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

A+| A| A-

India's Growing Services Sector: Database Problems and Issues

The increasing importance of services in the Indian economy prompted the organisation of a seminar, 'Growing Size of the Services Sector in the Indian Economy' , in 2006. After briefly touching on the problems of collecting data relating to the size of the sector, this article introduces the content of the papers that were presented at the seminar and that follow in this issue.

Services GDP: Issues in Estimation

India’s Growing Services Sector: Database Problems and Issues

The increasing importance of services in the Indian economy prompted the organisation of a seminar, ‘Growing Size of the Services Sector in the Indian Economy’ , in 2006. After briefly touching on the problems of collecting data relating to the size of the sector, this article introduces the content of the papers that were presented at the seminar and that follow in this issue.


he services sector has been the major contributor toward sustaining the 6 per cent-plus annual GDP growth story of India since the mid-1980s and has come to contribute more than 50 per cent of total GDP. This has also led to speculation whether India would chart out a unique growth path in which the country would leapfrog from a predominantly agricultural to a directly service-dominated economy by skipping the intermediate stage of rising share of industrial sector that was experienced by all the existing industrialised countries. Others raise doubts about the sustainability of service sector growth without concomitant growth of the real (or commodity producing) sectors of agriculture and industry.

In this connection, valid concerns have been raised about the adequacy, reliability, accuracy and timely availability of the data base for measuring the contribution of services. Services cover a very heterogeneous basket in terms of nature, character, mode of organisation and ownership structure. In terms of nature, they range from those which are market-oriented (like business services) to those operating through non-market channels (such as government administration, religious and community services). Their factor intensities vary widely from those which are capital intensive (real estate and housing) to those which are capital light (restaurants). The differences in their character range from those aimed directly at final consumption (hair cutting) to those serving intermediate consumption (banking, trade). The mode of organising productive activity ranges from tiny roadside own account enterprises to cooperatives to large corporations in terms of scale of operation, from public ownership (railways in India) to those operating mainly under private ownership (road transportation). Recent technological advances in information technology and audio-visual communications have added further dimensions to the inherent heterogeneity in terms of changing quality, rising productivity (though not possible to measure in concrete terms like in the case of commodities) and increasingly complex modalities of exchange. To capture these various dimensions in measurement for regularly compiling national accounts statistics poses difficult conceptual and database related problems.

In view of the rising importance of the services sector in the economy and in the context of difficult database-related problems, it was indeed a very welcome move on the part of the Indira Gandhi Institute of Development Research and the EPW Research Foundation to organise the database seminar on ‘Growing Size of the Services Sector in the Indian Economy’ on March 2, 2006 in Mumbai. The seminar provided a valuable forum for a fruitful face-to-face interaction between compilers and users of national accounts statistics. In all, eleven papers were presented and discussed in the seminar. The revised versions of the papers in the seminar are presented for wider circulation and debate in this issue. Their careful reading will also help better informed interpretation of the national accounts statistics and enrich the discussion of the alternative hypotheses regarding the role of services in Indian development.

Expert Opinions

Although the papers were written in their individual capacities by the authors, it is satisfying that the seminar attracted nine out of eleven papers from those in the three major data compiling and collecting agencies, namely, two from the Central Statistical Organisation (CSO) on the estimation of the contribution of services to GDP (Sharma, Hazra and Chitkara; and Kumar, Kar and Sanjay); three from the National Sample Survey Organisation (NSSO), out of which two focus on the reliability of the sample survey based estimates while the third suggests an alternative method (Giri, Roy and Mukhopadhyay; Banerjee, Baksi, and Roy; and Manna); four from the Reserve Bank of India, of which three deal with the estimation of GDP in the financial sector and the fourth with internationally traded services (Barman and Samanta; Chakraborty and Das; Rath, Nayak, Lakshmanan, Mandal, Rajesh and Fanai). The two remaining papers were contributed by veteran database experts: S L Shetty, formerly from the Reserve Bank of India and the initiator of the seminar, and M R Saluja (formerly with the Indian Statistical Institute) and Yadav. The collection thus offers a good feast for those who appreciate database intricacies in their applied research.

Two major methods are used in the compilation of national accounts depending on the condition of the underlying available database. In the case of the organised segments (namely, public sector providers of services and in some services, corporate sector units), the contribution to GDP can be traced through the budget documents and annually published accounts on a regular basis. These are the so-called “direct” estimates of GDP at current

Economic and Political Weekly September 15, 2007 prices as they reflect the situation in the year to which they refer. In the absence of any direct estimates of “price” or “productivity” of service, some physical indicator of volume of activity in that sub-sector is used to derive corresponding constant price estimates. Segments other than these directly estimated ones are treated as “unorganised” (in the national accounts terminology), where the method of GDP estimation is “indirect” in the sense that a benchmark estimate based on a periodically carried out sample survey is derived for some base year and extrapolated backward and forward to other years using some physical indicator of economic activity in the sub-sector. Consequently, the movements in the GDP contributions in the unorganised segment of a given service sub-sector from year to year are thus taken to be proportional to the movements in the physical indicator used. These constant price estimates are then “inflated” using some appropriate price index to derive current price estimates. The benchmark estimate itself is the product of the estimated number of workers multiplied by the independently estimated value added per worker (VAPW) derived from a sample survey. This is described as the labour input method of estimating GDP.

Turning now to the papers, 10 out of 11 papers can be classified into three distinct thematic groups.

Overall Perspective

The first group of three papers presents an overall perspective of services in differing details. The status paper by Shetty, after giving a valuable overview of the changing role of services as emerging from the national accounts statistics from 1950-51 to 2004-05, provides a helpful backdrop of the procedure of the labour input method of “indirect” estimation used in the GDP series with 1993-94 price base and raises several associated database related issues for discussion. The next paper by the CSO officers Sharma, Hazra and Chitkara discusses the changes in methodology and new databases (for the existing as well as the newly emerging services) in the application of the labour input method in the recently released new series of GDP with the base year 1999-2000. A useful comparison with the methods and database employed in the earlier (1993-94 price base) series is also provided at the detailed sub-sectoral level in this paper. The third paper by Saluja and Yadav covers the 1980-81 price base series in addition to those with 1993-94 and 1999-2000 price bases and deals with the methodology, data sources and limitations of the gross value added estimated for each category of services distinguished in the national accounts statistics, especially for the unorganised sector.

The second group of three papers from the NSSO officers deals with different aspects of the labour input method using the NSS data that enter (with modifications explained by Sharma, Hazra and Chitkara) in the national accounts statistics. While the new (1999-2000 price base) series uses the usual status based estimates of labour input drawing on the quinquennial employment-unemployment (EUE) survey, Giri, Roy and Mukhopadhyay suggest an alternative of using the weekly and daily status based estimates of labour input based on the EUE survey along with its strengths as well as limitations. They argue that their estimates of labour input provide a better approximation to the concept of “workers” used in the enterprise surveys from which VAPW is derived. The paper by Banerjee, Baksi and Roy assesses the reliability of the benchmark estimates of aggregate gross value added (GVA) for 28 homogeneous service activity groups given by the product [(estimated population)* (WPR)*(VAPW)] where WPR is the estimated worker-population ratio. They do this by combining the estimated relative standard error of WPR from the 55th round EUE survey and that of estimated VAPW from the 57th round enterprise survey. Manna’s paper carries out a similar exercise at a different level of disaggregation for the 57th round, confining himself only to the survey-based estimates of GVA per worker. The empirical results of these two papers are very different because of differences in methods of calculating relative standard errors.

Financial Sector

The next set of three papers is from officers of the Reserve Bank of India dealing with conceptual and database issues in their organisation’s major areas of interest, namely, the financial sector and balance of payments. With the deepening of the process of financial intermediation, liberalisation of interest rates and the emergence of a variety of financial institutions as well as instruments, newer issues of conceptual and empirical measurement of the contribution of the financial sector have emerged. Three out of this set of four papers deal with the financial sector. The paper by Barman and Samanta focuses on measurement of output and prices of banking intermediation services in the context of the new series of national accounts at 1999-2000 prices and, in particular, on the estimation of risk-free reference interest rate required for conceptually correct measurement of value added in the banking sector. Interestingly, the contribution of the banking sector to GDP in the new series has gone down with a change in methodology in comparison with the earlier series, despite the broadening of the concept of intermediation by including several other related services. In this connection, the paper by Chakraborty and Das examines the valuation of banking sector GDP by adjusting for non-performing loans. The third paper by Rath, Nayak, Lakshmanan, Mandal, Rajesh and Fanai suggests an alternative deflator for estimating the constant price value added in the banking sector by considering a weighted index of deflators of private final consumption expenditure and gross domestic capital formation from national accounts statistics on the lines of the existing practice in China.

The final paper by Kumar, Kar and Sanjay stands in a cate gory by itself. It raises an interesting conceptual issue of the boundary between manufacturing and services by considering the services embodied in manufactured products or what they call manufacturing services production and suggests an empirical approximation based on the unit level data from the 56th round of the National Sample Survey on unorganised manufacturing enterprises. They make a persuasive plea for a separate treatment of this component in the surveys and underline the need for their inclusion in the industrial classification system.

Needless to add, readers will find the individual papers richer in content depending on their appetite for database problems and their concern for making their applied research better informed with data-related judgments.



[This article is written in the author’s personal capacity and is not meant to reflect the views of the organisation with which he is associated. He is grateful to the organisers for inviting him to moderate the day-long discussions at the seminar and asking him to write this introduction.]

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