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

A+| A| A-

Measurement of GDP of Services Sector in the New Series of National Accounts Statistics

The rising share of the service sector in the gross domestic product marks a structural shift in the Indian economy and brings it closer to a developed economy. In view of its rising importance, an attempt has been made in this paper to present the salient features of changes in the methodology and coverage of this sector in the new series (base 1999-2000) of national accounts statistics.

Measurement of GDP of Services Sector

in the New Series of National Accounts Statistics

The rising share of the service sector in the gross domestic product marks a structural shift in the Indian economy and brings it closer to a developed economy. In view of its rising importance, an attempt has been made in this paper to present the salient features of changes in the methodology and coverage of this sector in the new series (base 1999-2000) of national accounts statistics.


I Introduction

ervices are crucial for the economies of both developed and developing countries. An efficient service infrastructure is a prerequisite for the economic performance of a country. Producers and exporters of textiles, cars or computers will not be competitive without access to efficient banking, insurance, accountancy, telecommunications or transport. Access to world class services helps exporters and producers in developing countries to capitalise on their economic strength. They lead to consumer savings, faster innovation and technology transfer. They contribute to long term investment. Moreover, the rapidly expanding service sector is contributing more to economic growth and job creation worldwide than any other sector.

The service sector has been India’s most dynamic in recent years, with telecommunications and information technology registering particularly rapid growth. The rise in the service sector’s share in GDP marks a structural shift in the Indian economy and brings it closer to a developed economy. The share grew from

40.6 per cent in 1990-91 to 49.3 per cent in 1999-2000, whereas the secondary sector’s share remained static around 24.5 per cent from 1990-91 to 1999-2000. The agriculture sector has shown a contrast to services with its share in GDP coming down to 27.6 per cent in 1999-2000 from 34.9 per cent in 1990-91, as shown in Table 1 and Figures 1 and 2.

II Service Sector in the Indian Economy

In the national accounts, the service sector activities include:

  • (1) trade; (2) hotels and restaurants; (3) railways; (4) other transport including tourist assistance activities as well as activities of travel agencies and tour operators; (5) storage;
  • (6) communication; (7) banking and insurance; (8) real estate and ownership of dwellings; (9) business services including accounting, software development, data processing services, business and management consultancy, architectural, engineering and other technical consultancy, and advertising and other business services; (10) public administration and defence;
  • (11) other services including education, medical and health, religious and other community services, legal services,
  • recreation and entertainment services; and (12) personal services and activities of extra-territorial organisations and bodies.

    III Service Sector: Public vs Private

    The contribution of the public and private sectors in the GDP of the service sector in 2003-04 is given in Table 2. It may be seen that public administration, defence and railways are totally in the public sector. Banking and insurance services follow, with 76 per cent share in the public sector. Two-thirds of sanitation services (government sector only) are provided by the public sector. Education services are equally shared by the public and private sectors. Services in respect of communication, health, trade, hotels and restaurants, transport other than railways and other services are fast moving into the private sector of the economy.

    IV Measurement of GDP of Service Sector – New Series

    The three major components influencing the new series of national accounts statistics include (1) revision of base year to a more recent year (for meaningful analysis of the structural changes in the economy), (2) complete review of the existing database and methodology employed in the estimation of various macroeconomic aggregates including choice of alternative databases on individual subjects, and (3) to the extent feasible, implementing the recommendations of the system of national accounts (1993 SNA) prepared under the auspices of the Inter Secretariat Working Group on National Accounts comprising the European Communities (EUROSTAT), International Monetary Fund (IMF), Organisation for Economic Cooperation and Development (OECD), United Nations and World Bank. Thus, the new series has been introduced by the CSO after a comprehensive review of both the databases and the methodology employed in the estimation of various macroeconomic aggregates.

    V Choice of 1999-2000 as the Base Year

    In the past, national accounts statistics have mostly been revised decennially, changing the base to a year synchronising with the

    Economic and Political Weekly September 15, 2007

    Figure 1: Service Sector-Growth (as Percentage of GDP) in the Indian Economy

    27.5 28.7 32.0 36.6 40.6 49.8 54.1 0 10 20 30 40 50 60







    2005-06 (Adv)

    year of the decennial population census. This was primarily because in the base year estimates, information on the workforce has played an important role. The workforce estimates were obtained from the population census, which is conducted decennially in the years ending with 1. As a sequel to this sequence, the previous series of national accounts statistics should have been with the base year 1990-91. However, it was observed that the data on worker participation rate (WPR) captured by the NSSO was better than that estimated through the population census. Accordingly, the CSO used the workforce estimates based on National Sample Survey (NSS) workforce participation rates from the NSS 1993-94 (50th round) survey results, and revised the base year of national accounts to 1993-94. In continuation with this practice, the new series of national accounts released on January 31, 2006 adopted 1999-2000 as the base year, as it has used the data on WPR from the NSS 55th round quinquennial survey on employment and unemployment, conducted in 1999-2000. In the new series, the WPR data has been used in conjunction with population data of the population census, 2001. The workforce estimates for the base year have been generated afresh as these play a crucial role in the GDP estimates for the service sector.

    VI Estimation of Workforce and Value Added per Worker in the New Series

    In the absence of annual enterprise surveys, the GDP estimates in respect of (1) unorganised segments of manufacturing and service sectors, and (2) for some segments of private organised service sectors are compiled through indirect methods, using the benchmark indicator procedure. In this procedure, the benchmark GDP estimates are initially prepared at detailed activity level for the base year of national accounts series using the estimated workforce engaged and the value added per worker (VAPW) in the activity. For subsequent years, the GDP estimates are extrapolated with appropriate indicators relevant to the economic activity.

    For arriving at appropriate workforce estimates for use in the new series, a working group (WG) on workforce estimation was appointed under the chairpersonship of Grace Mazumdar in December 2002. In its report, the WG recommended the detailed sources and methodology to be adopted for the new series for arriving at the estimates of workforce in various compilation categories, using the data available from the NSS 55th round and population Census, 2001.

    In view of the observations of the WG, further validation of estimates of workers for some compilation categories has been done for the new series using the limited single digit industry level national industrial classification (NIC 1998) workforce data obtained from the registrar general of India (RGI) (detailed information on workforce was not released by the office of the RGI at the time of the release of the new series), and this has been compared with the National Sample Survey (NSS) 55th round data obtained from the National Sample Survey Organisation (NSSO). While comparing the two sets of data, the number of workers in both sources has been adjusted to bring them to the reference date of October 1, 1999. The use of this data has been restricted to the doubtful cases mentioned in the working group report, besides a few other additions. These are (1) sale of motor vehicles, (2) wholesale trade except for motor vehicles and auctioning activities, (3) storage and warehousing, (4) sewage and refuse disposal, sanitation and similar activities, (5) activities of membership organisations, and social work with accommodation, (6) recreational, cultural and sporting activities, and (7) private households with employed persons. While the estimates of workforce in the public sector and the private organised sector are available annually from the directorate general

    Table 1: Sectoral Shares in GDP

    (In per cent)

    Year Primary Secondary Tertiary
    1950-51 59.2 13.3 27.5
    1960-61 54.7 16.6 28.7
    1970-71 48.1 19.9 32.0
    1980-81 41.8 21.6 36.6
    1990-91 34.9 24.5 40.6
    1991-92 34.1 23.9 42.0
    1992-93 34.2 23.7 42.1
    1993-94 33.5 23.7 42.8
    1994-95 32.9 24.4 42.7
    1995-96 30.6 25.5 43.9
    1996-97 30.9 25.4 43.7
    1997-98 29.0 25.2 45.8
    1998-99 29.0 24.5 46.5
    New Series Base
    1999-2000 27.6 23.1 49.3
    2000-01 26.6 23.6 49.8
    2001-02 26.6 23.0 50.4
    2002-03 24.2 23.7 52.1
    2003-04 24.4 23.5 52.1
    2004-05 (Quick Est) 23.0 23.8 53.2
    2005-06 (Adv Est) 21.7 24.2 54.1

    Table 2: Contribution (in Percentage) of Public and Private Sector in the GDP of Service Sector in 2003-04 at Current Prices

    Services Government NDCUs Public Private incl DCUs Sector Sector

    Public admin and

    defence 100 0 100 0 Railways 99.32 0.68 100 0 Banking and insurance 0 75.88 75.88 24.12 Sanitation 66.2 0 66.2 33.8 Education 50.42 0 50.42 49.58 Communication 6.11 41.86 47.97 52.03 Storage 0 37.48 37.48 62.52 Health 30.95 0 30.95 69.05 Transport by other means 1.47 15.33 16.8 83.2 Trade 0 2.19 2.19 97.81 Hotels and restaurants 0 1.05 1.05 98.95 Other services 0 0.46 0.46 99.54

    Figure 2: Percentage Shares of the Three Sectors in the Indian Economy

    1950-51 1999-2000 2005-06 (Advance Estimates)


    Primary Services

    28 per Services Primary

    28 per 54 per

    cent 49 per
    22 per cent cent



    Secondary Secondary

    59 per

    13 per 23 per Secondary cent cent cent 24 per cent










    of employment and training (DGE&T), those for the unorganised sector are derived as residual by deducting the organised sector workforce estimates from the total workforce of each compilation category.

    The estimates of VAPW for the unorganised and some private organised segments of manufacturing (non-SSI part) and service sectors are from the detailed data available from the NSS 55th round (1999-2000) on informal non-agricultural enterprises, NSS 56th round (2000-01) on manufacturing enterprises in the unorganised sector, and NSS 57th round (2001-02) on unorganised enterprises in the service sector (excluding trade and finance). The estimates of VAPW at the compilation category level have been arrived at as a ratio of the gross value added to the total employees in the particular compilation category. The following adjustments have been made while processing the data for estimating the VAPW:

  • (1) Rent on land and building has been treated as factor income, in the absence of data on rent on land.
  • (2) Wherever the survey results are for years other than 1999-2000, suitable adjustments have been made in the VAPWs using consumer price indices for agricultural labour and industrial workers to arrive at the estimates for the benchmark year 1999-2000.
  • (3) Wherever the unit level data has shown either zero or very low value of output, but has entries against inputs and salary and wages, they have been assumed to be outliers and have been removed from the estimation of VAPW. However, the number of such records removed from the data processing was very low.
  • VII Estimation of Labour Input

    Estimation of labour input requires estimates of workforce (number of workers) over economic activities or groups of economic activities. The workforce is then adjusted for multiple jobs in respect of each activity or activity group so as to measure the labour input (job counts). The method is called labour input method (LIM). The procedure can be formulated as below: Let Xij = Employment and unemployment survey (EUS) esti

    mate of ith compilation category of jth broad group (principal + subsidiary). Lij = Labour input estimate of ith compilation category of jth broad group. = EUS estimate of total population.


    = Population census (PC) estimate of total population.

    YP Lij = Xij (YP / XP )

    Principal and subsidiary activities are used as per NSSO definition of usual activity status;

    The category wise estimates have been calculated by multiplying workforce participation rate (WFPR) (principal and subsidiary activity together) with the population figure (as projected);

    The organised sector workforce estimates are provided by the DGET; and

    The private unincorporated sector workforce figures are obtained as residual from workforce estimates. The only deviation is “custom tailoring” for which workforce has been estimated from NSS 56th round enterprise survey on manufacturing. The activity has been treated as service for the new series and the workforce, which is merged with other manufacturing activities, could not be separated out from NSS 55th round based estimates.

    VIII Changes in the Estimation of GDP in the New Series

    The major changes in respect of the service sector adopted in the new series are enumerated in Table 3.

    IX GDP Estimates of Sub-sectors of Service Sector

    The estimates of GDP by economic activity for the service sector for the year 1999-2000, according to the new series and the 1993-94 series, are presented in Table 4. According to this table, the level of GDP has gone up by Rs 30,454 crore or 1.7 per cent, in the new series in 1999-2000. Of this, the GDP of the service sector has increased by Rs 38,367 crore or 4.5 per cent. In the service sector industries, the maximum upward revision in GDP data is witnessed in the ‘real estate, ownership of dwellings and business services’ sector, which is on account of increase in the number of dwellings from what was earlier projected for the year 1999-2000, as per the data available from the population census, 2001.

    X Estimates for Subsequent Years

    The estimates of value added for the subsequent years (other than base years) in respect of various industry groups are based

    Economic and Political Weekly September 15, 2007

    Table 3: Changes in the Estimation of GDP in the New Series

    Industry Methodology Followed

    Old Series New Series

    Trade GDP estimates were This sector now comprises
    prepared for the trade sector the following five categories,
    as a whole covering as per NIC 1998 classification:
    categories (NIC 1987 codes (1) 502+50404, ‘Maintenance
    of 6 excluding 69, 840, and repair of motor vehicles’
    841, and 890). (new item)
    (2) 50-502-50404, ‘Sale of
    motor vehicles’
    (3) 51+74991, ‘Wholesale
    trade except of motor vehicles
    + Auctioning activities’
    (4) 526, ‘Repair of personal
    and household goods’
    (new item)
    (5) 52-526, ‘Retail trade
    (except motor vehicles)’.
    The GVA estimates have been
    prepared separately for the
    public sector, the private
    organised sector and the
    private unorganised sector.
    Transport by GVA estimation for services GVA estimate has been
    means other incidental to transport prepared by multiplying the
    than railways (sub-sector of transport by workforce (duly adjusted for
    and storage other means) was on the the public sector) by GVA per
    basis of annual data on worker available from the
    commissions paid to booking enterprise survey 2001-02
    agencies by shipping results for the respective NIC
    companies and airline codes.
    companies. The economic activities of
    airport authorities have now
    been shifted to the sub-sector
    “service incidental to transport”.
    The economic activities of
    airport authorities were
    included under air transport
    Supporting service to water Supporting service to water
    transport was part of transport has been shifted to
    water transport. “service incidental to transport”
    Communication: The estimates of GVA of Separate estimates for cellular
    private private communication mobile, courier services, and
    communication services as a whole were public call office booths
    services prepared using NSS based (PCOs) have been compiled.
    estimates of workforce and
    value added per worker.
    Banking and The imputed value of The imputed value of services

    Insurance Mutual services in respect of UTI of UTI has been calculated

    Funds (MFs) was estimated in the same as total income on account of manner as that of banks dividend, interest, profit on and other financial sale/redemption of intermediaries. investment minus interest,

    dividend paid to the unitholders and undistributed profit.

    Non-government Current price estimates of Current price estimates of non-banking GVA and FISIM of this GVA of NGNBFCs have been financial sub-sector were prepared prepared using the RBI companies by analysing the data on studies of a particular year (NGNBFCs) income, expenditure and appearing in three studies.

    profits provided in the RBI The pooling of the sample sample study on companies has been done ‘Performance of non- to get a more stable estimate. government financial and investment companies’ published annually. The GVA for the population (excluding two major NBFCs namely, ICICI (prior to 2002-03) and HDFC) was obtained by blowing up the sample estimate. The ratio of the paid-up capital of the population to the paid-up capital of the sample formed the blowing-up factor.

    Industry Methodology Followed

    Old Series New Series

    Banking department of RBI

    Life insurance

    Real estate, ownership of dwellings, legal and business services Renting of machinery and equipment without operator Computer and related activities in private sector

    Research and development, market research and public opinion polling, business and management consultancy, architectural, engineering and other technical activities, advertising and business activities nec excluding auctioning Volume index prepared on The implicit price indices, as the basis of deflated observed from current and aggregate deposits and constant price GVA estimates credits of all scheduled of commercial banks have commercial banks was used been used to deflate current to move the base year GVA, price estimates of GVA of the for arriving at the constant banking department of RBI, to price estimates of the banking arrive at the constant price department of RBI. estimates of GVA for the

    banking department of RBI. Estimates of GVA at constant The annual changes in the life prices were prepared as fund and sum assured have cumulative amount of sum been deflated to get a volume assured + bonuses and life index for these two separately. fund at the end of the year, The simple average of two provided in the annual reports indices, one each for life fund of LIC, deflated separately with and sum assured, thus the help of WPI. Two volume prepared has been combined indices were constructed to get the volume index for separately and averaged moving the base year (simple average) to get a estimates of GVA of life combined volume index. insurance to obtain the This combined volume index estimates of GVA for was used to move the base subsequent years, at year value added of life constant prices. insurance for subsequent years. Research and scientific Research and scientific services was part of “other services have been services” sector included under ”real estate,

    ownership of dwellings and business services sector”. This activity has been covered for the first time.

    GVA estimates for The estimates of GVA have software activities in the been compiled for organised private sector were prepared and unorganised segments using data on production separately. The GVA estimates of software services made for the organised sector have available by NASSCOM been prepared using and the value added ratio NASSCOM’s data on (GVA to output ratio) production of software services estimated from the analysis and the value added ratio of annual reports of a few estimated from the analysis of software companies. annual reports of a few

    software companies. The estimates of GVA for the year 1999-2000 for the unorganised segment have been prepared using the data on workforce and VAPW. The estimates have been prepared separately for rural and urban areas.

    The estimates of research This compilation category has and scientific services been introduced after (NIC-87 code 922) were regrouping the activities under

    included under business services and other “other services”. services. In NIC-98, this Similarly, the GVA estimates activity is classified under of “press agency activities” code 73, which is under falling under NIC-87 “business services”, and as code 897 were being such in the new series this compiled, though activity has been included not exclusively, under the under “business services”. category of “business Now, this activity falls under services” in the 1993-94 “news agency activities”, which series. is classified under NIC-98

    code 922, and the GVA estimates for this activity have been included under “other services” in the new series. Thus, the comparison of GVA estimates of “business services” in the 1993-94 series and the new series has to be done keeping these aspects in view.

    on different procedures, depending upon the availability of data (Table 5).

    As discussed above, current data is used for estimating the GDP in respect of primary and secondary sectors, railways, banking and insurance, public administration and defence, the public sector part of various services and some components of the organised private sector. Although current data from the benchmark surveys has been used for estimating the unorganised part of manufacturing and service sectors in the base year, estimates for the subsequent years are based on proxy physical indicators (such as IIP for unorganised manufacturing, gross trading index in the case of trade, hotels and restaurants) or using inter-survey or inter-censal growth rates. It may also be mentioned at this stage that there is generally an upward revision in the GDP estimates whenever the base years are changed, indicating that the growth rates are not overestimated.

    The present weaknesses in the GDP estimates (as is well known to the users) are on account of lack of current data on the private corporate sector, as well as on the unorganised sector. In the case of the former, there is no agency in the country which provides global estimates of the private corporate sector, on either quarterly or annual basis. The only

    Table 4: Estimates of GDP at Factor Cost by Kind of Economic Activity, 1999-2000

    (Rs crore)

    Industry 1993-94 New Difference Weight Series Series

    1 Trade, hotels and restaurants 2,46,037 2,54,143 8,106 14.2

  • 1.1 Trade 2,29,004 2,31,878 2,874 12.9
  • 2.2 Hotels and restaurants 17,033 22,265 5,232 1.2
  • 2 Transport, storage and communication 1,24,307 1,31,754 7,447 7.4

    2.1 Railways 15,623 19,504 3,881 1.1

    2.2 Transport by other means 81,855 82,377 522 4.6

    2.3 Storage 1,262 1,427 165 0.1

    2.4 Communication 25,567 28,446 2,879 1.6

    3 Financing, insurance, real estate and business services 2,20,559 2,32,752 12,193 13.0

    3.1 Banking and insurance 1,19,075 1,05,662 -13,413 5.9

    3.2 Real estate, ownership of dwellings and business services 1,01,484 1,27,090 25,606 7.1

    4 Community, social and personal services 2,53,373 2,63,994 10,621 14.7

    4.1 Public administration and defence 1,16,686 1,19,671 2,985 6.7

    4.2 Other services 1,36,687 1,44,323 7,636 8.1

    Total service sector GDP at factor cost 8,44,276 8,82,643 38,367 49.3 Total GDP at factor cost 17,61,838 17,92,292 30,454 100

    Table 5: Procedure Adopted for Estimates of Value Added for Subsequent Years

    Sector Procedure adopted
    Agriculture, forestry, fishing and mining Manufacturing Electricity, gas and water supply Construction Trade, hotels and restaurants Service sectors Double deflation method Single deflation method Single deflation method Single deflation method Single deflation method Single deflation method

    source at present is the RBI’s company finance studies. However, due to its sample size being small, it is difficult to get reliable estimates at the economic activity level or at the state level. In the case of the unorganised sector, the surveys are conducted only once in about five years, by the NSSO. This, however, excludes trading activity, which is one of the major sectors of the economy. There is no other source giving data on the performance of the unorganised sector.

    Another weak area in the estimation of GDP is the absence of suitable price deflators for the service sectors. The recommended procedure for constant price estimates is the double deflation procedure, which means the value of output and value of inputs are to be deflated separately by their appropriate price indices and the gross value added is to be arrived at as the difference between output and inputs. It is well known that we do not have a service sector price index in the country, leave alone appropriate price indices separately for inputs and outputs or various components of GVA, such as salaries and wages and operating surplus.

    The ministry of company affairs has recently launched the MCA21 programme, which envisages online filing of accounts by companies. It is hoped that this programme would provide current and exhaustive data on the private corporate sector in the near future. Regarding the unorganised sector, a programme should be in place for collecting annual data on at least the larger establishments. The annual survey of non-manufacturing industries (ASNMI) was recommended by the National Statistical Commission (NSC) for implementation. This is similar to the annual survey of enterprises (ASE) or economy-wide survey (EWS), which most countries have in place. It is necessary to introduce ASE some time in the future. The strengthening of corporate sector statistics, introduction of ASE and the service sector price indices are the key to further improving the quality of GDP estimates in the country. These would also enable the country to have an index of service production (ISP) on the pattern of the index of industrial production (IIP).



    [The views expressed in this article do not necessarily represent those of the organisation to which the authors belong.]

    For the Attention of Subscribers and Subscription Agencies Outside India

    It has come to our notice that a large number of subscriptions to the EPW from outside the country together with the subscription payments sent to supposed subscription agents in India have not been forwarded to us.

    We wish to point out to subscribers and subscription agencies outside India that all foreign subscriptions, together with the appropriate remittances, must be forwarded to us and not to unauthorised third parties in India.

    We take no responsibility whatsoever in respect of subscriptions not registered with us.


    Economic and Political Weekly September 15, 2007

    To read the full text Login

    Get instant access

    New 3 Month Subscription
    to Digital Archives at

    ₹826for India

    $50for overseas users


    (-) Hide

    EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

    Back to Top