Identifying Employment Creating Sectors in India:

Transcription

1 ICDD Working Paers Paer No. 12 May 2014 Tulika Bhattacharya Meenakshi Rajeev Identifying Emloyment Creating Sectors in India: An Analysis of Inut-Outut Linkages The International Center for Develoment and Decent Work

2 Tulika Bhattacharya is a PhD scholar at the Institute for Social and Economic Change, Nagarbhavi, Bangalore tulika.0588@gmail.com Meenakshi Rajeev is a rofessor at the Institute for Social and Economic Change, Nagarbhavi, Bangalore meenakshi@isec.ac.in Editorial Board Khayaat Fakier (University of the Witwatersrand, South Africa) Béatrice Knerr (University of Kassel, Germany) The ICDD Working Paers are eer-reviewed. Contact Address Prof. Dr. Christoh Scherrer ICDD University of Kassel, Germany Mailing address: Mönchebergstr. 19, D Kassel Visitors address: Kleine Rosenstr. 3, D Kassel Contact: felmeden@icdd.uni-kassel.de Design / Layout: Nina Sangenstedt, gestaltvoll.de ICDD Working Paers ISBN: (rint) ISBN: (e-book) International Center for Develoment and Decent Work. All rights reserved. The material in this ublication may not be reroduced, stored or transmitted without the rior ermission of the coyright holder. Short extracts may be quoted, rovided the source is fully acknowledged. The views exressed in this ublication are not necessarily the ones of the ICDD or of the organization for which the author works. First ublished 10 / 2014 in Kassel, Germany Publishing House: kassel university ress GmbH

3 Tulika Bhattacharya Meenakshi Rajeev Identifying Emloyment Creating Sectors in India: An Analysis of Inut-Outut Linkages

4 Inhalt Abstract Introduction... 4 Review of Literature... 6 Methodology and Data Sources Inut-Outut Framework... 7 Data Sources Measuring Emloyment Linkage using I-O Tables Measurement of Emloyment Linkages at the aggregate level Measurement of Emloyment Linkages at the disaggregate level Conclusion References Aendices ICDD Working Paer Series

5 Abstract For a labour surlus economy like India emloyment generation remains a erennial concern. While growth of GDP is suosed to address this roblem, not all sectors of the economy are equally emloyment intensive. This aer attemts to identify the relatively more emloyment generating sectors within the major sectors of the economy by considering both forward and backward linkages of the sectors. A linkage based analysis is considered useful as it shows the creation of emloyment within the sector concerned as well as in the related sectors through its direct and indirect linkages. Thus this aer is an attemt to identify the emloyment generating sectors in terms of their emloyment linkages under the Inut-Outut framework for the time eriods and (the most recent eriod). The analysis utilizes the data rovided by the Central Statistical Organisation (CSO) on inut-outut matrix and aends this matrix by incororating emloyment data from the national samle survey organization (NSSO) in India. By considering the broad sectors of the economy the exercise shows that while the share of in-house emloyment from agriculture is declining, its emloyment generation caability through linkages with other sectors of the economy is increasing. The next attemt is to go to a more disaggregated analysis of the sectors. The exercise is exected to rovide imortant olicy inuts for the state that is trying to achieve inclusive growth. Keywords: Inut-Outut Analysis, Emloyment linkages, Outut linkages, Forward and backward linkages 3

6 1 Introduction Generation of emloyment remains an imortant issue in a labour surlus economy like India with a growing working age oulation. It is well-known that the official statistics on unemloyment in India do not take into account disguised unemloyment and thus rovides an underestimate of the unemloyment rate. Nonetheless to quote certain officials figures: according to the Labour Bureau, Government of India (2010), the unemloyment rate was at 9.4 ercent in , an increase of 1.1 ercent over the revious survey in The growth of labour, on the other hand, stood still at around 2 ercent for the decade between 1994 and 2005, which has added to growing unemloyment in the nation 1. Secondly, 93 ercent of the oulation working in the informal sector also shows the concerns India has on the emloyment front. Historically, since the mid 1970s emloyment generation has been considered as an inevitable develoment olicy in India, and several rograms for the creation of emloyment were subsequently taken u. Later, India adoted a series of reform measures during the 1990s, but what the eriod witnessed was a hase of jobless growth. One of the reasons for low emloyment growth couled with high outut growth may be because the high outut achieved was through the growth of certain sectors which were not articularly emloyment intensive. Subsequently, though rograms like National Rural Emloyment Guarantee Act (NREGA) were adoted by the government for the generation of emloyment, these rograms were not adequately linked with the roduction rocess. Therefore, if our collective concern is about inclusive growth, it is essential to identify the sectors that are not only growth oriented but also emloyment generating. In the existing literature, a number of studies have focused on the issue of emloyment by studying its trend, atterns, share across sectors and other such indicators. However, while examining the emloyment creation otentials of different sectors within the economy, the authors have mostly concentrated on the emloyment generated only within a sector. However, it is also imortant to cature how emloyment generation in one sector in turn hels to create emloyment oortunities in other sectors through linkage effects. More recisely, if outut of sector X rises, by imlication emloyment in the sector also increases through direct effect. In addition, emloyment of the sectors which suly inut to X would also rise due to higher inut demands; and the same rocess 1 Labour Bureau, Ministry of Labour and Emloyment, Government of India (2010), Reort on Emloyment and Unemloyment Survey , downloaded from htt://labourbureau.nic.in/final_reort_em_unem_2009_10.df, accessed on

7 of emloyment generation in turn goes on for the sectors that suly inuts to these first sets of inut sulying sectors and so on. Similar effects can be seen for the sectors that demand outut of sector X as their inut, and therefore the final icture needs to be evaluated considering all the linkage effects. This is an area where not many authors have worked in India and the current study attemts to fill this ga by examining emloyment generation otentials of major sectors of the Indian economy using both direct and indirect linkage effects of the sectors. Thus, this aer essentially incororates the emloyment asect within an Inut-Outut framework in order to understand emloyment linkages of individual economic sectors by estimating emloyment multiliers for the time eriods and To achieve this, the analysis uses National Samle Survey Organization (NSSO) data on emloyment and aends this in the standard inut-outut table rovided by the Central Statistical Organisation. This exercise we believe is of considerable imortance. This hels in the identification of sectors which have the otential to create emloyment through direct and indirect emloyment linkage effects. Further, an analysis is resented to show whether a highly linked sector in terms of outut (calculated by measuring the standard backward and forward linkage effects) is also emloyment generating or not, and this exercise is exected to rovide olicy inuts for fostering inclusive growth. Often a sector with high backward linkage is found to be deendent on caital intensive intermediate roducts. Thus, if olicy thrust is given to such sectors, the emloyment generation objective may not be accomlished. In this context, inclusion of the emloyment generation asect within an Inut-Outut framework serves a useful urose in the rocess of determining key sectors. Against this backdro the aer is organized as follows. Section 2 reviews the literature on emloyment asects within an Inut-Outut framework. Methodological details regarding estimation of emloyment multiliers and data sources are discussed in Section 3. In the enultimate section, the emloyment linkage of each sector is resented using the Inut-Outut tables for and both at the aggregate and disaggregates levels. Interretations of these results as well as the comarison between outut and emloyment linkages are also discussed in this section. A concluding section follows at the end. 5

8 2 Review of Literature As mentioned above, emloyment creation has been an issue of intense debate in academia. One set of literature in this context has focused on the role of emloyment in affecting economic growth, while the other strand of literature has incororated the emloyment asect within the Inut-Outut framework. By analyzing the emloyment trends of the Indian economy, Mehrotra et al. (2012) suggested that emloyment should be created in the non-agricultural sector, esecially in the organized manufacturing and services sector, with aroriate olicies. Thus, informal emloyment in the unorganized sector must be transformed to formal emloyment in non-agricultural organized sectors, which will in turn hel to ensure faster growth of the economy. Concentrating on the roblem of jobless growth, Mehrotra et al. (2013) calculate emloyment intensity of outut in the Indian context. In addition, Paola and Sahu (2012) have examined growth and structural changes in emloyment in the long and short term, emhasizing the ost-reform eriod of the Indian economy. They argue that good quality roductive emloyment should be created to ensure inclusive growth. Another set of literature has incororated the emloyment asect within an Inut-Outut (I-O) framework in order to relate the emloyment conditions with that of the change in final demands of several sectors. On the one hand, Gorg and Ruane (2000) have incororated emloyment within the I-O framework for the Irish economy considering average emloyment as a roxy for firm size. They exected this variable to have a negative relation with the firm s linkage. The reason for this negative relation may be that large firms have lower linkages than small firms because of their vertical integration for attaining self-sufficiency, and to rea the benefits of economies of scale. Even the trend of globalization and out-sourcing of activities also will suort the fact that the small firms have higher linkages within the domestic economy because they are more comfortable in local markets. On the other hand, Bin (2010) calculated emloyment elasticity and included it in the artial non-linear I-O framework in order to measure the emloyment effect of changes in final demand using the 2007 I-O table for China along with China s emloyment data for the years Using a non-linear outut-emloyment relationshi, he has found that emloyment loss in China due to the US sub-rime crisis will affect the outut as well as exort of China adversely, which will in turn reduce the growth of China s economy. Thus, he concludes that this non-linear formulation and emloyment elasticity estimated thereby will reflect the economic situation in a better way, and highlight the need for aroriate olicies for sustaining the growth of Chinese 6

9 economy. Using I-O tables, Valadkhani (2003) had identified the high emloyment generating industries for the Australian economy. Measuring sectoral emloyment elasticity, he showed that the sectors like retail trade, construction, health & community services, and education will be the most emloyment generating industries in future. In the Indian context, Sarma & Ram (1989) evaluated the emloyment, income and outut linkages for India s manufacturing industries using I-O tables. Measuring forward and backward linkage in outut and emloyment generation as well as grouing the manufacturing sectors into four broad categories (agro-based industries, non-agro based final goods industries, non-agro based intermediate goods industries and caital goods industries), their study identified the sectors that have relatively higher income, outut and emloyment otentials. Their results showed that agro-based industries had otential to generate relatively more income and emloyment than the others. The above review clearly reveals that the existing studies have either seen the emloyment trends through measuring emloyment elasticity or have incororated emloyment within the I-O framework in order to identify certain other firm related features, while identification of emloyment generating industries through direct and indirect linkage effects have received less attention. Needless to say, it would be insightful for olicy makers to relate outut growth and emloyment growth through identification of key sectors in terms of their outut linkage along with emloyment linkage. This is esecially valid for a labour surlus economy like India. However, the literature review resented reveals that only limited Indian studies have included emloyment generation caacity for identification of key sectors. In addition, there is no recent study on the toic in the Indian context even though the economy has witnessed significant change in the ost-reform era. In this context, the current research makes an attemt to examine whether the highly linked sectors are also able to create emloyment or not and thus seeks to fill the ga in the existing literature by roviding a new direction for determination of key sectors. 7

10 3 Methodology and Data Sources fotolia.com lamio Inut-Outut Framework A broad Inut-Outut model has been used for addressing the aforementioned objectives. In the famous Leontief Inut-Outut model (1936), i, j th element of the Leontief inverse ((I A) -1 ) can be interreted as the total effect (both direct and indirect) on the gross outut of the i th sector when the j th sector final demand changes by one unit. Now with each outut change, there will be an associated change in emloyment. Also following the dual-sector Lewis model, we have assumed that a change in outut is linked to a fixed roortional change in emloyment, that is, the roortion of labour consumed er unit of outut remains the same irresective of the scale of roduction (Pradhan, Saluja & Singh, 2006). This constant return to scale assumtion is considered standard in resect of an Inut-Outut framework. Though this assumtion aears somewhat restrictive, it is to be noted that the entire literature on inut-outut analysis is develoed and based on this assumtion. Following this assumtion, we get the fixed emloyment coefficients for each sector, noted as follows: 8

11 E i = L i / X i (1), (i = 1, 2,, n), where Li is the emloyment in sector i, X i is the gross outut and E i is the fixed emloyment coefficient. In other words, E i is the labour requirement er unit of gross outut, X i. We can, therefore, write: L i = Ê i * X i (2), where Ê i is a diagonalised matrix formed from the vector E, whose elements are defined by equation (1). The diagonalised matrix of Ê i clearly shows the labour requirement for each sector with resect to er unit of gross outut. Now, from our conventional I-O model, we have, X = (I A) -1 F (3) Substituting the relation of X from (3) in (2), we have got the following labour equation: L = Ê (I A) -1 F = KF (4), where Ê is the diagonalised matrix formed with elements of E i, (I A) -1 is the Leontief Inverse matrix, F is the vector comrising final demand, L is the emloyment requirement, K = [k ij ], the i, j th element of K, which measures emloyment created directly and indirectly in the i th sector when the j th final demand changes by one unit. Again, i k ij gives the emloyment multilier, thus measuring the total direct and indirect emloyment created throughout the economy, when the j th sector final demand increases by one unit (Pradhan, Saluja & Singh, 2006; Bulmer-Thomas, 1982). A mere consideration of emloyment multilier ( i k ij ) does not rovide adequate information as far as selecting a key emloyment generating sector is concerned. Thus, it would be more aroriate if one could calculate aroriate indices to cature emloyment forward and backward linkages, which, in turn, hels to identify a key emloyment generating sector. These indices as we consider are as follows: Emloyment Backward Linkage (Y j ) = [(1/n) k j ] / [(1/n 2 ) j k j ], (j = 1, 2,, n), Emloyment Forward Linkage (Z i ) = [(1/n) k i ] / [(1/n 2 ) i k i ], (i = 1, 2,, n). Based on these measures, we can identify key emloyment generating sectors of an economy (Bulmer-Thomas, 1982). 9

12 Data Sources For examining the above mentioned objectives, we have used Inut-Outut tables for the years and , as given by the Central Statistical Organization (CSO), India. Since the I-O table does not rovide information on the absolute emloyment numbers of resective sectors, we have used a large samle unit level data of the 61 st National Samle Survey (NSSO) round on the emloyment-unemloyment situation for and the 64 th National Samle Survey (NSSO) round on emloyment-unemloyment and migration articulars for From the unit level data, we have used the usual rincial as well as the subsidiary status (UPSS) of the samle observations as their emloyment status. From the very definition, the usual activity status relates to the activity status of a erson during the reference eriod of 365 days receding the date of the survey. The activity status on which a erson sent relatively longer time (i.e. major time criterion) during the 365 days receding the date of the survey is considered as the usual rincial activity status of the erson. In addition, a erson whose usual rincial status was determined on the basis of the major time criterion could have ursued some economic activity for a shorter time throughout the reference year of 365 days receding the date of the survey or for a minor eriod, which is not less than 30 days, during the reference year. The status in which such economic activity was ursued was the usual subsidiary economic activity status of that erson (NSSO, ). NSSO s major focus is to estimate the number of ersons gaining emloyment in different sectors in order to arrive at an emloyment or unemloyment rate. Thus, as er NSSO, if a erson belongs to both the categories rincial and subsidiary, he / she is counted only once (according to their rincial status) in order to avoid the double counting roblem. On the other hand, in the resent exercise we are not bothered with the double counting issue since our objective is to measure total emloyment generation caabilities (both through rincial and subsidiary status) of the major sectors. Subsequently, by incororating this information in an I-O table, we can identify the key emloyment generating sectors for the Indian economy by measuring their linkage effects. Measuring Emloyment Linkage using I-O Tables In this section, we will resent the emloyment linkages (both forward and backward) of each sector using the I-O tables for the years and at aggregate as well as disaggregate levels. Further we show a comarison of the outut linkage vs. emloyment linkage to examine whether the highly linked sectors (based on outut linkage) are also highly emloyment generating or not. 10

13 Measurement of Emloyment Linkages at the aggregate level Three broad sectors of the Indian economy (rimary, secondary and tertiary) have been considered at the aggregate level analysis. Further, the secondary sector is sub-divided into manufacturing and non-manufacturing sectors. It is to be noted that the non-manufacturing sector considers three sectors, mining & quarrying, electricity & water suly and construction, while the manufacturing sector includes other industries from food, beverages & tobacco to other miscellaneous manufacturing. With this secification, the emloyment forward and backward linkages have been comuted at the aggregate level for the years and Further, in order to comare the emloyment linkages across different time eriods, we have deflated I-O tables with resect to a base year, i.e, In addition, for measuring the emloyment linkage, we have taken the emloyment figures from the 61 st and 64 th NSSO survey on emloyment-unemloyment articulars, which requires a match between the sectors in the I-O table and that resented in the NSSO emloyment-unemloyment survey. Table A in the aendix shows such concordance between the I-O table and the NSSO survey as er NIC (National Industrial Classification)-04 codes. Finally, following the aforementioned methodology, we have calculated emloyment forward and backward linkages for broad sectors (as aggregated from 130 sectors in the I-O table) through measuring the emloyment multilier. Table 1 deicts the emloyment linkages of the broad sectors. Table 1: Emloyment Forward & Backward Linkage Coefficient for the years & (for broad sectors) SC Sectors in the I-O Table ES EB EF ES EB EF 1 Agriculture Manufacturing Non-manufacturing Services SC = Sector Code ES = Emloyment share EB = Emloyment Backward Linkage Co-efficient EF = Emloyment Forward Linkage Co-efficient Source: Author s estimation by using I-O table for (CSO, 2008) & (CSO, 2012). 11

14 Table 1 deicts the emloyment share along with emloyment forward and backward linkages for four broad sectors of the Indian economy for the years and The emloyment share shows how much in-house emloyment is generated by the sectors out of the total emloyment in the economy, while emloyment backward and forward linkage coefficients reresent the amount of emloyment generated by the sector both within itself as well as in other sectors through their linkage effect. Thus, Table 1 shows that the emloyment share of the agriculture sector is decreasing over time, but it is evident that there is an increase in emloyment linkages. This imlies that although its otentiality to generate in-house emloyment is decreasing, its contribution to generating emloyment outside the sectors through a linkage effect is increasing. Keeing in mind the existing argument of jobless growth (Mehrotra et al., 2012, 2013), this study also reveals that for the manufacturing sector, the share of generating in-house emloyment as well as its contribution to generating emloyment outside the sectors decreased between and In shar contrast, the non-manufacturing and service sector showed an increasing trend in the share of emloyment, however, it also shows a decreasing trend in linkage effect. The non-manufacturing sector s increase in in-house emloyment share may be due to the booming investment in the construction sector in recent years (Mehrotra et al., 2012, 2013). This is a generalized icture of emloyment for the broad sectors of the Indian economy. However, heterogeneity may exist in terms of linkage effects across different industry grous within manufacturing, non-manufacturing and services. After having resented the emloyment linkages for the aggregate sectors, we move on to comute the emloyment linkages of the sectors at the disaggregate level. Using I-O tables, we have combined similar sectors from 130 sectors before arriving at 21 sectors in order to identify the key emloyment generating sectors in a somewhat aggregate manner. It is to be noted here that we resent the emloyment forward and backward linkages for 21 sectors, and then carry out a comarison between outut linkage and emloyment linkage with a view to examining whether the highly linked sectors (based on outut linkage) ossess the otential to generate emloyment or not. Table B in the aendix shows the sectors merged together to resent a consolidated icture. Table 2 resents the forward as well as backward linkages for emloyment with resect to 21 sectors for the years and

15 Measurement of Emloyment Linkages at the disaggregate level Table 2: Emloyment Forward & Backward Linkage Coefficients for the years & (for 21 major sectors) SC Sectors in the I-O Table ES EB EF ES EB EF 1 Agriculture Mining and quarrying Food, beverages & tobacco Textiles Wood & wood roducts Paer, aer roducts Printing & ublishing Leather & lastic roducts Petroleum roducts Chemicals Non-metallic mineral roducts Metals Machinery Transort equiment & arts Other miscellaneous manufacturing Construction Electricity & water suly Transort, storage & communication Trade, hotels & restaurant Financing, insurance, real estate & business services Other comm., social & ersonal services SC = Sector Code ES = Emloyment share EB = Emloyment Backward Linkage Co-efficient EF = Emloyment Forward Linkage Co-efficient Source: Authors estimation by using I-O tables for (CSO, 2008) & (CSO, 2012). 13

16 Table 2 clearly shows the emloyment share along with emloyment forward and backward linkages across 21 sectors for the time eriods and It is to be noted that the sectors having more than unitary emloyment backward linkage are caable of creating above average (more than one unit) emloyment in other sectors, when final demand increases by unity. In contrast, the sectors having more than unitary emloyment forward linkage are caable of creating above average emloyment within that sector, when final demand for all the sectors increases by unity. Thus, an emloyment linkage analysis catures additional emloyment generation caabilities not catured through mere emloyment share, since emloyment share reresents the creation of only direct emloyment within the sectors, while a linkage analysis covers the creation of emloyment not only within the sectors, but also in other sectors. Focusing on Table 2, as exected we have found that the agriculture sector has the highest emloyment forward and backward linkages; thus it has the highest emloyment generation caacity in resect to both within the sector as well as outside the sector with a change in the final demand. With resect to an in-house emloyment share of the agriculture sector, we have found the emloyment share of agriculture decreased in as comared to although its emloyment linkage has increased over time, which again illustrates that considering only the emloyment share does not reflect the actual otential of a sector in terms of generating emloyment. Within the manufacturing sectors, textiles, wood & wood roducts also ossess high emloyment forward as well as backward linkages. Looking at the emloyment share of those sectors, it has been found that wood & wood roducts generates a lesser amount of in-house emloyment (in terms of absolute emloyment share), while it has the caacity to affect other sectors by way of creating emloyment through its high emloyment linkage. Even the aer, aer roducts sector creates a lesser amount of emloyment within it, while it has moderate emloyment linkage with others. Mining & quarrying holds the second osition in terms of emloyment forward linkage, thus creating more emloyment within itself in order to serve the demand from other sectors. In contrast, the agro based industry food, beverages & tobacco occuies the second lace with its emloyment backward linkage, thus creating roortionately more emloyment in other sectors through creating more demand. Thus, emloyment linkage rovides more insight into the sectors emloyment generation rather than their emloyment share. 14

17 Now, in order to identify the key emloyment generating sectors, it is useful to rank them in terms of their high emloyment forward as well as backward linkage coefficients. According to Diamond (1975), a key emloyment sector is the one which has high emloyment forward as well as backward linkage; more recisely, which has both emloyment forward as well as backward linkage coefficients greater than unity. Based on this criterion, we identify the key emloyment generating sectors for and (see Table 3). Table 3: Key Emloyment Generating Sectors for the years & (for 21 major sectors) Sectors in the I-O Table EB EF EB EF Agriculture Wood & wood roducts Textiles EB = Emloyment Backward Linkage Co-efficient EF = Emloyment Forward Linkage Co-efficient Source: Author s estimation by using I-O tables for (CSO, 2008) & (CSO, 2012). From Table 3, we can clearly identify the key emloyment generating sectors with both forward and backward linkages greater than unity. However, Diamond s secification imoses a strict restriction of both forward and backward linkages in that they have to be greater than unity. If we marginally relax this strict assumtion to consider the sectors which have at least one of the linkage coefficients (either emloyment forward or backward) greater than unity, we can arrive at a number of sectors with moderately high linkage effects. Table 4 lists those sectors. 15

18 Table 4: Key Emloyment Generating Sectors with relaxed norms* for & (for 21 major sectors) Sectors in the I-O Table EB EF EB EF Mining and quarrying Food, beverages & tobacco Textiles Paer, aer roducts Non-metallic mineral roducts Other miscellaneous manufacturing Trade, hotels & restaurants EB = Emloyment Backward Linkage Co-efficient EF = Emloyment Forward Linkage Co-efficient *Either forward or backward linkage coefficients are greater than unity. Source: Author s estimation by using I-O tables for (CSO, 2008) & (CSO, 2012). Finally, we attemt to examine whether the sectors that have been identified in terms of their high outut linkage also dislay a similar erformance in terms of generating emloyment or not. This necessitates calculation of outut forward and backward linkages of the sectors as well. This was done using the standard I-O technique 2. In Table 5 we resent both outut and emloyment linkage coefficients which aid in resenting a comarative icture. 2 Here, we have used the demand driven standard Leontief model for measuring the backward linkage coefficients, while the suly driven Ghosh model (1958) is used for measuring the forward linkage coefficients. For both these models, we have broadly used the Chenery-Watanabe method (1958) and the Rasmussen method (1956) in order to find the direct as well as indirect forward and backward linkage coefficients using the technical coefficient matrix (A), allocation coefficient matrix (B) and Leontief inverse matrix (I-A) -1, (I-B)

19 Table 5: Comarison between Emloyment & Outut Linkage coefficients for & (for 21 major sectors) Emloyment Linkage for Outut Linkage for Emloyment Linkage for Outut Linkage for Sectors EB EF EB EF EB EF EB EF 1. Agriculture Mining & quarrying Food, beverages & tobacco Textiles Wood & wood roducts, furniture & fixtures Paer, aer roducts Printing & ublishing Leather & lastic roducts Petroleum roducts Chemicals Non-metallic mineral roducts Metals Machinery Transort equiment & arts Other miscellaneous manufacturing Construction Electricity & water suly Transort, storage & communication Trade, hotels & restaurant Financing, insurance, real estate & business services 21. Other comm., social & ersonal services EB = Emloyment Backward Linkage Co-efficient EF = Emloyment Forward Linkage Co-efficient Source: Author s estimation by using I-O tables for (CSO, 2008) & (CSO, 2012). 17

20 Table 5 shows that divergence exists among the sectors in terms of their emloyment and outut generating caabilities. More recisely, it has been found that the sectors like agriculture, textiles, and wood & wood roducts are key emloyment generating sectors for both years, while aer & aer roducts, etroleum roducts, chemicals, and metals, are the sectors with a otential to influence others in terms of their high outut linkage. In order to get a clearer icture, we resent the sectors classified in terms of outut as well as emloyment linkage coefficients for the years and in a 2*2 format in Tables 6 and 7, resectively. Here we have incororated those sectors in the high category which have both forward and backward linkages greater than unity. The rests are resented in the low category. Table 6: Sectors classified using Outut & Emloyment Linkage Coefficients for (following the strict assumtion for high linkage) OUTPUT LINKAGE EMPLOYMENT LINKAGE HIGH LOW HIGH Nil Paer & aer roducts, etroleum roducts, chemicals, metals, other miscellaneous manufacturing, electricity & water suly LOW Agriculture, wood & wood roducts Mining & quarrying, food, beverages & tobacco, textiles, rinting & ublishing, leather & lastic roducts, non-metallic mineral roducts, machinery, transort equiment & arts, construction, transort, storage & communication, trade, hotels & restaurants, financing, insurance, real estate & business, other community, social & ersonal services Source: Author s estimation by using I-O table for (CSO, 2008). Table 7: Sectors classified using Outut & Emloyment Linkage Coefficients for (following the strict assumtion for high linkage) OUTPUT LINKAGE EMPLOYMENT LINKAGE HIGH LOW HIGH Nil Paer & aer roducts, chemicals, metals LOW Agriculture, wood & wood roducts Mining & quarrying, food, beverages & tobacco, rinting & ublishing, leather & lastic roducts, etroleum roducts, non-metallic mineral roducts, machinery, transort equiment & arts, other miscellaneous manufacturing, construction, electricity & water suly, transort, storage & communication, trade, hotels & restaurants, financing, insurance, real estate & business, other community, social & ersonal services Source: Author s estimation by using I-O table for (CSO, 2012). 18

21 Suorting the results from Table 5, Tables 6 and 7 also show that there is no sector that ossesses both high emloyment as well as outut linkage, although sectors like chemicals and metals exhibit a high outut linkage with low emloyment linkage for both years. In addition, agriculture, textiles, and wood & wood roducts show a high emloyment linkage with low outut linkage. Thus, focusing only on the outut oriented highly linked sectors may not fulfill certain imortant goals of the economy such as emloyment generation. As mentioned, the above classification is based on a strict assumtion that both forward and backward outut and emloyment linkage coefficients need to be greater than unity for a sector to be considered as highly linked outut and emloyment generating. This results in no matches. As a second ste it is useful to relax this strict assumtion in the manner described above and examine whether one can identify certain sectors which are imortant from the oint of view of generating both outut and emloyment through direct lus linkage effects. This exercise is resented in Tables 8 and 9 for and , resectively. Table 8: Classification of Sectors based on Outut & Emloyment Linkage Coefficients for (Relaxing the strict assumtion) EMPLOYMENT LINKAGE COEFFICIENTS Both greater than unity One greater than unity (either forward or backward) Both greater than 0.5 but less than one Either one or both less than OUTPUT LINKAGE COEFFI- CIENTS Both greater than unity 1 Other mis cel laneous manufacturing Paer & aer roducts Petroleum ro ducts, chemicals, metals, electricity & water suly One greater than unity (either forward or backward) 2 Wood & wood roducts Mining & quarry ing, food, beverages & tobacco, textiles, nonmetallic mineral roducts Printing & ublishing, transort, storage & communication Leather & lastic roducts, machinery, transort equiment & arts Both greater than 0.5 but less than one 3 Agriculture Trade, hotels & restaurants Construction Financing, insurance, real estate & business Either one or both less than Other community, social & erso - nal services Source: Author s estimation by using I-O table for (CSO, 2008). 19

22 Table 9: Classification of Sectors based on Outut & Emloyment Linkage Coefficients for (Relaxing the strict assumtion) EMPLOYMENT LINKAGE COEFFICIENTS Both greater than unity One greater than unity (either forward or backward) Both greater than 0.5 but less than one Either one or both less than OUTPUT LINKAGE COEFFI- CIENTS Both greater than unity One greater than unity (either forward or backward) 1 Paer & aer roducts 2 textiles Mining & quarrying, food, beverages & tobacco, nonmetallic mineral roducts Printing & ublishing, Other miscellaneous manu - facturing, construction chemicals, metals Petroleum roducts, Leather & lastic roducts, machinery, transort equiment & arts, electricity & water suly Both greater than 0.5 but less than one 3 Agriculture, wood & wood roducts Trade, hotels & restaurants Transort, storage & communication, other community, social & erso - nal services Financing, insurance, real estate & business Either one or both less than Source: Author s estimation by using I-O table for in CSO (2012). Tables 8 and 9 deict four sets of sectors from the oint of view of outut and emloyment generation caabilities. For both the time eriods, the sectors located at cell (1,2) (i.e., 1 st row, 2 nd column), cell (2,1) (i.e., 2 nd row, 1 st column) and cell (2,2) (i.e., 2 nd row, 2 nd column) are having reasonably good emloyment as well as outut linkage. Following this, for both time eriods, the sectors like mining & quarrying, food, beverages & tobacco, textiles, and non-metallic mineral roducts can be identified as the sectors caable of generating outut growth and sufficient emloyment oortunities. On the other hand, the second set of sectors like rinting & ublishing, construction, trade, hotels & restaurant, and transort, storage & communication (located at the cells 2,3; 3,2; 3,3) have moderate outut and emloyment linkage for both years, imlying less imortance as comared to the first set of sectors. Further, some sectors like chemicals, metals, etroleum roducts, and leather & lastic roducts (located at the cells 1,4 and 2,4) are not so imortant in terms of emloyment generation, although they have high outut linkage. Finally, the fourth set of sectors like financing, insurance, real estate & business services, and other community, social & ersonal services (located at the cells 3,4 and 4,3) have low outut as well as emloyment linkages. 20

23 4 Conclusion In this aer an attemt has been made to identify certain key sectors in terms of their emloyment generating caabilities. While highlighting such sectors, linkage effects are also taken into consideration. It has been found that sectors like agriculture, textiles, and wood & wood roducts are key emloyment generating sectors ossessing high emloyment forward as well as backward linkages for both the time eriods under consideration, and , although they have a low outut linkage. In contrast, etroleum roducts, chemicals, and metals, are the sectors that affect others in terms of their outut linkage for both time eriods, although they have a low emloyment linkage. In addition, discreancies also exist between the direct emloyment share and emloyment linkage of the sectors. For examle, the emloyment share of the sectors like agriculture, wood & wood roducts, and aer & aer roducts has decreased from to , even though their emloyment linkage has increased over time. Thus comutation of emloyment linkage rovides additional insightful information regarding the emloyment creation caabilities of a sector which only the share of emloyment is unable to fully cature. Policy makers, therefore, should focus on those sectors that are highly linked to others in terms of their outut linkage along with their caacity to generate direct and indirect emloyment. Marginally relaxing the strict assumtion of high linkage, our results have identified certain sectors like mining & quarrying, food, beverages & tobacco, textiles, and non-metallic mineral roducts which have a reasonably good outut as well as emloyment linkage for both the time eriods. Thus, olicy measures for boosting growth in some of these sectors can go a long way toward fostering inclusive growth in the economy. 21

24 5 References Bin, W. (2010): A Non-Linear Inut-Outut Model for Measuring the Emloyment Effect of Changes in Final Demand: An Aroach Based on the Emloyment Elasticity, 18th International Inut Outut Conference, downloaded from, accessed on Bulmer-Thomas, V. (1982): Inut-Outut Analysis in Develoing Countries, Sources, Methods and Alications, John Wiley and& Sons Ltd., New York. Central Statistical Organization (CSO): Government of India (2008), Inut-Outut Transaction table for Central Statistical Organization (CSO): Government of India (2012), Inut-Outut Transaction table for Chenery, H. B. and T. Watanabe (1958): International Comarisons of Structure of Production, Econometrica, 26(4): Ghosh, A. (1958): Inut-Outut Aroach in an Allocation System, Economica, 25(97): Gorg, H. and F. Ruane (2000): An Analysis of Backward Linkages in the Irish Electronics Sector, The Economic and& Social Review, 31(3): Labour Bureau, Ministry of Labour and Emloyment, Government of India (2010): Reort on Emloyment and Unemloyment Survey , downloaded from: htt://labourbureau.nic.in/final_reort_em_unem_2009_10.df, accessed on Leontief, W. W. (1936): Quantitative Inut and Outut Relations in the Economic System of the United States, Review of Economic Statistics, 18(3): Mehrotra, S., A. Gandhi, B. K. Sahoo and P. Saha (2012): Creating Emloyment in the Twelfth Five-Year Plan, Economic and& Political Weekly, 47(19): Mehrotra, S., A. Gandhi, P. Saha and B. K. Sahoo (2013):Turnaround in India s Emloyment Story: Silver Lining Amidst Joblessness and Informalisation?, Economic and& Political Weekly, 48(35): National Samle Survey Office, Ministry of Statistics & Programme Imlementation, Government of India (2010): Emloyment and Unemloyment Situation in India , NSS Reort No. 531(64/10.2/1). National Samle Survey Organization, Ministry of Statistics & Programme Imlementation, Government of India (2006): Emloyment and Unemloyment Situation in India , NSS Reort No. 515(61/10/1). Paola, T. S. and P. P. Sahu (2012): Growth and Structure of Emloyment in India: Long-Term and Post-Reform Performance and the Emerging Challenge, downloaded from: htt://isidev.nic.in/df/icssr_tsp_pps.df, accessed on Pradhan, B. K., M. R. Saluja and S. K. Singh (2006): Social Accounting Matrix for India: Concets, Construction and Alications, Sage ublications, New Delhi. Raa, T. T. (2005):The Economics of Inut-Outut Analysis, Cambridge University Press, Cambridge and New York. Rasmussen, P. N. (1956): Studies in Intersectoral Relations, North-Holland Publishing, Amsterdam. Sarma, A. and K. Ram (1989): Income, Outut and Emloyment Linkages and Imort Intensities of Manufacturing Industries in India, The Journal of Develoment Studies, 25(2): Valadkhani, A. (2003): Using Inut-Outut Analysis to Identify Australia s High Emloyment Generating Industries, Australian Bulletin of Labour, 29(3):

25 6 Aendices Table A: Concordance between I-O & NIC-2004 Sector Code 21 Sectors in the Consolidated I O Table I O Code of Sectors to Merge NIC 04 Code 1 Agriculture Mining and quarrying Food, beverages & tobacco Textiles Wood & wood roducts Paer, aer roducts Printing & ublishing Leather & lastic roducts Petroleum roducts Chemicals Non metallic mineral roducts Metals Machinery Transort equiment & arts Other miscellaneous manu Construction Electricity & water suly Transort, storage & communication Trade, hotels & restaurant Financing, insurance, real estate & business services , 123, Other comm., social & ersonal services , , Source: Author s estimation by using the I-O table for in CSO (2012) & 64 th round of NSSO emloyment-unemloyment survey ( ). 23

26 Table B: Merging Sectors Sector Code 21 Sectors in the Consolidated I O Table I O Code of Sectors to Merge 1 Agriculture Mining and quarrying Food, beverages & tobacco Textiles Wood & wood roducts Paer, aer roducts 57 7 Printing & ublishing 58 8 Leather & lastic roducts Petroleum roducts Chemicals Non metallic mineral roducts Metals Machinery Transort equiment & arts Other miscellaneous manufacturing Construction Electricity & water suly Transort, storage & communication Trade, hotels & restaurant Financing, insurance, real estate & business services , 123, Other comm., social & ersonal services , , Source: Author s estimation by using the I-O tables for (CSO, 2008) & (CSO, 2012). 24

27 ICDD Working Paer Series Vol. 1: Vol. 2: Vol. 3: Vol. 4: Vol. 5: Vol. 6: Vol. 7: Vol. 8: Vol. 9: Webster, Edward: Work and Economic Security in the 21st century. What Can We Learn from Ela Bhatt?, 17 ages Hagmann, Jonas: Oortunities and Constraints of Peri-urban Buffalo and Dairy Cattle Systems in Faisalabad, Punjab, Pakistan, 48 ages Marchetti, Sabrina:Together? On the Not-so-easy Relationshi between Italian Labour Organisations and Migrant Domestic Workers Grous, 23 ages Sinaga, Hariati / Scherrer, Christoh: Core Labor Rights: Cometitive Pressures and Non-Comliance, 29 ages Burchardt, Hans-Jürgen / Weinmann, Nico: Social Inequality and Social Policy outside the OECD: A New Research Persective on Latin America, 39 ages Beck, Stefan: Sozial verantwortliche Beschaffung von Informationstechnik. Socially Resonsible Public Procurement of Information Technology, ISBN , 40 ages Aufderheide, Mareike / Voigts, Clemens / Hülsebusch, Christian / Kaufmann, Brigitte: Decent Work? How Self-emloyed Pastoralists and Emloyed Herders on Ranches Perceive their Working Conditions, ISBN , 28 ages Bhattacharjee, Manojit / Rajeev, Meenakshi: Credit Exclusion of the Poor: A Study of Cultivator Households in India, ISBN , 22 ages Younas, Muhammad: The Dairy Value Chain: A Promoter of Develoment and Emloyment in Pakistan, ISBN , 22 ages Vol. 10: Erbach, Juliane: The Decency of Women s Working Conditions in Peri-urban Buffalo Production Systems in the District Faisalabad, Punjab, Pakistan, kassel university ress, ISBN , 45 ages Vol. 11: Schützhofer, Timm B.: Can Decent Work and Exort Oriented Growth Strategies Go Together? Lessons from Nicaragua s Exort Processing Zones, kassel university ress, ISBN , 52 ages Vol. 12: Bhattacharya, Tulika / Rajeev, Meenakshi: Identifying Emloyment Creating Sectors in India: An Analysis of Inut-Outut Linkages, kassel university ress, ISBN , 28 ages Vol. 13: Withanachchi, Sisira Saddhamangala / Houdret, Annabelle / Nergui, Soninkhishig / Ejarque i Gonzalez, Elisabet / Tsogtbayar, Ankhbold / Ploeger, Angelika: (Re) configuration of Water Resources Management in Mongolia: A Critical Geoolitical Analysis, kassel university ress, ISBN

28 The Global ICDD Network International Labour Organization Germany University of Kassel Civil Society Partners Mexico Universidad Autónoma de Yucatán Pakistan University of Agriculture Faisalabad Brazil Universidade Estaduale de Caminas Ghana University of Cae Coast Kenya Egerton University India Tata Institute of Social Sciences South Africa University of the Witwatersrand International Center for Develoment and Decent Work University of Kassel, Germany Phone: + 49 (0) felmeden@icdd.uni-kassel.de uuu