5.4.1 The Central Statistical Organisation (CSO) is responsible for the compilation and publication of the Index of Industrial Production (IIP) since 1950. The IIP is compiled as a simple weighted arithmetic mean of production relatives by using Laspeyre’s formula. The IIP is a quantum index, the production of items being expressed in physical terms. However, the unit of reporting in respect of certain items like machinery, machine tools, ship building, etc. is in value terms. The monthly figure of production value in such cases is first deflated by the Wholesale Price Index (WPI) of the corresponding categories, released by the Office of the Economic Adviser, Ministry of Industry. The scope of the IIP as recommended by the United Nations Statistical Office (UNSO) includes mining, manufacturing, construction, electricity, gas and water supply. But due to constraints of data availability, the IIP compiled in India has excluded construction, gas and water supply sectors.
Current Status of IIP (Base 1993-94)
5.4.2 The Commission discussed the following aspects of the IIP in detail:
Base Year
Scope
Coverage of items
Weighting diagram
Sources of data
Quick Estimates and Revisions
Base Year
5.4.3 As the structure of the Industrial Sector changes over time, it is necessary to revise the base period and the weighting diagram of the IIP periodically so as to measure the real growth in the industrial sector. The CSO made such periodical revisions by reviewing the coverage of items and industries and by improving, as far as practicable, the technique used in the construction of IIP and by shifting the base to a recent period. When the compilation of the index commenced in India, the base year adopted was 1946, and this was revised successively to 1951, 1956, 1960, 1970, 1980-81 and 1993-94. The base year, number of items, month of release, number of 2-digit industry groups and the period for which the indices were compiled are given in Annexe 5.19.
5.4.4 The United Nations has recommended that the base period of the index number should be revised quinquennially. The practice of revising the index five-yearly was followed till 1960. Thereafter, two revisions were made after a gap of about ten years each. The last revision of the base period (1980-81) was made after a lapse of 13 years. For the purpose of revising the IIP, the CSO generally constituted expert groups or committees to make suitable recommendations with regard to choice of the base year and necessary modifications in the weighting diagram. The revision of the 1980-81 series of the IIP was guided by a Technical Advisory Committee (TAC) constituted by the CSO in June 1995 to advise on the Compilation of Comparable State IIPs, the corresponding Composite All-India IIP and All-India IIP. The TAC submitted its report in June 1998. The main recommendations of the Committee along with their present status are set out in Annexe 5.20.
Scope
Sector
No. of Items
Mining and Quarrying
61
64
Manufacturing
290
478
Electricity
1
1
Total
352
543*
5.4.5 The scope of the index has been confined to the mining, manufacturing and electricity sectors, and does not cover gas, water supply and construction. The number of items included in the IIP with 1980-81 and 1993-94 as base years is given in Table 5.1:
Table 5.1: Sector-wise number of items in 1980-81 and 1993-94 series
Coverage of Items
5.4.6 In the revised IIP, the approach adopted for selection of items for inclusion in the item basket is as follows:
Each item should generally account for at least Rs. 80 crore of gross value of output at the item level and Rs. 20 crore of gross value added at the ultimate (4-digit) level of NIC 1987. The criteria have, however, been applied with flexibility in the case of industry groups which were not represented by a sufficient number of items;
In order to improve the representativeness of the item basket at the 2-digit level of NIC, some of the important items of the IIP series with base 1980-81 were also included.
5.4.7 The item basket so identified captured about 80 per cent of the output of the manufacturing sector. It was finalised after extensive discussions with source agencies for production data especially the Department of Industrial Policy and Promotion (DIPP), Textile Commissioner, Directorate of Vanaspati and Development Commissioner for Iron & Steel (DCI&S), keeping in view the distinctive character of the items and the availability of a regular flow of monthly production data. Also, to capture at least 60 per cent of the output at the 2-digit level for improving the representativeness of the items, the criterion for item selection was relaxed and some additional items of 1980-81 series were included in the basket.
5.4.8 In this section of the Report, the unregistered manufacturing sector is defined as those units which are not registered under the Factories Act, 1948. On account of non-availability of regular monthly production data, the unregistered sector has been represented by the units belonging to the Small Scale Sector only. Since the Office of the Development Commissioner, Small-Scale Industries (DCSSI), could not provide data on any of the additional items identified for the new series, all the 18 items of the Small Scale Sector included in the 1980-81 series were retained in the revised series.
Weighting Diagram
NIC Section
Description
Weight
1980-81
1993-94
1
Mining and Quarrying
114.64
104.73
2 and 3
Manufacturing
771.07
793.58
Registered
505.13
Unregistered
288.45
4
Electricity
114.29
101.69
Total
1000.00
1000.00
5.4.9 The weights of the mining, manufacturing and electricity sectors (1-digit level of NIC) have been allocated on the basis of gross value added for 1993-94 as published in the National Accounts Statistics. For the first time, the weighting diagram with base 1993-94 took into account the contribution of the unregistered manufacturing sector along with that of the registered sector. For estimating the 1993-94 gross value added for the unregistered sector, the data from the Follow-up Enterprise Surveys of the Economic Census namely, Directory Manufacturing Establishments (DME), Non-Directory Manufacturing Establishments (NDME) and Own-Account Manufacturing Enterprises (OAME), for the years 1989-90 and 1994-95 have been used. After interpolating the GVA at the 2-digit level for the year 1993-94, further allocation at the 3-digit and 4-digit levels of NIC has been done using the ratios of 1994-95 surveys, it being in close proximity to 1993-94. The total contribution at 2, 3 and 4-digit level of NIC has been arrived at by using registered sector data along with data for the unregistered sector for the corresponding groups and sub-groups. Further the allocation of weights to the items within the 4-digit industry groups has been done using Value of Output as available from ASI 1993-94.
5.4.10 The recommendation of the TAC on the issue of using Gross Value of Output instead of Gross Value Added as a criterion for allocation of weights for the compilation of the revised series of IIP was referred to a Special Committee. Following the use of GVA, as recommended by the Special Committee, the distribution of weights at the sectoral level for the IIP series with base year 1993-94 is given in Table 5.2.
Table 5. 2: Comparison of Weights at Sectoral level in 1980-81 and 1993-94 series
Sources of Data
5.4.11 For the 1980-81 series, the CSO used to get monthly production data from as many as eighteen source agencies. The source agencies, in turn, collect data from the production units spread all over the country. For the revised series with base 1993-94, the same set of source agencies has been retained except for Railways, for which the data are now being supplied in a consolidated form by the Railway Board, instead of by four agencies as was done earlier. This has reduced the number of source agencies to fifteen. In terms of the number of items covered, the largest source is the DIPP, which supplies data on as many as 213 out of 285 groups of items in the manufacturing sector, constituting more than 52 per cent in terms of weight of All-India IIP. The index relating to Mining and Quarrying sector is being supplied by the Indian Bureau of Mines, Nagpur, which is combined with manufacturing and electricity indices compiled by the CSO to arrive at the General Index of Industrial Production. The data on the electricity sector is furnished by the Central Electricity Authority. The source-wise number of items proposed by the CSO, finalised after discussion with source agencies, item groups, corresponding weights and the number of units in the frame is given at Annexe 5.21.
Quick Estimates and Revisions
5.4.12 By the Special Data Dissemination Standards (SDDS) of the IMF, the index for any reference month is to be released within six weeks from the end of that month. In order to achieve this norm for the IIP, all the fifteen source agencies are required to furnish data to the CSO within four to five weeks from the close of month. As all the production units do not furnish data within the stipulated time, the source agencies have to resort to estimation for the non-responding units and thus supply provisional data to the CSO. These provisional data are used for computing the quick index for a given month, which is released by the CSO in six weeks. On the basis of revised production data received later from the source agencies, this index is revised subsequently in the next month and finally, two months thereafter, along with the quick index.
Deficiencies
5.4.13 The current IIP has the following specific problems:
Non-response: The CSO expects to have at least 60 per cent response in terms of production for release of the Quick Estimate and 80 per cent response at the time of final revision of the IIP. The response rate is worked out on the basis of data supplied by the source agencies, which collect the data primarily through mail enquiry method. The response position is quite good for some of the source agencies, but not satisfactory in respect of others, particularly DCSSI, Directorate of Vanaspati and DIPP, which is a major source of data for the IIP. The response rates in terms of weighted item-wise production of responding units for the items on which the data was furnished by DCSSI, Directorate of Vanaspati and DIPP for the months of January, February and March 2000 were in the range of 9 to 16 per cent, 19 to 43 per cent and 37 to 54 per cent, respectively, at the time of quick estimates. The corresponding percentages at the time of final estimates were 56 to 64 per cent, 76 to 83 per cent and 62 to 66 per cent (see Annexe 5.22). Annexe 5.23 indicates distribution of the number of items by percentage of variation in the production reported at the time of quick and final estimates for the three months from January to March 2000. In the case of 73 items, production reported by DIPP varied by five per cent or more, on an average, over the three months. The number of such items were 10, 6, 3 and 2 for DCI&S, DCSSI, Textile Commissioner and Directorate of Vanaspati, respectively.
Repetition of Production data: The total production of an item depends on the number of active units registered with the source agencies for that specific item. It is quite unlikely that the production of an item, which itself depends on a number of endogenous and exogenous factors, will get repeated from one month to the next and that, too, at the national level. Such repetition of production figures for a particular item exposes the weaknesses in the system of collection of data along with inadequacy of the estimation procedure being used by some source agencies (see Annexe 5.24).
Nil and abnormally high or low Production: It is also quite unlikely that true production of a particular item at the national level would be ‘Nil’. But such data can appear when there are only one or two production units at the national level and they have not reported the production data. Again, it is not uncommon that abnormally high or low production figures are reported for an item compared to the previous months (see Annexe 5.25). On account of extreme fluctuations in the production data being reported by the producer to DIPP, the CSO has recently deleted 4 items namely, photosensitised paper, engines, chassis (assembly) for heavy commercial vehicles (bus, truck) and radio receivers from the item basket with effect from April 1998.
Very few units: Data on about 45 items reported by the DIPP are based on five or less units (see Annexe 5.26). For three items, there is only one production unit in the DIPP frame. Such a situation coupled with non-response may lead to extreme variation in estimates of production, thereby affecting the index value considerably.
Monitoring of Frame: Some shortcomings have been noticed in the frame of the units being maintained by the DCSSI, Office of Textile Commissioner, Directorate of Vanaspati and the DIPP. The frame of the units now in use at DIPP for reporting industrial production was earlier maintained by the then Directorate-General of Technical Development (DGTD). This frame covers all industrial establishments within the purview of the Industrial Development and Regulation (IDR) Act. However, it is observed that after liberalisation and the transfer of work to DIPP, the frame has become incomplete in respect of coverage and inclusion of new units. In the pre-liberalisation days, the DGTD was in a position to update the frame on the basis of licenses issued. Now, with de-licensing of certain sectors, the DIPP remains unaware of the setting up of new units in these sectors and can suitably update the frame only when new units inform the DIPP of the commencement of production. Under these circumstances, the frame is very likely to be incomplete. For the same reason, it would continue to include units, which have gone out of existence.
Inadequate Representation: The item basket for the All-India IIP with base 1993-94 was selected using detailed results of ASI, 1993-94. The provisional item basket contained 674 items capturing about 80 per cent of the total output of the registered manufacturing sector. The criteria for item selection were relaxed, wherever necessary, to ensure that the provisionally selected items in all the 2-digit industry groups capture at least 60 per cent of the value of output of the particular group. The overriding criterion for finalisation of the item basket was the regular flow of monthly production data from the source agencies. The final item basket contains only 478 items. The 196 items on which data could not be furnished by the source agencies mainly confined to the four agencies: DIPP (80 per cent), Development Commissioner for Iron & Steel (7 per cent), Directorate of Vanaspati (5 per cent) and Office of the Textile Commissioner (4 per cent). The shrinkage in the item basket on account of non-availability of data on nearly 30per cent of the items, contributing about 22 per cent of the total output of the manufacturing sector, has affected the representativeness of the index. While such a compromise has been considered necessary for the feasibility, as one may call it, of the monthly index, there is no denying the fact that the ability of the IIP to measure industrial growth has been severely affected. The problem will persist unless corrective steps are taken by the source agencies.
Representation of the Unregistered Sector: While the contribution of the unregistered sector was included in the weighting diagram, this sector was not adequately represented in deriving the production relatives as regular monthly production data remain hard to come by. Only 18 items representing the unregistered sector, which were included in the IIP in the 1980-81 series, continue to be included in the current IIP. Thus, the IIP has failed to adequately represent the unregistered sector in the overall index.
Weaknesses in the Data Collection Mechanisms of the Source Agencies: DCSSI, Directorate of Vanaspati, DIPP and the Office of Textile Commissioner do not have an appropriate data collection mechanism. This has resulted in a poor response from the manufacturing units included in their frame for monthly reporting of production data. Most of the source agencies are not geared for collection of data on all items that are being produced by the units within their purview. Ideally, the source agencies need to monitor the performances of all units under their jurisdiction and provide information on the complete set of items identified during the base year revision. But they have not been able to furnish information on most of the new items. Further, it appears that the source agencies do not have an adequate system for validating data before sending the same to the CSO.
5.4.14 The Working Group on Commerce, Industry and the Corporate Sector constituted in the Modernisation of Statistical System in India under the Chairmanship of Dr. Arun Ghosh, after careful consideration of all the issues pertaining to the IIP, felt that it would be useful if, alongside the index of production in the medium and large-scale sector, a separate index for the Small Scale Industrial Sector is computed. The Working Group, therefore, recommended compilation of:
An alternative IIP based on large manufacturing units (units having 200 or more workers); and
A separate quarterly index for the Small-Scale Sector.
5.4.15 According to the Working Group, the Index of Production in large and medium industries and the present IIP are to be run in parallel at least for a period of one year or till the revision of the base for the IIP is taken up. The CSO should also solicit the cooperation of the industry associations in the process of data collection and cross-validation. The Working Group was, however, not in favour of publishing two parallel indices, but felt a change-over could be made after one year of experimentation. The index so released should specifically clarify that it mainly pertains to the output of large and medium industries.
Use of Other Administrative Data
5.4.16 The Commission discussed the feasibility of using the Central Excise data for compilation of IIP. Almost all manufacturers are required to file the RT–12 return containing production data within 5 days from the close of a month to the Range Offices of Central Excise. The RT–12 return is being submitted by more than 100,000 units producing excisable products of which 45 to 50 per cent belong to Small Scale Sector and the remaining are from medium and large factories. At present, about 80 per cent of the RT–12 returns have been computerised and such data are available on the computer system since April 2000. The potential of this vast source of administrative data needs to be examined properly for the compilation of IIP. The feasibility of making use of data from other administrative sources, like sales tax, for compilation of IIP, at least at the State level, should also be explored.
Recommendations
5.4.17 Against the background of the above analysis of deficiencies in the IIP, the Commission makes the following recommendations which are grouped according to the areas of action:
Improvements in the Existing IIP
The item basket of the Index of Industrial Production (IIP) should be selected in such a way that the indices are representative of the growth in the Industrial Sector at least at the 2-digit level of NIC.
The source agencies should make available the data on the additional items to be included in the item basket. The agencies should expand their database to capture new units and new items.
To ensure the availability of data on new items, the Central Statistical Organisation (CSO) should regularly provide the source agencies with:
The list of items that are just below the cut-off criteria of item selection and likely to figure in the revised item basket on the basis of the current series.
Items identified on the basis of detailed results of Annual Survey of Industries (ASI).
The source agencies should also identify the important and fast-moving items for inclusion in their database for the purposes of administration and revision of IIP.
The base year of the Index should be revised quinquennially by the Central Statistical Organisation to adjust to the structural changes in the industrial sector. Strengthening of Source Agencies
The statistical set-up in DIPP needs restructuring in terms of statistical manpower and infrastructure, by creating a full-fledged statistical unit under the overall guidance of a professional statistician.
The proposed statistical unit should be vested with the responsibility of maintenance of the frame, timely supply of monthly production data with an adequate response rate, exploring of suitable methodologies for dealing with non-response and improving the overall quality of data.
The statistical set-up of other source agencies of the IIP also needs to be adequately strengthened.
Since such strengthening will take some time, the agencies should, in the mean time, attempt to achieve a minimum standard for ensuring the quality and reliability of the Index, by adopting a suitable monitoring mechanism to target a response rate in terms of production of at least 60 per cent in the first month and 80 per cent at the final revision.
The source agencies should correspond with the production units through fax, e-mail and telephone followed by a personal visit, if necessary, to minimise non-response. Cooperation from the Industrial Associations and State Governments should also be solicited in this context.
Source agencies should preferably avoid inclusion of items for which very few units (say, less than 5) are reporting production, in order to avoid extreme fluctuations in the production data due to non-response. If, however, it is necessary to include some such items, the source agencies should make all efforts to closely monitor and collect data for these items.
The problem of non-response needs greater and more detailed examination. Therefore, technical experts from the fields of industry and statistics should go into the question of whether statistical methods could be useful for the solution of this problem.
Additional All-India Index of Industrial Production
In view of the difficulties faced in the collection of data for compilation of IIP by the mail enquiry method, the possibilities of constructing an additional Index of Industrial Production by direct collection of monthly production data on selected items from factories with 200 or more workers should be explored. This should first be done for one year on an experimental basis.
For this purpose, the Field Operations Division (FOD) of the National Sample Survey Organisation (NSSO) should be entrusted with the task of collection of monthly production data under the existing legal provisions. The requirement of additional resources for this purpose should be met.
Before the above suggestion of compilation of the Index is taken up, a study should be conducted to compare the annual growth rate in production, based on ASI data of recent years (for factories with 200 or more workers) and current IIP. The findings should be examined by the Standing Committee on Industrial Statistics regarding the workability and adoptability of this approach for compilation of the Additional Index of Industrial Production.
Use of other administrative data
An exploratory study should be undertaken to examine the feasibility of using the production data as available with Central Board of Excise and Customs for compilation of an All-India Index of Industrial Production. The possibility of utilising data from other sources like sales tax for compilation of IIP, at least at the State level, should also be examined.
Comparable State-level Indices of Industrial Production
Introduction
5.4.18 Apart from the monthly IIP compiled by the CSO, a number of State Directorates of Economics and Statistics are engaged in the compilation of State-level Indices of Industrial Production (IIPs). However, these indices lack comparability due to differences in the choice of base year, periodicity of release, data sources used, etc. The need for comparable State-level IIPs has been felt for quite some time, for making comparative studies on the state-wise performance of the industrial sector. Recognising the need, a Technical Advisory Committee (TAC) was constituted for an in-depth examination of various technical issues relating to compilation of comparable State-level IIPs. The Committee submitted its report in June 1998. The main recommendations of the TAC pertaining to comparable State IIPs along with their present status are listed in Annexe 5.27.
Current Status
5.4.19 Some of the States have undertaken the compilation of monthly IIPs with base 1993-94. Tamil Nadu, Andhra Pradesh and West Bengal have released the monthly comparable State IIPs. Karnataka and Pondicherry have released the State-level IIPs on an annual basis with base 1993-94 and are making efforts to compile quarterly or monthly IIPs. Most of the States and Union Territories (UTs) have identified the item basket, prepared the weighting diagram and selected sample units from the frame of factories or units for collection of data for compilation of comparable State-level IIPs.
5.4.20 The State Governments desired that the regular monthly production data may be collected by the NSSO and supplied to them. The TAC was not in favour of a Central agency for collection of production data on behalf of the States. The Committee, therefore, recommended that States and UTs should make necessary augmentation in infrastructure for collection of regular data for the manufacturing sector. It was decided that for the mining sector, the requisite data for the preparation of the weighting diagram and regular production data for compilation of index may be made available to the States by IBM, Nagpur and for the electricity sector, the CSO may send the necessary data to the States after procuring them from the Central Electricity Authority.
Deficiencies
5.4.21 Most States find difficulty in regular collection of data for compilation of the State-level IIPs due to constraints of resources and manpower. The CSO has been emphasising the importance of comparable State-level IIPs and the need for action to augment the States’ resources for this purpose. Pending such augmentation, the CSO has advised the States to try the mail method of data collection. In this regard, the CSO has directed the States to impress upon the units, Industrial Associations and State Industrial Departments to cooperate in order to make the mail method a success for optimum data collection, as it is impossible to provide resources for data collection by the interview method, even in the case of very limited information on the production of only a few items from the factories.
5.4.22 Some State Governments have indicated that they are facing difficulties in the collection of monthly production data from the selected factories, as the response is not adequate. Under the Collection of Statistics Act, 1953, all the units registered under the Factories Act, 1948, are obliged to supply information required by the Government. The NSSO gets annual information from the factories under the above Act for the Annual Survey of Industries. Some States have pointed out that the DESs are not the authority under the Collection of Statistics Act and in the absence of any legal backing, it is difficult for them to collect production data from the factories. As such, it is necessary to delegate the authority for data collection under the Collection of Statistics Act to the State DESs for smooth and regular collection of monthly production data.
5.4.23 To some extent, the resource problem faced by the States in a monthly collection of data can be solved if a proper coordination mechanism between the source agencies of All-India IIP and various State DESs is established. A State or UT will then be able to use the information collected by the source agencies from units located in it for compilation of comparable State-level IIPs. This will reduce the requirement of additional resources by the States for compilation of State IIPs. However, for the purpose of compilation of the IIP, the States may need to augment their resources for collection of data on additional items from additional units, which are not included in the database of the source agencies.
5.4.24 The ASI, 1997-98 shows that about one-third of the units covered were making use of computers in their accounting work (see Annexe 5.14). This percentage is likely to be much higher in case of the Census Sector, which comprises large factories with substantial resources including IT facilities. For impressing upon the units the need to use computer and information technology for sending the information to the data collecting agencies, the cooperation of associations such as Confederation of Indian Industries (CII), Associated Chambers of Commerce and Industry (ASSOCHAM), Federation of Indian Chambers of Commerce and Industry (FICCI) and others may be solicited.
Recommendations
5.4.25 The Commission, therefore, recommends that:
The present practice of using Gross Value Added (GVA) for preparation of weighting diagram for All-India Index of Industrial Production and Gross Value of Output for State Comparable Indices of Industrial Production would result in incomparability in the two methodologies. GVA should, therefore, be used for the State IIPs.
Industrial Associations and State Industry Departments should be requested to impress upon their members and units to cooperate with the State and UT Governments in supplying regular monthly production data to the State Directorates of Economics and Statistics so that the Comparable State-level Indices can be compiled and released regularly.
Coordination mechanism needs to be established between the source agencies and the States and UTs for supply of regular production data.
For the purpose of collection of necessary data, the State DESs should be given authority as recommended by the Commission under 5.1.37(ii).
Vigorous efforts on the part of the States, with the help of industrial associations and State Industrial Departments, are required to try out the mail methods i.e. fax, e-mail, etc. for collection of regular monthly production data.
Additional resources to the States for compilation of State-level Comparable IIPs should be provided.
Reference period for ASI, 1998-99 was the accounting year of the factory ending on any day during the fiscal year 1998-99 and the survey period of ASI, 1998-99 was the year 1999-2000.
Directories of Establishments should contain identification particulars like name and code of: State or Union Territory, district, tehsil or town, village or block where the establishment is located; name and address of the establishment, if available; type of ownership of the establishment; agency with which registered; description and industry code of the major activity pursued by the establishment; number of workers employed; etc.
As sample selection for ASI, 1997-98 was not done in the form of two independent sub-samples, RSEs were approximated with the help of half-sampling technique by dividing the sample factories into two halves with odd and even orders of selection respectively.
Own Account Enterprises are the enterprises that work without employing any hired worker on a fairly regular basis, i.e. they do not employ hired workers during the major part of the period of operation during the last year preceding the date of survey.
Non-directory Establishments are the enterprises operating by employing at least one hired worker on a fairly regular basis but with total number of workers as five or less.
Directory Establishments are the enterprises operating by employing at least one hired worker on a fairly regular basis but with total number of workers as six or more.
SSI in India, The Growth Sector for the Millennium, August 2000.
