machinery and equipment 248033 282157 116839 115138
38 Waste collection, treatment and disposal activities;
materials recovery 97378 98254 90496 90814
58 Publishing activities 200898 207589 116658 153360
Others Other Industries 122851 132349 101804 126977
Total 123216 132929 85590 97774
Wage Rate (Direct Workers): Wages & Salary to Direct Workers / No. of Direct Workers
Wage Rate (Contract Workers): Wages & Salary to Contract Workers / No. of Contract Workers
*01: Includes only post harvest crop activities (0163) and seed processing for propagation (0164)
**08: Includes only the activities of extraction of salt (0893)
24 6 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 2: Estimates and Proportions of the Employed by Type of Employment at 2-digit

Industry Div. (NIC-2008) based on ASI 2013-14 (All India).
Factories Total Directly
in Total Contract Total Contract
NIC- Persons Emp.
Operation Workers Workers Workers Workers
2008 Engaged Workers
(no.) (no.) (no.) (%) (%)
(no.) (no.)
(1) (2) (3) (4) (5) (6) (7) (8)
01* 3153 97567 73096 45607 27489 74.92 37.61
08** 160 9072 8072 2525 5548 88.98 68.73
10 30130 1582527 1232679 897515 335165 77.89 27.19
11 1936 158507 121346 59186 62160 76.56 51.23
12 2452 444942 425799 327823 97976 95.70 23.01
13 13379 1496194 1267670 1057406 210264 84.73 16.59
14 7047 978709 823850 720475 103375 84.18 12.55
15 3330 311594 266153 209622 56531 85.42 21.24
16 3714 78981 60034 42704 17330 76.01 28.87
17 5626 248529 193026 134984 58042 77.67 30.07
18 3717 156988 100708 79477 21231 64.15 21.08
19 1327 109964 81641 39863 41778 74.24 51.17
20 9710 708401 494253 296487 197766 69.77 40.01
21 4315 618493 381664 198152 183512 61.71 48.08
22 10447 591001 466790 293938 172852 78.98 37.03
23 22089 970367 788819 308007 480813 81.29 60.95
24 9732 976196 748923 409285 339638 76.72 45.35
25 13685 678906 523947 302043 221904 77.18 42.35
26 2167 222987 152710 100069 52641 68.48 34.47
27 6191 513943 370071 219418 150653 72.01 40.71
28 10062 647199 441733 292959 148773 68.25 33.68
29 4908 792885 604693 343496 261196 76.26 43.19
30 1925 283498 223641 115233 108408 78.89 48.47
31 1212 60368 44394 29243 15151 73.54 34.13
32 2755 274129 212561 171255 41306 77.54 19.43
33 669 34906 24371 16383 7988 69.82 32.78
38 301 14202 10954 6326 4629 77.13 42.26
58 226 24243 11905 7422 4483 49.11 37.66
Others 9325 452816 288900 207319 81582 63.80 28.24
Total 185690 13538114 10444404 6934221 3510184 77.15 33.61
*01: Includes only post harvest crop activities (0163) and seed processing for propagation (0164)
**08: Includes only the activities of extraction of salt (0893)
 Employment by Industry Division in Manufacturing Sector : Table 3 24 7

Table 3: Estimates and Proportions of the Employed by type of Employment at 3-digit

Industry Group (NIC-2008), based on ASI 2013-14 (All India).
Factories in Directly
Operation Total Contract Total
NIC- Total Persons Emp. Contract
(no.) Workers Workers Workers
2008 Engaged (no.) Workers Workers (%)
(no.) (no.) (%)
(no.)
(1) (2) (3) (4) (5) (6) (7) (8)
016 3153 97567 73096 45607 27489 74.92 37.61
089 160 9072 8072 2525 5548 88.98 68.73
101 145 25607 19945 11007 8938 77.89 44.81
102 403 44178 37226 25974 11252 84.26 30.23
103 998 58331 45924 24406 21519 78.73 46.86
104 2654 107623 83115 54988 28127 77.23 33.84
105 1652 145601 107988 63412 44576 74.17 41.28
106 15703 319659 238365 147677 90688 74.57 38.05
107 7828 836742 668302 548845 119457 79.87 17.87
108 748 44786 31814 21205 10609 71.04 33.35
110 1936 158507 121346 59186 62160 76.56 51.23
120 2452 444942 425799 327823 97976 95.70 23.01
131 10060 1236688 1053632 891441 162191 85.20 15.39
139 3320 259506 214037 165965 48072 82.48 22.46
141 4320 713833 598482 510013 88469 83.84 14.78
142 6 615 499 311 188 81.14 37.68
143 2722 264261 224869 210151 14718 85.09 6.55
151 1628 104532 88941 58776 30164 85.08 33.91
152 1703 207062 177213 150846 26367 85.58 14.88
161 1188 8738 6091 5612 479 69.71 7.86
162 2526 70244 53944 37093 16851 76.80 31.24
170 5626 248529 193026 134984 58042 77.67 30.07
181 3699 156430 100441 79210 21231 64.21 21.14
182 18 558 267 267 0 47.85 0.00
191 567 28876 22733 15830 6903 78.73 30.37
192 760 81088 58908 24033 34875 72.65 59.20
201 3766 279851 198967 103404 95563 71.10 48.03
202 5788 399828 272518 174868 97650 68.16 35.83
203 156 28722 22768 18216 4553 79.27 20.00
210 4315 618493 381664 198152 183512 61.71 48.08
221 2230 218754 174510 104351 70159 79.77 40.20
222 8217 372247 292281 189588 102693 78.52 35.14
231 611 61978 50650 27865 22786 81.72 44.99
239 21478 908389 738169 280142 458027 81.26 62.05
241 4768 650680 499885 271286 228599 76.83 45.73
242 1336 107454 82764 39652 43112 77.02 52.09
243 3628 218062 166275 98347 67928 76.25 40.85
251 3843 230178 172830 90558 82272 75.09 47.60
252 75 2153 1426 1131 295 66.23 20.69
259 9768 446576 349691 210355 139337 78.30 39.85
261 974 76619 55266 40503 14763 72.13 26.71
262 113 24594 15515 8834 6682 63.08 43.07
263 256 41562 27164 15521 11643 65.36 42.86
264 191 26327 19817 10449 9368 75.27 47.27
265 484 45892 29183 20792 8392 63.59 28.76
266 94 6003 4225 2502 1723 70.38 40.78
267 53 1951 1512 1451 61 77.50 4.03
268 3 39 28 18 10 71.79 35.71
271 2342 198774 132980 84701 48278 66.90 36.30
272 373 43871 35374 21589 13786 80.63 38.97
24 8 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 3: Estimates and Proportions of the Employed by type of Employment at 3-digit

Industry Group (NIC-2008), based on ASI 2013-14 (All India).
Factories Total Directly
in Persons Total Emp. Contract Total Contract
NIC-2008
Operation Engaged Workers Workers Workers Workers Workers
(no.) (no.) (no.) (no.) (no.) (%) (%)
(1) (2) (3) (4) (5) (6) (7) (8)
273 1219 95439 73573 36696 36876 77.09 50.12
274 416 52601 42682 21677 21005 81.14 49.21
275 845 53623 36427 24143 12284 67.93 33.72
279 996 69635 49036 30612 18424 70.42 37.57
281 4828 338964 235526 154078 81448 69.48 34.58
282 5235 308235 206206 138881 67325 66.90 32.65
291 163 176523 120786 84899 35887 68.43 29.71
292 510 50210 37712 19533 18179 75.11 48.20
293 4235 566153 446194 239064 207130 78.81 46.42
301 106 28018 22963 5623 17340 81.96 75.51
302 289 27252 20578 12420 8158 75.51 39.64
303 85 10864 7335 5197 2138 67.52 29.15
304 36 654 480 214 266 73.39 55.42
309 1409 216710 172287 91781 80506 79.50 46.73
310 1212 60368 44394 29243 15151 73.54 34.13
321 1170 157630 122188 104480 17708 77.52 14.49
322 11 269 214 203 11 79.55 5.14
323 151 8540 6491 6367 124 76.01 1.91
324 52 2162 1727 1396 331 79.88 19.17
325 450 41540 31083 24860 6224 74.83 20.02
329 921 63989 50858 33949 16909 79.48 33.25
331 598 29790 20857 14569 6288 70.01 30.15
332 71 5115 3514 1814 1700 68.70 48.38
381 24 799 550 108 442 68.84 80.36
382 104 7855 5958 2878 3080 75.85 51.70
383 173 5548 4446 3339 1107 80.14 24.90
581 226 24243 11905 7422 4483 49.11 37.66
Others 9325 452816 288900 207319 81582 63.80 28.24
Total 185690 13538114 10444404 6934221 3510184 77.15 33.61
For NIC 2008 detailed description, may visit the URL given below:
http://mospi.nic.in/Mospi_New/site/inner.aspx?status=2&menu_id=129
http://www.csoisw.gov.in/CMS/En/1026-national-industrial-activity-classification.aspx
 All India ASI Data Based on Units with 100 and more Employees : Table 4 24 9

Table 4: Selected Characteristics of Factory Sector (100 and more employees) by 2-

digit Industry Div. (NIC-2008) for all-India based on ASI 2013-14
(Values in Rs. Lakh unless otherwise mentioned)
2-Digit Industry Div: NIC-2008
Characteristics Total ASI
All 01 08 10 11 12 13
1 Number of Factories (no.) 224576 30411 194 34 4116 439 423 3079
2 Fixed Capital 237371903 208361832 150111 24732 10650139 2277802 474282 18120322
3 Physical Working Capital 101083632 81019014 372196 8497 11469139 832018 603790 5446129
4 Working Capital 66268577 49251115 280356 7891 5370752 454417 501172 4065888
5 Invested Capital 338455535 289380846 522308 33229 22119278 3109820 1078072 23566451
6 Rent Paid 1527272 1189913 3354 379 91156 20215 5865 29934
7 Outstanding Loan 122209355 99261824 58096 12456 6264305 2438807 157728 6965493
8 Interest Paid 15485061 12142216 14331 3580 1170016 121118 34882 1083152
9 Rent Received 297051 160052 62 1 15075 2177 1741 11559
10 Interest Received 2333143 1924011 2403 222 152578 16621 4165 87427
11 Gross Value of Plant & Machinery 216613848 197718577 120921 18267 9934541 2054142 536456 14478402
12 Value of Product and By-Product 570344878 473444346 1557075 38396 45936682 4748083 2463135 25461418
13 Total Output 655525116 529202509 1768217 45135 52610598 5265185 2627958 29800805
14 Fuels Consumed 29850770 24778180 27502 2532 1364635 217940 27986 2345082
15 Materials Consumed 423046161 346135179 898246 3861 38268828 3084257 1157512 18724824
16 Total Input 549013952 438598334 1170075 18827 47549652 4061648 1529846 24817955
17 Gross Value Added 106511164 90604175 598141 26309 5060946 1203536 1098112 4982850
18 Depreciation 16976977 14384044 17001 1766 838492 222318 52478 1092704
19 Net Value Added 89534187 76220130 581141 24542 4222454 981218 1045634 3890147
20 Net Fixed Capital Formation 18396832 16670420 -2475 -1391 590633 89391 35613 356941
21 Gross Fixed Capital Formation 35373809 31054465 14526 375 1429126 311709 88091 1449645
22 Total workers (no.) 10444404 7864151 22615 5972 854331 102868 395152 1054630
23 Total Persons Engaged (no.) 13538114 10047793 29916 6538 1075519 132841 405786 1220337
24 Wages to Workers 12649644 10406101 17105 3630 943050 124699 133501 1091596
25 Emoluments to Employees 27241503 22564407 42125 5074 1816224 293307 180925 1696404
26 Gross Capital Formation 42184321 36009096 50748 1163 1558456 287260 294312 1766304
27 Income 75152048 64972064 565921 20807 3128935 858683 1010793 2876047
28 Profit 43956552 38932357 520651 15165 1079833 526057 806663 963744

Table 4 (cntd.): Selected Characteristics of Factory Sector (100 and more employees)
by 2-digit Industry Div. (NIC-2008) for all-India based on ASI 2013-14
(Values in Rs. Lakh unless otherwise mentioned)
2-Digit Industry Div: NIC-2008
Characteristics 14 15 16 17 18 19 20 21
1 Number of Factories (no.) 2722 906 106 486 462 176 1599 1382
2 Fixed Capital 1512157 609303 277319 3780032 863537 18615019 14334216 7478162
3 Physical Working Capital 1445138 642874 237753 1052386 309841 10569476 5710965 3794790
4 Working Capital 1106052 335819 142686 437065 -845582 -327066 4505498 6528515
5 Invested Capital 2957296 1252177 515072 4832419 1173378 29184495 20045180 11272953
6 Rent Paid 80327 28341 2572 10269 18395 25787 71469 91615
7 Outstanding Loan 1024123 181698 172246 1378747 312657 4032360 3642547 2355999
8 Interest Paid 174253 63078 32156 204314 59276 674734 969121 356551
9 Rent Received 5671 2847 193 3120 13278 2337 10593 7806
10 Interest Received 22174 6362 2292 17571 17255 101992 115417 222642
11 Gross Value of Plant &
Machinery 895964 408519 249320 4161690 904059 21714769 20443539 6567210
12 Value of Product and By -Product 8487471 3071844 806514 5415134 1174628 1.08E+08 40119388 16655118
13 Total Output 9594126 3412651 1033934 5692067 1957918 1.09E+08 43394210 18847709
14 Fuels Consumed 168733 70932 37270 686603 50471 1564697 3308560 715118
25 0 The Journal of Industrial Statistics, Vol. 5, No. 2

Table 4 (cntd.): Selected Characteristics of Factory Sector (100 and more employees)
by 2-digit Industry Div. (NIC-2008) for all-India based on ASI 2013-14
2-Digit Industry Div: NIC-2008
Characteristics 14 15 16 17 18 19 20 21
15 Materials Consumed 4770980 2159394 548893 3445541 1048684 93049596 26909998 8719972
16 Total Input 7644390 2767299 865866 4614398 1488745 97490390 35131691 12191811
17 Gross Value Added 1949736 645352 168068 1077669 469174 11092566 8262519 6655897
18 Depreciation 143759 65821 31960 270087 92843 1173626 1229718 629659
19 Net Value Added 1805978 579531 136108 807582 376331 9918940 7032801 6026239
20 Net Fixed Capital Formation 142026 40323 -7339 199615 3359 2743409 998330 653634
21 Gross Fixed Capital Formation 285784 106144 24621 469702 96202 3917035 2228048 1283293
22 Total workers (no.) 706701 217539 25045 111127 49093 66204 366982 327262
23 Total Persons Engaged (no.) 829035 250524 31516 140204 84068 86480 531500 534997
24 Wages to Workers 624468 196838 25999 159817 80789 260238 587258 460377
25 Emoluments to Employees 1052448 321334 53119 308113 237867 541556 1508995 1611329
26 Gross Capital Formation 431021 177669 43759 628135 150155 4545229 2530557 1730207
27 Income 1579243 497321 103865 613689 329193 9322749 6118221 5808520
28 Profit 407024 133856 45762 247715 63868 8641095 4317906 3978498

Table 4(cntd.): Selected Characteristics of Factory Sector (100 and more employees) by
2-digit Industry Div.(NIC-2008) for all-India based on ASI 2013-14
(Values in Rs. Lakh unless otherwise mentioned)
2-Digit Industry Div: NIC-2008
Characteristics 22 23 24 25 26 27 28 29
1 Number of Factories (no.) 1347 2027 1624 1670 505 1190 1399 1551
2 Fixed Capital 5058511 13138496 51314439 3902754 3412570 3837391 5225966 13481274
3 Physical Working Capital 2170564 2649380 12653924 2587973 1717391 3162133 3630699 4026996
4 Working Capital 1877303 1951120 8319590 1180842 2222065 3276626 3419393 -395651
5 Invested Capital 7229075 15787876 63968363 6490727 5129960 6999524 8856665 17508270
6 Rent Paid 32598 67790 63691 21501 117018 62491 76595 88960
7 Outstanding Loan 2364387 5351313 32027904 6063797 1636511 1773528 2263296 5973063
8 Interest Paid 308914 553128 2697954 300859 207062 439908 404602 556136
9 Rent Received 4334 3935 11415 6830 5960 3188 10458 20589
10 Interest Received 32115 63162 340099 42936 111722 59437 99394 156100
11 Gross Value of Plant &
Machinery 5459700 14173527 41267382 3311245 2483911 3834864 4710787 14577610
12 Value of Product and By -
Product 14439797 14227439 59804584 10219127 7543721 15993952 16908258 34513237
13 Total Output 15274177 15275496 66187068 12012821 10724839 17695748 19196078 37445852
14 Fuels Consumed 677356 3372715 5952457 411287 109686 273125 299947 2244668
15 Materials Consumed 10147691 5862904 39564972 6738834 5719666 11561920 11209685 24939964
16 Total Input 12139754 11064810 53096720 9546536 8678922 14271650 14618946 31659415
17 Gross Value Added 3134423 4210686 13090348 2466285 2045917 3424099 4577132 5786437
18 Depreciation 461113 979768 2258893 347514 219230 387492 515399 1686898
19 Net Value Added 2673310 3230918 10831455 2118771 1826688 3036607 4061732 4099539
20 Net Fixed Capital Formation 353338 648116 5682259 144003 64310 192770 338500 1033542
21 Gross Fixed Capital Formation 814450 1627884 7941152 491517 283540 580262 853900 2720441
22 Total workers (no.) 327978 424635 614662 343590 121019 283216 310464 530221
23 Total Persons Engaged (no.) 402958 525117 793539 433870 176238 390386 451692 693954
24 Wages to Workers 418256 439944 1299273 444674 241707 465120 553830 919995
25 Emoluments to Employees 829694 938007 2497918 927997 868939 1172172 1602326 2047947
26 Gross Capital Formation 1183718 1618769 8837381 588799 620405 733092 757661 2786478
27 Income 2368247 2677098 8421325 1846178 1620290 2596833 3690387 3631132
28 Profit 1430030 1596209 5449742 794984 638582 1235329 1816184 1195887
 All India ASI Data Based on Units with 100 and more Employees : Table 4 25 1

Table 4 (cntd.): Selected Characteristics of Factory Sector (100 and more employees)
by 2-digit Industry Div.(NIC-2008) for all-India based on ASI 2013-14
(Values in Rs. Lakh unless otherwise mentioned)
2-Digit Industry Div: NIC-2008
Characteristics 30 31 32 33 38 58 Others
1 Number of Factories (no.) 539 150 752 89 33 75 1336
2 Fixed Capital 3070267 223639 921037 288200 108095 245661 24966400
3 Physical Working Capital 1429338 184443 2410824 89176 31022 50589 1729568
4 Working Capital 782432 927902 2233949 111977 -4275 62405 721972
5 Invested Capital 4499605 408082 3331861 377376 139117 296250 26695968
6 Rent Paid 23807 7493 29271 5558 325 3125 110014
7 Outstanding Loan 1544952 122017 981166 190718 86176 619727 9266007
8 Interest Paid 218823 25354 174845 10698 9726 10606 1263043
9 Rent Received 3097 1965 797 389 164 1293 9178
10 Interest Received 48343 1281 85748 3563 8871 2377 99741
11 Gross Value of Plant &
Machinery 2490035 166964 709961 131839 91261 272355 21549336
12 Value of Product and By-
Product 11839703 916908 11304826 122834 106510 182788 11724172
13 Total Output 12827799 1065574 14622686 483699 203697 548906 21004599
14 Fuels Consumed 212656 27384 67914 11669 10474 8370 510413
15 Materials Consumed 8826261 642804 10110106 104418 102793 188228 7624346
16 Total Input 10408020 880860 13403097 300263 154432 268927 16763390
17 Gross Value Added 2419779 184714 1219589 183436 49265 279979 4241209
18 Depreciation 261495 19057 90600 27449 14009 24431 1228465
19 Net Value Added 2158284 165657 1128989 155987 35256 255548 3012744
20 Net Fixed Capital
Formation -20519 -6610 86455 9178 -5300 4987 2303321
21 Gross Fixed Capital
Formation 240976 12447 177055 36628 8709 29418 3531786
22 Total workers (no.) 201605 28069 176010 15501 5509 9202 166946
23 Total Persons Engaged (no.) 252821 38839 222495 21194 6597 19337 259495
24 Wages to Workers 323073 36015 246922 43074 5972 17718 241163
25 Emoluments to Employees 651190 101908 441996 82016 9683 84949 638845
26 Gross Capital Formation 310418 31620 567797 28036 3976 34179 3711791
27 Income 1967095 136056 1011419 143684 34240 245487 1748606
28 Profit 1221432 24439 528080 50902 23511 148843 1030366
25 2 The Journal of Industrial Statistics, Vol. 5, No. 2

2-digit NIC Division and Description

NIC- Description
2008
01 CROP AND ANIMAL PRODUCTION, HUNTING AND RELATED SERVICE ACTIVITIES
02 FORESTRY AND LOGGING
03 FISHING AND AQUACULTURE
05 MINING OF COAL AND LIGNITE
06 EXTRACTION OF CRUDE PETROLEUM AND NATU RAL GAS
07 MINING OF METAL ORES
08 OTHER MINING AND QUARRYING
09 MINING SUPPORT SERVICE ACTIVITIES
10 MANUFACTURE OF FOOD PRODUCTS
11 MANUFACTURE OF BEVERAGES
12 MANUFACTURE OF TOBACCO PRODUCTS
13 MANUFACTURE OF TEXTILES
14 MANUFACTURE WEARING APPA REL
15 MANUFACTURE LEATHER AND RELATED PRODUCTS
16 MANUFACTURE OF WOOD AND OF PRODUCTS OF WOOD AND CORK, EXCEPT FURNITURE;
ARTICLES OF STRAW AND PLAITING MATERIAL
17 MANUFACTURE OF PAPER AND PAPER PRODUCTS
18 MANUFACTURE OF PRINTING AND REPRODUCTION OF RECORDED MEDIA
19 MANUFACTURE OF COKE AND REFINED PETROLEUM PRODUCTS
20 MANUFACTURE OF CHEMICALS AND CHEMICAL PRODUCTS
21 MANUFACTURE OF BASIC PHARMACEUTICAL PRODUCTS AND PHARMACEUTICAL PREPARATIONS
22 MANUFACTURE OF RUBBER AND PLASTICS PRODUCTS
23 MANUFACTURE OF OTHER NON -METALLIC MINERAL PRODUCTS
24 MANUFACTURE OF BASIC METALS
25 MANUFACTURE OF FABRICATED METAL PRODUCTS, EXCEPT MACHINERY AND EQUIPMENT
26 MANUFACTURE OF COMPUTER, ELECTRONIC AND OPTICAL PRODUCTS
27 MANUFACTURE OF ELECTRICAL EQUIPM ENT
28 MANUFACTURE OF MACHINERY AND EQUIPMENT N.E.C.
29 MANUFACTURE OF MOTOR VEHICLES, TRAILERS AND SEMI -TRAILERS
30 MANUFACTURE OF OTHER TRANSPORT EQUIPMENT
31 MANUFACTURE OF FURNITURE
32 OTHER MANUFACTURING
33 REPAIR AND INSTALLATION OF MACHINERY A ND EQUIPMENT
35 ELECTRICITY, GAS, STEAM AND AIR CONDITIONING SUPPLY
36 WATER COLLECTION, TREATMENT AND SUPPLY
37 SEWERAGE
38 WASTE COLLECTION, TREATMENT AND DISPOSAL ACTIVITIES; MATERIALS RECOVERY
39 REMEDIATION ACTIVITIES AND OTHER WASTE MANAGEMENT S ERVICES
41 CONSTRUCTION OF BUILDINGS
42 CIVIL ENGINEERING
43 SPECIALIZED CONSTRUCTION ACTIVITIES
45 WHOLESALE AND RETAIL TRADE AND REPAIR OF MOTOR VEHICLES AND MOTORCYCLES
46 WHOLESALE TRADE, EXCEPT OF MOTOR VEHICLES AND MOTORCYCLES
47 RETAIL TRADE, EXCEPT OF MOTOR VEHICLES AND MOTORCYCLES
49 LAND TRANSPORT AND TRANSPORT VIA PIPELINES
50 WATER TRANSPORT
51 AIR TRANSPORT
52 WAREHOUSING AND SUPPORT ACTIVITIES FOR TRANSPORTATION
53 POSTAL AND COURIER ACTIVITIES
55 ACCOMMODATION
56 FOOD AND BEVERAGE SERVICE ACTIVITIES
58 PUBLISHING ACTIVITIES
59 MOTION PICTURE, VIDEO AND TELEVISION PROGRAMME PRODUCTION, SOUND RECORDING AND
MUSIC PUBLISHING ACTIVITIES
 2-digit NIC Division and Description 25 3

2-digit NIC Division and Description (Contd.)

NIC- Description
2008
60 BROADCASTING AND PROGRA MMING ACTIVITIES
61 TELECOMMUNICATIONS
62 COMPUTER PROGRAMMING, CONSULTANCY AND RELATED ACTIVITIES
63 INFORMATION SERVICE ACTIVITIES
64 FINANCIAL SERVICE ACTIVITIES, EXCEPT INSURANCE AND PENSION FUNDING
65 INSURANCE, REINSURANCE AND PENSION FUNDING, E XCEPT COMPULSORY SOCIAL SECURITY
66 OTHER FINANCIAL ACTIVITIES
68 REAL ESTATE ACTIVITIES
69 LEGAL AND ACCOUNTING ACTIVITIES
70 ACTIVITIES OF HEAD OFFICES; MANAGEMENT CONSULTANCY ACTIVITIES
71 ARCHITECTURE AND ENGINEERING ACTIVITIES; TECHNICAL TESTING AND ANALYSIS
72 SCIENTIFIC RESEARCH AND DEVELOPMENT
73 ADVERTISING AND MARKET RESEARCH
74 OTHER PROFESSIONAL, SCIENTIFIC AND TECHNICAL ACTIVITIES
75 VETERINARY ACTIVITIES
77 RENTAL AND LEASING ACTIVITIES
78 EMPLOYMENT ACTIVITIES
79 TRAVEL AGENCY, TO UR OPERATOR AND OTHER RESERVATION SERVICE ACTIVITIES
80 SECURITY AND INVESTIGATION ACTIVITIES
81 SERVICES TO BUILDINGS AND LANDSCAPE ACTIVITIES
82 OFFICE ADMINISTRATIVE, OFFICE SUPPORT AND OTHER BUSINESS SUPPORT ACTIVITIES
84 PUBLIC ADMINISTRATION AND DEFENCE; COMPULSORY SOCIAL SECURITY
85 EDUCATION
86 HUMAN HEALTH ACTIVITIES
87 RESIDENTIAL CARE ACTIVITIES
88 SOCIAL WORK ACTIVITIES WITHOUT ACCOMMODATION
90 CREATIVE, ARTS AND ENTERTAINMENT ACTIVITIES
91 LIBRARIES, ARCHIVES, MUSEUMS AND OTHER CULTUR AL ACTIVITIES
92 GAMBLING AND BETTING ACTIVITIES
93 SPORTS ACTIVITIES AND AMUSEMENT AND RECREATION ACTIVITIES
94 ACTIVITIES OF MEMBERSHIP ORGANIZATIONS
95 REPAIR OF COMPUTERS AND PERSONAL AND HOUSEHOLD GOODS
96 OTHER PERSONAL SERVICE ACTIVITIES
97 ACTIVITIES OF HOUSEHOLDS AS EMPLOYERS OF DOMESTIC PERSONNEL
98 UNDIFFERENTIATED GOODS - AND SERVICES -PRODUCING ACTIVITIES OF PRIVATE HOUSEHOLDS
FOR OWN USE
99 ACTIVITIES OF EXTRATERRITORIAL ORGANIZATIONS AND BODIES
25 4 The Journal of Industrial Statistics (2016), 5 (2), 254 - 260

New Initiatives in Annual Survey of Industries (ASI)

Various new initiatives have been taken to improve the coverage and quality of ASI data.
Based on the detailed deliberations of the Standing Committee on Industrial Statistics
(SCIS), the following decisions were taken:
1. Issue of coverage of frame of Annual Survey of Industries (ASI)
1.1 After detailed discussions relating to the issue of coverage of ASI and also
inclusion of Company Identification Number (CIN) in ASI schedule and discussions based
on the report of the study of database maintained by M/o Corporate Affairs & the list of
units available in the frame of the 6th Economic Census, the followings decisions were
taken:
1.2 The coverage of ASI may be extended beyond the purview of the Section 2m (i)
and 2m (ii) of the Factories Act, 1948 and the Bidi & Cigar Workers (Conditions of
Employment) Act, 1966 as recommended by the Sub-Group on Sampling Design of ASI.
For this purpose, Business Register of Enterprises(BRE) prepared for the respective
states and Directory of Establishments based on Sixth Economic Census will be used by
CSO (IS Wing). It was decided that initially the coverage should be restricted to units
having 20 or more employees not registered under Section 2m (i) and 2m (ii) of the Factories
Act, 1948 but registered under Companies Act, which is identified and validated in the
Business Register of Establishments (BRE).
1.2.1 To start with the implementation of the augmented frame, units with 100 or more
employees not registered under Section 2m (i) and 2m (ii) of the Factories Act, 1948 but
included in the BRE of the respective states would be included in ASI frame. For this, BRE
of Andhra Pradesh (AP) readily available with National Accounts Division (NAD) was
included in the frame of Andhra Pradesh for ASI 2014-15 after verification of such units by
FOD because inclusion of full set of units from the BRE would have suddenly increased the
size of the frame of AP considerably and consequently sample size for AP would have to be
increased appreciably, apart from verification of such units in the field before actual selection.
In ASI 2015-16, most of the states had sent the list of units not registered under Factories
Act, 1948 and also having 100 or more workers and also registered under any of the Seven
Acts included in BRE for inclusion in ASI 2015-16 Frame. Thus, ASI 2015-16 survey scheduled
to be launched from October, 2016 would include units not registered under Factories Act,
1948 in the respective States. This is a significant departure from past practices and it is an
improvement in coverage of registered manufacturing sector.
2. New Sampling Design of ASI to be implemented from ASI 2015-16
2.1 Report of the Sub-Group of SCIS on Sampling Design of ASI under the
Chairmanship of Dr. G. C. Manna, the then ADG, ESD (CSO), was placed in the meeting of
the SCIS for discussion and approval. Detailed discussion was held on various issues
related to the existing sampling design of ASI and the recommendations of the Sub-Group
and accordingly, the following decisions were taken:

2.2 It was decided that the existing employment cut-off of 100 employees to define
Census Sector might be changed to 75 for six (06) States (namely, Jammu & Kashmir,
Himachal Pradesh, Rajasthan, Bihar, Chhattisgarh and Kerala) and to 50 for three (03)
States/UTs (namely, Chandigarh, Delhi and Puducherry). For other States/UTs, the existing
 New Initiatives in Annual Survey of Industries (ASI) 25 5

cut-off of 100 employees may be continued. The recommendation of the Sub-Group relating
to the formation of strata based on District X 3-digit level of NIC was also approved by the
SCIS. Considering the above points, the new sampling design of ASI, which would be
implemented from ASI 2015-16, is noted as follows:

2.3 ASI sample is divided into two parts – Central Sample and State Sample. The
Central Sample consists of two schemes: Census and Sample. Under Census scheme, all
the units are surveyed.

(1) Census Scheme:

(i) All industrial units belonging to the seven less industrially developed States/
UTs viz. Arunachal Pradesh, Manipur, Meghalaya, Nagaland, Sikkim, Tripura and
Andaman & Nicobar Islands.

(ii) For the States/ UTs other than those mentioned in (i),

(a) units having 75 or more employees from six States, namely, Jammu &
Kashmir, Himachal Pradesh, Rajasthan, Bihar, Chhattisgarh and Kerala;

(b) units having 50 or more employees from three States/UTs, namely,

Chandigarh, Delhi and Puducherry;

(c) units having 100 or more employees for rest of the States/UTs, not
mentioned in (a) and (b) above and;

(d) all factories covered under ‘Joint Return’ (JR), where JR should be allowed
when the two or more units located in the same State/UT belonging to the
same industry (3-digit level of NIC) under the same management. It may be
noted that the principle of JR is applicable only before the selection of units
before the survey and unit(s) belonging to the “Census Scheme” will not be
joint with unit(s) of “Sample Scheme” (defined below) at the field stage or all
units belonging to the “Sample Scheme” should not be joint among
themselves at the field stage even if the conditions of JR are satisfied.

(iii) After excluding the Census Scheme units, as defined above, all units belonging
to the strata (State x District x Sector x 3 digit NIC-2008) having less than or
equal to 4 units are also considered under Census Scheme. It may be noted that
strata are separately formed under three sectors considered as Bidi, Manufacturing
and Electricity.

(2) All the remaining units in the frame are considered under Sample Scheme. For all the
states, each stratum is formed on the basis of State x District x Sector x 3-digit NIC-2008.
The units are arranged in descending order of their number of employees. Samples are
drawn as per Circular Systematic Sampling technique for this scheme. An even number of
units with a minimum of 4 units are selected and distributed in four sub-samples. It may be
noted that in certain cases each of 4 sub-samples from a particular stratum may not have
equal number of units.
(3) Out of these 4 sub-samples, two pre-assigned sub-samples are given to NSSO (FOD)
and the other two subsamples are given to State/UT for data collection.
25 6 The Journal of Industrial Statistics, Vol. 5, No. 2

(4) The entire census units plus all the units belonging to the two sub-samples given to
NSSO (FOD) are treated as the Central Sample.
(5) The entire census units plus all the units belonging to the two sub-samples given to
State/UT are treated as the State Sample. Hence, State/UT has to use Census Units (collected
by NSSO (FOD) and processed by CSO (IS Wing)) along with their sub-samples while
deriving estimates for their respective State/UTs based on State Sample.
(6) The entire census units plus all the units belonging to the two sub-samples given to
NSSO (FOD) plus all the units belonging to the two sub-samples given to State/UT are
required for obtaining pooled estimates based on Central Sample and State Sample with
consequent increase in sample size.
2.4 It was decided that the list of units having status code “02: Closed” and “03: Non-
Operating” in the ASI Frame might be prepared based on the latest updated frame, which
were not selected in the current survey of ASI, and may be shared with NSSO (FOD) for
verification of the units considering its existence, activities etc., so that, the units, which
are not in existence at all, existing but shifting its line of activities from manufacturing to
others, which are not associated in the line of manufacturing activities, may be deleted from
ASI frame by NSSO (FOD) before sample is drawn for the subsequent survey. This is
expected to increase the effective sample size used for estimation.
2.5 It was also decided to exclude Govt. Departmental units from the ASI frame and
CSO (IS Wing) were entrusted with the responsibility to identify such units based on the
current survey, so that the same may not be selected in the subsequent surveys of ASI
starting from ASI 2015-16
3. Modifications made in ASI schedule for ASI 2015-16
3.1 The schedule for the year 2015-16, as compared to the ASI schedule for the year
2014-15, are given in the following table:
Block Row, As per ASI As per ASI Remarks
Column 2014-15 2015-16
A No change in ASI 2015-16
B 3 Type of Company
ownership Identification
(code) Number (CIN)
8 Does your unit Whether the As the existing item 8 of Block
have share capital of – B has no relevance in the era
computerized the company of advancement of IT and such
accounting includes share information is not tabulated, it
system? (yes – of foreign was decided to replace the
1, no – 2) entities (yes – 1, existing item with the
no – 2) information related to share
capital of foreign entities.
9 Can your unit Any R&D unit This information is considered
supply ASI data in your factory? useful in today’s context.
in computer (yes &
media? (yes – 1, registered with
no – 2) DST/DBT – 1,
yes & registered
with others – 2,
no – 3)
 New Initiatives in Annual Survey of Industries (ASI) 25 7

Block Row, As per ASI As per ASI Remarks

Column 2014-15 2015-16
C, D No change in ASI 2015-16
&E
F 7,3 Total expenses Expenses on Research Si nce t he “ total” is auto
(1 to 6) & D e v e l o p m e n t calculated field without any
(R&D) s ig n ifi c anc e , f or min i mal
changes in ASI Web Portal, the
item “Total expenses (1 to 6)”
was replaced with the item
“Expenses on Research &
Development (R&D)”.

G 1,3 Income from Receipts from Because of growing importance

services (industrial/ manufacturing of units engaged in 'manufac-
non ind ust rial s e r v ic e s ( i n c l ud in g turing services'(job work), the
i n c l u d i n g wo r k work done for others existing item G1,3 is split into
done for others on on materials supplied 'manufacturing services' G(item
materials supplied by them and sale value 1, col 3) and 'non-manufacturing
by them and sale of waste left by the services' G(item 2, col 3).
value of waste left party)
by the party)

2,3 Variation in stock Receipts from non-

of semi-finished manufacturing
goods (col. 4 minus services (including
col. 3 against item 5 non-industrial
in Block D) services)

7,3 Total Receipts Variation in stock of

(1 to 6) semi-finished goods
(col. 4 minus col. 3
against item 5 in Block
D)

H&I No change in ASI 2015-16

J Col. 11 Distributive Subsidy (-) Due to incorporation of new
expenses (Rs.) - column relating to Subsidy, the
Total c o l . r e l a t i n g t o “ To t a l
Distributive Expenses” may be
replaced with the col. relating to
“Subsidy”.
K, L, No change in ASI 2015-16
M, N,
Part A
&B
Part II 11 Type of N.A. Due to omission of the item, the
ownership item description is modified
(code) accordingly the position of
entering the code is disabled, so
that, no code may be recorded.
The item may be utilized in
future, if new information are
required to be collected.
.
25 8 The Journal of Industrial Statistics, Vol. 5, No. 2

4. Type of Organisation and Ownership codes used in ASI schedule

4.1 Organisation and Ownership codes used in Block – B of ASI schedule were reviewed
and finally it was decided that the “Type of Ownership” code collected in item 3 of Block –
B might be excluded from the ASI schedule, as the proposed “Type of Organisation” code,
which would be collected in item 2 of Block – B, would take care of the ownership aspect of
the unit also. The modified codification structure of “Type of Organisation” along with the
concepts, definitions and terminologies to be used is given below:
Block B: item 2: type of organisation: This item is to be recorded in terms of the following
codes:
Individual Proprietorship -1
Partnership -2
Limited Liability Partnership -3
Government Company-Public -4
Government Company-Private -5
Non-Government Company-Public -6
Non-Government Company-Private -7
Co-operative Society -8
Others (including Joint Family (HUF),
Trusts, Wakf Boards, Handlooms,
KVIC, etc.) -9
Type of Organisation: The following characteristics may be noted for recording entries in
this block:
Type of
Description
Organisation
Proprietary Here, an individual is the sole owner of the enterprise.

Partnership It means relation between persons who have agreed to share the profits of a
business carried on by all or any one of them acting for all.

Limited Liability A Firm incorporated under the Limited Liability Partnership Act, 2008. LLPs
Partnership(LLP) are given by ROC a unique 8-digit identification number called LLPIN.

It is a company where paid-up share capital of the appropriate Government

Government
(Central/ State/ Local) is not less than 51% and number of shareholders is at
Company-Public
least 7 and no upper limit for number of shareholders.
It is a company where paid-up share capital of the appropriate Government
Government
(Central/ State/ Local) is not less than 51% and number of shareholders
Company-
(including the Government) is at least 1 and maximum number of
Private shareholders is 200.
Non- It is a company where paid-up share capital of the appropriate Government
Government (Central/ State/ Local) is less than 51% and number of shareholders is at least
Company-Public 7 and no upper limit for number of shareholders.
Non- It is a company where paid-up share capital of the appropriate Government
Government (Central/ State/ Local) is less than 51% and number of shareholders is at least
Company- 1 and maximum number of shareholders is 200.
Private
 New Initiatives in Annual Survey of Industries 25 9

Type of
Description
Organisation
It is a society formed through the co-operation of a number of persons
(members of the society) to benefit the members. The funds are raised by
Co-operative members' contributions/ investments, and the members share the profits. The
society government or government agency can also be a member or shareholder of a
registered co-operative society but this fact cannot render the society into a
public sector enterprise for the purpose of the survey.
Others These are the enterprises not falling under any of the above categories.

5. Other changes related to conceptual issues in ASI Schedule

5.1 Block C, item 6: Pollution control equipment/environment improvement equipment:
This refers to machinery installed for pollution control as well as environment improvement.
5.2 Convertible debentures were included as part of ‘Outstanding Loan’ in Block D.
Convertible debentures had been considered as part of outstanding loan or long-term debt,
as it remains part of long-term liability of the factory until the date of redemption of the
debenture.
5.3 Omission of Non-operating expenses (Block F, Item 4) in ASI Schedule. Non-
operating expenses have been merged in ‘Operating Expenses’ and now all the expenses of
a factory have to be shown as Operating expenses (Item 3 Block F) only.
5.5 Rationalization of reporting of Value of Own Construction (Block G item 4). The
net balance of capital work-in-progress i.e. (Col. 7 – Col.3) of item 9 in Block C, if positive,
would be transferred to Value of Own Construction (Block G item 4). One new item has been
included in Block F item 4 as “Expenses on raw materials and other components for own
construction” to include the input cost for value of own construction.
5.6 Treatment of Subsidy/Rebate in ASI schedule. A new column has been added as
Column 12 in Block J with heading ‘Subsidy’ under Distributive expenses to include subsidy
on product. The value in this column has been added to Gross Sale Value (Col.7) to arrive
at the ex-factory value. The production subsidies (i.e. as distinct from ‘Subsidies on
Products’) have been included in existing Block G, Item 12 and hence new item has been
described as ‘Other Production Subsidies’ as per guidelines in SNA.
5.7 Specific Reporting of Sale value in Block J. Output value of the entire amount of
finished products/by-products have to be calculated at cost price of producing the item
and reported in Block J as ex-factory-value in cases where a factory has not sold any
quantity of a finished product/by-products produced during the current reference period.
As an alternative to the above, sale value has not been counted in the current reference
period in cases where a factory did not produce any quantity of finished product/by-
product during the current reference period and sold only from last year stock.
5.8 Recording of R&D Expenditure in ASI Schedule. A separate item in Block F has
been created to report R & D expenditure. One question has been formed in Block B ‘Any
R&D unit in your factory?’ with options 1) ‘Yes & Registered with DST/DBT’ 2) ‘Yes &
Registered with Others’ 3) ‘No’.
26 0 The Journal of Industrial Statistics, Vol. 5, No. 2

5.9 Reframing the existing questions in Block K: Information and Communication

Technology (ICT) block as per UN guidelines.
Block K: Information and Communication technology (ICT) usage
Sl. ICT indicator yes-1,
No. no-2

1. Did the factory use computer/s during <reference period>?

2. Did the factory use the Internet during <reference period>?
3. Does the factory have a web site as on the date of survey?
4. Did the factory receive orders via the Internet during <reference period>?
5. Did the factory place orders for business purpose via the Internet during
<reference period>?
6. Did the factory connect to the Internet either by a. Narrowband or b. Fixed
broadband or by c. Mobile broadband during <reference period> ?
7. Does the factory have a local area network (LAN) as on the date of survey?

Explanatory Notes
A computer refers to a desktop or a laptop computer. It does not include equipment with some
embedded computing abilities such as mobile cellular phones, personal digital assistants (PDA)
or TV sets.
The Internet is a worldwide public computer network. It provides access to a number of
communication services including the World Wide Web and carries email, news, entertainment
and data files, irrespective of the device used (not assumed to be only via a computer - it may also
be by mobile phone, games machine, digital TV, etc.). Access can be via a fixed or mobile
network.
A web presence includes a website, home page or presence on another entity's website (including
a related business). It excludes inclusion in an on-line directory of any other webpages where the
business does not have control over the content of the page.
Orders received include orders received via the Internet whether or not payment was made
online. They include orders received via websites, specialized Internet marketplaces, extranets,
-
EDI over the Internet, Internet-enabled mobile phones and email. They also include orders
received on behalf of other organizations – and orders received by other organizations on behalf
of the business. They exclude orders that were cancelled or not completed.
Orders placed include orders placed via the Internet whether or not payment was made online.
They include orders placed via websites, specialized Internet marketplaces, extranets, EDI over
the Internet, Internet-enabled mobile phones and email. They exclude orders that were cancelled
or not completed.
Narrowband includes analogue modem (dial-up via standard phone line), Integrated Services
Digital Network (ISDN), Digital Subscriber Line (DSL) at speeds below 256 kbit/s, and mobile
phone and other forms of access with an advertised download speed of less than 256 kbit/s.
Narrowband mobile phone access services include CDMA 1x (Release 0), GPRS, WAP and
imode
Fixed broadband refers to technologies such as DSL, at speeds of at least 256 kbit/s, cable
modem, high speed leased lines, fibre-to-the-home, powerline, satellite, fixed wireless, Wireless
Local Area Network (WLAN) and WiMAX.
Mobile broadband access services include Wideband CDMA (W-CDMA), known as Universal
Mobile Telecommunications System (UMTS) in Europe; High-speed Downlink Packet Access
(HSDPA), complemented by High-Speed Uplink Packet Access (HSUPA); CDMA2000 1xEV-
DO and DCMA 2000 1xEV-DV. Access can be via any device (mobile, cellular phone, laptop,
PDA, etc.)
A LAN refers to a network connecting computers within a localized area such as a single
building, department or site; it may be wireless.