Ministry of Statistics and Programme Implementation
Government of India
New Delhi
(Secretariat Provided by the Reserve Bank of India)

Published by Rekha Misra for the Reserve Bank of India, Shahid Bhagat Singh Marg, Fort, Mumbai-400001 and designed and printed by her on behalf of the Government of India at Alco Corporation, A-2/72, Shah and Nahar Industrial Estate, Lower Parel (West), Mumbai-400013.

Dr. C. Rangarajan Member of Parliament (Rajya Sabha)

AB-13TilakMarg New Delhi-110001 Ph: 2686 6478 (Off) 2338 3842 (Res) Fax: 26512703,26852548 Email: c.rangarajan@nic.in

I have the pleasure of submitting herewith the Report of the High Level Committee on Estimation of Savings and Investment set up by the Government of India in December 2007. The issues relating to the estimation of gross domestic savings have assumed immense significance in the recent times, when the Indian economy has begun its journey on a high growth trajectory and structural changes are taking place at rapid pace with respect to savings behaviour of the economic agents. In the contemporary global setting, issues related to savings and investment are also attracting attention in the academic and policy circles due to significant policy implications. Current account imbalances are seen largely as the outcome of domestic proclivities of savings and investment rather than trading patterns. Therefore, the criticality of precise estimation of savings and investment needs hardly any emphasis. Against this setting, the estimation procedures pertaining to savings and capital formation are revisited from time to time in view of the critical role played by savings and investment in the growth process.

In India, the methodology currently adopted for estimating savings and investment has evolved over the years and has been acknowledged as sound by experts. The conceptual framework as per which savings and investment estimates of the household sector are estimated is an 'indirect and residual' method. After due thought, the Committee considered it appropriate to continue the estimation of savings with this framework. However, the Committee suggests that a Comprehensive Income -Expenditure Survey may also be attempted by the National Sample Survey Organisation for 'pure households' along with Enterprise Survey. The Survey will provide an alternative data base and methodology to cross-validate the household savings estimated through the residual method. Besides, the Committee made extensive recommendations to improving the quality of the household sector financial savings through a Worksheet Approach. The Committee made recommendations aimed at improvising the estimates for organised sectors like private corporate sector by shifting to the extent possible to a census-based procedure from the existing sample-based procedure by deploying the data resources of the Ministry of Corporate Affairs (MCA). The Committee also undertook the major initiative of sensitising the apex financial bodies for assisting the Government of India and Reserve Bank of India to provide information pertaining to different financial instruments under their regulatory purview regularly to ensure smooth flow of data from diverse primary sources for developing alternative data bases.

In its endeavour to delineate the areas of shortcomings in the data sources and estimation procedures, a surprising revelation for the Committee has been that significant shortcomings have been found not only in respect of the unorganised sectors but also for the organised sectors. Such weak statistical foundations distinctly identified in the organised sector areas relate to (i) aggregates of corporate sector statistics; (ii) industrial output series as depicted by the Annual Survey of Industries (ASI) and those used for the construction of the index of industrial production (IIP); (iii) operations of large segments non-banking financial companies (NBFCs); (iv) capital market institutions and their operations; and (v) cooperative credit and non-credit institutions.

There have been multiple reasons for the deterioration in the statistical foundations of the above sectors, which have been addressed in the body of this Report rather extensively, and various recommendations have been advanced for correcting them. But, I wish to take this opportunity and reiterate one aspect of our recommendations, which has long-term implications and which deserves urgent attention at the highest level in the Government. This is regarding the low priority assigned to statistics and statistical basis of policy making in the concerned ministries, Departments and institutions not only at the State level but also, more significantly, at the Central Government level and national level institutions. This low priority to statistics and statistical analysis has been traced to the inadequate provision made for manning the statistical wings with adequate number and quality of technical staff. There is, therefore, the urgent need for conferring adequate attention to strengthening staff positions in the statistical wings of the Department of Company Affairs, the Department of Industrial Policy and Promotion and in the Central Statistical Organisation, as also in the RBI, SEBI and NABARD. We sanguinely hope that the Government Departments and institutions take our recommendations seriously and adopt the necessary steps to improve staff strength and correct their weak statistical foundations. I may add that the Committee's recommendations concern the issues germane to the estimations of national income, saving and investment and flow-of-funds accounts, but we are convinced that the drawbacks in data base extend to wider horizons.

In respect of unorganised sector too, the Committee has made a number of recommendations. A pivotal one concerns the need for undertaking a comprehensive integrated survey of household income, expenditure and saving on a quinquennial basis so that some direct estimates of household saving and investment are possible, apart from providing data on income distribution and such other insightful information on the evolving structure of the economy. The Committee was, of course, conscious of the fact that such an integrated household survey conducted once in five years would only provide some benchmark estimates and not annual or quarterly series as are necessary for a sound statistical system. What is more, even the benchmark estimates are possible only if the integrated household surveys are accompanied by enterprise surveys so as to cover all unincorporated enterprise components of the presently defined "household" sector. All of these call for strengthening the National Sample Survey Organisation (NSSO) as well as the Economic Census and Surveys Division of the Central Statistical Organisation (CSO). While the NSSO has agreed to consider undertaking the integrated household income-expenditure-saving survey after a while, the Committee was convinced that the NSSO's and CSO's staff and infrastructure resources have been stretched to the brim. Probably because of that the global estimates of the existing surveys have begun to produce two key drawbacks: (i) substantial sampling and non-sampling errors; (ii) more importantly, their inability to produce reasonably representative or dependable state-level estimates. The Committee's interactions with the National Statistical Commission (NSC) have helped to reinforce the validity of our findings. Therefore, in the reformed statistical system, it is imperative that we place an added focus on field surveys and considerably strengthen the staff and resource positions of the NSSO and the CSO as these institutions are concerned with nation-wide surveys of immense analytical and policy significance.

The recommendations proposed by the Committee are expected to improve the quality of estimates, when implemented. On behalf of the Members of the Committee and on my own behalf, I thank you for giving us the opportunity to critically review the procedure for estimation of savings and investment and make recommendations for improving the system.

With warm personal regards,

Yours sincerely,

(C.Rangarajan)

Shri G.K. Vasan
Minister of State (Independent Charge)
Ministry of Statistics and Programme Implementation
Government of India
Sardar Patel Bhavan
NEW DELHI-110 001

Acknowledgement

The High Level Committee (HLC) takes this opportunity to thank all Members of the Sub-committees for their valuable contributions. The Sub-committees were benefited by their interaction with various experts and the Apex Regulatory Institutions. The HLC is indeed grateful to all of them. The HLC specially acknowledges the involvement of Shri U.C. Sarangi, Chairman, NABARD for taking initiatives to streamline the cooperative data, which form an important input in the savings estimation of the household and private corporate sector. Acknowledgements are also due to Shri S. Sridhar, Chairman, NHB, Dr. T.C. Nair, Whole-time Member, SEBI, Shri Ganesh Tawte, Chief General Manager, NABARD, Ms. Mamta Suri, Deputy Director, IRDA and Shri Lalit Kumar, NHB.

CSO Officials Shri Ramesh Kolli, Shri Naresh Kumar, Shri Girish Chandra, Smt. Rajeswari, Smt. Anandita Sinha Ray provided considerable technical assistance. In particular, the HLC acknowledges and places on record the contribution made by Shri Ramesh Kolli who was a Member in the HLC as well as in all the four Sub-Committees. The Secretariat was immensely benefited by the contributions made by Shri Kolli and his team. Special thanks are due to Dr. Rajiv Mehta, Shri G.C. Manna and other officials from NSSO for their contributions. Dr. Sunita Chitkara, Director, Ministry of Corporate Affairs provided considerable assistance to the Committee.

The HLC must make a special mention of the dedicated efforts put in by the Committee’s Secretariat provided by the Reserve Bank of India. Earlier, before his retirement from the RBI’s service, its Executive Director, Dr. R.B. Barman had laid the foundation for the Committee’s work as its first Member-Secretary; he had played a pivotal role in ensuring close coordination between the RBI and the Ministry of Company Affairs. Dr. Barman remained a Member of the Committee till the end and guided its deliberations. Shri K.U.B. Rao, Adviser, Department of Economic Analysis and Policy, who was already associated with the work of the Committee, ably took over the Member-Secretary’s role from Dr. Barman and saw through the varied deliberations and coordination tasks and organised the preparation of initial drafts of the Report in collaboration with Shri Ramesh Kolli of the Central Statistical Organisation (CSO). Shri K.U.B. Rao’s firm professional commitments has helped immensely the completion of the tasks of coordination and getting the Report finalised. The HLC places on record its profound thanks to Shri K.U.B. Rao and also to Dr. R.B. Barman.

The logistics support received from the Regional Directors of the Reserve Bank of India namely, Shri R. Gandhi, New Delhi, Smt. Suma Verma, Hyderabad, Shri B. Mahapatra, Kolkata, Shri Franklin Joseph, Chennai and Shri A.K. Hirve, O-in-C, Bangalore is gratefully acknowledged for helping to organise the meetings and enabling the HLC to have stimulating discussions with organisations at these places. Smt. S.Chinngaihlian and Shri R.L.Taneja, DEAP, RBI, New Delhi provided exemplary support in organising various HLC’s meetings.

The HLC sincerely thanks the Reserve Bank of India for providing the Secretariat to this Committee. The HLC appreciates the professional skills and expertise of all the officers of the RBI who were associated in the preparation of the Reports of the Sub-Committees and the main Report. Dr. Deba Prasad Rath has ably supervised and guided the team of officials involved in this exercise from the Secretariat. In particular, the HLC acknowledges the rich contributions made by Shri P.K. Nayak, Dr. Jayanthi Anand, Smt. Gunjeet Kaur, Smt. Rekha Misra, Shri L. Lakshmanan, Shri Ajay Prakash, Smt. Atri Mukherji, Shri Raj Rajesh, Shri Naveen Kumar, Shri Rakesh Kumar, Dr. Ramesh Golait, Shri S.M. Lokare, Shri S. K. Takle and Shri Pankaj Kumar for providing competent secretarial assistance. The HLC also acknowledges the assistance provided by RBI officials Dr.A.S. Rama Sastri, Dr.Seema Saggar, Dr.Abhiman Das, Dr. G.P. Samanta and Mr. Ramesh Jangili. But for the tireless efforts of all these officers, this work could not have taken the present shape. The Committee records its deep appreciation for the service and professional support rendered by the Secretariat.

(C. Rangarajan)
Chairman
December 24, 2008

Contents

SECTION I: BACKGROUND TO THE CONSTITUTION OF THE HIGH LEVEL COMMITTEE

I.1 In view of the critical role played by savings in the growth process, it is widely recognised that it is important to have reliable and timely estimates of domestic savings and investment. The issues relating to the estimation of gross domestic savings and household savings, in particular, have assumed immense significance in the recent times. These issues are very relevant when the Indian economy has commenced its journey on a high growth trajectory. The Eleventh Five Year Plan (2007-08 to 2011-12) has set high growth targets, which pre-supposes robust savings and investment rates.

I.2 In India, the methodology adopted for estimating savings and investment has evolved over the years in tune with the international guidelines and improvements in the domestic statistical system. Nonetheless, it is felt that there is a need to critically review the available estimates of savings and investment in the Indian economy with respect to data base, methods of estimation, reliability and interpretational significance. The compilation of savings of the household sector continues to pose a challenge in view of the heterogeneity and residual character of this sector in the National Accounts Statistics. While countries like the US and the UK have established mechanism to directly estimate the savings of their respective household sector through income-expenditure surveys, India follows a unique way of estimating the household savings indirectly by a method, which is a mix of the flow-of-funds and commodity-flow approaches, where the household sector’s share is extracted as residual. There are pros and cons in the direct as well as in the indirect approaches. However, India’s present practice is considered conceptually sound and it has been endorsed by the earlier Committees, namely, K.N. Raj Committee and Raja Chelliah Committee in addition to some reputed economists. Notwithstanding the conceptual strength of the estimation practice followed by India, it is conceded that the present method is recognised to have weaknesses emanating mainly from data quality, data gaps and estimation problems.

I.3 It is felt inter alia that while financial deepening has been taking place at a rapid pace in the Indian economy, there is a view that the consequent effects are not being captured in the estimates of household financial saving. There is literature that argues that financial innovations may relax borrowing constraints for households and informal enterprises through several ways: it can give more households access to credit (sometimes termed as democratisation of credit), it can increase the amount of credit available to households that already have some credit and it can reduce the cost of borrowing (Dynan et al, 2007). Illustratively, in the US, such innovations have made it much easier for households to borrow against their housing wealth and accentuate the effect of housing price on debt. In respect of household financial savings, there is need to assess whether current state of financial deepening - an outcome of financial sector liberalisation - is being adequately captured by the data across various financial instruments.

I.4 In this context, it needs to be emphasised that in the indirect system of estimation financial savings can only be estimated in ‘net’ terms where the financial liabilities of the household sector are deducted from the gross financial assets, while physical savings are invariably financed by such financial liabilities. Many eminent economists have been pointing to various lacunae in the present practice. Some of them, in particular Professor T. N. Srinivasan, strongly, feels that India should devise a system of direct estimation of household savings based on integrated household surveys providing estimates on income, expenditure and savings of the households. Once such attempt albeit partial and on a pilot basis was made by National Sample Survey Organisation (NSSO), but the estimates thus obtained were not found to be entirely realistic and reliable. With the changing times, with the enormous experience that the country has in sampling surveys, and with better design of integrated household survey schedules, the feasibility of directly estimating household savings through integrated income and expenditure surveys merits consideration. In respect of the private corporate sector, there is a felt need to review the current methodology used in the light of better availability of corporate sector data due to widespread computerisation of the corporate balance sheets and income expenditure accounts and expand the scope of the samples used in estimating the corporate sector savings. It is also felt in some quarters that there is need to examine whether it would be appropriate to make corporate savings estimates on ‘marked-to-market basis’ (MTM) or to go by the ‘present book value method’. In respect of the public sector, the savings and investment estimates can be further strengthened by improving the coverage to include municipalities, city corporations, gram panchayats and other local Government bodies, on the one hand and increased private participation in public investments, on the other. Similarly, there is need to examine the current methodology of estimation of capital formation from various perspectives.

I.5 In the contemporary global setting, the estimation and interpretational issues pertaining to savings and investment have become critical. As alluded to above, savings estimates can be direct or indirect, it could be survey based or flow-of-funds based. However, it is recognised that problems related to savings are not specific to India alone, as one finds the interplay of these issues in different countries, including developed countries like the US, where debates are persisting with regard to private saving and inclusion of capital gains in the savings estimates. What is important, therefore, is that the estimates of savings need to be reasonably robust and these estimates have to be cross-validated through estimates emerging from different data bases and methodologies to use them as policy inputs.

Constitution of the High Level Committee

I.6 Against this backdrop, the Reserve Bank of India (RBI) suggested to the Government of India that the appointment of a High Level Committee will go a long way to review and place the estimation of household savings on a sound footing and to initiate an exercise to improve the methodology of savings and investment estimation. Recognising the criticality of these issues, the Government of India appointed the present High Level Committee on Estimation of Savings and Investment (referred to as HLC hereafter) under the Chairmanship of Dr. C. Rangarajan, the then Chairman of the Prime Minister’s Economic Advisory Council on December 12, 2007. The Committee was expected to critically review the existing methodologies used to estimate savings and investment aggregates for the Indian economy and suggest measures for improvement. The terms of reference assigned to the HLC are wide-ranging.

Terms of Reference (ToRs)

The ToRs assigned to the High Level Committee are as follows:

i In the light of the higher growth path of the economy, to undertake a critical review of the available estimates of domestic and national savings and investment in the economy, both in the aggregate and its components with respect to data base, methods of estimation, reliability and interpretational significance;

ii.To examine if rapid financial deepening in the economy is getting duly reflected in the estimates of financial savings and suggest improvements, if needed;

iii. To examine the feasibility of directly estimating household savings through integrated income and expenditure surveys;

iv.To examine the feasibility of arriving at separate estimates for pure households, household enterprises and unincorporated bodies through a suitable method;

v. To examine savings in the farm sector in relation to investments;

vi.To examine if corporate savings estimation should be done on marked-to-market basis or the present book value method;

vii. To suggest improvements in the methods and procedures used in the estimation of corporate investment and savings;

viii. To recommend methods of strengthening public sector savings and investment estimates by taking account of municipalities, city corporations, gram panchayats and other local Governments on the one hand and increased private participation in public investments on the other;

ix.To examine the empirical methods and procedures used in the estimates based on commodity-flow-method and flow-of-funds method and suggest improvements therein;

x.To suggest new data bases, if any, to be devised/built-up for improving the reliability or checking validity of the estimates; and

xi. To review the existing methodology and suggest improvements in the estimation of capital formation at the regional level.

Composition of the Committee

The HLC constituted by the Government of India has the following composition.

Constitution of the Sub-Committees

I.7 The first meeting of the HLC on Estimation of Savings and Investment was held on January 24, 2008 at RBI, New Delhi. After due deliberations, the Chairman suggested that for the purpose of addressing various terms of reference, the emphasis of the Committee should be on a review of the existing methodology used in estimation of savings and investment, a review of data sources to assess the data quality and ascertain data gaps and verification and validation of data.

I.8 For this purpose, the Chairman constituted four Sub-Committees, dealing with household sector savings, public sector savings, private corporate savings and capital formation. It was decided that the Sub-Committees may co-opt experts in the subject as per the need. On completion of their areas of investigation, the Sub-Committees would report to the HLC, which would deliberate on all relevant issues across these sectors. The HLC was expected to submit its Report to the Government within six months from the date of its Notification, which, however, was extended for a period of another six months till December 26, 2008.

In pursuance of the decisions taken in the first meeting of the HLC, the following four sub-committees were constituted:

I.9 It was decided that the ToRs assigned to the four Sub-Committees would include all the ToRs of the High Level Committee on Estimation of Savings and Investment relevant to their respective specific sectors and areas of estimation.

Secretariat of the High Level Committee

I.10 The Reserve Bank of India provided the Secretariat to the HLC. The Department of Economic Analysis and Policy (DEAP) coordinated the organisation of all the HLC meetings, consolidation of the four Sub-Committees’ Reports and assisted the Chairman in drafting of the main Report. The Department of Statistics and Information Management (DSIM) was responsible for the Report of the Sub-Committee on Private Corporate Sector. The DSIM also assisted DEAP in going through the draft of the HLC Report.

Scheme of the Report

In pursuance of the above approach and keeping the ToRs in view, the Report is organised in four sections as shown below
:
I.11 Section I, spread over five chapters, provides the background to the formation of the High Level Committee with introductory observations on an overview of the trends in savings and investment in the Indian economy and their interpretational significance, an overview of the existing methodology in India and select country practices in estimation of savings and investment and the major issues in the estimation of savings and investment. It sets out the broad approach adopted by the HLC and a summary of its deliberations.

I.12 Section II, spread over five chapters, forms the core part of the Report in terms of providing the HLC’s analysis on the estimational issues of savings for the three constituent sectors, viz., the household sector, private corporate sector and public sector and then on the estimation of capital formation. This forms the Sub-Committees’ work on the relevant ToRs assigned to each of them. From the point of view of the National Statistical Commission’s future work, the HLC’s observations and recommendations made in these four chapters pertain mainly to the aspects of accuracy, reliability and timeliness underlying the estimates of savings and investment. For operational convenience, these have been put together as Annex to the Report at ANSC 1.

I.13 Section III, spread over five chapters, puts together the HLC’s analysis on the other issues covering the remaining terms of references with regard to financial deepening and its reflection in the financial savings estimates, a review of the empirical methods and procedures used in the estimates based on flow-of-funds method, farm sector savings in relation to its investment, feasibility of arriving at separate savings estimates for pure households (consumer households), household enterprises and unincorporated bodies. It also deals with one of the issues on the impact of high current transfers on savings as it is recently being debated in certain circles.

I.14 Section IV provides the following: (i) Executive Summary, (ii) Annexes to the chapters including the recommended formats for data supply from apex financial bodies to the compiling agencies, namely RBI and CSO, (iii) Annex for the NSC, (iv) Statistical Annex, (v) list of abbreviations used, and (vi) select references used in the Report.

Chapter II : An Overview of the Trends in Savings and Investment in the Indian Economy and their Interpretational Significance¹

II.1 Savings and investment are important drivers in taking the economic growth process forward. Economic history suggests that countries that were able to accumulate high levels of domestic investment largely financed by domestic savings achieved faster rates of economic growth and development. There is an ongoing debate on the role of savings and investment in promoting economic growth. While the Harrod-Domar Model identified investment as the prime contributory factor, the Solow Model emphasised on savings. According to the conventional perception, savings contribute to higher investment and, hence, higher GDP growth in the short-run (Bacha, 1990; Jappelli and Pagano, 1994). The central idea of Lewis’s (1955) traditional development theory was that increasing savings would accelerate growth. Conversely, several more studies have concluded that economic growth contributes to savings (Sinha and Sinha, 1998; Salz, 1999; Anoruo and Ahmad, 2001). Caroll, Overland, and Weil (2000) demonstrated that “if utility depends partly on how consumption compares to a habit stock determined by past consumption, an otherwise-standard growth model can imply that increases in growth can cause increased savings.” Bacha (1990), Otani and Villanueva (1990), and Jappelli and Pagano (1994) concluded that a higher savings rate led to higher economic growth. Furthermore, a study of 32 countries by Krieckhaus (2002) notes that a higher level of national savings led to higher investment and consequently, caused higher economic growth. Besides, macroeconomic stability, factors like inflation, public investment, exchange rate policy, income and wealth play a vital role in the determination of savings and investment.

II.2 In an international perspective, India has had a relatively high savings rate as compared to many other countries, except those in East Asia (Athukorala and Sen, 2001). Inter-country experiences with regard to the links amongst savings, investment and growth appear complex, divergent and country specific with the conclusion that the positive effect of savings on growth is more straight-forward, and higher savings raises the growth rate of output by increasing capital accumulation. The relationship between savings rate and growth has been found to be bi-directional and positive for south-east and south Asia (Dholakia et al, 2008). The process of economic growth hinges critically on the generation of greater savings and its channelisation into productive investment. There are a host of channels for such a process to operate leading to a virtuous cycle of savings, investment and growth that is said to be operating in India now. Firstly, savings affects growth positively in those countries that are not too close to the technological frontier, but does not affect it at all in countries that are close to the frontier. Growth may result from innovations that allow the domestic sectors to catch up with the frontier technology. Secondly, lagged savings is significantly associated with productivity growth for poor but not for rich countries. This effect operates entirely through total factor productivity (TFP) rather than through capital accumulation. Thirdly, savings is significantly associated with higher levels of FDI inflows and equipment imports and that the effect that these have on growth is significantly larger for poor countries than rich (Aghion Philippe et al, 2006).

II.3 Over the decades, the secular uptrend witnessed in domestic growth is clearly associated with the consistent trends of increasing domestic savings and investment. A review of the performance of the Indian economy suggests that there is a quantum jump in the real GDP growth rate in the post-reform period. Manufacturing and services sector have emerged as the growth drivers. Gross domestic savings have increased continuously from an average of around 10.0 per cent of GDP during the 1950s to almost 33 per cent of GDP at present; over the same period, the domestic investment rate has also increased continuously from around 11 to 33 per cent (Table 2.1). The behaviour of the savings rate and economic growth in India during the reform period seems to suggest that the high growth phase is associated with higher order of increase in domestic savings.

II.4 The secular uptrend in growth is clearly associated with consistent trends of domestic savings and investment over the decades, with the latter also explaining the turning points from the traditional low growth to contemporary high growth phases since the 1980s. An empirical exercise conducted in RBI (2001) showed that the impressive growth story of India, particularly in the aftermath of reforms seems to have been facilitated by the improvement in the rate of aggregate domestic savings. Empirical relationship between the changes in GDP and incremental savings provides the evidence of a bi-directional causality - highlighting the role of the feedback effects emanating from savings to economic growth and vice versa (RBI, 2001). The behaviour of the savings rate and economic growth in India during the reform period seems to suggest that the high growth phase is associated with higher order of increase in domestic savings. A noteworthy feature of these trends is that Indian economic growth has been financed predominantly by domestic savings. The recourse to foreign savings – corresponding to the current account deficit – has been rather modest in the Indian growth experience.

¹The data used in this Report is sourced from NAS 2008 and the back series (base 1999-2000) published by CSO, GOI, New Delhi.

II.5 The restructuring measures initiated by the industry in a competitive environment, improvement in the corporate profitability, conducive investment climate and favourable monetary and fiscal policies have led to a significant pace of growth in the Indian economy.

II.6 A look at the trends in savings (Table 2.2) shows that the rate of gross domestic savings (GDS) as a proportion of Gross Domestic Product at current market prices (GDPCMP) has more than doubled from an average of around 10 per cent in the 1950s to around 23.0 per cent in the 1990s and to 29.2 per cent in the 2000s so far. It scaled a peak of 34.8 per cent in 2006-07, the highest saving rate ever achieved in India since 1950-51. Similarly, the rate of gross domestic capital formation (GDCF) i.e. GDCF as a proportion of GDPCMP, has more than doubled from an average of around 11 per cent in the 1950s to around 24.4 per cent in the 1990s and to 29.1 per cent in the 2000s so far.

Some of the salient features of savings and investment trends in India are as set out below:

• More than 98.0 per cent of investment is funded by domestic savings. As a result, the savings-investment gap has remained relatively in a narrow range reflective of modest reliance on foreign savings.
  
  
• There is a turn-around in the public sector savings, which increased from (-) 2.0 per cent of GDP in 2001-02 to 3.2 per cent of GDP in 2006-07.
  
  
• With a rise in corporate profitability the rate of private corporate savings has more than doubled to 7.8 per cent in 2006-07 from 3.7 per cent in 2001-02.
  
  
• Correspondingly, the rate of private corporate investment has almost tripled to 14.5 per cent in 2006-07 from 5.4 per cent in 2001-02.
  
  
• The rate of household financial savings increased to 11.3 per cent 2006-07 from 10.8 per cent in 2001-02. According to RBI’s preliminary estimates, the rate of household financial savings is placed at 11.2 per cent for the year 2007-08.

Sectoral Savings in India

II.7 As stated earlier, India continues to remain one of the high savings economies among the emerging market economies. Gross Domestic Savings (GDS) of the Indian economy constitutes savings of public, private corporate and household sectors. In the recent period, the high growth performance of the Indian economy is driven by rise in savings and investment. The buoyant trend in the gross domestic savings is powered by savings in household sector until recent past, more recently by the corporate sector and the public sector. It is significant to note that the increasing trend in gross domestic savings as a proportion of GDP since 2001-02 has continued with the savings rate rising sharply from 26.4 per cent in 2002-03 to 34.8 per cent in 2006-07 (Table 2.2 and Statistical Annex).

II.8 At a disaggregated level, it is the household sector, which occupies a position of dominance over the other institutional sectors like private corporate sector and the public sector in terms of generating savings (Chart 2.1). Savings by public sector and private corporate sector are improving in recent years. From 1999-2000 to 2002-03, the declining trend of the public sector savings (from 0.6 to -2.0 per cent) was a cause of concern. However, during the recent period it witnessed improvement. The rate of savings in private corporate sector increased from 4.4 per cent 2003-04 to 7.8 per cent in 2006-07 (Chart 2.2).

i) Household Sector

II.9 In India, household sector occupies a position of dominance over the other institutional sectors like private corporate sector and the public sector in terms of generating savings. Household savings is composed of both financial and physical savings. As a percentage of GDP at current market prices, the rate of savings of the household sector increased from around 7.0 per cent in the 1950s to over 18.0 per cent in the 1990s. In 2006-07, it stood at 23.8 per cent (Chart 2.2). Within the household sector savings, the rate of savings held in financial assets steadily increased during this period. Since 2000-01 the household sector has shown a preference for savings in the form of physical assets, which could be attributed partly to the soft interest regime in recent years. Increase in the rate of household savings in physical asset in recent years reflects booming construction activities mainly of housing and accelerated industrial activities requiring machinery and equipments.

ii) Private Corporate Sector

II.10 The rate of savings of the private corporate sector witnessed a steady increase from 1.0 per cent of GDP in the 1950s to 1.7 per cent in the 1980s, to rise to 3.8 per cent in the 1990s and further to 7.8 per cent in 2006-07. In terms of composition, the share of the private corporate sector savings in the GDS also increased from 10.2 per cent in the 1950s to 16.3 per cent in the 1990s. By 2006-07, this increased further to 22.4 per cent. The rate of savings in private corporate sector since the last three years followed upward momentum, reflecting higher retained earnings resulting from higher profits. It may be mentioned that the savings rates of private corporate sector had been stagnant during the 1950s to 1990s.

iii) Public Sector

II.11 On account of sharp deterioration in the savings of the Government Administration, the rate of savings of the public sector, which witnessed an increasing trend till the 1970s, started declining thereafter, and turned negative since 1998-99. However, from 2003-04 onwards savings of public sector turned positive, reflecting mainly the outcome of the implementation of Fiscal Responsibility and Budget Management Act (FRBM Act), 2003. Public sector savings rate increased to 2.2 per cent in 2004-05 from 1.0 per cent in 2003-04, resulting from lower dis-savings by public authorities as well as improvement in non-departmental enterprises savings. Savings of public non-departmental undertakings is a major component of public sector savings, which witnessed steady improvement since 1970s till the recent times (Table 2.2).

II.12 In sum, there is an across-the-board rise in savings in all the constituent sectors in line with the growth performance in the recent years.

• Important structural changes are discernible within the household sector savings with clear cut preference pattern changes year-to-year in financial savings and physical savings, and both have responded to a host of factors.
  
  
• Increasing role of corporate sector has driven it to the centre-stage in savings and investment.
  
  
• Important phenomenon within public sector includes FRBM which has impacted favorably the savings performance in the Government Administration.
  
  
• Non-departmental companies continue to remain significant in generation of savings within the Government sector.

Trends in Capital Formation

II.13 Chart 2.3 shows a co-movement between savings and investment at the overall level. At the disaggregated level, Table 2.2 provides the trends in capital formation since the 1950s for the institutional sectors. The increase in the rate of capital formation registered between the middle of the 1950s and in the middle of the 1960s was mainly in the public sector. This was on account of adoption of industrialisation in a big way by huge investments in public sector undertakings since the second five year plan onwards. The adoption of mixed economy after independence implied that the dominant source of investment would be public sectors with private sector playing only supportive role. The reliance on public sector for rapid industrialisation was mainly on account of absence of mature private corporate sector. The investment remained predominantly, thus, in the sphere of the public sector.

II.14 However, the economic reforms have reversed this trend with the private corporate sector overtaking the public sector in the realm of investment in the economy since 1994-95. Significant structural change is discernible in the investment behaviour of the economy in the recent past in terms of the change in relative shares of public and private investment. Years of reforms have marked a significant break from the previous trends in terms of increasing investment with major role being played by the private corporate sector in the last five years. The investment (gross capital formation) has increased sharply from 24.2 per cent of GDP in 2001-02 to 36.0 per cent during 2006-07 i.e., increase of 11.8 percentage point within five years.

II.15 It is observed that during the period (1999-2000 to 2006-07) the share of gross investment by household sector witnessed increasing trends from 1999-200 to 2003-04 (from 10.5 per cent to 13 per cent) (Chart 2.4). However, it was stagnant in the last two years.

II.16 In the same corresponding period, the investment in private corporate sector trebled from 5.4 to 14.5 per cent of GDP, thus, accounting for the overall increase in investment in the economy (Chart 2.5). The rapid increase in private sector investment in the aggregate investment is in large part a reflection of the impact of the reforms initiated in the 1990s, which reduced restrictions on private investment and created a more favourable investment climate. It reflects the fact that the private corporate sector has responded positively with improvement in the investment climate. The reduced requirement by the Centre for meeting budgetary mismatches, and for overall public sector financing has improved the availability of resources for the private sector considerably. Furthermore, the private corporate sector has responded to increased global competition by improving its productivity and efficiency through increased application of technology. The economic reform process has helped greatly in making the policy environment more conducive for more efficient entrepreneurial activity. The corporate tax rate was steadily reduced from 45 per cent in 1992-93 to 30 per cent by 2005-06 and was kept stable thereafter. Financial restructuring of firms has also led to reduction in overall debt-equity ratios in the corporate sector. The substantial reduction in debt servicing costs has thereby added to the private corporate sector’s competitiveness and profitability.

II.17 Another notable feature of performance of the private corporate sector in the recent period is the progressive increase in retained profits, which as a share of net profits [Profits after tax (PAT)], increased from 30.9 per cent in 2001-02 to 73.6 per cent in 2005-06. The improved profitability, reflecting improved productivity and lowering of tax rates, enabled corporates to deleverage their balance sheets. The improved corporate financial performance resulted in a more than doubling of the private corporate sector savings rate (from 3.4 per cent in 2001-02 to 7.8 per cent in 2006-07), which has also contributed to the pick-up in the overall savings rate. From a long-term perspective, it is interesting to observe that in addition to the rise in savings of the private corporate sector, higher retained profits along with availability of resources from the banking sector facilitated by the lower financing requirement of the Government and the increased access to the domestic and international capital markets led to a sharp increase in the investment rate of the corporates.

II.18 To sum up, significant structural changes characterise the trends in savings and investment in the economy over the period since the structural reforms were undertaken in the economy since the 1990s. While the households continue to retain their status of being financially the surplus sector, in combination with the private corporate sector, the two sectors, namely the private sector, has strengthened its relative position in the sphere of savings and investment vis-à-vis the public sector.

Interpretational Significance of the Savings and Investment Estimates

II .19 In the light of the higher growth path of the economy in recent years, a critical review of the available estimates of domestic and national savings and investment in the economy would be useful with respect to their interpretational significance. Savings and investment estimates are attracting centre stage attention globally from the standpoint of their implications for contemporary issues such as persisting global imbalances. As mentioned earlier, savings can accrue to the economy from three sectors - namely, household, corporate and public sectors. More importantly, interpretational aspects and measurement issues have come to the fore, particularly in household savings, which is a dominant savings sector in India where a number of issues related to the estimation are still unsettled. Interpretational issues have remained endemic to the savings estimates, surfacing from time to time leading thereby to formation of Committees to address these concerns.

Savings from a Macro-Economic Perspective

From a macroeconomic perspective, three important measures of savings are private savings, Government savings and national savings.

(a) Private Savings = Sp = Private disposable income – Consumption
= (Y + NFP +TR – T + INT) – C ——————————————————(1)

(b) Government Savings = Sg = Net Government Income – Government purchase
= (T – TR – INT) – G ——————————————————————(2)
Where Y = Gross Domestic Product, NFP = Net Factor Payment from Abroad, TR = Transfers received from the Government, INT = Interest Payment on the Government’s Debt, T = Taxes (including income from entrepreneurship and property), C = Consumption and

G = Government purchases (purchase of goods and services and compensation of employees).

(c ) National Savings = S = Sp + Sg

= (Y + NFP +TR– T + INT) – C + (T – TR – INT) – G
= Y + NFP – C – G
= C + I + G + NX + NFP – C – G
= I + NX + NFP
= I + CA ———————————————————————————(3)
Where NX = Net Exports and CA = Current Account Deficit.
Sp + Sg = Ip + Ig + CA—————————————————————(4)

In the Indian context, the gap between the savings and investment (Gross Capital Formation) by different sectors and its distribution for the year 1999-2000 (the latest base year for the National Accounts Statistics) is given in Table 2.3.

II.20 One analytical use of the national savings identity is to bring out an important short-run, long-run dichotomy. In the short-run, international borrowing can go up or down depending on cyclical movements, exchange rates, and a host of other factors. Very few countries do without any international borrowing or lending in a particular year. But over a longer period of time, international borrowing kept close to zero stabilises a country’s ratio of international obligations to its gross domestic product (GDP). When this liability ratio is stable, a nation’s savings is implicitly financing most of its domestic investment.

II.21 Secondly, persistent borrowing, keeping investment above savings, may be possible, but it is not common, and in any case it automatically implies a build-up of international liabilities relative to its national output. If a nation wants to have its investment and not pay increasing shares of income in interest or dividends, it has to finance this investment by its own national savings. Alternatively, turning the equation around, high national savings will raise future living standards, whether it finances investment directly or reduces international borrowing.

II.22 Thirdly, the identity (4) ties savings to budget and current account deficits. Both deficits are subtraction items.

II.23 The savings-investment gap in India is relatively low reflective of modest reliance on foreign savings. Among the domestic sectors, contribution-wise, the private corporate sector savings, of late, has turned out to be a significant driver in domestic savings and together with household sector, is contributing to the generation of private savings. Although the public sector savings remained in a dis-savings mode for a long time, there is a turn-around recently. Significantly, these changes are occurring when the economy remained on a high growth trajectory in the last five years.

II.24 Cross-country differences in the evolution of gross corporate savings over the past half decade have tended to reflect to a large extent those of the gross operating surplus. Such differences across countries are likely to be due to countries’ exposure to global factors (accelerated globalisation and technological progress), presumably depending on institutional framework conditions (such as product and labour market regulations), as well as the sectoral composition of their economies.

II.25 As mentioned earlier, there is an across-the board rise in savings by all the constituent sectors in a period of high growth particularly in the post-reform period of the 1990s and 2000s. To the extent there are compositional shifts within the components of savings happening in conjunction with the structural transformation of the Indian economy, these trends have been recognised as structural in nature in the literature. Within the constituent sectors, the household financial savings, the non-departmental undertakings within the Government sector and the non-financial joint stock companies within the private corporate sector contribute predominantly to the generation of domestic savings in the era of reforms.

Interpretational Difficulties that have led to Frequent Questions on Estimation

II.26 Amongst the components of national accounts statistics (NAS), it is the savings and investment estimation on which the authorities have cast doubts at frequent intervals on the grounds of interpretational difficulties faced by them in explaining some aspect of macroeconomic development or the other. Working Groups/Expert Groups were appointed to enquire into the questions and they have all come out with recommendations for improvements in the methodology of savings and investment estimation.

II.27 It may be recalled that Raj and Chelliah Expert Groups were asked to look into some of the interpretational issues then existing. First, there was the K.N. Raj Working Group (1982), which was set up when it was found that,

II.28 “The rates of gross capital formation and of gross saving have gone up significantly since the early years of the 1970s, to about 24 per cent of the gross domestic product by 1978-79; even net capital formation and savings were touching rates of around 20 per cent of the national income. This has raised a number of questions, in regard to both the estimates themselves and their interpretation, particularly since there has been evidently no corresponding improvement in growth momentum within the economy and the problems of mobilisation of resources for development seem still as severe as ever before” (RBI, 1982).

II.29 Second, there was the Chelliah Expert Group appointed in the context of the contrary development when savings fell and growth improved. It was found that:

II.30 “An important reason for this lagged recovery in private savings is the decline in the absolute amount of household savings in 1992-93 as estimated by the CSO. Since this decline is observed for a year in which the GDP growth rates rises from 0.9 per cent to 4.3 per cent and GDP growth from agriculture shows a sharp turnaround from (-) 2.5 per cent to plus 5.3 per cent, there appear to be grounds for reviewing the methodology for estimating, saving and capital formation in the economy [Economic Survey 1994-95:3].

II.31 The Chelliah Expert Group itself argued that there were large discrepancies in the estimates of saving and investment, with saving being far in excess of investment – “the order of discrepancy being almost 2 per cent of GDP in 1994-95. Indeed, this has raised questions about the quality of the estimates, and the methodology adopted for collecting data” (Department of Statistics 1996:5).

II.32 More recently, the Report of the Working Group on Savings for the Eleventh Five Year Plan, 2007-08 to 2011-12 (RBI, 2007) argued that “the procedures currently adopted for estimation of household savings and its projection needs a re-look as we move along”. This is because the estimates of household financial savings are showing a decline, despite signs of financial deepening - a puzzle for the Working Group, which has impliedly made out a case for fully measuring household income, expenditure and savings through integrated nation-wide field surveys so as to capture the complex interplay of a number of variables impacting on household savings.

II.33 The important observations that the above Committees have made on the interpretational issues of savings and investment estimates are given below.

II.34 First, the relatively small accretion to the savings rate during the ‘eighties’ underscored the need for undertaking measures to raise this rate or/and lower the incremental capital output ratio (ICOR) for effecting improvement in the overall growth rate. This position changed subsequently with rise in growth since the eighties.

II.35 Second, a sharp increase in the preference for financial assets as against physical assets in the 1990s implied that this preference is essentially a reflection of financial widening in the sense of availability of a larger menu of financial assets, and of a sharp reduction in financial controls. The growing flexibility that investors enjoyed in terms of choice of assets in the increasing financial portfolios and in terms of optimisation of rates of return with risk minimisation furthered by financial sector reforms. The Chelliah Committee (1996) noted that as the reforms are intensified, financial savings are bound to play a more crucial role in economic development than hitherto.

II.36 Third, apart from undertaking structural and institutional measures for influencing productivity, it is imperative that the investment rate in the economy is higher than the GDS rate essentially to ensure that savings of a relatively underdeveloped country such as ours is not exported. One of the aspects of the estimates of savings and investment in India is the discrepancy in the estimates of GDS and gross domestic capital formation (GDCF) in most of the years. The Chelliah Group believed that GDCF estimates are subject to errors in several respects, and methodologies need to be, therefore, improved. The savings in the form of physical assets of the household sector being a part of the estimate of capital formation of the economy would reflect the errors, which creep in the estimates of total capital formation in the economy. As physical assets of the household sector are common and critical to both the aggregates, the Chelliah Group thought it desirable to have this component independently estimated.

II.37 Fourth, the problem of obtaining the data on a regular basis is severe in the case of the household sector. This sector comprises not only farm households engaged in agricultural production, but also individual households and unincorporated enterprises engaged in industry, trade, transport, finance, private trusts, etc. The estimates for the household sector are worked out on the basis of the available data from various censuses, sample surveys, and research studies and assumed relationships. The estimates, therefore, would have errors.

II.38 Fifth, as is known, number of ratios and norms are used in the preparation of estimates of capital formation for the economy as a whole and in respect of the household sector. These ratios are based on data of somewhat remote past and their use may be unwarranted in the light of the structural changes taking place in the economy. If these ratios are updated, the estimation of capital formation would considerably improve in qualitative terms and the relative size of errors and omissions would probably come down.

II.39 Sixth, according to Mulheisen (1997), the interpretation of Indian savings trends is complicated by a number of weaknesses in the methodology for measuring both investment and savings. The most important shortcomings are:

• The estimate for physical household savings is set equal to household investment, which itself is calculated only indirectly as a residual. Measured physical household savings has been highly volatile
  
  
• There are errors and omissions in the estimates of both savings and investment, but adjustments are made only to investment. The underlying assumption is that the savings estimates are more reliable (based on the greater accuracy of public, financial, and corporate savings data) and, therefore, investment is adjusted to equal the sum of domestic and foreign savings. Reversing the present CSO practice, when domestic savings is adjusted to include errors and omissions, so that the sum of adjusted domestic savings and foreign savings equals the original investment estimates. This yields a much smoother, more plausible path for domestic savings (Mulheisen, 1997).
  
  
• The estimates of corporate savings and investment are based on a small, unrepresentative sample, and rely largely on voluntary responses from enterprises.

• Poddar and Yi (2007) find that a 1 percentage point decrease in the dependency ratio adds 0.8 to the national savings ratio.
  
• Hiroko Oura (IMF 2007) finds that 10 per cent increase in the savings ratio between 2005 and 2025 is on account of fall in dependency ratio.

II .45 The household savings are also driven in recent years by the physical savings, especially since 2000-01 when the household sector has shown a preference for savings in the form of physical assets. This could be attributed partly to the soft interest regime in recent years. Increase in the rate of household savings in physical asset in recent yearsreflects booming construction activities mainly of housing and accelerated industrial activities requiring machinery and equipments.

II.46 While countries like the US and UK have established a mechanism to directly estimate the savings of their respective household sector through income-expenditure surveys, India follows a unique way of estimating the household savings indirectly by a method, which is a mix of flow-of-funds and commodity flow approaches, where household sector’s share is
extracted as residual.

II.47 The estimates of financial savings in respect of households including non-profit institutions and unincorporated private business are estimated by changes in the net financial assets held by them such as: currency, deposits with financial institutions, shares and debentures, claims on Government, net equity in the life funds, provident and pension funds net of changes in financial liabilities. The estimates in respect of various financial instruments are arrived at as a residual after duly accounting for such instruments held by public and private corporate sectors. In estimating the financial savings of the households, increments in their holdings of financial assets are calculated net of increments in their financial liabilities. These annual flows are compiled instrument-wise. As the household sector is an unorganised sector and its balance sheets are not available, its financial flows are either estimated on the basis of firm information obtained from the accounts of counterpart institutions engaged in transactions with the households or as a residual after deducting the accounted financial information for the other sectors from the financial totals or are based on the information collected on sectoral distribution either directly or on the basis of surveys. Select methodological problems are the following.

II.48 First, in absence of comprehensive income-expenditure surveys, instrument-wise savings estimates are prepared in an indirect manner through the residual treatment accorded to household sector in the flow-of-funds accounts. The difficulty here is the absence of cross-validation from a direct survey. Savings is estimated either from the income-expenditure account or from the flow-of-funds account, as earned net worth, being the difference between change in asset and liabilities, adjusted for capital gains and losses. Theoretically, these two measures of savings should yield identical results. In US and UK, both the estimates are available, which help in cross-validating the income-expenditure survey based direct estimates of savings. But in India, only flow-of-funds based estimation is available for household financial savings and commodity flow based estimation is available for household physical savings.

II.49 Second, in the absence of income-expenditure survey, there is no direct estimation of household savings and as a result there is no cross validation of the indirect (residual) estimates of savings.

II.50 Third, the estimates of domestic savings are presently prepared only at current prices. The estimates at constant prices have not been attempted so far as the suitable procedure of estimation is yet to be evolved.

Some Marked Features of Trends in the Savings in India

Predominance of Household Savings

II.51 A remarkable feature of the Indian macro-economic story since independence has been the continuous rise in household savings over the decades. Since the 1950s, the household sector has remained the predominant source of the GDS, and it contributed, on an average, around 74.0 per cent to total domestic savings during the last five years ending 2006-07. Conversely, private corporate sector accounting for 20.0 per cent, occupies second place followed by the public sector (6.0 per cent) (Chart 2.6).

Structural Shift in the Household Savings

II.52 A salient feature of the 1990s was the rising trend in the household sector savings. Within the household sector savings, the rate of savings held in financial assets steadily increased during this period. Financial liberalisation has an important bearing on financial savings as it involves the creation of newer instruments and avenues of savings, and also reduces intermediation costs (McKinnon, 1973). In India, the phenomenon of improvement in financial savings could be attributed to more efficient financial intermediation, greater opportunities for diversification across financial assets and emergence of market related returns.

II.53 Over time, although both financial and physical savings have recorded an increase, the composition of household savings has seen a shift in favour of financial savings reflecting the spread of banking and financial services across the country. The share of household savings in physical assets in the total household savings declined from more than 70.0 per cent in the 1950s to around 53.5 per cent in last five years ending 2006-07. Concomitantly, the share of household financial savings in the total household savings increased from around 25.0 per cent in the 1950s to around 47.0 during the same period. The rate of financial savings increased from less than 2.0 per cent in the 1950s to over 11.0 per cent during 2003-04 to 2006-07 (Chart 2.7). The spread of the financial sector, in particular of bank branches, post office savings and the likes, seems to have helped in mobilising household financial savings.

II.54 However, since 2000-01, the household sector has shown some preference for savings in the form of physical assets, which could be attributed partly to the soft interest rate regime in recent years, substantial growth in self-employment giving rise to a large number of informal sector activities as well as the rising expectation arising from rapid appreciation of the value of investments in housing. The ongoing financial deepening is facilitating larger access of bank credit for the households. As a result, household financial savings have increased only marginally from an average of 10.3 per cent during 1997-2003 to 11.3 per cent of GDP during 2003-2007. On the other hand, with increased availability of housing finance, household sector’s investment rate (physical savings) increased from 10.5 to 12.7 per cent during the same period. Thus, the widening of Savings-Investment gaps of the public and private corporate sectors combined was partly financed from household financial savings and partly by foreign savings.

Changing Composition of Household Financial Savings

II.55 The rate of household savings increased from 18.4 per cent in 1990-91 to the peak level at 24.4 per cent in 2003-04 and subsequently it declined to 23.8 per cent in 2006-07. At the average level, the rate of household savings is rising from 17.3 per cent during the five year period (1990-95) to 18.1 per cent during 1996-2000 and further to 23.2 per cent during 2006-07.

II.56 ‘Bank deposits’ is an instrument of great significance in that it constitutes the largest proportion of household financial savings (Table 2.4). Its share which fell during the 1980s, has been recovering since the 1990s. Notwithstanding financial innovations, bank deposits continued to be the most important instrument of financial savings among the households during the period of reforms. The buoyancy in bank deposits over the recent years partly reflects some migration from small savings; as this signifies only a shift in the asset portfolio composition of households.

II.57 Another important feature of household savings during the reform period has been the increasing importance of savings held in insurance funds, and provident and pension funds (together described as contractual savings). Within financial assets, preference has shifted from savings in the form of shares and debentures and deposits to claims on Government and contractual savings. A major component of contractual savings, viz, savings in life insurance funds, amounted to around 14.6 per cent of total savings during 2003-04 to 2006-07 much higher compared to the earlier periods (Table 2.4). Demographic variables like life expectancy, literacy rate and dependency ratios have emerged as key determinants of savings in addition to traditional variables like real interest rate, growth, per capita income, spread of banking facilities and rate of inflation (Athukorala and Sen, 2001). Apart from ensuring assured rates of return with tax exemptions, contractual savings also provides old age security. As disposable income rises and life expectancy continues to increase, concerns regarding old age security can be expected to result in increasing share of financial savings held in contractual instruments. Going forward, improvement in financial savings would depend on the further deepening of the financial sector, particularly through the continuation of insurance through participation of private insurance companies and pension sector reforms.

II.58 ‘Shares and debentures’ is another instrument of financial savings that was expected to get a boost from the reforms. Measures undertaken for developing the capital market were expected to divert savings from the traditional financial instruments to the capital market instruments. Reflecting this, the proportion of household savings in ‘shares and debentures’ (inclusive of investment in mutual funds) rose steeply to over 8 per cent of total financial savings in the initial years of the 1990s from 4 per cent in the 1980s. This was due to a shift away from the relatively safer modes of savings, such as small savings instruments (included under ‘claims on Government’). Following the irregularities in stock market in 1992 and the associated price uncertainty that prevailed in the subsequent period, the proportion of household financial savings held in ‘shares and debentures’, however, witnessed a decline to reach a low of around 4.0 per cent in the subsequent period. The pattern of financial savings that emerges during the decade of reforms, thus, indicates a continued preference of households for relatively safer instruments with assured returns (RBI, 2001-02).

II.59 In sum, the rise in household savings is attributed to both household financial andb physical savings. Furthermore, at the average level, there are clear changes in the preference pattern as seen in the rise in the rate of contractual savings (mainly life insurance funds) and decline in the rate of net deposits, while as per the related literature, the quantum of the household savings tracks a host of theoretically well-established determinants, the prime among them being the growth in per capita income, financial liberalisation, greater opportunities for diversification across financial assets and market related returns. A succinct discussion can be found in Athukorala and Sen (2001), EPWRF-NCAER study (2003) and Report of the Working Group on Savings for the Eleventh Five Year Plan (2007). Financial savings have gained increasing importance and to that extent, the existing methodology has captured these critical economic changes rather well as seen from the savings estimates.

Quality and Limitations of Data Base

II.60 It may be stated that there is the possibility of under/over estimation bias in the present estimates of household sector’s savings. Some of the contributing factors have been identified in the Sources and Method, CSO, 2007. According to CSO, the savings estimates suffer from a number of limitations, mainly from deficiency of data. These are:

• Data on Co-operative societies are not available for all the years. There is a time lag of four to five years.
  
  
• Estimates for the household sector are built up in two parts (i) savings in financial assets, and (ii) savings in physical assets.
  
  
• In case of savings in financial assets, the estimates of household savings in currency are estimated as 93 per cent of total currency on the basis of past trends of currency holding.
  
  
• The estimates of provident fund of non-departmental enterprises are obtained by doubling the employers’ contribution as available in the books of accounts. No adjustments are made for withdrawals, interest, etc., since the data on these items are not available.
  
  
• The non-Government educational institutions are being covered under the Employees Provident Fund Act, 1952 since 1981-82. However, the coverage of private educational institutions has been incomplete as many of these institutions are not making any returns to the CPFC. Thus, the estimates of PF can be improved only when the coverage of private educational institutions improves and reliable data on withdrawals, interest, etc., in respect of non-departmental enterprises becomes available.
  
  
• The results of the RBI survey on ‘Growth of Deposits with Non-banking Companies and Ownership of Capital of Joint Stock Companies’ are not available beyond 1995-96. Estimates in respect of most of the other financial assets are based on current data, though, they mainly refer to total savings under each of these categories and households’ contributions are obtained as residuals.
  
  
• The households’ savings in physical assets is also estimated by using the residual approach. The residual approach adopted for deriving savings estimates for households is undoubtedly not very satisfactory in all cases and has enough scope for improvement. The only approach through which this can be achieved is by conducting independent household surveys to collect direct data on income and expenditures , as far as pure households are concerned and by conducting annual surveys on household enterprises.

Private Corporate Savings

II.61 Private corporate sector, primarily characterised for its capital-intensive nature, contributed substantially to the rise in the overall domestic savings rate. The share of private corporate sector in overall savings increased markedly over the years and it stood at 22.4 per cent in 2006-07 as against average 9.5 per cent in the 1980s and 17.1 per cent in 1990s. It is interesting to observe that the rate of savings of the private corporate has increased from around 1.0 per cent in the 1950s, 1.5 per cent in the 1960s and 1970s, 1.7 per cent in the 1980s and 3.8 per cent in the 1990s to 7.8 per cent in 2006-07.

II.62 Generally higher growth in revenues vis-a-vis a slower growth in expenditure resulted in marked improvement in financial performance of the business enterprises in last few years. The reform process has helped the corporate sector favourably for more efficient business activity. The corporate tax rate is steadily reduced from 45.0 per cent in 1992-93 to 30.0 per cent by 2005-06 and was kept stable thereafter. The peak custom duty on non-farm goods was reduced gradually from 150.0 per cent in 1991-92 to just 10.0 per cent in 2007-08. Effective monetary policy has contributed to the sustained moderation of inflation leading to a reduction in nominal interest rates and, therefore, the interest burden.

II.63 The strong growth in profits reflective of improved productivity, increased efficiency through increased application of technology and lower tax rates, enabled corporates to de-leverage their balance sheets resulting in the sharp decline in the debt-equity ratio. Financial restructuring of firms also led to the reduction in overall debt equity ratios. The substantial reduction in debt servicing costs thereby added to the corporate sector’s competitiveness and profitability. Notable feature of this improved performance of the corporate sector in the recent period has been the progressive increase in retained profits, which as a share of net profits (PAT) increased from 30.9 per cent in 2001-02 to 78.3 per cent in 2006-07. The private corporate sector has financed a large part of its investment in the on-going long capex cycle from such retained earnings. Strong financial performance translated into a more than doubling of the private corporate sector savings rate from 3.4 per cent in 2001-02 to 7.8 per cent in 2006-07.

Public Sector Savings

II .64 The trends in savings of the public sector in the country in terms of rates (of GDP at market prices), since 1950-51 are presented in the following Table 2.5. From this table, it can be seen that there has been a gradual increase in the savings of the Nom-Departmental Commercial Undertakings (NDCUs) over the years, since 1950-51. However, in the case of administrative departments, till 1980s, there has been positive savings, but since then the savings of these departments has been negative, although the dis-savings is reducing gradually following the FRBM Act. The savings of Departmental Commercial Undertakings (DCUs) has more or less stagnated at about 0.6 per cent over the years, although there were decreases and increases in between.

II .65 The estimates of savings from the public sector by its various entities, since 1999- 2000, year-wise, are presented in Table 2.6. There were some changes made in the procedure/ analysis of the economic transactions of the budget analysis in the current NAS series. One such major change is identification of capital expenditure incurred under object head“office expenses”. In consultation with the State Governments and central ministries, it has been given to understand that about 25 per cent of the total expenditure incurred under the head “office expenses” relates to purchase of machinery and equipment like photocopiers, ACs, furniture, Computers, etc. Due to this, the CSO included part of purchase of goods services into other capital outlay, machinery outlay and software outlay, which in turn, resulted in additional gross fixed capital formation from current expenditure of goods and services. The consequent changes are that there is decrease in the final consumption expenditure and increase in savings as well as GFCF of public sector.

II .66 The second major change made in analysis of budgets in the new series of national accounts (base year 1999-2000) is treatment of operating losses of the DCUs as imputed subsidies. The rationale behind is that the DCUs incur losses on account of administrative policies of the Government and the losses are met through the Governments’ current revenues. This treatment was only restricted to irrigation departments earlier, but in the new series it has been extended to all DCUs incurring losses.

II.67 The role of public sector in propping up the domestic savings of the Indian economy, however, declined since the 1970s. The share of the public sector in gross savings of the country declined to 19.9 per cent in the 1980s, which further went down to 9.6 per cent in the mid-1990s. The public sector savings, however, turned negative for the first time in 1998-99. This declining trend continued for five years till 2002-03, after which the public sector savings become positive. Since 2003-04, the public sector savings have turned positive.

II.68 As per the sources of public sector savings, it is found that Government Administration accounted for a bulk of the savings in the 1970s – accounting for nearly 60 per cent of gross savings. However, the savings of Government Administration received a setback in the 1980s and since the 1990s, they have turned into huge dis-savings, which, in turn has led to declining savings in the public sector.

II.69 However, in the last two decades or so, a rise in the savings of the non-departmental enterprises has been instrumental in keeping the public sector savings positive except during 1998-99 to 2002-03, when huge dis-savings of Government Administration led to negative public sector savings. While the savings of the departmental enterprises have broadly remained lower.

II.70 The savings of the Government Administration had been positive since the 1970s, but it turned negative since 1987-88. What led to this? Possibly, fiscal profligacy in the late 1980s was responsible for this. On examination of the trends in revenue deficit, it is observed that it also started worsening from the year 1987-88. Trends in revenue deficit and savings of the Government Departments reveal that the dis-savings of the Government Administration were reflected in the rising revenue deficit (Chart 2.8). A rise in revenue deficit led to the widening of the dis-savings of the Government Administration departments.

II.71 The dis-savings of the Government Administration further worsened since the mid-1990s till 2001-02. This could have also been on account of the implementation of the Fifth Pay Commission award during 1997-98, which revised up the salaries of Government employees, which enhanced the Government expenses by about Rs.17,000 crore (Chart 2.9).

II.72 In the recent years in India, however, efforts at fiscal consolidation has led to the buoyancy in the revenues following revenue augmenting strategy encompassing moderation in tax rates, broadening the tax base, removal of exemptions, and some improvement in tax administration. This has led to improvement on revenue deficit since 2003-04. The trend is also reflected in the dis-savings of the Government Administration, which has also declined.

II.73 Since the year 1983-84, the savings of non-departmental undertakings has become the major component of public sector savings.

II.74 Much before the public sector savings turned negative in 1998-99, the savings ratio of Government Administration turned negative much earlier in 1986-87. It is the non-departmental enterprises’ (NDEs) savings that has kept the public sector savings positive for over a decade.

II.75 The savings of the NDEs have seen a consistent improvement on account of improvement in the functioning of the public sector enterprises (PSEs). They have exhibited continued and steady improvement in their commercial functioning since the early 1990s. Consequently, since 2003-04 onwards, total public savings have turned positive again.

II.76 The improvement in the functioning of the PSEs is on account of the following two factors, which led to improvement in the savings of the NDEs. First, the public sector enterprises have benefitted from the autonomy given to them since the mid-1990s. In July 1997, the Government identified nine PSEs viz., BHEL, BPCL, HPCL, IOC, IPCL, NTPC, ONGC, SAIL and VSNL as Navratnas to turn these firms into global giants. This Navratna status granted of autonomy to the Boards of these PSEs. Second, there has been a decline in employment in the PSEs, which has led to improvement in the efficiency of operation of the PSEs. The employment in the public sector had been rising on a secular basis till the mid-1990s. Public sector employment reached a peak of 19.6 million in 1996-97 and since then, however, it has been declining (Chart 2.10). The public sector employment declined from 19.6 million in 1996-97 to 18.0 million in 2004-05, which represents a decline of about 8.2 per cent.

Public Sector Savings - Data Related Issues

II .77 The present savings estimates of the public sector broadly reflect the broad set of changes that occur in the economy. For instance, a rising revenue deficit leads to either lower/decline in public sector savings, which is clearly reflected in the available savings estimates. Besides, the declining public sector employment leading to reduction in wage cost, which has led to some improvement in savings estimates for Government Administration.

Nonetheless, there is a need to improve the quality of savings estimates of the public sector by taking on board the related data from the local authorities and quasi-Government bodies, which might further refine the savings estimates of the public sector.

Investment

II .78 Indian economy being trapped in ‘low level equilibrium trap’ during the 1950s with subsistence level of economic activities and lack of industrialisation implied that the savings would be perpetually low for some years resulting in very small level of gross fixed capital formation (investment) in the country. Gross domestic savings have increased continuously from an average of 9.6 per cent of GDP during the 1950s to almost 35 per cent of GDP at present; over the same period, the domestic investment rate has also increased continuously from 11.2 per cent in the 1950s to close to 36 per cent by 2006-07 (Chart 2.11 and Table 2.2). A very significant feature of these trends in savings and investment rates is that Indian economic growth has been financed predominantly by domestic savings. The recourse to foreign savings – equivalently, current account deficit – has been rather modest in the Indian growth process.

II.79 The dominant mode of investment for public and household sectors remained the construction activities with some variations over the years. The investment trends for public sector witnessed secular deterioration in the share of construction over years, while the share of machinery and equipment picked up over the same period (Chart 2.12). However, after reforms the trends have reversed with construction share rising and share of machinery and equipment falling. For private corporate sector dominant share for investment remained machinery & equipment over the years (Chart 2.13).

II.80 The increase in the rate of capital formation between the middle of the 1950s and in the middle of the 1960s was mainly in the public sector. This was on account of adoption of industrialisation in a big way by huge investments in Public sector undertakings since second plan onwards. The savings mobilised in the economy was invested by the Government in setting various projects. During this period (which coincided with the Second and Third Plans) inflows of foreign savings were considerable, ranging from 2 to 4 per cent of GDP. Since the early 1960s there has been some decline in the rate of capital formation in the private corporate sector, but this has been more than compensated for by increase in the household sector.

II.81 The adoption of mixed economy after independence implied that the dominant source of investment would be public sectors with private sector playing only supportive role. The reliance on public sector for rapid industrialisation was mainly on account of absence of mature private corporate sector. The investment remained predominantly in the sphere of public sector. However, economic reforms have reversed this trend with private corporate sector overtaking the public sector in realm of investment in the economy since 1994-95 (Chart 2.14).

II.82 The long-term upward trends in savings and investment have been interspersed with phases of stagnation. In particular, during the 1980s, the inability of Government revenues to keep pace with the growing expenditure resulted in widening of the overall resource gap. Accordingly, the public sector savings investment gap, which averaged (-) 3.7 per cent of GDP during the period 1950-51 to 1979-80, widened sharply during the 1980s culminating in a high of (-) 8.2 per cent of GDP in 1990-91 (Chart 2.15). The resultant higher borrowing requirements of the public sector led the Government to tap financial surpluses of the household sector through enhanced statutory pre-emption from financial intermediaries at below market clearing interest rates. As fiscal deficits widened, periodic increases in the statutory liquidity ratio (SLR) were resorted to finance the rising fiscal gap, indicative of the financial repression regime in place.

II.83 The 1991 brought a break from planning era with economic reforms ushering structural changes in the economy. The investment trends since then have been impacted by the various changes brought in the economy. The success of structural adjustment programme (SAP) reform in bringing out a sustainable recovery in economic activity in the economy depended crucially on the behaviour of investment in the aftermath of the reform process. Since the expansion of public investment is usually constrained as a part of fiscal austerity measures embodied in SAP reform, the required recovery of investment came largely from the private sector. Investment by private corporate sector plays a significant role in sustaining the industrial growth of an economy. The economic reforms in the 1990s had a significant impact on the Indian economy. At a generic level, the reforms played a facilitative role and created a competitive environment for Indian’ companies. Industrial de-licensing provided flexibility to firms in investment decisions. This led to a surge in private investment.

II.84 Since the early 1990s, growth impulses appeared to have gathered further momentum in the aftermath of comprehensive reforms encompassing the various sectors of the economy. There was some loss of the growth momentum in the latter half of the 1990s, which coincided with the onset of the East Asian financial crisis, setbacks to the fiscal correction process, quality of fiscal adjustment, slowdown in agriculture growth affected by lower than normal monsoon years, and some slackening in the pace of structural reforms. The slowdown could also be attributed to the excessive enthusiasm and optimism in regard to investment plans in domestic industry following deregulation, which was followed by significant problems experienced in viability and competitiveness. The investment during this period remained more or less stagnant with income growth being also moderate. Thus, during the 1990s the average economic growth was moderate and the investment improved only marginally over the 1980s rates.

II.85 An important structural change in the investment behaviour of the economy in the recent past is the change in relative shares of public and private investment. Years of reforms has marked a significant break from previous trends in terms of increasing investment with major role being played by private corporate sector in last five years. The investment (gross capital formation) has increased sharply from 24.2 per cent of GDP in 2001-02 to 36.0 per cent during 2006-07 i.e., increase of 11.8 percentage point within five years (Table 2.7).

II.86 In the same period, the investment in private corporate sector trebled from 5.4 to 14.5 per cent of GDP, thus, accounting for the overall increase in investment in the economy (Chart 2.16). The rapid increase in private sector investment in the aggregate investment is in large part a reflection of the impact of the reforms initiated in the 1990s, which reduced restrictions on private investment and created a more favourable investment climate. It reflects the fact that the private sector has responded positively with an improvement in the investment climate. The reduced requirement by the centre for meeting budgetary mismatches, and for overall public sector financing has improved the availability of resources for the private sector considerably. Furthermore, the corporate sector has responded to increased global competition by improving its productivity and efficiency through increased application of technology. The economic reform process has helped greatly in making the policy environment more conducive for more efficient entrepreneurial activity.

II.87 From the long-term perspective, it is interesting to observe that the rate of savings of the private corporate sector has increased from around 1 per cent in the 1950s, 1.7 per cent in the 1980s and 3.8 per cent in the 1990s, to almost 8 per cent at present. Higher retained profits along with availability of resources from the banking sector facilitated by the lower financing requirement of the Government and the increased access to the domestic and international capital markets led to a sharp increase in the investment rate of the corporates.

Inadequacies in Investment Estimates

• The turnaround in public sector savings in recent years has not reflected corresponding improvement in public sector investment levels (Table 2.8). The public sector investment estimates might be underestimated particularly with reference to defence capital expenditures, local bodies and autonomous Government institutions.
  
  
• One of the aspects of the estimates of savings and investment in India is the discrepancy in the estimates of GDS and gross domestic capital formation (GDCF). While the Raj Committee categorically mentioned that no adjustments be made to the estimates of capital formation on account of the ‘errors and omissions’, the Chelliah Group believed that GDCF estimates are subject to errors in several respects, and methodologies need to be, therefore, improved. The savings in the form of physical assets of the household sector being a part of the estimate of capital formation of the economy, would reflect the errors which creep in the estimates of total capital formation in the economy.
  
  
• Non availability of current and reliable data on output of capital goods in respect of unregistered manufacturing sector, which is essential in the estimation of capital formation for the entire economy through the commodity-flow approach.
  
  
• Lack of annual enterprise surveys and reliable data from benchmark enterprise surveys in respect of expenditures made by various industries on acquiring capital goods, as also on inventories, which is essential for industry-wise and State-wise estimates of capital formation.

• While estimating the fixed capital formation by the ‘Commodity Flow Method’, a number of rates and ratios are being used in absence of direct data. Some of these rates and ratios are based on old surveys or studies, while some of them are updated using the data available from Input Output Tables at the time of revising the base year of national accounts series.
  
  
• The estimates in respect of local authorities are not based on actual annual expenditure data. The estimates relating to Quasi-Government bodies are prepared using the workforce estimates and the estimated value added per worker obtained from the annual reports of the research and scientific institutions. There is a need, therefore, to improve the quality of estimates of public sector with regard to local authorities and quasi-Government bodies, as also the emerging PPPs and SPVs.
  
  
• There is under-coverage of the private corporate sector in the analysis in respect of new companies under setting up, besides the large time-lag in the availability of detailed results from the NABARD on cooperative sector. The under-coverage of capital formation in the public and private corporate sectors gets reflected in the household sector, as they are derived as residual from the overall estimates of capital formation, which are compiled through the commodity flow approach.
  
  
• A number of ratios and norms are used in the preparation of estimates of capital formation for the economy as a whole. Some of these ratios are based on very old norms and may not be appropriate in the current economic scenario.
  
• The estimates of capital formation by type of assets suffer from limitation. The proportions used for estimating the production of capital goods out of partly capital goods and part of capital goods are based on data on household durable goods as obtained from NSS survey results.
  
  
  
• Estimates of GFCF by industry of use are prepared mainly following expenditure approach. A critical evaluation of the estimates in the context of data base shows that for a large percentage share of the aggregate capital formation (about 70 per cent) originating in organised industry groups like registered manufacturing, electricity, gas, railways and public administration sector etc in public sector, the estimates are based on direct annual data. For the remaining sectors (namely, private corporate and household sectors) the estimates are based on either benchmark estimates or analysis of balance sheets of selected companies in the corporate sector which leads. The estimates for some of these industry activities where private share is dominant, therefore, need to be based on more satisfactory information and current data in order to improve their reliability.

A Summary of HLC’s Observations

II.a Despite the fact that the NAS provide a consistent and coherent set of macro-economic accounts, often interpretational issues arise while analysing these data. The following points can be made in regard to the interpretational significance of the savings and investment estimates:

II.b Contribution-wise, with regard to the generation of savings, the major drivers are household sector, corporate sector and public sector in that order and for investment; they are corporate sector, household sector and public sector. While households have traditionally remained major savings sector in Indian economy reinforced by demographic dividends in turn reflecting lower dependency, of late corporate savings have emerged as savings generator responding inter alia to robust sales and corporate profitability with the economy on a high growth trajectory recently. Cross-country differences in the evolution of gross corporate savings over the past half decade have tended to reflect to a large extent those of the gross operating surplus, which in turn are likely to be due to countries’ exposure to global factors (accelerated globalisation and technological progress), presumably depending on institutional framework conditions (such as product and labour market regulations), as well as the sectoral composition of their economies. With Indian economy being increasingly gloabalised, the interplay of domestic and external factors in enhancing corporate savings can be envisaged. Within the public sector, the sources of savings are characterised by certain differential performance with the Government Administration being on a consistent deficit mode since the 1990s and the non-Departmental enterprises and Departmental enterprises in that order compensating by way of positive savings reflective of their steady improvement in commercial functioning. Significantly, these changes are occurring when the economy remained on a high growth trajectory in the last five years.

II.c As highlighted in the aforementioned paras, there are limitations in the estimation of savings and investment at the level of each of the institutional sectors. This is understandable given the predominance of unorganised sector in India. Even for the organised sectors, analysts have pointed out the data gaps and quality issues for those data that are available. Interpretational issues have, thus, remained endemic to the savings and investment estimates, surfacing from time to time leading thereby to formation of Committees to address these concerns.

II.d A point mentioned in the past Committees regarding correct interpretation and a clear understanding of the available estimates of domestic savings and capital formation, it is essential to first keep in mind the various data sources used and the current methodology of estimation of savings and the capital formation. In doing so, it would be possible to indicate the areas where errors in estimation could arise.

II.e Interpretational issues with regard to estimates of savings and investment are important to be assessed from time to time, as is the practice so far in India. In this connection, it is desirable to develop alternative data bases for these estimates for the constituent sectors so that savings estimates could be cross-validated with the help of appropriate alternative data bases, methodologies and surveys.

II.f Notwithstanding the limitations, which are well accepted, the HLC observes that the existing methodology for estimation of savings has broadly captured the changes that are taking place in the economy in general and various sectors like the household, corporate and public sectors in particular. The estimates are reasonably reliable reflective of the trends in the economy. The estimates may be reasonably reliable on account of the following factors:

• To the extent increase in GDS is spread across all constituent sectors, the improvement in the GDS rate in the recent decades is essentially a reflection of the fundamental transformation of the structure of the economy. The rise during 2000s simultaneous with the buoyant growth performance corroborates the positive cycle of growth, savings and investment, such that as growth enhances savings, which in turn further enhances growth. The rise in savings rates in India appears to be mainly structural and one can expect savings rates to increase further with rising income and declining dependency ratio.

• A look at the time series of ‘errors and omissions’ relative to GDP indicates that the magnitude of this item is narrowing over time which is a welcome development.
  

• Seen vis-à-vis the efficiency of capital use (ICOR) trends, there is a clear movement downwards, which has enabled the same dose of capital formation leading to higher GDP growth contributed to a larger extent by services sector which is less capital intensive in nature. It may be mentioned that at the economy level, the ICOR has decreased to 4.2 during the Tenth Five Year Plan period as compared to 4.5 during the Ninth Five Year Plan.

II.g The savings and investment estimates are crucial in the decision making and planning process. They are used to compile incremental capital-output ratios (ICORs). The ICOR is a summary measure of the productivity of investment in the economy or the output generating capacity of incremental capital (ICOR = (capital employed - previous year capital employed) / (value of output - previous year value of output)). That is, the ratio of incremental capital to incremental output or the amount of incremental capital that goes to generate one unit of incremental output. The recent increases in the savings and investment rates indicate to some extent the lowering of the ICORs. As in the case of aggregates which hide inherent deficiencies at the detailed industry levels, the data on industry-wise deployment of capital formation is crucial in further identifying the key sectors of the economy, data on which are currently provided through national accounts statistics.

II.h Increasing evidences of a rise in efficiency in the economy is seen by the declining share of inputs - final use of commodities and services in gross value added (GVA). The share of inputs in the gross value added has come down to 49.7 per cent in 2003-04 as compared to 60.3 per cent in 1968-69.

II.i Another issue that has come into focus in the recent literature was that if the corporate sector’s savings is going up, the household sector’s savings would come down, to the extent

that unincorporated entities, constituting a part of the household sector, get incorporated and such savings are reflected as corporate savings. An examination of the data since the 1950s shows a declining trend in the share of unincorporated sector. If this trend continues, as to be expected, the share of household savings as a percentage of GDP may decline correspondingly with a rise in corporate sector savings (Table 2.9 and Chart 2.17).

II.j In sum, the interpretational issues with regard to estimates of savings and investment are important to be assessed from time to time, as is the practice so far in India. In this connection, it is desirable to develop alternative data bases for these estimates for the constituent sectors so that savings estimates could be cross-validated with the help of appropriate alternative data bases, methodologies and surveys. In the absence of income-expenditure survey of households and household enterprises, there is no direct estimation of household savings and as a result, there is no cross validation of the indirect (residual) estimates of savings at present which is expected to be taken up in due course enabling cross-validation of savings of the predominant household sector. With regard to private corporate sector and the public sector, the issues of inadequate data coverage and local bodies and autonomous Government institutions, respectively, are the weak areas that need to be resolved.

Chapter III : Broad Approach adopted by the Committee

III.1 The approach adopted by the HLC in addressing various ToRs is distinctly different than those of the earlier Committees appointed to review the estimation of savings and investment. Unlike the approach followed by the past two Committees, namely Raj and Chelliah Committees, keeping in view the relatively large number of ToRs assigned to it and the intricacies involved in the estimational issues of savings and investment, the HLC set up four Sub-Committees for examining the estimational issues pertaining to the savings of the household sector, private corporate sector, public sector and one Sub-Committee for investment. The approach followed by the HLC is briefly explained below.

III.2 At the outset, the HLC examined the existing methodology adopted to estimate the savings and investment and whether household savings estimates can be based on any other alternative methodology. Perhaps this is first such attempt by any formal committee. It was recognised that the present method needs to be strengthened though updation of various rates and ratios and development of alternative data base for constant cross-validation of the estimates. Accordingly, a ‘Worksheet Approach’ was followed, wherein each and every component of financial savings in the financial savings worksheet was examined with respect to data base, data quality and methodology. A comprehensive review of the rates and ratios used in the estimation procedure was undertaken.

III.3 Secondly, keeping in view the number of rates and ratios underlying the present estimation procedure, the HLC sought the involvement of apex financial bodies for the different financial instruments as well as sectors, which provide the required data for estimation of savings and investment. This is also a maiden effort, which is expected to enable development of alternative databases for robust estimation of savings estimates and also enable alignment of the present procedures to the procedures as laid down in the National Accounts - ‘Sources and Methods, 2007’ of the CSO.

III.4 Thirdly, an effort was made for the first time to involve and elicit the views of the NSSO on the possibility of introducing a comprehensive Survey of Income and Expenditure of the household sector. Although the reference to estimate savings by direct method through a comprehensive income-expenditure household survey was made in the earlier Committees’ Reports, there was no progress on that front. The NSSO has agreed in principle to undertake such a Survey for the households initially with a pilot survey on the basis of its interaction with the HLC. When undertaken and stabilised, such a survey is expected to provide an alternative database for direct estimation of household savings, which can be cross-validated with the present estimates.

III.5 In sum, the principal approach adopted by the HLC was to examine in detail the conceptual aspects of the estimates of savings and investment and refine the procedure by eliminating the data gaps and thereby improving the data quality. Efforts were directed at strengthening the databases presently used for estimation of savings and investment. It was felt useful to involve the apex financial bodies and NSSO for development of appropriate alternatives, which could achieve eventually the desired level of accuracy at disaggregated and sectoral level.

Brief Description of the Deliberations in the HLC Meetings

III.6 The first meeting of the HLC on Estimation of Savings and Investments was held on January 24, 2008 at RBI, New Delhi. The Chairman, Dr. C. Rangarajan, in his opening remarks explained that recently, a sharp increase has been observed in the savings and investment estimates and, therefore, they require a fresh look. At the present juncture, the issues related to methodology, nature and procedure of data collection and accuracy of collected data require critical examination. In the process of estimation, the blow-up factors or ‘Rates and Ratios’ currently used require to be examined afresh in the light of evolving developments. It was, thus, decided that the Committee has to examine three aspects of Savings and Investment estimates, viz., (a) methodology; (b) procedure of data collection and identification of data gaps; and (c) validation of the data. In order to accomplish the task assigned to the Committee, it was decided to set-up four Sub-Committees that would critically examine these aspects for each of the three sectors for savings estimates and the fourth that of capital formations. The Sub-Committees were advised to report to the HLC, after due deliberations. The Sub­Committees constituted were namely, Sub-Committee on Household Sector Savings, Sub­Committee on Private Corporate Sector Savings, Sub-Committee on Public Sector Savings and Sub-Committee on Investment.

III.7 In the second meeting of the HLC held on April 18, 2008 the Committee discussed the draft reports submitted by the Sub-Committee on Household Sector and Sub-Committee on Private Corporate Sector. In this context, HLC had broadly agreed with the major recommendations made by the two Sub-Committees and suggested improvements in certain areas. The HLC’s discussions focused on the “worksheet” approach adopted by the Sub­Committee on Household Sector, wherein each and every constituent items for household sector’s financial savings compilation was examined with respect to (a) sources of data and the underlying conceptual issues (focusing mainly on Raj Committee, Chelliah Committee, National Statistical Commission Reports wherever applicable), (b) methodology adopted for each of the parameter along with the limitations and problems, (c) prescribed procedure for estimation, (d) procedure followed in practice, (e) deviations and constraints in practice, and (f) lacunae identified in data and methodology. In this meeting, it was felt essential to involve the Apex Financial Bodies for developing an alternative database and facilitating cross-validation of the present estimation. Accordingly, it was explained that the Sub-Committee on household sector had a separate meeting with the apex financial bodies namely, SEBI, IRDA, NABARD and NHB, which were invited for an interface with the Sub-Committee on Household Sector Savings for supply of authentic, appropriate and timely data on the household financial savings instruments that come under their regulatory purview. The representatives from the apex financial bodies agreed in-principle to provide regular information, henceforth, to the RBI as per the format to be provided by the RBI.

III.8 As regards the Report of the Sub-Committee on Private Corporate Sector Savings, it was explained that four inter-related topics: (i) making a critical review of present methodology; (ii) assessing the available databases in terms of scope, definitional issues, coverage of sub-sectors and coverage of companies/entities under each sub-sector; (iii) suggesting alternative methodology; and (iv) identifying data-gaps and proposing a framework for improving data quality and timely collection of data were undertaken by them. In this meeting the HLC deliberated on the issue of treatment of ‘marked-to-market’ also. However, conceptual understanding on integrating ‘marked-to-market’ concept in estimating corporate savings is not traceable in existing literature and also there is hardly any cross-country experience available. The Sub-Committee felt that it might not be a right time to explore the issue, though future research may be undertaken to address the issue.

III.9 On the issue of methodology adopted to estimate corporate sector savings, it was explained to the HLC that the Sub-Committee reviewed the existing methodology of estimating savings of the entire sector using estimates obtained from samples. It was further explained that availability of representative samples and suitable auxiliary variables for blowing-up the sample estimates to the savings for the entire sector were two important inputs in the present method. Though this method is conceptually appealing, the estimates could suffer from low precision if samples lack representativeness of the sector and/or the auxiliary variable shares a non-proportional or poor relationship with target variable savings. Thus, review of existing methodology addressed two issues, namely, (i) representativeness and coverage of database used at present in estimating savings for the entire private corporate sector; and (ii) suitability of auxiliary variable used in the current blowing-up mechanism. The Sub-Committee felt that much of the problem could be sorted out had the data coverage been adequately large. In this context, the Sub-Committee examined the usefulness of MCA21 database, a new corporate database being collected online by the Ministry of Corporate Affairs (MCA) that features fairly large coverage. Accordingly, the Sub-Committee first assessed the consistency and comparability of the relevant data fields being reported in MCA21 with the corresponding data in company’s annual accounts available with RBI for some common set of non­Government financial and non-financial companies. Based on this exercise, the Sub-Committee enumerated further initiatives with regard to scope and definitions that might be undertaken to strengthen and improve the MCA21 database for the purpose of estimating savings for the corporate sector. The Sub-Committee also deliberated upon the features of the recent global initiatives of eXtensible Business Reporting Language (XBRL) and examined whether this platform would further improve and strengthen the required corporate databases for estimating savings of the corporate sector of the economy.

III.10 The third meeting of the HLC was held on July 17, 2008 at RBI, New Delhi under the Chairmanship of Dr. C. Rangarajan to consider the draft reports of the Sub-Committees on Public Sector Savings and Capital Formation. The Chairman remarked that while public sector savings were based on sound data sources like budget documents and annual reports of public enterprises, the estimates of investment are based on indirect approaches and use of a number of rates and ratios.

III.11 In the meeting, the following issues were deliberated with respect to the Report on the Public Sector Savings: (i) the discrepancy between the revenue deficit shown in the budget documents of Centre and States and that of the savings of the public authorities shown by the CSO is, to some extent, due to non-inclusion of local bodies in the budget documents; (ii) on the capital transfers made by the Centre and States, a separate table may be included in the Report showing the capital transfers made to the Quasi Government bodies; (iii) treatment of defence capital outlay; (iv) the issue of capital formation made by the public-private- partnerships (PPP) and Special Purpose Vehicles (SPV). It was observed that the problem in accounting for the investment made by these entities is similar to that of companies under construction, the HLC recommends that the CSO may try to collect information on these PPP/SPVs through nodal agencies and send the same to the RBI for analysis and compilation of estimates of savings and capital formation.

III.12 The fourth meeting of HLC on Estimation of Savings and Investment was held on October 7, 2008 at RBI, New Delhi under the Chairmanship of Dr. C. Rangarajan to consider the draft report of the Sub-Committee on Capital Formation. In the meeting, Chairman remarked that apart from the three estimates of capital formation that are being presented in the National Accounts Statistics, there is yet another estimate of savings from the income and outlay account. The Chairman enquired whether this estimate could be considered as an independent estimate of the other savings estimate being used in the estimation of gross capital formation; and also the difference between these two. Regarding the estimate of savings from the income-outlay account, it was explained that in the NAS, disposable income of the nation is estimated as sum of (i) GDP, (ii) net factor income from abroad and (iii) net current transfers from abroad. From this disposable income, if private final consumption expenditure (PFCE) and Government final consumption expenditure (GFCE) are subtracted, we arrive at a residual, which could be termed as an estimate of savings. However, since this estimate differs from the independent estimate of savings, the difference is shown as‘statistical discrepancy’ in the NAS. This is of the order of about 1.0 per cent of GDP at present. Normally, this discrepancy is attributed to PFCE, as the savings estimated in the NAS by institutions is treated as firmer estimate. In the meeting the following issues were also deliberated: (i) The issue of strengthening data base for capital formation; (ii) issue related to State level capital formation; (iii) the issue related to data on companies under construction; (iv) the issue of deteriorating quality of ASI data, particularly at the State level; (v) the issue of updation of rates and ratios, through alternative agencies like agro-economic research centres, etc. While the Sub-Committee for Capital Formation recommended that the estimates of capital formation arrived through three different methods are to be shown separately without any adjustments, the HLC observed that estimates of capital formation arrived from the savings side could be treated as firmer estimates relative to estimates based on commodity flow/expenditure approach, as those estimates are based on a number of rates and ratios evolved from various ad hoc sources. It was, therefore, recommended that for operational convenience, only one figure of capital formation arrived at through savings route may be used for compiling rates of capital formation as is being done at present.

III.13 Further, there was also comprehensive discussion on the treatment of ‘valuables’ and ‘errors and omissions’. Regarding valuables, one point of view was that they were not contributing to the production process at all; therefore, should not be treated as capital formation. The other point of view was that ‘valuables’ were not treated either as intermediate consumption or as final consumption expenditure, therefore, should be part of gross capital formation, but outside the gross fixed capital formation, as they were not used in the production process and are kept as store of value. According to 1993 SNA, the data on expenditures made on net acquisition of valuables on precious items like gold has been included under Gross Capital Formation (GCF). It was decided that the present practice of showing net acquisition of valuables under GCF, as a separate category distinct from the Gross Fixed Capital Formation (GFCF) and Change in Stocks, may continue. There was also a detailed discussion on the issue of treatment of ‘errors and omissions’. It was decided that different independent estimates of capital formation as obtained by the commodity flow method and estimates of capital formation as obtained from the savings estimates and estimates of capital formation as obtained by the expenditure approach, should be presented as such. The issues of (i) arriving at pure household savings; (ii) examining the savings of farm sector in relation to investment; and (iii) basing savings estimates on marked to market basis were discussed in details. It was felt that at this juncture it is not possible to address these issues in the Report as no solution could be found with respect to these issues.

III.14 In the fifth meeting of the HLC at Chennai on November 20, 2008, the draft of the Report was discussed threadbare. Among the major issues discussed was the likely underestimation as per the current practice of household sector’s physical assets. It was agreed that this could be because of the fact that while household financial liabilities are increasing, the same is not reflected in the pace of expansion in the household physical assets. In this discussion, the Chairman stressed that the concept of ‘transferable savings’ is more relevant and in this context, it is useful to look at ‘gross financial savings’ and not ‘net financial savings’. As economic agents are living in an era of highly leveraged society and financial liabilities are important as the households borrow for a variety of purposes – both financial and physical investment, increased financial liabilities being a subtraction item lead to depressed net financial savings. Another issue discussed was that to be able to find out the estimates of savings for the different segments of the household sector (as per India’s national accounts framework), it is desirable that the survey of income and expenditure of ‘pure households (consumer households)’ needs to be taken up concurrently with the Enterprise Survey.

III.15 In the last meeting of the HLC held in New Delhi on December 24, 2008, the final draft of the Report was discussed and adopted.

Chapter IV : An Overview of the Existing Methodology in Estimation of Savings and Investment

An Overview of the Existing Methodology

IV.1 The methodology currently adopted for estimating savings and investment are explained in the ‘Sources and Methods, 2007’ brought out by the Central Statistical Organisation (CSO). Savings represent the excess of current income over current expenditure. It is the balancing item of the income and outlay accounts (as per 1968 SNA) and use of disposable income account (as per 1993 SNA) of producing enterprises and households, Government administration and other final consumers. The concept emphasises the current nature of receipts and payments, thus, ruling out the inclusion of ‘flows out of past assets’ and ‘future liabilities’.

IV.2 For preparation of the estimates of domestic savings, the economy has been divided into three broad institutional sectors viz., (i) public sector, (ii) private corporate sector and (iii) household sector. The estimates of domestic savings are presently prepared only at current prices. The estimates at constant prices have not been attempted so far since the suitable procedure of estimation is yet to be evolved. In pursuance of the recommendations of the Working Group on Savings and Expert Group on Savings and Capital Formation, set up by the then Department of Statistics under the Chairmanship of Prof. K.N. Raj and Prof. Raja J. Chelliah, respectively, in 1981 and 1995, the CSO and RBI agreed on a uniform methodology and data base. It was decided that each agency would be responsible for compilation of estimates of specified sectors and supply the data to the other agency. Accordingly, the CSO now prepares the estimate of public sector and household savings in the form of physical assets, life funds, provident and pension funds, while the RBI is responsible for estimating the savings of the private corporate sector and that of the household sector in other instruments of financial savings. Thus, the estimates of gross domestic savings released by the two organisations are based on uniform methodology and same data sources. However, the two sets of estimates of gross domestic savings may differ slightly due to difference in timings of the release by the two organisations. The RBI normally releases its Annual Report in September i.e., three months before the release of the official estimates of national income and related aggregates based on the data available up to June end, whereas the estimates released by the CSO in the following January incorporate data supplied by the RBI and other agencies by December end.

IV.3 Public sector covers Government administration, departmental enterprises, quasi-Government bodies and non-departmental enterprises. The non-departmental enterprises comprise Government companies, statutory corporations and port trusts. Government administration & departmental enterprises: The gross savings of Government administration, quasi bodies and departmental enterprises is defined as the excess of current receipts over current expenditure. The current expenditure includes: (i) final consumption expenditure; (ii) interest on public debt; (iii) subsidies and (iv) current transfer, while the receipts comprise of (i) income from entrepreneurship and property; (ii) direct taxes; (iii) indirect taxes; and (iv) miscellaneous receipts. The Issue Department of the RBI, which is considered to be more akin to administrative activities of the Government, is also included here. The gross savings of the non-departmental enterprises including Banking Department of RBI is estimated from the results of the analysis of annual accounts of these companies and corporations. In the case of Life Insurance and UTI mutual funds, the savings arising out of life insurance business and management of funds respectively is included in the household sector.

IV.4 Private Corporate Sector comprises all non-Government financial/non-financial corporate enterprises, co-operative institutions and quasi corporate bodies. Non-Government non-financial enterprises include public and private limited companies registered as joint stock companies under the Companies Act, 1956. These include foreign direct investment companies as well. Non-Government financial institutions constitute all scheduled and non-scheduled commercial banks in the private sector, other financial and investment companies engaged in activities such as financing of hire-purchase, transactions in shares and commodities & financing of loans or investment in securities, housing finance companies and insurance companies. Co-operative institutions comprise all co-operative banks, co­operative credit and non-credit societies. Estimates of saving for the non-Government non-financial, financial and investment companies (excluding commercial banks) are prepared based on the analysis of balance sheets and profit and loss (P/L) accounts of select companies. Net saving is computed as retained profits adjusted for non-operating surplus/deficit. Retained profits are that proportion of total net profits that are ploughed back into business (either transferred to reserves or carried forward to balance sheet) and after making commitments to financial institutions (in the form of interest payments), Government (tax provisions), share-holders (dividend, etc.) and after making depreciation provision for various fixed assets. The savings of non-banking financial and investment companies is estimated on the basis of RBI studies on performance of financial and investment companies following the same procedure as relating to non-financial companies. The savings of private commercial banks is estimated as addition to the reserve funds. The transfers to reserve funds include net amount carried to reserves, depreciation provision, amount allocated for other special purposes and amount carried forward to next year’s account net of surplus/deficit of the previous year brought forward as available from the details of “Earnings and Expenses of Commercial Banks” published in the RBI’s ‘Statistical Tables Relating to Banks in India. Since the details of appropriation of profit are not available for the later years, the saving has been estimated on the basis of trend observed in the profit of these banks. In case of co-operative societies, data on statutory reserves/funds, bad debt reserves, and other reserves are available in the ‘Statistical Statements Relating to Co-operative Movement in India’. The gross savings is taken as the increase in statutory funds and other reserves/ funds. In the absence of relevant data, the estimates of savings are prepared on the basis of trends observed in the value added of banks and trade sector.

IV.5 The household sector comprises, apart from individuals, all non-Government, non­corporate enterprises like sole proprietorships and partnerships owned and/or controlled by individuals and non-profit institutions, which furnish educational, health, cultural, recreational and other social and community services to households. The savings of the household sector is taken as the sum of its investments in various instruments of financial savings and in the form of physical assets. Since direct annual data on household income expenditure are not available, the savings of the household sector is worked out by following the residual method except savings in the form of life insurance funds and provident & pension funds. The investment in financial savings comprises currency, net deposits, shares and debentures (including mutual funds), net claims on Government in the form of small savings, investment in Central and State Government securities, life insurance funds and provident & pension funds. Household savings in the form of currency is estimated as a residual by deducting the amount of currency held by the private corporate sector and public sector enterprises from the total currency with the public. The data on total currency with the public (i.e., notes in circulation plus circulation of Rupee Coins and Small Coins minus cash in hand with banks) for each year are available in the RBI Bulletin. On the question of cash in hand, the Advisory Committee on National Accounts in its meeting held in June 1987 had suggested that the present procedure of estimating the currency with the households by deducting from the total currency, the currency held by various institutional sectors may be examined. It had further suggested that a simple procedure of possibly taking a proportion of the total currency in circulation might serve the purpose. Accordingly, the RBI on the basis of past trends of currency holding of the households and non-households sectors, estimated this proportion to be around 0.93 of the ‘currency with public’, which has been used for estimating the currency holding of the households from 1985-86 onwards.

IV.6 Gross capital formation (GCF) refers to the aggregate of gross additions to fixed assets (i.e., fixed capital formation), increase in stocks of inventories, hereinafter referred to as change in stocks during a period of accounting and net acquisition of valuables. Broadly, two types of fixed assets namely construction and machinery and equipment (including transport equipment, software and breeding stock, draught animals, dairy cattle and the like) are covered. Construction for military purposes (other than construction or alteration of family dwellings for military personnel), defence equipments, durable goods in the hands of the households and increase in the stocks of defence materials are excluded from the scope of gross capital formation. However, capital outlays of defence enterprises on ordnance and clothing factories are included. Software which was earlier included in the machinery and equipment, is now segregated from machinery and equipment and shown as a separate item. Besides the above, a new category ‘valuables’ which covers the expenditures made on net acquisition of valuables, has been included for the first time in compilation of the gross capital formation in the New Series released with base year 1999-2000, in accordance with the recommendations of 1993 SNA.

IV.7 Construction activity covers all new constructions and major alterations and repairs of buildings, highways, streets, bridges, culverts, railroad beds, railroads, subways, airports, parking areas, dams, drainages, wells and other irrigation sources, water and power projects, communication systems such as telephone and telegraph lines, land reclamations, bunding and other land improvements, planting and cultivating new orchards (tea, coffee, rubber, mango, cashew nut, areca nut, coconut, citrus, grapes and Sapota plantations), afforestation projects, installation of wind energy systems etc.

IV.8 Machinery and equipments comprise all types of machineries like agricultural machinery, power generating machinery, manufacturing, transport equipment, furniture and furnishings. Also included are increments in livestock in respect of breeding stock, drought animals, dairy cattle and other animals raised for wool clippings. Additions to livestock other than the said fixed assets are accounted in change in stock estimates.

IV.9 The estimates of GCF for the economy as a whole include all relevant items of new capital goods which are produced domestically (exclusive of exports) and new and second­hand imported goods. According to SNA, imports should include the outlays in foreign countries of the embassies, consulates and other extra-territorial establishments of the Government of the given country on new fixed assets reduced by the net sales of their second-hand and scrapped assets. Similarly, the net outlays of foreign embassies, consulates and other diplomatic bodies located in a country, on fixed assets produced within the country are to be recorded in the exports and not in the gross fixed capital formation (GFCF) of the given country. The fixed capital outlays of international agencies located in the country are to be treated in an identical manner. Due to non-availability of data, it has not been possible to account for these. The estimation, therefore, follows the domestic concept, whereby, only additions made to stocks of fixed assets and inventories within the geographical boundary of the country have been taken into account. Accordingly, the GCF refers to gross domestic capital formation (GDCF).

IV.10 Additions to non-reproducible tangible assets such as land, mineral deposits and the natural growth of standing timber or crops have not been included in GCF. However, outlays on improvement of land and development or extension of mining sites, timber tracts and plantations are taken as part of capital formation. The outlays on incomplete construction works incurred during the period are included in the estimates of fixed assets.

IV.11 Inventories consist of materials and supplies, work-in-progress and finished products and goods in the possession of producers and dealers. Stocks of strategic materials, grains and other commodities of special importance to the nation in the possession of Government are also included in the estimates of change in stocks.

IV.12 The estimates of GFCF for the country are prepared both by type of assets viz., construction, machinery and equipment, software and by industry of use.

IV.13 The methodology adopted to estimate savings and investment as explained above can be summarised into the following four basic steps explaining the estimation procedure followed at present.

Basic Steps in Estimation of Savings and Investment

IV.14 Basic steps in estimation of savings and investment as per the existing methodology are summarised and given in Table 4.1 below:

IV.15 It is recognised that the estimates of savings suffer from a number of limitations, mainly from deficiency of data. This particularly refers to the estimates from the private corporate and the household sectors. Estimates for the private corporate sector are based on the analysis of the sample companies undertaken by the RBI and blown up on the basis of the ratio of paid up capital (PUC) of the sample companies to all companies. The PUC may not be appropriate for blowing up the sample estimates of savings, but in the absence of any other global indicator one has to use this indicator. Data on co-operative societies are not available for all the years. There is a time lag of four to five years. Estimates for the household sector are built up in two parts (i) savings in financial assets and (ii) savings in physical assets. In case of savings in financial assets the estimates of household savings in currency are estimated as 93 per cent of total currency on the basis of past trends of currency holding. This percentage may undergo a change as and when more data based on the survey results of the RBI become available. The estimates of provident fund of non-departmental enterprises are obtained by doubling the employers’ contribution as available in the books of accounts. No adjustments are made for withdrawals, interest etc., since the data on these items are not available. The non-Government educational institutions are being covered under the Employees Provident Fund Act, 1952 since 1981-82. However, the coverage of private educational institutions has been incomplete as many of these institutions are not making any returns to the Central Provident Fund Commissioners (CPFC). Thus, the estimates of PF can be improved only when the coverage of private educational institutions improves and reliable data on withdrawals, interest etc., in respect of Non-departmental enterprises becomes available. The results of RBI survey on Growth of Deposits with Non-banking Companies and Ownership of Capital of Joint Stock Companies are not available beyond 1995-96. Estimates in respect of most of the other financial assets are based on current data, though, they mainly refer to total savings under each of these categories and households, contributions are obtained as residuals. The households’ savings in physical assets is also estimated by using the residual approach. The residual approach adopted for deriving savings estimates for households is undoubtedly not very satisfactory in all cases and has enough scope for improvement. The only approach through which this can be achieved is by conducting independent household surveys to collect direct data on income and expenditures. Similarly, the coverage of public institutions is limited while estimating the public sector savings. The validity of ‘rates and ratios’ used in estimating the capital formation remain doubtful. Therefore, it is critical to periodically review the methodology used to estimating savings and investment.

Select Country Practices in Estimation of Savings and Investment

IV.16 It is well known that the estimation of savings and investment is a complex exercise, particularly because savings is not a directly measurable macroeconomic variable and that it has to taken as an aggregation of the constituent sectors (Table 4.2) and that within the constituent sectors household sector is characterized by heterogeneity in its composition (Table 4.3). The national accounts define this aggregate as a balance. More precisely, savings is the balance item of the “Use of income account”. Income and expenditure are calculated from different data sources. Because household income and household expenditure are large aggregates, small changes to either are capable of producing a large change in gross savings.

IV.17 A brief treatment of estimation of savings and investment methodologies of select countries is placed below.

Estimation of Savings in Select Countries

1. United Kingdom (UK)

Personal Savings

IV.18 The household savings in the United Kingdom is estimated by Office of National Statistics (ONS). Besides, the Department of Work and Pension also does surveys such as Family Resources Survey¹, which capture the profile of savings and assets across the region in the UK.

IV.19 To define households sector, the new European System of Accounts (ESA 95) in UK uses the SNA 1993 definition. Accordingly, household sector covers ‘individuals or groups of individuals as consumers and possibly also as entrepreneurs producing market goods and non-financial and financial services (market producers) provided that, in the latter case, the corresponding activities are not those of separate entities treated as quasi-corporations’. This sector covers people living in traditional households, as well as those living in institutions, including those living in retirement homes, hostels, boarding houses, hotels or prisons. The household sector also includes sole traders, any non-profit institutions serving households (NPISH) which do not have separate legal status and private trusts.

IV.20 The financial account of UK shows the acquisition and disposals of financial assets and liabilities by households. Office for National Statistics (ONS) and the Bank of England work together closely to produce the financial account data. They do not produce separate accounts for households and Non Profit Institutions Serving Households (NPISH), since there is virtually no direct financial information collected from households. Therefore, the estimates are made from counterpart information, residually.

¹The latest available survey was conducted in 2003-04.

General Government Savings

IV.21 This is the surplus or deficit on the use of disposable income account, and is the amount available for acquisition of physical or financial assets, or the reduction of liabilities. It represents the surplus or deficit of current resources over current uses. The public sector surplus on current budget equals public sector net savings (the surplus on the use of disposable income account after allowing for depreciation) plus receipts of capital taxes.

2. United States (US)

IV.22 Domestic capital account presents information on savings and investment for the US economy. For savings, a breakdown by sector is shown for corporate, personal, and Government savings. For investment, because of source data limitations, the breakdown is shown for private fixed and inventory investment and for Government fixed investment. The savings and investment account (also known as the capital account) records the sector’s gross savings and gross investment, where gross investment is net acquisitions of assets less net increase in liabilities.

Personal Savings

IV.23 The Bureau of Economic Analysis produces the standard measure of personal savings as part of its National Income and Product Accounts (NIPA). The savings measure is derived from the NIPA measure of disposable personal income, which is defined as personal income after the payment of payroll and income taxes. Personal savings is calculated as disposable personal income less personal outlays. It is the current savings of individuals (including proprietors and partnerships), non-profit institutions that primarily serve households, life insurance carriers, private non-insured welfare funds, private non-insured pension plans, publicly administered Government employee retirement plans, and private trust funds.

Corporate Savings

IV.24 The difference of income and expenditure is the corporate savings. Interest receipts on assets and net operating surplus are sources of private enterprise income. In the uses of income as income payments on assets (such as holders of financial liabilities and equity claims of other businesses), business current transfer payments, and income that accrues to the owners of business (namely proprietors’ income, rental income of persons, and corporate profits). Corporate profits, a widely used measure in the United States, is distributed to Government (taxes on corporate income) and to shareholders (net dividends) or is retained (undistributed profits, which can be thought of as a measure of corporate savings).

Government Savings

IV.25 Net Government savings is the difference between current receipts and current expenditures of the Federal Government and State and local Governments. Estimates of Federal Government current receipts, by component are published annually in NIPA tables along with expenditures and net Federal Government savings. The quarterly estimates are based on data from the Federal budget. The current tax receipt is one of the major items of current receipt while the Government consumption expenditure and Government social benefit are the major items of the expenditure.

3. Australia

Households’ Net Savings

IV.26 Net savings of households is equal to gross household disposable income less household final consumption expenditure and consumption of fixed capital. Household savings is estimated as the balancing item in the households income account. It includes savings through life insurance and superannuation funds (including net earnings on these funds), increased equity in unfunded superannuation schemes and the increase in farm assets with marketing boards.

Corporations’ Net Savings

IV.27 Net corporations’ savings is equal to the gross income receivable by corporations less income payable and consumption of fixed capital. Income receivable by corporations includes gross operating surplus, property income and current transfers receivable. Income payable includes property income and current transfers (including income taxes) payable.

General Government

IV.28 Net savings of general Government is the surplus of general Government gross income over current use of income. Current use of income includes final consumption expenditure and current transfers (interest and other property income payable, social assistance benefits payments to residents, transfers to non-profit institutions, subsidies, etc.).

4.Japan

IV.29 In Japan, the household savings is compiled through family survey and FoF method. In Japan, flow and stock data of the household sector are provided by the flow of funds accounts on a quarterly basis. The Bank of Japan has compiled the accounts since 1954 and a historical revision was made in 2000 to introduce the recommendations of the 1993 SNA. The household sector is determined by allocating a certain portion of residual amounts of financial assets and liabilities after allocating them to the financial, general Government, and rest of the world sectors. Therefore, the household sector depends on various financial data, rather than its own balance sheet data.

Household Savings

IV.30 Some of the methodological issues and rates and ratios such as currency holding of household sector match with India. In Japan, such allocation was made by using a fixed ratio on an assumption that 90 per cent of residual amount is held by the household sector and 10 per cent of that is held by the private non-financial corporations’ sector. This ratio is based on a study conducted by the Bank of Japan on the holding sectors of currency. As regards the issue of household sector exposure towards stock market and the impact of capital gains on the household savings, the estimation of the unlisted shares is a major challenge as the source data for the holding sectors and the market-equivalent value of those shares are limited.

Estimation of Investment in the Select Countries

1. Australia

IV.31 Gross capital formation refers to the gross additions to national wealth that result from three categories of investment:

• gross fixed capital formation, measured by the total value of a producer’s acquisitions, less disposals, of fixed assets during the accounting period;

• changes in inventories, equal to the value of inventories acquired by an enterprise less the value of inventories disposed of during the accounting period; and

• acquisitions less disposals of valuables.

IV.32 Fixed capital formation estimates are shown on a ‘gross’ basis (i.e. deductions have not been made for the consumption of existing assets during the production process). However, the estimates are net of the sale of second-hand capital assets.

IV.33 Gross fixed capital formation is made up of the outlays of producers on commodities which do not add to their inventories or enter into the intermediate consumption for the period. The fundamental point of distinction between intermediate consumption and gross fixed capital formation is whether commodities are used up during the course of a particular period or whether they yield benefits beyond that period. In the case of households as consumers, all expenditure except the purchase of dwellings is treated as final consumption expenditure, whether or not it yields future benefits. Therefore, a purchase of a motor vehicle by a household (but not by an associated unincorporated enterprise) is treated as final consumption expenditure, whereas the same purchase by a business would be classified to gross fixed capital formation.

2. United Kingdom (UK)

Gross Capital Formation

Gross capital formation is made up of three components:

• Gross fixed capital formation

• Changes in inventories

• Acquisitions less disposals of valuables.

These components are very different in nature and magnitude.

IV.34 Gross fixed capital formation (GFCF) relates principally to investment in tangible fixed assets such as plant and machinery, transport equipment, dwellings and other buildings and structures. However, it also includes investment in intangible fixed assets, improvements to land and also the costs associated with the transfer of assets. The investment relates to assets which are used repeatedly in the production process for more than one year.

Change From Former System of Accounts

IV.35 Previously in the UK accounts GFCF had been described as GDFCF (Gross domestic fixed capital formation). The dropping of the term ‘domestic’ is of itself no significance as regards the concept and coverage of this variable. However, the new accounts extend the range of items to be regarded as capital assets. The main additions, which were previously included in intermediate consumption, are intangible fixed assets. However, also included in GFCF, all within tangible assets, are cultivated assets, such as livestock, which are used repeatedly to produce goods and services; expenditure on military structures and equipment (other than weapons), previously mostly in Government final consumption; and historic monuments.

3. Japan²

Measurement of Capital in Japan

IV.36 The Economic and Social Research Institution (ESRI), the producer of the Japanese national accounts, publishes two main estimates for capital stock. The first estimate is net capital stock, which is described in the balance sheet of the Japanese national accounts. The second estimate is Gross Capital Stock of Private Enterprises (GCSPE), which is the main data source for analysis of production by industry. In addition, the ESRI irregularly publishes gross capital stock for infrastructure.

Gross Capital Stock

IV.37 For industry analysis of the Japanese economy, the GCSPE is the main data source for the Japanese capital stock. The GCSPE covers all fixed assets, excluding residential buildings owned by private corporations and unincorporated enterprises and fixed assets owned by private non-profit institutions. The GCSPE is sometimes used as a measure of the productive capacity of the private sector. The Japanese National Wealth Survey (NWS) in 1955 as the benchmark stock. In comparison with the JSNA-NCS, whose benchmark stock is based on the 1970 NWS, this methodology is close to the perpetual inventory method. To estimate productive capacity, the GCSPE intentionally excludes the residential capital owned by private sectors.

² Koji Nomura (March 25, 2005) “Turn the Tables! Reframing Measurement of Capital in Japanese National Ac­counts”, Keio University, Tokyo, Japannomura@sanken.keio.ac.jp

4. China

IV.38 In china, gross fixed capital formation refers to the value of fixed assets³ acquired minus those disposed of by all resident units during a given period. It can be categorized into gross tangible capital formation and gross intangible capital formation. The gross tangible capital formation includes the value of the construction projects, installation projects completed and the equipment, apparatus and instruments purchased (minus those disposed of), the value added from the sale of commercial house, the value of land improved, the value of newly increased draught animals, breeding stock, animals for milk, wool and for recreational purpose, and the newly increased forest with economic value during a given period. The gross intangible capital formation includes the prospecting of minerals, the acquisition of computer software, recreation, literature and art originals minus the disposal of them.

IV.39 For the initial accounting and the initial check of gross fixed capital formation, China calculates the total investment in fixed assets in the whole country as per the regular statistics requirements. For those not included in the regular statistical requirements, it is being calculated by the following:

• For the fixed assets formed by the investment in fixed assets of a small amount that is less than 0.5 million Yuan, these are calculated by the related quarterly information.

• For the increased fixed assets of the trial-produce of new products, these are calculated according to the expenses in the trial manufacture of new products on the basis of the expenditure information of the financial budget.

• For the value added from the sale of commercial house after its being built, these are calculated by using the value added from the sale of that in the same period of the previous year, times the development speed, i.e. the ratio the floor space of commercial house sold in the current period divided by that in the same period of the previous year.

IV.40 In sum, it may be stated that various countries’ experience in estimating savings and investment suggests that the estimation procedures vary.

• Broadly, the savings is estimated for three sectors namely, personal savings, corporate
and Government savings.

•The flow-of-funds accounts provide useful information on the household sector. In the absence of balance sheet data for the household sector, the estimation of that sector tends to be complex. Flow-of-funds compilers need to explore the use of various statistics to improve the accuracy of data. For the international transactions and positions of the household sector, cooperation with the compilers of balance of payments and international investment positions statistics is very important.

• In some countries there is virtually no direct financial information collected from households and, therefore, the estimates are made from counterpart information, residually. In some other economies the savings measure is derived as measure of disposable personal income, which is defined as personal income after the payment of payroll and income taxes. Others estimate household savings as the balancing item in the households income account or is compiled through family survey and FoF method.

• As regards gross capital formation, most of the countries are doing it through expenditure side of the GDP and broadly follows the UN SNA 1993 recommendations for estimation of gross capital formation. Therefore, there is no unique method that is followed universally in estimating the savings and investment.

³The fixed assets acquired include those purchased, transferred-in and self-produced, and the disposed of refer to those sold and transferred-out.

Chapter V : Major Issues in the Estimation of Savings and Investment

A Brief History of Estimates of Savings in India

V.1 The first systematic attempt at estimating domestic savings by an official agency was first made by the Reserve Bank of India around 1959 for the period 1950-51 to 1957-58. National Council of Applied Economic Research (NCAER) also generated the estimates of savings separately for rural and urban households based on sample surveys, but the objective of the exercise had been to gather information on motivational and behavioral aspects of household savings. The resulting estimates of the NCAER survey, however, have not yielded a time series, which could throw some trend. The first set of estimates consistent with the national income series published by the CSO in 1961 covered the period since 1950-51 and since then had been a regular exercise for annual data dissemination.

Survey of Literature and an Overview of Important Issues pertaining to Savings and Investment Estimation

V.2 Even before setting up of the Raj Committee, K.N Raj (1962) himself had pointed out that the estimates of savings in the physical component are too crude. Before the setting up of Raj committee in 1982, once again V.K.R.V.Rao (1980), in his Dr. Zakir Hussain Memorial Lecture at the Aligarh Muslim University, questioned the reliability of the household sector savings data in physical assets. The Issue of indirect estimates of household physical savings was well addressed in the Raj Committee Report. Subsequently, Raja Chelliah Expert Group’s Report on Savings and Capital formation also endorsed the view points of the Raj Committee on having an independent survey for estimating household savings. Raja Chelliah Expert Group Report made a suggestion in the context of the need to cross-check and validate the indirect estimation of household savings through direct independent survey. Noted commentator Prof. T. N. Srinivasan also argued for an integrated household survey for compiling household savings.

V.3 Dr S.L. Shetty from the EPW Research Foundation has extensively commented and critically reviewed the estimation methodology and data gaps in the estimation of household savings. For instance, issues related to the estimation of deposits of non-banking financial companies (NBFCs) in the overall deposit; (ii) Share of informal sectors’ financial assets; (iii) Share of currency holdings with the public (93 per cent of which is treated as part of household savings since 1985-86); and (iv) Shares of companies for want of regular surveys on ownership of company shares etc., have been highlighted by Dr. Shetty. Apart from that he had also highlighted the need for addressing other pertinent issues pertaining to savings estimates such as treatment of valuables as a part of gross capital formation without any counter entry of it in savings side of the national accounts and treatment of capital transfers.

V.4 Dr. Saumitra Chaudhuri (2005) commenting on the high rate of savings for 2002-03 and 2003-04 felt that a close scrutiny of data does not indicate much scope for savings rate to have been significantly over estimated. However, on the investment side there might be underestimation. He feels that the increase in the savings rate of households appears consistent with the record of economic growth in recent years and past behavior.

V.5 Amongst the components of National Accounts Statistics (NAS), it is the savings and investment estimation on which the authorities have cast doubts at frequent intervals, primarily on the grounds of interpretational difficulties faced by them in explaining some aspects of macroeconomic development and appointed Working Groups/Expert Groups to enquire into the questions and come out with recommendations for improvements in the methodology of savings and investment estimation.

V.6 The debate on various issues involved in the estimation of savings and capital formation, however, remains unsettled. A number of studies have commented on the inadequacies in the current method adopted in India to estimate household sector’s savings. A large number of scholars associated with two institutions, namely, Institute of Economic Growth and Delhi School of Economics – like late Prof. V.K.R.V. Rao, late Prof. K.N. Raj, Profs. Suresh Tendulkar, B.L. Pandit, V.N. Pandit, K. Krishnamurthy, D.U. Sastry, P.D. Sharma and many others – produced technical papers on the interpretative significance of the savings behaviour and the behaviour of household savings. Most of the authors have contributed to the debate on various lacunae associated with current methodology adopted to estimate household savings and some refinements.

V.7 This section gives an overview of the major issues brought into focus in the writings of experts and official committees, as together they form substantial literature on the subject. Before this, let us first give the genesis of savings estimates in India.

Issues brought out by Reports of the Earlier Committees

V.8 There are two earlier well known formal exercises, which reviewed the savings estimation i.e., K. N. Raj Committee Report and the Raja Chelliah Expert Group, which have examined a variety of issues relating to the savings and investment estimation procedure.

K. N. Raj Committee Report on Capital Formation and Savings in India 1950-51 to 1979-80 (February 1982)

V.9 First, the K.N. Raj Working Group (1982), which was set up as “the rates of gross capital formation and of gross savings had gone up significantly since the early years of the 1970s, to about 24.0 per cent of the gross domestic product by 1978-79; even net capital formation and savings were touching rates of around 20 per cent of the national income”. This has raised a number of questions, in regard to both the estimates themselves and their interpretation, particularly since there has been evidently no corresponding improvement in growth momentum within the economy and the problems of mobilisation of resources for development seems still as severe as ever before. Second, there was the Chelliah Expert Group appointed in the context of the contrary development when savings fell and growth improved. The Government made out a case in the following words: “An important reason for this lagged recovery in private savings is the decline in the absolute amount of household savings in 1992-93 as estimated by the CSO. Since this decline is observed for a year in which the GDP growth rates rises from 0.9 per cent to 4.3 per cent and GDP growth from agriculture shows a sharp turnaround from minus 2.5 per cent to plus 5.3 per cent, there appear to be grounds for reviewing the methodology for estimating, savings and capital formation in the economy (Economic Survey 1994-95:3). The mandates of the Group were three-fold: (a) critically review the available estimates of both the aggregates and components of savings and capital formation; (b) evaluating their interpretational significance; and (c) recommend improvements in methods and procedures of estimation including suggestions as to the possibility of building up regional estimates.

Approach of the Working Group

V.10 The Working Group was mostly concerned with the differences in the estimates of savings and investment brought out by the two official agencies, CSO and RBI. The Working Group examined the differences in detail and made recommendations concerning improvements in data sources and the methodology of measurement with respect to individual sectors of the economy. The Group generally agreed that the database for the estimation of financial savings was comparatively firm. The Group felt, however, that there could be certain margin of error arising from the aggregation of financial accounts of diverse sets of financial corporations, co-operatives and banks with differing accounting years, which cannot be avoided. However, the series of steps in estimation, particularly those concerning the derivation of the ‘residual’ attributable to the household sector by deducting the estimates for the public sector and private corporate sector from the totals, involve the use of a number of ratios and proportions emerging from diverse sources and these may introduce estimation biases at various stages unless extra care is taken to ensure the accuracy of those ratios and update them regularly. Though these biases may individually appear to be of minor nature, all of them put together might result in a noticeable degree of error. In this respect, from the point of view of achieving as much of accuracy as possible in the estimation of household savings in the form of financial assets, the Working Group observed that it is necessary that the estimation procedures are tightened.

V.11 Some of the reforms suggested by the Raj Committee were: (i) for Currency with the Households, to improve the estimation, the Committee proposed to have census studies of the corporate sector as well as Government companies, particularly at the State level, at least once in five years on a regular basis; (ii) Need for a more disaggregated and quicker flow of current data, like the ratio of total bank deposits as revealed by the surveys of ownership of deposits (BSR-4) undertaken by the RBI on a biennial basis, also emphasising on the need to provide a more disaggregated and detailed classification of the ‘ownership of deposits’ of the household sector into farm households, unincorporated enterprises, and households proper in non-farm sector; (iii) Improving the quality and timeliness of statistical surveys, like the RBI surveys on ‘Growth of Deposits with Non-banking companies; (iv) on Security deposits with trusts, housing boards and Electricity Boards, such security deposits should be treated as household savings in the form of deposits, (v) on Provident fund (PF) deposits in non-Government educational institutions, the Committee suggested that the ratio of deposits to loans and advances to the employees applicable for the Government institutions should be applied to the PF of the educational institutions of the private sector; (vi) Bonus shares: The Committee observed that care should be taken to ensure that the estimates in regard to bonus shares are not duplicated at any stage; (vii) Loans from local authorities: The Committee, suggested that the State Statistical Bureaux should consolidate statistics of local bodies in this regard; and (viii) Loans from cooperatives: The Committee found that there is a considerable lag in the consolidation of data on loans and advances to households extended by the primary credit and non-credit co-operative societies. The Committee opined that the RBI may be requested to expedite such consolidation so that appropriate data for the latest year are included in the CSO’s quick estimates.

V.12 Accordingly, the reforms that were implemented are: the survey on ownership of deposits with commercial banks, which was biennial, has been made annual. Similarly, the survey results relating to growth of deposits with non-banking companies are being made available periodically, coordination between the RBI and CSO was improved, the survey on ownership of Government debt has been revived, security deposits with trusts, housing boards and Electricity Boards: Household deposits with Electricity Boards are now being covered.

Prof. Raja J. Chelliah Expert Group on Savings and Capital Formation (1996)

V.13 Given the fact that the Indian economy has gone through many changes as a consequence of major economic reforms with rapid financial liberalisation and advancement of information technology, a need was felt for radical change in the management of information. Hence, the Expert Group on Savings and Capital Formation under chairmanship of Prof. Raja J. Chelliah was set up in 1995 to undertake a critical review of the available estimates of savings and capital formation in the economy and recommend improvements in the methods and procedures of estimation with particular emphasis on the household and private corporate sectors. Prior to Chelliah Expert Group, K.N.Raj Expert Group had also favoured uniformity in estimation methodology of various agencies involved. Accordingly, the estimates of gross domestic savings released by the RBI and CSO were based on a uniform methodology and the same data sources. The estimation work for household savings is divided between these two agencies. While the CSO prepares the household sector savings in the form of physical assets, life funds, provident fund and pension fund, the responsibility for estimating financial savings of the household sector in a number of instruments, namely, Currency, Deposits, Shares and Debentures, and Claims on Government lies with RBI. The Chelliah Expert Group was also mandated to look into desirability of continuing with the residual system of estimation and to recommend a system of conducting appropriate Studies/surveys on a regular basis to determine important ratios and proportions used in the estimation of capital formation and savings.

Issues Identified and Reforms Suggested

V.14 The problem of obtaining the required data on a regular basis is severe in the case of the household sector. Given its coverage, the estimates for the household sector are worked out on the basis of the available data from various censuses, sample surveys, research studies and assumed relationships, and as such, the estimates, therefore, would have errors. Estimation with regard to household sector is based on the residual approach and not independently made. Thus, the errors in the estimation of aggregate investment in the economy and that occurring at the stage of estimation for the organised segment would be reflected in the estimates of the household sector.

V.15 Select reforms suggested were the following: (i) For estimating the currency component of the household sector savings, the Group recommended the use of flow of funds data for estimation household savings in the form of currency and also suggested for taking an average (moving) ratio for the past three years for the years for which flow of funds data are not available; (ii) the Group recommended for showing separately the consumer credit by banks and non-banking financial companies in the NAS Statement: ‘Financial assets and liabilities of the households sector’; (iii) In the light of no feasible alternatives to residual method, the Group suggested for exploring the feasibility of conducting independent income expenditure surveys on a periodic basis for validation of estimates generated by CSO; (iv) The Group suggested that enterprise surveys on the lines of All India Debt and Investment Survey (AIDIS) to have the estimates of the complete household sector and also indicated the need for having to conduct such surveys preferably every five years; and (v) The Group suggested for taking into account in the estimation of savings all new financial instruments as and when the relevant data become available.

National Statistical Commission, 2001

V.16 The National Statistical Commission in 2001 under the Chairmanship of Dr C. Rangarajan stated that the estimates of savings suffer from a number of limitations, mainly from deficiency of data. On a review, the Commission listed the following observations which appeared in the Chelliah Committee Report and are still valid and need attention of the concerned agencies for implementation:

‘Discrepancies’ and ‘Errors and Omissions’ in National Accounts Statistics and an Alternative Estimate of Savings from Flow Accounts assuming PFCE to be Firmer Estimate

Discrepancies

V.17 The discrepancies presented in the national accounts statistics refer to the differences between the GDP estimated by economic activity marked up by net indirect taxes (indirect taxes less subsidies) to arrive at GDP at market prices; and the GDP estimated from the expenditure side. These discrepancies are the sum total of (i) the statistical discrepancy in the National Disposable Income and its Appropriation Account,

(ii) errors and omissions in the Capital Finance Account (in the estimation of gross capital formation), (iii) adjustment of merchandise exports to the change of ownership basis in the External Transactions Account, and (iv) adjustment of merchandise imports to the change of ownership basis, also in the External Transactions Account. The component-wise break-up of the discrepancies is available at the time of releasing the Quick Estimates of National Income, Consumption Expenditure, Saving and Capital Formation on 31st January, when all the above accounts are compiled based on detailed data given in the balance of payments statistics. The estimates of Saving and Gross Capital Formation (which is estimated using the funds-flow approach as sum of saving and net capital inflow from abroad) and their rates in terms of GDP at market prices are also released at the same time, as these are compiled using the detailed data available from central and State Government accounts and on the private corporate sector from the Reserve Bank of India.

Errors and Omissions

V.18 The gross capital formation in the NAS is arrived at as sum of Savings and Net capital inflow from abroad (current account balance). This indicates that the finances are available for gross capital formation. A second estimate of gross capital formation, arrived at from the commodity flow approach, gives the availability of capital goods for investment in the country. The link between the two estimates, however, is the household sector capital formation (also taken as savings of households in physical assets), which is arrived at as residual from the commodity flow approach capital formation estimates, after accounting for public and private corporate sector’s investment, data on which is directly available. The difference between the two estimates is termed as ‘errors and omissions’.

Alternative Estimate of Savings from Flow Accounts

V.19 If one were to estimate overall savings as excess of consumption expenditure (private final consumption expenditure and Government final consumption expenditure) from the disposable income of the nation, the discrepancy with the present estimates of savings as shown in National Accounts Statistics will be of the order of 1.6 per cent in 2006-07. This discrepancy is presently adjusted against PFCE while compiling the input-output tables for the Indian economy, since the input-output tables do not have any provision for discrepancies.

V.20 The above table can further be disaggregated by the institutional sectors, Government and households as well as for the total economy in the following tables:

Select Other Major Issues in the Estimation of Savings and Investment

• Due reflection of financial deepening in financial savings estimates: During the 1990s, the process of financial deepening has strengthened in India as judged from a host of indicators. A sharp increase in the preference for financial assets as against physical assets in the 1990s implied that this preference is essentially a reflection of financial widening in the sense of availability of a larger menu of financial assets, and of a sharp reduction in financial controls. The growing flexibility that investors enjoyed in terms of choice of assets in the increasing financial portfolios and in terms of optimisation of rates of return with risk minimisation has been furthered by financial sector reforms. Chelliah committee, 1996 noted that as the reforms are intensified, financial savings are bound to play a more crucial role in economic development than hitherto. There are doubts as to whether this deepening has fully reflected in financial savings of the household sector contrary to the a priori expectation that financial development should spur economic growth and savings.

• Validity of rates, ratios and norms: Presently, a number of ratios and norms are used in the preparation of estimates of capital formation for the economy as a whole and in respect of the household sector. Some of these ratios are based on five-yearly input-output transactions tables (IOTTs) of the CSO. If these ratios are updated more frequently, the estimation of capital formation would considerably improve in qualitative terms and the relative size of errors and omissions would probably come down.

•Capturing data on various constituents of the household sector: The problem of obtaining the data on a regular basis is severe in the case of the household sector. This sector comprises not only farm households engaged in agricultural production, but also individual households and unincorporated enterprises engaged in industry, trade, transport, finance, private trusts etc. The estimates for the household sector are worked out on the basis of the available data from various censuses, sample surveys, and research studies and assumed relationships. The estimates, therefore, would have errors.

• Periodical surveys and type studies be conducted to collect income and expenditure data on NPISHs: The residual and indirect approach for the household sector has been adopted due to the absence of income-expenditure surveys for pure households (consumer households) and Non-Profit Institutions Serving Households (NPISHs) estimates. The National Statistical Commission has recommended that periodical surveys and type studies be conducted to collect income and expenditure data on NPISHs, since estimation through direct surveys for pure household sector is not enough for national accounting purposes.

• Timely compilation of flow of funds data as basis for savings estimates: The flow of funds data should be used for estimating household savings in the form of instruments of financial savings, namely currency. For the years for which flow of funds data are not available, the average ratio for the past three years should be applied. A reasonably expeditious system needs to be evolved to reduce the time lag in making available the flow of funds accounts.

• Estimation of ‘currency’ component: At present household share is taken as 93 per cent of total currency with public. This proportion is being used for a long time and needs a revision. It has been seen that some companies club their cash holdings with bank balance. Therefore, bank balance should be excluded while estimating currency holdings. Further, a more serious question is whether the total amount of currency should be treated as savings or a part of it. Cash with household may be used for consumption, for expenditure to be incurred in the near future and speculation. Expenditure on consumption cannot be part of savings.

• Estimation of deposits component: Household savings in the form of deposits consists of deposits with banks, non-banking companies, co-operative societies, electricity boards and trade debt (net). Deposits with non banking companies are based on RBI survey on growth of deposits with non-banking companies 1995-96. There is a time lag of 10 years. Similarly, there is a time lag of four to five years in respect of co­operatives societies’ data. Furthermore, deposits with electricity boards, housing boards, MTNL etc., are not available separately.

• Estimation of shares, debentures and bonds: The estimates are based on total paid up capital and the RBI Survey of ownership of capital of joint stock companies. From this survey, ratio of household share is derived and applied to the total paid up capital. The survey has been discontinued after 1995. Thus, the ratio that is being applied for current year estimates relate to 1995.

• Estimation of net claims on Government: Net claims on Government include household investment in Government securities, small savings, bearer bonds, capital investment bonds, etc. The household share in investment in Government securities is derived on the basis of RBI survey on ownership of Central and State Government securities, 1992. The survey needs to be updated. Others such as the provident fund, etc., invested in Government securities and small savings should be seen whether the same is adequately covered in the survey otherwise there are chances of duplication.

• Netting of liabilities in individual financial instruments: In practice, the household savings in individual instruments is presented in net terms by deducting the liabilities incurred by the sector. In this respect, Raj Committee opined that this way the method tends to under-estimate the size of financial savings. Further, there may be a distinction between the classes of financial savers and debtors within the household sector. The Committee, therefore, suggests furnishing a disaggregated as well as total savings in financial instrument so that for each instrument three sets of data may be generated viz. (a) savings in financial assets in gross terms; (b) increase in financial liabilities; and (c) net savings in financial assets.

• Issue of impact of high current transfers: In the context of household savings estimation, an issue that has been noted by Dr. S.L. Shetty of the EPW Research Foundation is as to what is the implication of high ‘current transfers’ from abroad on household savings, particularly financial savings? He argues that the current practice includes capital transfers also as a part of the financial savings thereby overestimating the same. It was clarified that private transfers shown in the balance of payments statistics are taken into account for the estimation of disposable income of the nation. Therefore, these resources should have corresponding entries on the financial savings of the household sector.

• Capital inflow and rise in savings: Given the necessarily complex way in which savings and investment are estimated in an economy like India with a large household sector, including unincorporated enterprises, the question arises as to whether the sudden increase in the estimated savings is more statistical than actual. In fact, since the current savings surge coincides with a massive increase in the inflow of portfolio FII investment in India’s stock and debt markets, it had been argued (see S.L. Shetty in Economic and Political Weekly, February 12, 2005,) that some of this capital may have been lodged with the banks and got erroneously recorded as constituting household savings invested in financial assets. It was clarified that in the present method of compilation of household financial savings, there is no scope for such erroneous recording.

• The major weak areas in the estimation of public sector savings are, (i) coverage of local bodies, (ii) coverage of quasi-Government (QG) bodies, (iii) the emerging entities, the public-private-partnerships and the Special Purpose Vehicles (PPP/SPVs). Besides these, there are conceptual issues relating to (a) treatment of defence capital expenditures and (b) treatment of capital transfers to quasi-Government bodies and private non-Governmental organisations.

• The major limitations in estimating capital formation are common to those in respect of Savings, namely, (i) coverage of local bodies and quasi Government bodies in the public sector, (ii) coverage of new companies, the companies executing projects through public private partnerships (PPPs) and the cooperatives in the corporate sector, and (iii) lack of direct data on savings and capital formation in respect of household sector. Besides these, the other limitations are: (i) non availability of current and reliable data on output of capital goods in respect of unregistered manufacturing sector, which is essential in the estimation of capital formation for the entire economy through the commodity-flow approach; (ii) lack of annual enterprise surveys and reliable data from benchmark enterprise surveys in respect of expenditures made by various industries on acquiring capital goods, as also on inventories, which is essential for industry-wise and State-wise estimates of capital formation; and (iii) while estimating the Fixed Capital Formation by the Commodity Flow Method, number of rates and ratios are being used in absence of direct data. Some of these rates and ratios are based on old surveys or studies, while some of them are updated using the data available from Input Output Tables at the time of revising the base year of national accounts series.

• The private corporate sector savings estimation is also beset with data gaps and limitations. The selection of companies in the RBI studies is largely governed by the availability of balance sheets and is not drawn using any statistical sampling technique. While other parameters like sales, total net assets, etc., may have high correlation with various financial parameters, use of paid-up capital (PUC) as the characteristic for blow up procedure is chosen mainly on consideration of the data availability at population level and partly on account of its relative robustness. Another limitation of using PUC comes from the fact that while the selection of companies for the studies exclude non-operating, defunct, and under-construction companies, the population PUC includes data on such companies. It may be stated that in the absence of the population PUC numbers at a time of estimating saving, these figures are estimated based on certain assumptions, such as average growth of population PUC in last 3 to 5 years. However, on availability of the population PUC subsequently, the population estimates of saving and capital formation undergo revision on account of change in the multiplier used for arriving at the population estimates.

V.21 To sum up, notwithstanding the conceptual strength of India’s savings estimation practice, a critical review of the present method shows limitations mainly emanating from source data quality, data gaps and estimation problems. Various authors have identified different areas for closer examination and proposed reforms like methodology to be adopted for estimation of savings, various parameters that go into the estimation of savings, data sources, data gaps etc, for close scrutiny. What comes out of the literature is that there is near unanimity in their thinking and writings that we need to take a relook at the entire procedure and data related issues pertaining to the estimation of savings at regular intervals as there is scope for further improvement in the procedures and method followed for savings estimation. This is the background against which, the High Level Committee on Savings and Investment (Chairman: Dr.C.Rangarajan) has been constituted by the Government.

Chapter VI : Estimation of the Household Financial Savings

VI.1 Household sector occupies a pivotal position in the generation of savings in the economy. The approach adopted by the HLC in reviewing the methodology of estimation of household sector’s savings and suggesting the reforms is ‘consultative’ keeping in view the fact that the data required for estimating the household savings has to be collected from several divergent sources including the apex institutions. The HLC focused on addressing the following relevant ToRs, while dealing with the household sector’s savings estimation.

i. To examine the feasibility of directly estimating household savings through integrated income and expenditure surveys;
ii. In the light of the higher growth path of the economy, to undertake a critical review of the available estimates of household savings and investment in the economy, with respect to data base, methods of estimation, reliability and interpretational significance;
iii. To examine if rapid financial deepening in the economy is getting duly reflected in the estimates of financial savings and suggest improvements, if needed;
iv. To examine the feasibility of arriving at separate estimates for pure households, household enterprises and unincorporated bodies through a suitable method;
v. To examine the empirical methods and procedures used in the estimates based on flow-of-funds method and suggest improvements thereon; and
vi. To suggest new data bases, if any, to be devised/built-up for improving the reliability or checking validity of the estimates.

VI.2 This Chapter examines the feasibility of directly estimating household savings through integrated income and expenditure surveys. It also undertakes a critical review of the available estimates of household savings in the economy, with respect to data base and methods of estimation, and provides recommendations in these regards.

Approach adopted by the HLC

VI.3 The approach adopted by this HLC is comprehensive and different as compared to the earlier committees. Keeping in view the entire gamut of issues germane to estimation of household savings, as overviewed in the preceding Chapter, the approach to be adopted by the HLC to address them was conceived along the lines as indicated below:

(i) At the outset, the HLC has examined the feasibility of introducing a direct estimation of savings. From that perspective, the HLC examined if preparation of household savings estimates can be based on an integrated income-expenditure survey as the alternative methodology and data base. As explained subsequently in this Section, it came out from the HLC’s interaction with NSSO that this cannot be done immediately but possible in due course. The HLC recognised that even when this comes into operation and stabilises in future, this method has to complement the present method of estimation of household savings as is the practice in several other countries. It was, thus, considered appropriate that the present flow of funds based indirect method needs to be strengthened as it will have to be used in conjunction with the direct estimation method, which will be eventually available for use.

(ii) The second component of the approach was to strengthen the present data base and methodology and for this purpose, the worksheet approach was followed wherein each and every item in the work sheet was examined with respect to data base, quality and methodology as also deviation from the accepted methodology. The focus of this exercise has been to improve the data quality.

(iii) In the context of the worksheet framework, the third component of the HLC’s approach was to examine, if an alternative and more accurate data base can be created, keeping especially in view the dated rates and ratios underlying the present estimation procedure, and the need for improvement of data coverage and evolution of an alternative data base. Towards this pursuit, it was felt that the different financial instruments can be sourced from the apex financial bodies under whose regulatory purview, the instruments are covered. The HLC’s interaction with the apex financial bodies, namely, NABARD, SEBI, NHB and IRDA helped in sensitizing them about their role in putting in place a statistical system which will enable regular flow of data to RBI and CSO for preparation of household savings and flow of funds with updated and accurate data. The outcome of the effort undertaken by the HLC in this regard is encouraging and due mention has been made in the Report at appropriate places under different instruments in the Section pertaining to the Recommendations (Box 6.1).

(v) Finally, in order to pave way for eventual introduction of periodical comprehensive survey, the HLC interacted with the NSSO to examine the possibility of introducing such a survey. The outcome of this interaction has turned out to be favourable in the sense that the NSSO has agreed to undertake such a survey for households with a pilot survey to begin with, in the year 2010-11. The survey will provide an alternative data base for direct estimation of household savings for cross-validation of the present estimates.

The Issue of Direct Estimation of Savings through an Integrated Comprehensive Income and Expenditure Survey

VI.4 One of the ToRs specifies that the HLC needs to examine the feasibility of directly estimating household savings through integrated income and expenditure surveys. The current practice of residual method of estimation of household savings is in vogue since

Box 6.1: HLC’s Interaction with Apex Financial Bodies to Streamline Data Flow

One of the major objectives of the HLC on Estimation of Savings and Investment in Indian Economy is to examine how the savings estimates for different constituent sectors, namely the public, corporate and the household sectors can be improved upon. Another related ToR of the HLC is to examine the alternative data base to be formed which could cross-validate the present estimates of savings. Towards these pursuits, the HLC examined the present data collection procedure adopted for estimating household financial savings and arrived at a consensus that in the existing scheme of things, although some apex bodies (namely, the NABARD and the NHB) are already involved, there is a possibility of consulting with other apex bodies such as the IRDA and the SEBI for the purpose of building up an alternative and more accurate database in respect of financial instruments under their regulatory purview and flow-of-funds accounts. The view of the HLC on the household sector savings was, thus, to engage with the major apex financial bodies namely, SEBI, NABARD, IRDA and NHB in an attempt to improve the data in respect of financial savings instruments coming under the purview of their regulatory operation and also flow of funds for the institutions under their purview on a consolidated basis. This forms a major shift in the approach of the HLC, vis-à-vis earlier such Committees, which will go a long way in putting in place, a data system for strengthening of the present estimates of savings and flow of funds for India.

No doubt, all the apex bodies in the financial sector have a significant role to ensure that the quality of data emanating from them is reliable which finally get incorporated in the household financial savings through the underlying Flow of Funds method. The underlying logic is: the household financial savings is estimated instrument-wise and among them, different instruments are presently estimated from the data received from individual institutions, thus, giving a scope for further development in association with the apex bodies.

Towards this pursuit, SEBI ’s involvement was envisaged in respect of household investment in shares, debentures, mutual funds and commercial bonds, IRDA’s in respect of ‘life funds’ of insurance companies, NABARD’s in respect of deposit, credit and investment data pertaining to co-operative banks, credit and non-credit societies and NHB’s in respect of household deposits with the housing finance companies. The HLC recommends creation of a regular data supply mechanism from the apex bodies to RBI. Going forward, it is also envisioned that all the apex bodies are engaged in the ongoing review of the data on household financial savings from the following standpoints:

(i) Identification of the database used;
(ii) Methodology prescribed and in practice used;
(iii) Validating the data as well as results; and
(iv) Comment on changes required in the procedure.

Annex 6.5 provides the details with regard to the institutions and the information base in respect of the items that need to be built upon:

(i) SEBI - Data on the household financial savings in respect of ‘shares, debentures and commercial bonds’ as per Annexed Format A6.5.1.
(ii) SEBI - Data on the flow-of-funds for registrars and broker houses as per Annexed Format A6.5.2.
(iii) NABARD - Data on the household financial savings in respect of cooperatives – banks and credit as well as non-credit societies in respect of deposits, loans as well as investment in shares and debentures as per Annexed Format A6.5.3.
(iv) NABARD - Data/projections for years for which flow-of-funds data are not available for cooperative credit as well as non-credit societies as per Annexed Format A6.5.4.
(v) IRDA – Data on insurance sector – both public and private – on household financial savings in respect of ‘life funds’ as per Annexed Format A6.5.5.
(vi) IRDA – Data on insurance sector – both public and private – on flow-of funds data for consolidated insurance sector as per Annexed Format A6.5.6.

The above information, when made available and incorporated in the estimation, would give a reliable estimate of financial savings of the household sector, as well as enable timely compilation of flow-of-funds accounts of the Indian economy.

the inception of estimation of savings in India, and the recourse to this method continues, for want of a better and acceptable practical method. Further, earlier Committees like Raj Committee and Chelliah Committee as also reputed economists and statisticians have endorsed India’s present practice as conceptually sound. It may be noted that Chelliah Committee (1996) recognised that as large scale income-expenditure surveys have so far not been successful in India, there seems to be no possibility of obtaining a direct estimate of savings of the household sector from an alternative method.

VI.5 The HLC considered this issue by reviewing certain analytical issues as they emerge from an examination of select country experiences (US in particular, which disseminates savings data based not only on income expenditure survey, but also FoF) in this regard and relevant for consideration in the context of a switch-over to the direct method of estimation of household savings from the indirect method that is currently in vogue in India. The HLC’s interaction with NSSO in this regard is documented later in this chapter.

VI.6 The countries like the US and the UK have established the mechanism to directly estimate the savings of their respective household sector through income-expenditure surveys, while India follows a unique way of estimating the household savings indirectly by a method, which is a mix of ‘flow-of-funds’ and ‘commodity flow’ approaches, where household sector’s share is extracted as residual. In the latter, the data on domestic investment, savings and their disposition into household, corporate and Government components are not based on direct observation but arrived through indirect estimation.

VI.7 Savings of an economic unit can be estimated either from the income account as earned surplus, being the difference between current income and current consumption and taxes (as followed in the US for households) or from the balance sheet as earned net worth, being the difference between changes in financial assets and liabilities (as followed in India for households’ financial savings). In the US and the UK, both the estimates are available, which help in cross-validating the income-expenditure survey based direct estimates of savings. But in India, only flow of funds based estimation is available for household financial savings and commodity flow based estimation is available for household physical savings. In the absence of income-expenditure survey, there is no direct estimation of household savings.

VI.8 Many eminent economists have been pointing to various lacunae in the present practice. Some of them, in particular, Professor T. N. Srinivasan, Yale University strongly feel that India should devise a system of direct estimation of household savings based on integrated household surveys based on data relating to income and expenditure of the households. We understand that once such attempt was made by the NSSO but the data, thus, obtained were not reliable. So the idea was abandoned. But with the changed time and with better design of an integrated household survey schedule, it is felt that the income and expenditure survey might be worth giving one more fair trial. Against this background, the need is to examine the feasibility of adopting an income-expenditure survey based estimation of household savings.

VI.9 The proposal for such an integrated survey has been in the air now for quite some time. As stated earlier, T. N. Srinivasan (1994) examined this issue at length, using the observations of Angus Deaton (1989) which described the extraordinarily complex problems found in estimating income in developing countries through a survey. Deaton (1989) had brought out how even in developed countries the measurement of self-employment incomes was notoriously inaccurate. Apart from the growing size of self-employment in India, Srinivasan (1994) argued that the upper income groups do not respond to surveys or are likely to understate their incomes. Srinivasan (2003) revisited the subject while reviewing the National Statistical Commission (NSC 2001) (Chairman: Dr. C. Rangarajan) recommendations and contended that “in the Indian context, it is virtually impossible to collect reliable data on incomes and receipts from household surveys” (p.306). The Chelliah Committee also addressed this issue and suggested that feasibility of conducting independent income and expenditure surveys on a periodic basis should be explored, “only for validation of the estimates generated by the CSO”. Also, as the all-India debt and investment survey (AIDIS) covers only pure households, the Chelliah Committee suggested that the enterprise surveys on the lines of AIDIS may also be designed and conducted.

US Practice

VI.10 Table 6.1 provides the difference between the US personal savings estimates made by the National Income and Product Accounts (NIPA) and Flow-of-funds Accounts (FFA). The sharp decline in the personal savings rate, as measured by NIPA, has kindled renewed interest in alternative measures of savings and their relative merits. In particular, savings as measured in the capital accounts prepared by the flow of funds of the Board of Governors of the Federal Reserve System has often been cited as an alternative to the income-expenditure system based on NIPA measure. As per the relevant data, the difference between the US personal savings estimates made by NIPA and FFA vary.

VI.11 The main difference between these savings ratios is that one is derived by treating expenditure on consumer durables as investment, while the other ones are compiled by considering them as final consumption expenditure. It has been argued that the fact that US uses different official savings ratios, which highlight the usefulness of this approach, and it takes into account the actual behavior of the households (ECB, 2007).

Measurement Issues

VI.12 Certain measurement issues identified in the cross-country literature on issues of estimation of savings are being discussed from the standpoint of reassessing the trends of observed fall in savings rates in the context of countries like US and UK (Chamberlin Graeme and Sumit Dey-Chowdhury, 2008). Some of these arguments and their relevance for India have been identified in Rath (2007) as below:

• Treatment of employers’ contributions to pension and insurance funds: Employers’ contributions to pension and insurance funds, excluding the benefits of these funds also form part of personal income of households. Conceptually, while this makes sense as an accounting entity and fits in well with the rest of the NIPA methodology, it fails to reflect the actual cash flow to households in any given period. To an extent, the recent decline in the savings rate in the US may reflect that not only people retire earlier and but they also live longer on their benefits.

• Treatment of consumer durable goods: Savings is roughly equal to post-tax income less consumption, so the measurement of savings depends critically on whether certain expenditures are classified as consumption or investment. A defining feature of net investment – or increments to wealth net of capital gains and depreciation – is that it increases the future consumption possibilities of households whereas current consumption expenditures does not. However, the NIPA treats net purchases of consumer durable goods, which also provide a stream of services over a period of years, as consumption rather than as investment. Therefore, acquisition of a durable good increases future consumption possibilities in much the same way that the acquisition of a financial asset or housing does, and for this reason many have argued that spending on durable goods should be treated as investment rather than consumption. Some recent research findings on the issue of the impact of capitalizing consumer durables on household savings are worth-commenting here. ECB (2007) estimates the impact of capitalising consumer durables on the Euro area household savings ratios and disposable income for the first time. It finds that the effect of capitalising consumer durables on savings ratios is moderate and is estimated between 1.0 and 1.8 per cent. This is lower than in the US, where the effect has been estimated to vary from 1.0 to 3.0 per cent. This shows that treating expenditure on consumer durables as investment increases the savings ratio in the EA. However, these issues find little relevance in the context of national accounts, as these accounts have a clear distinction of producers and consumers and include only those fixed assets which are used in repeated production of goods and services in the hands of producers. If one were to include the consumer durables as fixed assets, then the output resulting from these assets (such as cooking meals in the house, personal transportation, etc.) should also get included in the GDP of the country, and the persons associated with this production will be treated as ‘employees’, resulting in no unemployment in the country.

• Effect of inflation on measured personal savings rates: Another issue that arises is whether the savings should be stated in real or in nominal terms. The personal savings rate in the NIPA is calculated from nominal values of income and consumption. To the extent that inflation simply scales up the value of income and consumption, it will have little effect on the savings rate. Inflation tends, however, to raise interest income and outlays by more than the change in the general price level. As a result, savings rates vary with the rate of inflation.

• Treatment of capital gains and capital gains taxes: In the NIPA, personal income excludes capital gains (and losses) because they do not derive from current production. As a result, large capital gains realised during the stock market boom of the 1990s failed to boost personal savings. Indeed, they effectively reduced measured personal savings over that period because taxes paid on those gains are included in personal tax payments, which are deducted from personal income in calculating disposable personal income. Some have argued that the NIPA treatment of capital gains is inconsistent with its treatment of capital gains taxes. If capital gains are not part of income, then taxes on those gains should not be counted against income as personal tax payments. However, it was clarified that although the capital gains are not included in the NIPA, the disposable income includes dividends received in the form of property incomes; hence, there is no inconsistency as far as payment of capital gains taxes are concerned. Similarly on the production side, the change in the value of inventories during the year is taken into account in the measurement of current year’s output. The inventories are measured through perpetual inventory method, in which the items of goods are valued at the time when they enter inventories and again valued at the prices prevailing at the time when they are removed from inventories. The capital gains which are not recorded in the flow accounts refer to the capital gains on assets, which get recorded only in the other changes in assets account.

• Treatment of human capital: According to Roy H. Webb (1993) (Personal Savings Behavior and Real Economic Activity, Federal Reserve Bank of Richmond Economic Quarterly, 1993), another definitional problem is dividing private spending between consumption and investment. NIPA investment is defined as the purchase of physical assets. If a person acquires productive capabilities through additional schooling, any payments for tuition, textbooks, and related items are defined as consumption. Many economists, however, see a strong analogy between physical capital, the tangible assets that can be used for future production, and human capital, the skills and abilities that people can use for future production. Since future production can be boosted by either physical or human capital formation, and since the purchase of either human or physical capital involves a trade-off of consumption today for future productive capacity, it is somewhat arbitrary to label spending for one as investment while labeling spending for the other as consumption. Again, it was informed that the System of National Accounts provides a comprehensive accounting framework within which economic data can be compiled and presented in a format that is designed for purposes of economic analysis, decision-taking and policy-making. The issues of treatment of human capital, etc., therefore, do not find place in the asset boundary or the production boundary of the SNA.

India’s Migration from Indirect to Direct Method – Practical Issues

• An important issue in the process of migration from indirect to direct method would be the conduct of an integrated income-expenditure survey, as far as pure households are concerned. The major issues confronted by the US while adopting and estimating savings through direct method have already been explained in the previous paragraphs. Needless to mention, issues of similar nature are likely to be encountered in India also.

• The estimates under the flow method vis-a-vis the present method: Table 6.2 compares the estimates of India’s household savings that would emerge from an application of two methods – the existing method and an alternative method where household savings refers to the difference between personal disposable income and private final consumption expenditure. This estimate roughly corresponds to the direct method. The difference between the two estimates is in the range of (–) 0.6 to 2.7 per cent of GDP during the period 1999-2000. This discrepancy is adjusted by the CSO in PFCE in the IOTT.

• Sampling and Non-sampling Errors: The literature has widely commented on the limitations of the survey findings on household savings. According to the Bureau of Labour Statistics, the US Consumer Expenditure Survey is subject to two types of errors, non-sampling and sampling. Non-sampling errors can be attributed to many sources, such as differences in the interpretation of questions, inability or unwillingness of the respondent to provide correct information, mistakes in recording or coding the data obtained, and other errors of collection, response, processing, coverage, and estimation for missing data. The full extent of non-sampling error is unknown. Sampling errors occur when the survey data are collected from a sample and not from the entire population.

Discussion at the NAS Advisory Committee (ACNA) Meeting in September 2007

VI.13 A proposal by RBI to the National Statistical Commission was circulated among the Members of the ACNA to give their views on the proposed income-expenditure survey to estimate savings. While responding on the issue of direct estimation of savings, there was a near unanimity among members on the RBI’s proposal on the desirability for an independent income-expenditure survey as is the practice in many countries – developed and developing. Among others, it was stated that such surveys would provide insight on the distribution of consumption, savings and income, while providing cross-check for the present household savings estimates. However, the Committee noted a range of problems that need to be deliberated upon and resolved before such a Survey could yield direct estimates of savings and Capital formation for the Household sector and/or provide a credible cross-check for the current estimates. First, the household sector in the national accounts comprises besides households viewed as a pure consumption unit (this forms the universe for the current NSSO Consumer expenditure Surveys), all un-incorporated enterprises and, Non-Profit Institutions (NPIs) and, without a full and unduplicated coverage of all three components, residual approach as at present might have to be adopted. In respect of NPISHs, there was no full and complete listing of such institutions which could serve as the frame for a survey of their income and expenditure. In respect of the un-incorporated enterprises, survey of all economic activities in the same year was needed – which is not attempted currently for a number of valid reasons. Further, as is well-known, a very large proportion of such enterprises do not maintain proper accounts so that the estimates of value added and capital formation derived from the NSSO Enterprise Surveys have not been found adequate for direct use in the National Accounts. These difficulties are directly carried forward in respect of all the self-employed households which derive their income from one or more such enterprises-including those in partnership with other households. Even if all the above issues are disregarded, it needs to be recognised that currently, canvassing the standard Consumer Expenditure schedule in the NSSO Consumer Expenditure Surveys takes close to or more than three hours per household. And, combining an Income Survey with NSS Consumer Expenditure Survey by adding one or more blocks to the already heavy consumer expenditure schedule carries with it sizeable risks of adverse consequences for the quality of response in respect of both income and expenditure. This is over and above the question of the ability of self employed households in reporting their incomes from own-account enterprises for which no accounts are maintained.

VI.14 In view of the past experience in the efforts of the NSSO with regard to household surveys, it was thought essential to conduct further pilot surveys to arrive at a satisfactory methodology. Brief profiles of the past such efforts from the NSSO are now in order. There were earlier efforts made by the NSSO, in the form of pilot studies ‘with the specific objective of evolving an operationally feasible and technically sound methodology for the household income surveys’ to facilitate the direct estimation of household savings. Earlier experience suggests that while the survey results threw considerable light on the problems of conducting household income surveys in India, the reported data were not found to be of high quality. Overall, the experience was disappointing as many problems associated with the direct estimation of income, consumption and savings could not be resolved.

VI.15 Another pilot survey on income, consumption and savings was originally proposed to be conducted during January-December 1992, ‘but this scheme could not be implemented owing to financial constraints’ (NSSO, 1995).

Earlier Attempts by the NSSO in Conducting Household Income-Expenditure Survey

VI.16 The need for conducting large-scale, nation-wide, comprehensive and integrated surveys on household income, consumption and savings in India had been acutely felt in official circles right since the early 1950s. It is reported that in the 9th (1953-54) and 14th (1959-60) rounds of NSS survey, attempts were made to collect information on household incomes along with consumer expenditure data on experimental basis, but the results were not satisfactory as no proper methodology for systematic data collection on household incomes had yet been evolved. Information on household receipts and disbursement were collected in the 19th (1964-65) and 24th (1969-70) rounds through an integrated household schedule. Significantly, during these rounds NSS estimates of consumer expenditure were found to be substantially lower than in other rounds. It was inferred that this was due to prolonged interaction effect. The long questionnaire integrating data collection on income, expenditure and savings involved protracted hours of interrogation, which may have led to respondents’ fatigue thereby adversely affecting the quality of response on data; could have also led to under-reporting.

VI.17 With a view to obviating the difficulties experienced in the above integrated household surveys, NSSO conducted a Pilot enquiry in 1983-84 with the specific objective of evolving an operationally feasible and technically sound methodology for the household income surveys. It adopted two different approaches for the measurement of household income. First, there was a household income schedule to collect directly income from different sources of earning; and second, there was a schedule on household consumption and savings drawn up aiming at obtaining household income as the sum of household consumption and savings. One-third of the households surveyed were subjected to enquiries on both the schedules to enable the study of prolonged interaction effects and response fatigue, any. This survey focusing on methodology was carried out on a limited scale covering only five States and four metropolitan cities (NSSO, 1993 and 1995).

Limitations of the Pilot Study Results

VI.18 The results presented in the NSSO Reports (August 1993 and September 1995) threw considerable light on the problems of conducting household income surveys in India. The experience is, however, somewhat disappointing, for the results show that many problems remain to be resolved. One needs to conduct further full-scale pilot surveys for finally arriving at a satisfactory methodology for household income surveys. The reasons may be stated as follows:

VI.19 Reported data were not of high quality. This was realised at the stage of data processing. Income and savings are sensitive subjects of enquiry and the long interviews made data collection really difficult. The weaknesses of primary data stood in the way of reaching firm conclusions in many respects.

VI.20 The sample size proved to be inadequate for highly variable items like savings and its components. The study of savings was not, however, the major objective of the pilot survey.

VI.21 The most important limitation of the study was that the schedules of enquiry did not seem to have been very well designed. Insufficient care was taken to collect data on non-monetised components of income and consumption. This is the most likely explanation of the large difference C+S minus Y in the rural sector, especially for non-wage/salaried households.

VI.22 The NSSO Report (September 1995) States that, “It should be clear from the overview of results presented above that the Pilot Survey on Income, Consumption and Savings did not settle all issues connected with the methodology to be followed for household income surveys in India. It is, therefore, essential to conduct further pilot surveys for resolving these issues with a view to arriving at a satisfactory methodology”. The NSSO Report (September 1995) further confides that the Governing Council of the NSSO had approved the Scheme of the Second Pilot Survey on Income, Consumption and Savings, which was originally proposed to be conducted during January-December 1992, “but this scheme could not be implemented owing to financial constraints (ibid: 40).

The HLC’s Interaction with the NSSO

VI.23 Against the above mentioned prior understanding of the HLC on related survey issues in respect of NSSO, the Members of the HLC interacted with the NSSO on May 26, 2008 in Kolkata to elicit from them their earlier experiences in designing and conducting comprehensive integrated income and expenditure survey. The purpose of this interaction with the NSSO was to explore the possible constraints and obstacles that one may face in conducting such surveys. Besides, it was also discussed as to what should be the parameters on which surveys may be designed so that it can truly be an integrated and comprehensive survey encompassing a wide variety of parameters related to income, consumption, savings and capital formation and other related aspects.

The NSSO’s Feedback on the Direct Estimation of Household Savings

VI.24 The NSSO in its presentation to the HLC mentioned that they have endeavoured to estimate the indicators of financial flows in the household sector. It has generally avoided collection of data on household income in view of the inherent difficulties of the task. Instead, it has been more emphatic on measuring household consumer expenditure. NSSO data on consumer expenditure have been widely used for studies on economic inequality. In conjunction with savings data, they have been used to estimate the size distribution of income.

VI.25 The NSSO has been conducting different surveys, the last being 61st Round (2004-05) quinquennial survey: So far as the household surveys are concerned, 19th to 25th rounds of NSSO (1964-1971) surveys cover the data on receipts and disbursements of the household and the data were collected in an Integrated Household Survey schedule. Since 27th rounds of NSSO (1972-73) Consumer Expenditure Survey has been broadly stabilized and has become the integral part of the quinquennial surveys’ Rounds viz, 27th, 32nd, 38th, 43rd, 50th, 55th and 61st .

VI.26 The NSSO in their presentation enumerated the design and problem faced in Income and Expenditure Survey, canvassed in Rounds 9-14. In the process , the NSSO discussed the details of the 10th round, wherein framework of Receipts & Disbursements blocks were used with 78 items on both the sides. Due to too detailed survey schedule, the results so obtained were not reliable as it might have resulted in correspondent’s fatigue.

Later Endeavour: 19th to 25th Rounds of NSSO (1964-1971)

VI.27 The later exercise was done for 19th to 25th round of NSS (1964-1971). In these rounds the receipts and disbursements blocks were much more condensed, with only 16 items on each.

• The experienced gained from this round reveals that the integrated approach to data collection on income, expenditure and savings from the same household necessarily led to a long questionnaire, causing informant fatigue.

• Lower estimates for consumer expenditure were obtained during the 19th to 24th rounds (in which the Integrated Household Survey schedule was canvassed) as the reporting of consumption expenditure could be affected by collection of data on income and savings from the same household.

• Therefore, the results so obtained did not stand to the statistical scrutiny and, hence, were not reliable.

Pilot Survey on ‘Income, Consumption and Savings’, 1983-84

VI.28 As comprehensive results were not satisfactory, another Pilot Survey on Income, Consumption and Savings, 1983-84 was conducted with the objective of development of an appropriate methodology for conducting comprehensive surveys of household income, which could overcome the difficulties faced in the earlier survey.

• For this pilot survey, strategy was to draw up two schedules of Income, Consumption, and Savings and to divide sample households into 3 sets of income, consumption and savings and both.

• Each sample household was visited twice. This enabled information on income and savings to be collected separately for two halves of a year. Information on consumption was collected in each visit with a reference period of 30 days, each sample household was visited twice.

• This enabled information on income and savings to be collected separately for two halves of a year.

• The pilot survey was conducted in 5 States, namely, Haryana, Maharashtra (including Bombay), Tamil Nadu (including Madras), Orissa and UP – and the cities of Delhi and Calcutta .With the Sample size of 100 villages and 80 urban blocks (Total).

• A sample of 24 households (three matched sets of 8 households each) was selected for survey from each sample village/block.

• All estimates were generated for three sectors of population: rural (R), urban non-metropolitan (U), and metropolitan (M).

• The objective of this pilot survey was to ensure reliability and validity of income data collected through direct enquiry on household income (Y) vis-à-vis income estimated as the sum C+S and reliability and validity of data on household savings (S).

Overall Assessment of the Pilot Survey

• The report of the pilot survey described the results as disappointing.

• This may be attributed to the fact that: (i) income and savings are sensitive subjects of enquiry and the long interviews made data collection really difficult. The weaknesses of primary data stood in the way of reaching firm conclusions, (ii) the sample size proved to be inadequate for highly variable items like savings and its components and (iii) it also revealed the problems of respondent resistance, which have accentuated over time.

• The Pilot survey of 1983-84 revealed some possibilities for improvement (e.g., in schedule design and sample size) and further full-scale pilot surveys were needed for arriving at a satisfactory methodology for household income surveys.

• However, no further surveys on income have been conducted since 1983-84.

Issues and Problems to be addressed in undertaking the Income and Expenditure Survey

Enquiries on Household Income

• As the pattern of income financial flow in the household is more complex than the consumption financial flow it requires the respondent to make much more computational efforts, than consumer expenditure survey.

• It arouses suspicion and meets with evasive / hostile response from the respondents.

Combined Income-Expenditure Enquiry

• A combined income-expenditure enquiry is expected to have the problems of a consumer expenditure enquiry in addition to those faced by an income survey.

• The reporting of consumption expenditure could be affected by collection of data on income and savings from the samehousehold.

Attempts to Collect Savings Data

• In case of collection of savings data it requires a substantial recall effort from the informants, especially in the reporting of stocks of products of agricultural activity.

• Like income enquiries, savings enquires arouse suspicion and hostility (e.g., in collection of information on bullion, ornaments, etc.,).

• It generally meets with evasive responses thereby resulting in unavoidable under­reporting. However, the savings data is less complex than income data.

Problems with Sampling

• In the pilot survey of 1983-84, the sample size was inadequate for highly variable items like savings and its components.

• Normally, the estimate of sampling error indicates how much the sample size needs to be increased to obtain estimates with tolerable variability.

• But if non-sampling errors like non-response predominate, they affect the estimate of sampling error and no statistical results exist regarding the effect of increasing sample size.

Resistance from Respondents

• Of all surveys of the NSS, the consumer expenditure survey is the one that meets with the most resistance from respondents.

• The detailed information sought (the schedule has over 340 items: not collected in terms of codes).

• The difficulty of informants in reporting quantities and values since normal people do not have these details on their fingertips.

• Nevertheless, there is a consistent time series of indicators derived from the consumer expenditure surveys based on established survey instruments.

Insights from Some Existing Surveys and their Results

VI.29 As against the experience of the NSSO in conducting income and expenditure survey, the NCAER has been doing similar surveys, the latest for which was for the year 2004-05 (Table 6.3).

(i) NCAER and Max New York Life for 2004-05

VI.30 One of the surveys conducted jointly by NCAER and Max New York Life on ‘How India Earns, Spends and Saves’ was for 2004-05. The Survey covered 342 towns and almost 2,000 villages across 250 districts and 2,255 wards. The sample size includes 63,016 households living in rural and urban areas. Its salient findings are:

• There were 205.9 million households in the country in 2004-05, of which 30 per cent (61.4 million) lived in urban areas and the rest (144.5 million) in rural areas. An average household in India has annual income of Rs. 65,041 and expenditure of Rs. 48,902, leaving it with a surplus of Rs.16,139 to save and invest.

• This survey shows that the households saved roughly 25.2 per cent of their income as in 2004-05, as against the present NAS estimate of household savings at 29.0 per cent for household sector in 2004-05 (based on personal disposable income). The difference could partly be due to the fact that the NCAER survey is a ‘pure household’ survey as compared to the combined household sector which includes unincorporated bodies, etc.

(ii) Invest India Income and Savings Survey, 2007

VI.31 The 2007 edition of Indian Retail Finance Markets is coproduced by Invest India Market Solution (IIMS) Data works and CNBC TV18. The Report is drawn from the findings of the Invest India Incomes and Savings Survey 2007 that was completed in June 2007. This is based on in-depth interviews with of 100,000 respondents aged 18 to 59 years with cash incomes and supported by a household listing sample of one million.

VI.32 The report provides a profile of India’s 321 million working age population with cash incomes. This includes commentaries and analysis for 141 million retail bank customers, 105 million life insurance customers, 77 million retail credit customers, 5.3 million mutual fund investors, 3.5 million stock market investors and 16 million residential property investors. Commentaries and analysis are provided also on 18 million gold investors and 30 million members of the informal savings sector, which are of interest in the Indian context.

VI.33 The latest survey conducted by IIMS provides the data of savings by various instruments across the States for 2006-07. The survey suggested that the State-level profile of savings instruments indicates the preference of savings instrument varies across the States quite significantly.

VI.34 However, the HLC notes that there is no evidence of such privately-run surveys contributing to an improvement in the estimation procedure for savings and capital formation.

HLC’s Perceptions

VI.35 The subject of direct measurement of household savings along with the various issues involved generally in improving the savings estimation in India, has been extensively studied in a collaborative publication christened Household Savings and Investment Behaviour in India (September 2003) jointly brought out by the EPW Research Foundation (EPWRF) and the National Council of Applied Economic Research (NCAER) in 2003. Some perceptions that come out from this analysis are as below.

VI.36 The need for undertaking household surveys on income, consumption and savings is self-evident but to assume that such surveys will necessarily provide firm direct estimates of household savings is not correct. The American and other experiences have shown that using flow-of-funds accounts are not necessarily inferior to using the field survey results in so far as financial savings are concerned. Both sets of estimations have to be undertaken so that they serve as cross-checks on each other. Therefore, it is noted that both the methods are complimentary to each other but not substitutes.

VI.37 There are a series of imponderables in expecting that the household survey results can give firm estimates of savings in India on annual basis. Even when a firm basis is established for conducting the surveys, they may provide bench-mark results on a decennial or at best on a quinquennial basis. There will have to be separate surveys for unincorporated enterprises as well as trusts and non-profit organisations which constitute a part of the household sector except for some designated as quasi-corporations. Above all, income, consumption and savings sample surveys are invariably known to be finding difficulties in providing appropriate representation to the extreme sections of households – the very poor and the very rich. Likewise, there is the respondents’ fatigue cited earlier due to long questionnaire and prolonged interaction effect. But, these cannot be intractable problems for the sampling theorists and survey practitioners, whose expertise is considered as unique in India.

VI.38 Despite the plethora of problems associated with household income surveys, there is a strong case for venturing into initiating such surveys on a quinquennial basis. Direct methods of household savings estimation for the benchmark year (the survey year) may be considered as only one of the objectives of such regular survey. More important objectives would be to get insights into income and savings distributions of households in rural and urban areas by size groups as well as by socio-economic categories and, thus, help build a more meaningful social accounting matrix (SAM) framework for India.

VI.39 A possible roadmap would be as follows: At the outset, the NSSO may initiate a pilot survey on household income, consumption and savings and be prepared for launching regular quinquennial surveys on the subject soon thereafter. This may be a workable option than to firm up the whole arrangement to put in place a system of regular income, consumption and savings studies for the country. From the point of view of a full coverage of the household sector, Enterprise Surveys have to be so designed as to cover the savings and investment activities of unincorporated enterprises which will remain covered in pure household surveys including the AIDIS.

VI.40 The problems that the proposed survey designers are expected to face is the difficulty in capturing the incomes and savings of richer households in urban areas. To obviate that problem, one method would be to segregate income-tax paying assesses in urban areas above a cut-off limit of say, Rs.5 crore and study their income-tax returns to arrive at estimates of their incomes and savings separately. The results may have to be supplemented with separate sample studies on such households. It would be useful to have the income tax revenue statistics published.

VI.41 While there are no two views about the desirability of such a survey in India, the feasibility of conducting an income-expenditure survey needs to be examined as certain difficulties could creep in designing and conducting such a survey in India. Keeping in view the fact that such an experiment in the past did not yield meaningful results, both direct and indirect methods have their own merits and shortcomings. In view of the issues in the context of a migration to direct estimation of household savings in India, the need is to have a better design of an integrated household survey schedule and examine the feasibility of adopting an income-expenditure survey based estimation of household savings, given the desirability of such a survey in India.

VI.42 During the HLC ’s discussion with the NSSO, it emerged that the NSSO is in favour of conducting the survey, but the question is how to address various practical issues. Some of them were: (i) enquiries on household income requires the respondent to make computational efforts, much more than consumer expenditure survey, while it arouses suspicion and meets with evasive / hostile response from the respondents, (ii) Attempts to collect savings data involves a substantial recall effort from informants, especially in the reporting of stocks of products of agricultural activity, while it generally meets with evasive responses and resulting in unavoidable under-reporting. However, the savings data may be less complex than income data, (iii) Problems with sampling especially with regard to the presence of sampling and non-sampling error, and (iv) Resistance from respondents, etc.

VI.43 The NSSO pointed out that the 66th Round of survey is slated for 2009-10 (employment, wage and consumption expenditure survey), making modification in the survey schedule to accommodate integrated income expenditure survey will not be feasible at this point of time. The NSSO can undertake a pilot survey on integrated income and expenditure of household in the year 2010-11.

Recommendation

VI.a As mentioned earlier, in order to pave way for eventual introduction of periodical comprehensive survey, the HLC interacted with the NSSO to examine the possibility of introducing such a survey. The outcome of this interaction has turned out to be favourable in the sense that the NSSO has agreed to undertake such a survey for households with a pilot survey to begin with, in the year 2010-11. The survey will provide an alternative data base for direct estimation of household savings for cross-validation of the present estimates. The HLC recommends that the NSSO should initiate at the earliest, the task of launching a comprehensive income-expenditure survey for household sector. While initiating such a survey, the NSSO may give a careful thought to the design issues through consultation with all appropriate institutions, like the CSO and the RBI. The NSSO should come out with a concrete action plan in this regard. To be able to find out the estimates of savings for all components of the household sector (as per national accounts statistics framework), and also generate reliable estimates of savings, consumption, investment and income at the State level in India, appropriate sampling design needs to be done. It is imperative that the survey of income and expenditure of ‘pure households’ needs to be taken up concurrently with the Enterprise Survey.

Review of the Existing Methodology

VI.44 As stated already, India follows a unique way of estimating the household savings indirectly by a method, which is a mix of flow-of-funds and commodity flow approaches, where household sector ’s share is extracted as residual. The estimates of financial savings in respect of households including non-profit institutions and unincorporated private business are estimated by changes in the net financial assets held by them such as: currency, deposits with financial institutions, shares and debentures, claims on Government, net equity in the life funds, provident and pension funds net of changes in financial liabilities. The estimates in respect of various financial instruments are arrived at as a residual after duly accounting for such instruments held by public and private corporate sectors. In estimating the financial savings of the households, increments in their holdings of financial assets are calculated net of increments in their financial liabilities. These annual flows are compiled instrument-wise. As the household sector is an unorganised sector and its balance sheets are not available, its financial flows are either estimated on the basis of firm information obtained from the accounts of counterpart institutions engaged in transactions with the households or as a residual after deducting the accounted financial information for the other sectors from the financial totals or are based on the information collected on sectoral distribution either directly or on the basis of surveys.

VI.45 The HLC felt that no alternative is possible to the present method of estimation of savings for the time being. At the same time, it is recognised that the present method needs to be strengthened through updation of rates and ratios and development of alternative data base for constant cross-validation of the estimates. Secondly, the envisaged system of conducting a regular and comprehensive income and expenditure survey of the household sector, once completed and stabilized, will complement the present procedure and provide a cross-validation for the savings estimates. Thereafter, both the methods would be used in the estimation and would complement each other.

VI.46 As of the current methodology, it is admitted that there is considerable scope for improving it. Accordingly, a thorough review of the existing procedure has been undertaken by the HLC in this context through a careful scrutiny of the worksheet and its components.

Recommendation

VI.b The HLC is of the opinion that no alternative is possible to the present method of estimation of savings but it is recognised that the present method needs to be strengthened through regular updation of rates and ratios and development of alternative data bases for constant cross-validation of the estimates. The envisaged system of conducting a regular and comprehensive income and expenditure survey of the household sector once in five years, when stabilise, will complement the present procedure and provide a cross-validation for the household savings estimates. Thereafter, both the methods would be used in the estimation and would complement each other.

A Review of the Worksheet pertaining to the Household Sector Savings and the Estimation Procedure Thereof

VI.47 In view of the above recommendation, the HLC undertook a detailed examination and discussion of each of the savings instruments in terms of conceptual and methodological issues pertaining to each of the instruments. Accordingly, the HLC has taken up a very close scrutiny of the worksheet employed to estimate the financial savings at present. Such comprehensive re-look at every step of the estimation procedure has brought out clues to the current inadequacies in data coverage. Especially, the usage of primitive surveys related various parameters that go into the estimation of financial savings, the antiquated rates and ratios that are used in the absence of latest surveys and the consequent under and over estimation at various stages of the estimation procedure would throw light on the current limitations and provide guidance to undertake a proposal for a comprehensive overhaul of the current system.

VI.48 While suggesting that the existing methodology may continue, the HLC is fully aware of the current problems in estimating various components of the household sector savings due to a variety of reasons and constraints, which result in over (under) estimation of various components of household savings in turn leading to inflating (deflating) the estimates of aggregate savings. Therefore, the HLC felt the need for a ‘worksheet approach’ to analyse the issues involved in each of the instruments of household financial savings. This has enabled the HLC to examine in detail various concepts and methodology associated with the estimation of household savings in India, keeping in view the existing data sources, data flows at regular interval, data gaps that persist and the consequent need to use rates/ ratios computed on the basis of certain periodical surveys.

VI.49 The details of the presently used underlying ‘rates and ratios’ are given as an Annex at A6.1 at the end of this Report for illustration purposes. Basic worksheet form employed is given at Annex A6.3.

Approach adopted by the HLC to examine the Worksheet Pertaining to Estimation of Household Financial Savings

The approach followed by the HLC to examine the worksheet comprised the following stages:

VI.50 The worksheet approach, wherein each and every constituent items are examined with respect to (a) sources of data and the underlying conceptual issues (focusing mainly on Raj Committee, Chelliah Committee, National Statistical Commission Reports wherever applicable), (b) methodology adopted for each of the parameter along with the limitations and problems, (c) prescribed procedure for estimation, (d) procedure followed in practice, (e) deviations and constraints in practice, and (f) lacunae identified in data and methodology.

I. The present practice is that either the data on various household financial savings instruments is collected from individual institutions or is estimated. The HLC recognised that the effort should be to keep the actual methodological practice in alignment with the description in National Accounts Statistics – Sources and Methods, 2007. Towards this pursuit, all the apex bodies in the financial sector have a significant role in ensuring the regularity, reliability and quality of data emanating from them. As such, all the apex bodies were requested to examine the data on the household financial savings instrument from the following standpoints:

i. Identification of the database used;
ii. Methodology prescribed and in practice used;
iii. Validating the data as well as results; and
iv. Comment on changes required in the procedure.

II. Discussions in the meetings on the lacunae; and
III. Recommendations.

Conceptual and Methodological Issues

Currency held by Households

VI.51 Currency is held by different economic agents and significantly by households. At present, a fixed ratio of 93.0 per cent of the ‘currency with public’ is used as the household savings in the instrument of currency. The members were of the opinion that the fixed ratio of 93 per cent used in the estimation hitherto needs periodical re-examination. There is a need to look into the validity of the ratio at periodic intervals.

Raj Committee

VI.52 Raj Committee has made the following recommendations regarding measurement of currency held by the households.

“From the total currency with public, currency held by private corporate business, cooperative institutions, Government undertakings and Government treasuries and local authorities (including port trusts) are taken out to estimate the currency held by the households. In the case of private corporate business, the estimate of currency held by the sample companies studied by the RBI is blown up on the basis of the ratio of paid-up capital of the sample companies to all companies. As the companies selected by the RBI are generally of big size, at least in the case of currency, the blown-up estimate may turn out to be smaller than the actual if worked out on the basis of the analysis of all the companies. It is likely to be so because in the case of smaller companies, the ratio of currency to paid-up capital would be relatively larger. With the rapid growth of smaller companies in recent years, errors between the estimate and the actual would be growing. This is likely to give an upward bias in the estimates of currency held by the households. As the number of companies expands, the ratios based on paid up capital may give an unrealistic picture. In this context the Working Group’s recommendation to have census studies of the corporate sector at least once in five years on a regular basis assumes significance. This is equally true of Government companies, particularly at the States’ level.

In the case of currency holdings of the cooperative sector, as separate figures of currency and bank balances for the primary and central land development banks are not available, the proportion of currency holding to total of cash and bank balances obtained for the central cooperative banks is applied to the total of cash and bank balances of the primary and central land development banks. Here again, the ratios obtained for the higher bodies are applied to the small bodies. In respect of bigger societies, the ratio of currency to total of cash and bank balances must be smaller than in the case of smaller cooperative societies. Thus, there is a built-in bias in favour of small cash holding of the primary credit societies which are scattered throughout and command the largest number. This situation is likely to give an upward bias in the estimation of household savings in the form of currency. ”

Chelliah Committee

The Chelliah Committee made the following recommendation.

“On the basis of the past sectoral pattern of currency holdings as revealed by the FoF data, the currency held by the household sector is presently estimated at 93 per cent of ‘currency with public’. The currency with the public is obtained from the monetary data for the last day of the financial year. While agreeing to the present proportion (i.e., 93 per cent), the Group recommends that for the years for which the FoF data are not available, an average (moving) ratio for the past three years should be applied in order to reduce somewhat the possible errors of a fixed ratio.”

VI.53 It may be noted that the National Statistical Commission has generally recommended the updation of various ratios and rates on the basis of surveys and studies with wide geographical spread, which is applicable to currency ratio as well.

Identification of Data Source

VI.54 The data on total currency with the public is obtained from DEAP, RBI. This data is used for estimation of currency with households.

Methodology Prescribed

VI.55 Household savings in the form of currency is estimated as a residual by deducting the amount of currency held by the private corporate sector and public sector enterprises from the total currency with the public. The data on total currency with the public (i.e., notes in circulation plus circulation of Rupee Coins and Small Coins minus cash in hand with banks) for each year are available in the RBI Bulletin. The institutions/sectors other than households in respect of which the amount of currency holdings deducted are: (i) private corporate business, comprising non-Government financial and non-financial companies registered under the Companies Act, (ii) co-operative institutions (other than co-operative banks), (iii) Government companies and statutory corporations (both financial and non-financial) and (iv) Railways and Central and State Government treasuries.

Methodology Used

VI.56 On the question of cash in hand, the Advisory Committee on National Accounts in its meeting held in June 1987 had suggested that the present procedure of estimating the currency with the households by deducting from the total currency, the currency held by various institutional sectors may be examined. It had further suggested that a simple procedure of possibly taking a proportion of the total currency in circulation might serve the purpose. Accordingly the RBI, on the basis of past trends of currency holding of the household and non-household sector estimated this proportion to be 0.93, which has been used for estimating the currency holding of the households from 1985-86 onwards.

Deviations from Prescribed Procedures

VI.57 The currency ratio of 93.0 per cent was adopted by RBI on the recommendation of Advisory Committee on National Accounts, whose meeting was held in June, 1987. RBI decided on this ratio on the basis of past trends observed from 1985-86 onwards. The ratio, however, was not subjected to a detailed check from time to time because of unavailability of data.

Discussion in the HLC Meetings

VI.58 The discussion in the HLC centred around two issues, namely, the appropriateness of (i) including currency as part of household savings and (ii) the presently used proportion of 93.0 per cent of ‘currency with public’ used as household currency holding.

VI.59 On the first issue, a view has been expressed to the effect that cash holdings of households may be used for consumption or may be intended for similar expenditure to be incurred in the near future or may be for speculative purposes; such cash holdings, therefore, cannot be part of household savings.

VI.60 Such a perception is conceptually incorrect. All additions to currency with the public, whether they are intended for acquiring consumption goods or for speculation or whether they are held as idle cash, constitute the counterpart of the expansion in net central bank credit to the Government (the so-called deficit spending by the State) or to the commercial sector directly, or through the commercial and cooperative banks; they could also be the counterpart of the net flow of foreign currency assets from abroad on account of merchandise, invisibles and capital transactions. The resources thus released to the public and private sectors represent the active use of private savings - that part of savings which the modern monetary medium of exchange obliges the public to hold in the form of additional currency. Such increases in currency may be akin to increases in inventories of goods forming part of physical capital formation (or savings). In a wider sense, all assets of the monetary system in the form of investments in Government securities and bank credit to the public and private sectors represent the active use of savings affected through additions to monetary assets – through additional currency holdings as much as through additional bank deposits. There can be constant switching as between assets, from currency to bank deposits or to other financial assets. The switch could also be in favour of consumption, but all incremental financial assets held as such retain the character of savings until they are used up in some form of current consumption.

VI.61 On the second issue, the currency ratio of household savings in currency as proportion of ‘currency with public’ used at present is 93 per cent and the ratio was determined in the year 1985-86. This is dated and needs to be revised to take care of the structural changes in the economy and capturing the quantum of currency holdings with households. The members were of the opinion that the fixed ratio used in the estimation needs re-examination. The HLC noted that financial innovations are lowering the use of cash in the transactions made by the households in recent years. Therefore, it may be expected that the ratio should be lower in the recent years. Further, there is an urgent need to look into the validity of this ratio at frequent and regular intervals.

VI.62 The HLC noted that as per the earlier Committees, like Raj Committee, it was recommended that in order to find out the household currency holding residually from the currency with public, census studies of the corporate sector at least once in five years on a regular basis need to be conducted. The underlying reasoning for this recommendation was that the blown-up estimate of sample companies may turn out to be smaller than the actual if worked out on the basis of the analysis of all the companies. It is likely to be so because in the case of smaller companies, the ratio of currency to paid-up capital would be relatively larger. With the rapid growth of smaller companies in recent years, errors between the estimate and the actual would be growing. This is likely to give an upward bias in the estimates of currency held by the households. As the number of companies expands, the ratios based on paid up capital may give an unrealistic picture. This is equally true of Government companies, particularly at the States’ level.

VI.63 Similarly in respect of bigger co-operatives, societies, the ratio of currency to total of cash and bank balances must be smaller than in the case of smaller cooperative societies. Thus, there is a built-in-bias in favour of smaller cash holding of the primary credit societies which are scattered throughout and command the largest number. This situation is likely to give an upward bias in the estimation of household savings in the form of currency.

VI.64 It was indicated by the HLC members that the sample studies of the corporate sector finances published by the RBI shows a substantial rise in the cash holdings of corporate bodies. The same is true for the Government companies. This means that there should be a corresponding decline in the cash holdings of the households. Moreover, there have been structural changes in the economy during the recent years. Therefore, a ratio determined in 1985-86 needs to be revised to take care of the structural changes. The use of credit and debit cards and ATMs are lowering the use of cash in the transactions made by the households. Therefore, it is expected that the ratio should have become lower in the recent years.

VI.65 On the issue of implication of adoption of electronic alternatives on cash usage, it may be of interest to take into account the perception emerging from international experience. A stylised fact borne out from research on this area in advanced countries is that despite the strong growth in the adoption of electronic payment instruments, the stock of currency in circulation has not declined as fast (Amromin, Gene and Sujit Chakravorti, 2007). “On the basis of examination of data from 13 OECD countries for the period 1988-2003, it has been concluded that despite growth in adoption of electronic payments instruments, particularly debit cards, cash usage remains significant in most OECD countries suggesting that electronic alternatives have not succeeded in mimicking all the benefits of cash. It is also shown that electronic alternatives to cash reduce the demand for cash of certain denominations. While debit card usage has reduced the demand for low denomination notes and coins, demand for high denomination notes is generally less affected suggesting that these denominations are also used for non-transactional purposes”.

VI.66 Another point that comes out from international practice of reporting the flow of funds data in respect of different sectors (including households) is that ‘currency and deposits’ are usually reported together and not separately (Source: Points on International Comparison of the Flow of Funds Accounts, Research and Statistics Department, Bank of Japan, December 28, 2000). Important examples are UK, US, Germany, France etc. In cases where they are reported separately, usually a fixed proportion is adopted, example being Japan [where 90 per cent of currency holding is allocated to household sector and remaining 10 per cent to the private non-financial corporations sector (Hagino, Satoru, 2007)].

VI.67 What emerges, thus, from a comparative perspective of Indian approach to flow of funds and household financial savings vis-à-vis select country practices is that our practice of using a fixed proportion and reporting ‘currency’ and ‘deposits’ separately stands on a better footing, in so far as it confirms to detailed flow of funds accounting. Secondly, this practice can be further strengthened by periodic revalidation of the sectoral currency holding patterns, which has been attempted by the HLC subsequently here.

VI.68 However, there was no unanimous opinion about the method of ascertaining the accurate ratio. While some members were of the opinion that to re-examine the ratio there needs to be a comprehensive survey of cash holdings, other members of the HLC felt that since conducting such a survey is costly in terms of time and manpower the latest available flow of funds data might be used in this regard. However, flow of funds data suffers from the problems mentioned by Raj Committee. Moreover, the latest flow of funds data available up to 2000-01 exhibits an average 93.0 per cent ratio for households’ cash holdings.

VI.69 It was decided to work out the household share after deducting from the total currency with public, the shares of organised sectors (like, Government, corporate sector and others) on the basis of RBI data on currency with public as on end-march and the data that CSO provided in respect of ‘currency holding with the Government sector’ and DSIM/ RBI provided in respect of ‘currency holding with the corporate sector’.

VI.70 The revalidation exercise was conducted for 2003-04, 2004-05 and 2005-06 for finding out the percentage of currency that is held by households (Table 6.4). It comes out that the average household currency holding during these three years works out to an average of around 95 per cent of the currency with public.

VI.71 The simulated results in respect of the impact that different assumptions of household currency holding as proportion to currency with public (assumed at 85.0, 88.0, 90.0, 93.0, 95.0 and 98.0 per cent, respectively under the scenarios I to VI) would have on household financial savings and gross domestic savings are given in Annex A6.2. It can be seen that the impact is marginal with the difference between household savings in currency to GDP ratio vis-à-vis the present estimates (household currency holding assumed at 93.0 per cent of currency with public) ranging -0.1 to 0.1 per cent during the years 1995-96 to 2006-07. The impact is even more marginal when the present currency ratio (at 93.0 per cent of currency with public) is compared with the scenario with the currency proportion at 95.0 per cent.

Recommendations

VI.c Rather than using a fixed ratio like 93.0 per cent, which is a rough approximation, the HLC recommends that the currency held by households as per the latest available flow-of-funds data be considered for the purpose of estimating the currency component for the household sector ’s financial savings estimation. As the latest flow-of-funds data is available up to 2000-01 and significant shift in behavioural ratios (such as currency holdings with the public) might have happened since then with structural changes in the economy, the HLC recommends compilation of flow-of-funds data in a more timely manner to consider different ratios including the ratio of households’ cash holding on a more realistic basis. Therefore, the HLC recommends that the flow-of-funds accounts of the Indian economy need to be compiled regularly and in a more up-to-date manner. There is a need to closely examine the current data gaps and constraints coming in the way of compiling the flow-of-funds accounts regularly.

VI.d Till the flow-of-funds accounts compilation is updated on a regular basis, alternatively, the HLC recommends periodic review of the assumed proportion of the household share of ‘currency with public’. While estimating the household financial savings for a particular year, it may be appropriate to deduct the sectoral currency holdings (to the extent they are available for the organised sectors) from the ‘currency with public’. In case the former is not available, estimates for the year may be approximated by using the latest 2 to 3 years average holdings. This practice may enable to move away from the assumed proportion of the household share of ‘currency with public’ in a phased manner. Therefore, in future, household currency holding can be estimated residually following the method as stated in Table 6.4 for the purpose of estimation of the household financial savings. For this purpose, data on currency held by the corporate sector can be sourced from DSIM, RBI and currency holding by the Government sector can be sourced from the CSO.

Bank Deposits

VI.72 In a typically bank-dominated financial system as India is, bank deposits constitute a prominent instrument of financial savings for households. The relevant instruments here are deposits from (i) commercial banks, (ii) cooperative banks and credit societies and (iii) cooperative non-credit societies.

(i) Deposits with Commercial Banks

Raj Committee

VI.73 The recommendation of the Committee in this regard was as follows.

“The current estimates of the increase in bank deposit holdings of unincorporated enterprises and individuals are based on their ratio to total bank deposits as revealed by the surveys of ownership of deposits (BSR-4) undertaken by the RBI now on a biennial basis. Apart from expediting the survey results, the RBI needs to attempt a classification of the ‘ownership of deposits’ in a manner that would enable us to get a fairly disaggregated picture of the sub-categories owning deposits under the category of ‘individuals’. The data show that the sub-category called “others (including unclassified)” owned as much as 40.7 per cent and 32.1 per cent of commercial bank deposits, in March 1976 and March 1978, respectively. A meaningful classification of such deposits is necessary, particularly in the context of this Working Group’s recommendations on the need to provide a more detailed classification of the household sector into farm households, unincorporated enterprises and households proper in the non-farm sector.”

Chelliah Committee

VI.74 The Chelliah Committee has made the following observation.

“For estimating household savings in commercial bank deposits, aggregate deposits with scheduled and non-scheduled commercial banks are added. Data on aggregate scheduled commercial bank deposits as on the last day of March are specially collected by RBI from these banks. Non-scheduled commercial bank deposits are also obtained as on the last Friday of March. The aggregate commercial banks deposits so obtained are disaggregated into current, savings and fixed deposits by applying the ratios obtained from Form X Statement (which is submitted by all scheduled commercial banks to RBI). Thereafter, non-resident deposits, inclusive of interest thereon, are deducted to obtain domestic deposits, since non-resident deposits constitute foreign savings inflows.

The ratios of holdings of the household sector in current, savings and fixed deposits, excluding the deposits of the banking sector and non-resident deposits, are calculated on the basis of ownership pattern obtained from the Basic Statistical Return (BSR 4) which is concerned with the “Ownership of Bank Deposits by Type and Economic Sector”. The household ownership ratios are applied separately for current, savings and fixed deposits. The household deposits so obtained are then aggregated.”

Regarding the status of BSR 4 surveys, the Chelliah Committee has made the following observations:

“The survey on ownership of deposits with commercial banks which was biennial has been made annual with the latest survey pertaining to March 1993. Similarly the survey results relating to growth of deposits with non-banking companies are available up to March 1992.”

Identification of Data Source

• Data on total deposits with scheduled commercial banks are obtained from Division of Money and Banking, DEAP, RBI.

• Form X data are obtained from Division of Money and Banking, DEAP, RBI.

• Information about deposits under foreign currency, non-residents deposits are obtained from Division of International Finance, DEAP, RBI.

• BSR survey on “Composition and Ownership Pattern of Scheduled Commercial Bank Deposits” from DSIM, RBI.

Methodology Prescribed

VI.75 Data on commercial banks ’ deposits for each year are available as on last reporting Friday of the respective financial year. Deposits under foreign currency and/or non-resident deposits are excluded from bank deposits. These bank deposits are then bifurcated into current, savings and fixed deposits on the basis of relationship of these categories of deposits observed in Form X, which contains data on liabilities and assets of scheduled commercial banks in India. The share of household deposits in current, savings and fixed deposits is estimated on the basis of RBI annual survey on “Composition and Ownership Pattern of Scheduled Commercial Bank Deposits”. This survey is undertaken every year. From this survey, the table on ownership of bank deposits by type and economic sectors is used for working out household deposits. The household deposits consists of (i) Individuals (including Hindu Undivided Families), (ii) Trusts, associations, clubs etc., (iii) Proprietary and partnership concerns, (iv) Educational Institutions, (v) Religious Institutions and (vi) Others (not elsewhere classified).

Methodology Used

VI.76 The methodology used is generally the same as prescribed.

Validating the Data as well as Result

VI.77 No fixed ratio is used for estimation under this item. The necessary ratios are calculated every year.

Deviations from Prescribed Procedures

VI.78 Since the ownership of deposits by sectors is not available for the year for which estimation of financial savings is done and there is usually a year’s lag, the ratio for the year of the estimation is projected and to that extent there is deviation from the usual practice.

Discussion in the HLC Meetings

VI.79 The HLC noted that the data on deposits with scheduled commercial banks (SCBs) are based on the Survey of Composition and Ownership of Deposits, conducted as a part of BSR system (BSR 4). A stratified sampling design is used for selection of branches of banks for the survey. All the branches of the SCBs in the country are first stratified into basic strata based on State/Union Territory, population group of the centre where bank branch is located, and bank group. The population groups are (i) rural, (ii) semi-urban, (iii) urban, and (iv) metropolitan. Five bank groups, viz., (i) State Bank of India and its Associates, (ii) Nationalised Banks (including IDBI Ltd.), (iii) Regional Rural Banks, (iv) Other Indian Scheduled Commercial Banks or Indian Private Sector Banks and (v) Foreign Banks, are considered for the purpose. Thereafter, each stratum is sub-stratified into 3 size classes of deposits (currently the size classes are ‘up to Rs. 25 crore’, ‘Rs.25 crore and above and up to Rs.100 crore’ and ‘Rs.100 crore and above’) to form the ultimate strata. All branches having deposits of Rs.100 crore and above are included in the sample with certainty. From the other deposit classes, sample is drawn using the simple random sampling (SRS) technique.

VI.80 For example, following the above procedure, sample for March 2006 survey consisted of 10,531 branches of SCBs, out of 68,681 reporting offices as per quarterly BSR-7 return as on March 31, 2006. 2,759 branches, each with deposits above Rs.100 crore were selected with certainty and these branches accounted for 44.8 per cent of deposits as at end-March 2006. 7,772 branches were selected randomly as per the sampling design and these branches accounted for 7.8 per cent of total deposits. Thus, the 10,531 selected branches covered 52.6 per cent of total deposits.

VI.81 The State-wise coverage of sample (both in terms of number of branches and amount of outstanding deposits) is found to be adequate when compared to population level shares (Annex A6.6). The HLC was informed that differences in shares (between population and sample) in certain States are reflection of structure (in terms of size of deposits) of the branches in those States. For example, in the case of Delhi and Maharashtra, where many large branches (with deposits of Rs.100 crore or above) are present, the share of deposits in sample is higher than that in population. This is because all such branches are included in the sample with certainty. Similarly, in smaller States (with fewer branches or with branches with low deposits), the coverage in the sample is higher than in population, to ensure adequate representation of such States in the sample.

VI.82 It is a considered opinion of the HLC that representative nature of the sample study on ownership of deposits with scheduled commercial banks may be given a re-look in view of the predominance of the branches having deposits of Rs.100 crore and above in the sample. Details of the estimation procedure for each spectrum may be clearly stated. The HLC while acknowledging the survey to be a well-designed one considers it necessary that the possible margin of error needs to be provided. It was suggested that there is a need to re-look at the estimation procedure.

VI.83 DSIM, RBI was advised by the HLC to look into the formula and sampling issue. DSIM, in response, assured the HLC that the estimation procedure for future rounds of the survey will be modified so as to generate estimates at the ultimate stratum level (i.e., strata formed by classifying at State/UT X Population Group X Bank Group X Size Class). It also assured that the standard error of estimates will be computed for future rounds of the survey.

VI.84 It was suggested that farmers and non-Government organisations (NGO) should be included as separate categories. Further, it was also suggested by the members that bifurcation of NRE deposits between savings and fixed deposits is available with the Monetary Policy Department, RBI. There is no need to apply a historical ratio to bifurcate this. FCNR deposit can be entirely taken as fixed deposit.

Recommendations

VI.e RBI’s annual survey on ‘Composition and Ownership Pattern of Scheduled Commercial Bank Deposits’ needs to be looked at for further refinement in terms of :

• Representative nature of the sample

• Margin of error
  

• Reduction in time-lag

VI.f In view of the emerging importance of farmers and institutions like Non­Government Organisations (NGOs) and self-help groups (SHGs) in the financial system and also from the perspective of policy issues, the HLC recommends that such unincorporated non-profit institutions be treated as separate categories under households. Accordingly, there is need to capture these entities in the returns for deposits provided by banks to the RBI.

VI.g Since the ownership of deposits by sectors is not available for the latest year for which estimation of financial savings is undertaken and there is usually a year’s lag, the ratio for the year of the estimation needs to be projected. In this pursuit, an appropriate statistical basis for arriving at the ratios of household ownership in deposits (like using the average for the latest 2 to 3 years) needs to be arrived at through a consultative approach among various institutions such as the CSO, RBI, etc.

VI.h Presently, as per the existing method in the compilation of household financial savings, use is made of Section 42 data (as of the last reporting Friday of the year) for data on commercial banks’ deposits for each year while Form X data on liabilities and assets of scheduled commercial banks in India for the bifurcation of deposits into current, savings and fixed deposits, the latter being on end-March basis, and the Ownership Pattern of Scheduled Commercial Bank Deposits on an end-March basis. There is, thus, a need to standardise the database for the estimation of household deposits by using consistently the March 31 figure in place of the last reporting Friday of March. This is available for every year with a lag of three to four months.

(ii) Deposits with Co-operative Banks and Credit Societies

Raj Committee

VI.85 There is no specific suggestion by the Raj Committee for this particular instrument.

Chelliah Committee

VI.86 The Chelliah Committee has made the following suggestions.

“Data on assets and liabilities of all the co-operative societies are published in ‘Statistical Statements Relating to Co-operative Movement in India – Part I – Credit Societies, Part II – Non-Credit Societies’ by the National Bank for Agriculture and Rural Development (NABARD). Deposits with primary societies are taken as household deposits, while in the case of co-operative banks and other credit societies the household deposits are estimated using the information available on ownership pattern which gives details of such deposits. Individual deposits are treated as household deposits. Similar treatment is given to deposits with non-credit societies. As NABARD publications referred to above become available generally with a long time lag, the total deposits of the co-operative banks and credit societies are estimated for recent years using the deposit figures of co-operative banks, as available in NABARD’s summary publications, viz., ‘Important Items of Data, Credit and Non-credit Co-operative Societies’. The deposits are derived on the basis of the household’s share in total deposits as obtained through additional returns received by RBI from co-operative banks and credit societies. The flow of such deposits during a year represents household savings in such deposits. As regards non-credit societies, the estimation of household deposits is worked out on the basis of available data.”

Identification of Data Source

VI.87 The source of data for this item is the publication ‘Statistical Statements Relating to the Co-operative Movement in India’ (NABARD).

Methodology Prescribed

VI.88 Data on deposits of the households in co-operative societies and banks available annually from the publication ‘Statistical Statements Relating to the Co-operative Movement in India’ (NABARD) is made use of. The data is directly available in the above publication for different tiers of co-operative banks.

Methodology Used

VI.89 The data is estimated on the basis of past trends in growth rate or some other judgmental basis because of non-availability of those publications on time. Generally, there is a lag of six to seven years in the publication.
Validating the Data as well as Result

VI.90 Because of the unavailability of the NABARD publications in time, the item needs to be estimated on the basis of past trends in growth.

Deviations from Prescribed Procedures

VI.91 Because of the unavailability of the NABARD publications on time, there is deviation from the prescribed methodology. Till NABARD publications are made available on time, there is a little scope to improve on this account.
Discussion in the HLC Meetings

VI.92 The HLC felt that the data provided by NABARD to RBI as per the sectoral break-up indicated in the “Statistical Statements Relating to the Co-operative Movement in India” is appropriate from the point of view of computation of household financial savings. The timeline is an issue, which was discussed with the NABARD representative in the HLC’s meeting with Apex Bodies on March 14, 2008.

VI.93 According to NABARD, the data for the “Statistical Statements Relating to the Co­operative Movement in India” is compiled on the basis of audited data submitted by the State level and District level agencies. The data pertaining to the State Cooperative Banks (SCBs), District Central Cooperative Banks (DCCBs), State Cooperative Agriculture and Rural Development Banks (SCARDBs) and ICBs are compiled and forwarded by the respective agencies and the State level and national level consolidation for each of the agencies is done by NABARD. In the case of all other primary credit societies and all non-credit societies, the data is consolidated at the State level by the Registrar of Cooperative Societies (RCS)/ Functional Registrars/Concerned Directorates of individual States, and the national level consolidation is done by NABARD for each category of society.

VI.94 Collection of data is through a set of structured tables (51 input tables) from the SCBs/DCBs/ICBs/SCARDBs/RCS. These 51 formats used for collection of data have been adopted on the basis of the recommendations of the Review Committee on Co-operative Statistics, Department of Agriculture and Co-operation, Government of India and have been distributed to all the reporting agencies along with the guidelines for filling up the formats. Validation of data is done through a variety of approaches which include rechecking the data in the Annual Reports and Balance Sheets of the respective agencies and comparing the previous year’s tables. In case of discrepancies, clarification is sought from the concerned agencies.

VI.95 NABARD held the view that although data is submitted by the RCS to NABARD in a standard format, no standard system has been evolved for collection of data from primary society level for further consolidation at different levels. Effective and timely collection and compilation of co-operative statistics would be possible if all the co-operative organisations are required to submit the data in a prescribed format, annually. As various other agencies viz. National Federation of State Co-operative Banks (NAFSCOB), National Cooperative Union of India (NCUI) are publishing some basic data on credit and non-credit societies, Government of India may be requested to consider modification of the formats and make them uniform with those used by other agencies, so that the reporting agencies could submit the data in the same format to all the compiling agencies.

VI.96 The HLC has learnt that the delay with respect to co-operative data is on account of consolidation of data at the district, State and national level. Furthermore, the delayed submission of data is due to delay in the audit of the societies. It is felt by the HLC that provisional data may be published without much delay with subsequent correction after audit of the data.

VI.97 In this regard, a conclave with the RCS was organised by NABARD and held at Bankers’ Institute of Rural Development (BIRD), Lucknow on May 9, 2008 for sensitizing the Registrars of Cooperative Societies RCS about the issue of data from the cooperative sector where they play a major role. In this meeting, the Chairman, NABARD highlighted that the statistical database of the cooperative sector is extremely important not only for NABARD, but also RBI and Government, and drew the attention of RCS to reduce the time lag and that the lag should not be more than a maximum of one-year in publication of ‘Important statistics relating to cooperative movement in India’, the latest available being for the year 2001-02.

VI.98 It was clarified by Dr. Shetty, who was representing the HLC, that presently the weak areas of data base are being examined by the HLC, which has identified cooperative sector as one such weak area and for which NABARD as the apex body is expected to play a major role in making the data available to the agencies. A clear-cut action point emerged in the conclave that RCS will ensure posting of suitable staff for posting of data and their timely submission so that auditing can also be taken up early.

VI.99 An issue that there is heterogeneity in the pattern of receipt of audited Statements from States, such that while some States submit in time, some States delay in submission resulting in a situation where NABARD at any point will have in their possession some information. In such a situation, under the most ideal situation, the lag of one year in the NABARD publication may be difficult to achieve. Hence, the HLC feels that NABARD may supply the provisional estimates/projections based on the latest receipt position of cooperative data from RCS every year to RBI for household financial savings (for instruments like, deposits, credit and investment in shares and debentures all for households) and flow of funds and to CSO for national income accounting, as these estimates are to be prepared year after year in a time-bound manner.

VI.100 As a follow-up measure, Chairman of NABARD (vide his letter DO. No. NB.IDD/824/ CSD-32/2008-09 dated July 1, 2008) has informed the HLC that NABARD has advised all RCS to sensitize the officials at the grass root level through a capacity building (training) module to ensure that the quality data on cooperatives are made available to NABARD at least with a lag of one year. NABARD’s regional offices/training establishments are being instructed to get in touch with RCS as well as their district-level functionaries to ensure their participation in capacity building programme organised by NABARD, which in turn will help in bridging the data gaps in content, coverage and timeliness in respect of cooperative societies.

Recommendations

VI.101 While the HLC recognised the reasons for delay in the NABARD publications, it was pleased to have favourable response from NABARD in their effort to streamline the data issues. Accordingly, the HLC recommends the following:

VI.i In the light of genuine difficulties expressed by NABARD in regularly compiling the comprehensive data, NABARD may endeavour undertaking “projections”, which could be revised when firm data become available. The data issues need to engage the attention of the highest level in NABARD. In this context, the HLC recommends that NABARD should be involved in providing a projection of household deposits in co-operative banks and credit societies to get around the problem of considerable time-lag in the publication such as ‘Statistical Statements Relating to the Co-operative Movement in India (Credit and Non-credit societies)’.

VI.j All efforts should be undertaken to improve the time gap in the publication ‘Statistical Statements Relating to the Co-operative Movement in India (Credit and Non-credit societies)’, a process that is said to be underway at the NABARD.

(Action: NABARD)

VI.k NABARD should continue in its role as the nodal agency for the co-operative sector and provide data for the compilation of national accounts, savings estimates and FoF accounts in so far as the cooperative sector is concerned (Annexes A6.5.3 and A6.5.4). In this regard, NABARD may collaborate with National Federation of State Co-operative Banks (NAFSCOB), National Cooperative Union of India (NCUI) as these agencies are publishing some basic data on credit and non-credit societies.

(Action: NABARD, NAFSCOB and NCUI)

(iii) Deposits with Co-operative Non-credit Societies

Methodology Prescribed

VI.102 Data on deposits of the households in co-operative societies and banks available annually from the publication ‘Statistical Statements Relating to the Co-operative Movement in India’ (NABARD) is made use of. The data is directly available in the above publication for different tiers of co-operative banks.

Methodology Used

VI.103 The data is estimated on the basis of past trends in growth rate or some other judgmental basis because of non-availability of those publications in time. Generally there is a lag of six to seven years in publication.
Validating the Data as well as Result

VI.104 Because of the unavailability of the NABARD publications in time, the item needs to be estimated on the basis of past trends in growth.

Deviations from Prescribed Procedures

VI.105 In view of the non-availability of the NABARD publications in time, there is deviation from the prescribed methodology. Till NABARD publications are made available in time, there is a little scope to improve on this account.

Discussions in the HLC Meetings

VI.106 Discussions in the HLC meetings were on the lines as described above in the case of deposits with co-operative banks and credit societies.

Recommendations

VI.1 Considering the similarity in the nature of issues, the HLC’s recommendations pertaining to cooperative credit societies may be applied to the cooperative non-credit societies as well.

(Action: NABARD)

Non-Banking Deposits

VI.107 Non-banking companies consist of financial and non-financial Government and non-Government companies and electricity boards.

Raj Committee

VI.108 The observations/recommendations made by the Raj Committee with respect to non-banking deposits are as under.

“Estimates of household deposits with non-banking companies are made by the CSO on the basis of the RBI surveys on ‘Growth of Deposits with Non-banking Companies’. Serious doubts have been expressed regarding the adequacy of the coverage of non-banking companies, particularly in the non-financial category, which accept deposits from the public. Studies in the RBI have shown that in respect of the non-financial companies, the total public deposits even for the sample companies turned out to be higher than the total deposits for all companies revealed in the above surveys. This is an important area of household savings, particularly in the context of assessing household preferences for financial assets with higher yield. Therefore, it is imperative that steps are taken in the RBI to ensure adequate coverage of the non-banking companies in the survey of their deposits. Also, recently the fully Government-owned public sector undertakings began accepting deposits from the public. Arrangements may be made in the RBI to cover them also in the above surveys. Security deposits kept by households with various improvement trusts, housing boards and electricity boards should be treated as household savings in the form of deposits and should be accounted for as such in the estimation.”

Chelliah Committee

“Chelliah Committee held the view that information regarding deposits of households with Housing Boards and Telephone companies (similar to the treatment given to electricity boards) need to be collected and included as part of financial savings.”
Instruments

VI.109 The instruments are deposits from financial and non-financial companies, both for Government and non-Government companies.

Identification of Data Source

• Public deposits data, in respect of financial companies, are obtained from the Department of Non Banking Supervision (DNBS), RBI.

• Housing financial companies from National Housing Bank.

• Studies on Finances of Public Limited Companies and Finances of Private Limited Companies, DSIM, RBI.

• Returns from State electricity boards.

Methodology Prescribed

VI.110 The non-banking financial companies are broadly classified into (i) registered and regulated by RBI (ii) not registered with RBI but RBI issues direction relating to deposits acceptance activity and (iii) exempted from RBI regulations including requirement of registration. Under category (i) loan companies, investment companies, equipment leasing companies, hire purchase finance companies and residuary non-banking finance companies are included. Mutual benefit finance companies (notified Nidhis), Mutual benefit companies (potential Nidhis) and miscellaneous non banking companies (chit fund companies) are included under second category. Category (iii) includes insurance companies, stock exchange, stock brokers, merchant banking, housing and micro finance companies.

VI.111 The data source for estimation of household sectors’ deposits with non-banking companies, excepting SEBs, was used to be the RBI surveys on ‘Growth of Deposits with Non-banking Companies’ and studies on company finances published in the RBI Bulletin from time to time. As the survey has been discontinued after 1995-96, public deposits data, in respect of financial companies, are obtained from the Department of Non-Banking Supervision (DNBS) of RBI and for housing finance companies from National Housing Banks. The break­up of such deposits into Government and non-Government is obtained based on 1995-96 survey. In respect of non-Government non-financial companies, public deposits data available in the RBI studies on Finances of Public Limited Companies and Finances of Private Limited Companies is used. For non-financial Government companies, estimates are made based on the ratio as per the survey on Growth of Deposits with Non-banking Companies. That is, on the basis of DNBS survey, the ratio of Government to non-Government non-financial companies is calculated and this ratio is applied to non-Government non-financial companies’ data. The total security deposits of consumers for supply of power available from the annual reports of SEBs do not provide separate details of deposits by the households. The share of household deposits in the total security deposits is worked out by allocating it on the basis of the household share in total consumption of electricity.

Methodology Used

VI.112 The methodology for the non-banking financial companies is same as per the prescription. However, Studies on Finances of Public Limited Companies and Finances of Private Limited Companies, by DSIM, RBI are not available on time. More serious problem is that the blow-up factor for arriving at the global figures from the sample has not been available with DSIM for recent years. Therefore, the deposits of non-banking non-financial companies are estimated on the basis of last three years growth. The data on State electricity boards are obtained with a lag. Therefore, this item is also estimated on the basis of past trends in growth.

Validating the Data as well as Result

VI.113 The estimation of non-banking deposits of financial companies are based on actual data. In the absence of a blow up factor it is not possible to validate the data on non-banking non-financial companies. The State electricity boards’ data are validated at the time of compilation of flow of funds.

Deviations from Prescribed Procedures and Reasons thereof

VI.114 As noted above, there is no deviation for non-banking deposits of financial companies. However, estimation on the basis of past trend is worked out for non-banking non-financial companies and State electricity boards. Moreover, the break-up for Government and non-Government sectors are worked out through historical ratio obtained from the Survey of 1995-96.

Discussion in the HLC Meetings

VI.115 The HLC noted that the data source for estimation of household sectors’ deposits with non-banking companies, excepting SEBs, was used to be the RBI surveys on ‘Growth of Deposits with Non-banking Companies’ and studies on company finances published in the RBI Bulletin from time to time. As the survey has been discontinued after 1995-96, public deposits data, in respect of financial companies, are obtained from the DNBS of RBI and for housing financial companies from National Housing Bank. There is a need that RBI survey on ‘Growth of Deposits with Non-banking Companies’ is redesigned, keeping in view the fact that the unregistered companies are not covered in the data reported to the Department of Non-banking Supervision.

VI.116 The HLC observed that data gaps exist for a part of non-banking financial companies, which are partially covered at present. Non-banking financial companies can be grouped into three categories, viz., (i) companies registered with the RBI, (ii) those applied for Certificate of Registration but rejected by the RBI, and (iii) companies registered with Department of Company Affairs but exempted for registration with the RBI. According to the Department of Company Affairs, Government of India, there were 65,382 NBFCs as on March 31, 1999. The comprehensive regulatory framework for NBFCs was introduced by the RBI in January 1997 and it made it compulsory for all financial companies to register with the RBI. As many as 37,274 companies, which were conducting financial business, applied for Certificate of Registration (CoR) with the RBI as on June 30, 2000. Of these, 679 companies were approved for registration with permission to accept public deposits, 8451 were approved for registration without authorisation to accept public deposits, and 14,986 were rejected for issue of Certificate of Registration. Applications of 13,158 companies were at various stages of consideration (RBI, 1999-2000). Thus, as per the available data, only 57 per cent of the companies have applied for CoR and all those, which have applied have not been given the certificates as they have not met the prescriptions set by the RBI for registration.

VI.117 Amendment to the Reserve Bank of India Act, 1934 in 1997 gave comprehensive powers to the Reserve Bank to regulate the NBFCs. The amended Act, inter alia, provided for compulsory registration of all NBFCs. NBFCs can be classified into two broad categories, viz., (i) NBFCs accepting public deposit & (ii) NBFCs not accepting/holding public deposit. Total number of NBFCs registered with the RBI, consisting of NBFCs-D (Deposit taking NBFCs), RNBCs, mutual benefit companies (MBCs), miscellaneous non-banking companies (MNBCs) and Nidhi companies, declined from 13,014 at end-June 2006 to 12,968 at end-June 2007. The number of NBFCs-D declined from 428 as at end-June 2006 to 401 as at end-June 2007. Of the 401 deposit taking NBFCs, 362 NBFCs filed annual return for the year ended March 2007 by the cut-off date of September 30, 2007 (RBI, 2006-07).

VI.118 According to the Department of Non-Banking Supervision (DNBS), RBI, all NBFCs, which are registered with the Department of Company Affairs, need to apply to RBI for CoR for undertaking the banking/finance activity. Therefore, an NBFC cannot exist or function as a financial company without the CoR from RBI. This, however, is the technical or legal position. In reality, there are small size companies carrying on financial activities without explicit authorisation; they even accept deposits from the public (Report of the Committee on Informal Financial Sector Statistics, RBI, March 2001). There also exists a set of companies, which are rejected by RBI for issue of CoR on grounds of not complying with the extant prescription within a stipulated period, including the extension period and have to close down their business activity. These companies are also supposed to submit annual Statements on liabilities and assets in the prescribed Form. But, the response from the rejected companies is very less as only about 25 per cent of the companies submit the return.

VI.119 All NBFCs incorporated under the Companies Act are supposed to be registered with the Department of Company Affairs (DCA), Government of India. There is, however, a big chunk of financial companies at present, which are not registered with the RBI. For instance, of the 65,382 companies, only 37,274 companies applied for CoR as on June 30, 2000, implying that 28,108 companies exist without applying for registration with the RBI and conducting financial activities. Further, a set of companies have been exempted from registration with the RBI, which are (i) engaged in micro-financing activity, (ii) licensed under 25 of the Companies Act, 1956, (iii) not accepting public deposits; and (iv) mutual benefit companies in existence as on January 9, 1997 and having net owned funds of Rs.10 lakhs.

VI.120 The HLC is of the view that in respect of companies which have either not applied / been rejected for issue of CoR or are exempted from registration, there is a need to firm up the database. As observed by the Committee on Informal Financial Sector Statistics, RBI, March 2001, these companies are carrying on financial activity including accepting of public deposits. There is a need to ascertain an estimate of the quantum of household savings channelising into these entities.

VI.121 It was deliberated at length in the HLC meetings that estimation related to the deposits with non-banking companies is an important area of concern. It was conceded that there exists a large data gap in the estimation of household savings in the form of deposits with non-banking companies. This data gap is because of non-inclusion of a wide array of unregistered companies, which do not report to the Department of Non-Banking Supervision. Another issue discussed in the HLC was with regard to household deposits with Mahanagar Telephone Nigam Limited (MTNL) and Housing Boards, which according to the Chelliah Committee were to be incorporated in household financial savings. The HLC discussed the appropriateness of such a methodological change and took the view that these do not fall under the category of savings; they are rather advances against future consumption of the respective services in certain cases.

VI.122 Presently, data on deposits from the public are in the form of FD, RDs, etc., and deposits received by a public company from its shareholders (shown under ‘regulated deposits’) and money received from employees of the company by way of security deposits or advance from purchasing, selling or other agents in the course of company’s business (shown under ‘exempted borrowings’) of housing financial companies. These data are provided by NHB regularly, which is used in the household savings estimates. The HLC discussed with NHB the feasibility of NHB providing the information on household deposits with housing boards. It was clarified that housing boards are under the control of the State Governments and there exists no consolidated database of household deposits with housing boards as such. The HLC was also told that although the housing boards have been authorised to collect household savings under the I.T. Act, it was gathered from a few of the leading housing boards (e.g. Tamil Nadu, Karnataka, Andhra Pradesh, Rajasthan and Himachal Pradesh) that no public deposits have been accepted by them till date. Though NHB has no direct regulatory powers to collect such information from the housing boards, NHB may request all the housing boards in the country to get the information even in the case of ‘Nil’ information for the benefit of the Committee. The NHB agreed to make an attempt on a ‘best effort’ basis.

Recommendations

VI.m As the discontinuation of the survey on “Growth of Deposits with Non-banking Companies” since 1995-96 puts constraints on a reliable estimation of household deposits with non-banking companies, the HLC recognises that there exists a large data gap in the estimation of household savings in the form of deposits with non-banking companies, because of non-inclusion of a wide array of unregistered companies. Hence, the HLC recommends that a census should be conducted on a regular basis, say once in five years, covering all companies incorporated with the Department of Company Affairs. The census should collect data on assets, liabilities, income and expenditure of NBFCs including the unregistered/exempted category of NBFCs. The census should be conducted by DSIM, RBI in collaboration with the Department of Company Affairs, Government of India, if necessary by outsourcing the same. It may be noted that this is in consonance with the recommendations made earlier by the Committee on Informal Financial Sector Statistics, 2001 (Chairman: P. Venkataramaiah).

(Action: RBI)

VI.n The HLC recommends that the data emerging from the census on the assets/ liabilities position of the NBFCs should be appropriately incorporated in the flow-of-funds accounts also.

(Action: RBI)

VI.o In respect of companies registered with the RBI, the RBI should provide the necessary details of deposits of public. Data for registered companies have to be firmed up with appropriate consolidation of data and follow up if response is weak. For the unregistered companies, sample studies will have to be undertaken and their results will have to be used after establishing proper blow-up factors.

(Action: RBI)

Life Insurance Funds

VI.123 It is recognised that contractual savings are a prominent instrument preferred for deploying savings by households. Life insurance is essentially contractual in nature. Household savings through insurance schemes is measured as changes in the life fund. In case of the Life Insurance Corporation, such savings strictly arising out of its life insurance business in India (net of transfer to Government), attributable to the policyholders is estimated. The life insurance business is primarily considered as a conduit of savings of the household sector and as such the savings of the LIC arising strictly out of its life insurance business (net of the transfer to Government) is regarded as accruing to the policy holder and, therefore, included in the estimates of savings of the household sector.

Instruments

VI.124 The instruments are life fund of LIC, life fund of private insurance companies, postal insurance and life annuity funds, central and State Government insurance fund.

Identification of Data Source

• Annual report of LIC,
• Special returns obtained from private insurance companies,
• Relevant data from CSO.

Methodology Prescribed

VI.125 Households ’ savings in the case of LIC is estimated as net accruing liability or an increase in the life funds of the LIC and bonus to policy holders excluding Government share in profit, capital gains and old claims. Similarly households’ savings in the case of private life insurance companies is estimated as an increase in the life funds. Savings in the form of postal life insurance, State Government’s life insurance and Central Government employees’ group insurance scheme is estimated as the excess of receipt over payments. Receipts include subscription realised, interest accrued, etc., while payments comprise payment of loans to policy holders, payment of insurance amounts after maturity or death and payment for other miscellaneous charges. The data on postal life insurance, central Government employees’ group insurance scheme and State insurance are available from the annual reports of the Department of Post and Telegraph and the budget documents of the Central/State Governments, respectively. However, the final figures for these items are made available by the CSO.

Methodology Used

The methodology used is same as that has been prescribed.

Validating the Data as well as Result

The data are on the basis of actual figures taken directly from the flow.

Deviations from Prescribed Procedures

There has been no deviation from the prescribed procedure.

Discussion in the HLC Meetings

VI.126 The household financial savings in life-funds of different schemes floated by insurance companies – both in the public and private sector – are emerging as an important instrument for household savings. It was felt that IRDA should be engaged in providing data relating to life funds operated both in the public and private sector.

VI.127 Presently, the information on life funds is collected from individual institutions. The HLC felt that the requirement of data is of very diverse kind for an accurate estimation and there is also a need to cross-validate the information for individual institution with information from apex bodies. In this regard, the following issues were discussed:

VI.128 Life fund is a contractual savings wherein a policyholder makes regular premium payments towards a policy that will mature when he or she reaches a certain age, or, alternatively, in the event of the policyholder’s death before maturity of the scheme. Such payments are also typically made over a long period of time, and also give rise to an accumulated fund that can be invested to generate further returns for the policyholder.

(i) As per the Chelliah Committee, in the case of LIC, this covers increase in the life fund and bonus to policy holders excluding Government share in profit, capital gains and old claims. Presently, these data are collected from individual institutions. The HLC felt that this estimate should be cross-validated with the estimate of the apexbody, i.e., IRDA, and accordingly there was discussion with IRDA on the format for collection of such data from Insurance companies containing three sub-heads, like (a) Life Funds; (b) Bonus to Policyholders and (c) Loans and Advances from the insurance sector.

(ii) According to the IRDA representative, given the nature of the non-life industry, it automatically was outside the purview of the analysis. Over recent years, there has been a shift in the manner in which insurance has been underwritten in the country and there has been a perceptible shift towards unit linked insurance products. Thus, households’ savings through these avenues too needs to be captured. As the manner of capturing the data is stock position, the funds under management (AUM) under both the traditional products and the ULIPs would need to be captured.

(iii) On the issue of the drawals through loans, it was indicated by the IRDA that the Policyholders ’ loans are captured as a separate line item. The net position of savings through the AUM less the policyholders’ loans (borrowings from the insurance companies) would, thus, reflect the net savings by the household sector in a given year. In such a scenario, the bonus component also gets built into the information on AUM. In addition, what is required is to get the data on the insurance institutions’ ‘loans to their staff.’

Recommendations

VI.p The HLC recommends that the IRDA should be involved in the household financial savings estimation by way of providing regular information on ‘life funds’ consolidated for all insurance institutions – both in the public and private sector in the format prescribed in this Report (A6.5.5).

(Action: IRDA)

VI.q To net out the household liabilities against assets in respect of the insurance sector, the IRDA may arrange to get the required data on the insurance institutions’ ‘loans to their staff’ in addition to the policy-holders.

(Action: IRDA)

VI.r The RBI may approach the IRDA to provide combined balance sheet for all insurance institutions – public and private - as per the flow-of-funds format. This will go a long way in the IRDA providing an alternative database not only for ‘life fund’ but also for flow-of-funds compilation at the RBI from time to time (A6.5.6).

(Action: IRDA)

Provident and Pension Funds

VI.129 Provident and Pension funds belong to the contractual category of instruments of financial savings often preferred by households for deployment of their savings.

Raj Committee

VI.130 The relevant observations/recommendations of the Raj Committee with respect to household savings in provident fund are given below.

“In the estimation of savings of the household sector in net financial assets, the loans and advances taken by the employees of the non-Government educational institutions are not taken into account; only their contribution to the provident funds is estimated. This tends to inflate savings in this medium. To obviate this and to arrive at net accretions to provident funds, it is necessary that the ratio of the deposits to loans and advances to the employees applicable for the Government institutions, should be applied to the provident funds of the educational institutions of the private sector.

In the case of some private institutions, the provident funds of the employees are deposited with the Post Offices in the form of small savings. Care should be taken to avoid any duplication in this regard.”

Instruments

VI.131 The instruments are: (i) Central Government Provident Fund, (ii) Public Provident Fund, (iii) State Government Provident Fund, (iv) Non-Government Provident Fund, and (v) Pension Funds and New Pension Scheme.

Identification of Data Source

• Data from CSO,

• Special returns and annual reports from non-Government provident funds like port trusts, dock labour boards etc,

• Special return from Department of Government and Budget Accounts (DGBA), and

• Special returns from Ministry of Finance for new pension scheme.

Methodology Prescribed

VI.132 Employees ’ savings in the form of provident fund (PF) is through PF schemes which are either contributory or non-contributory in nature. In the former scheme the employers and the employees both contribute to the PF. The total contributions minus withdrawals plus interest credited to PF account constitute the savings of the employees. Information on PF contributions relating to Central and State Governments and Public PF Scheme is culled out from the budget documents and also available on request from CSO. For want of requisite data on PF of employees of local authorities, the same is estimated as 6 per cent of wages and salaries paid to the employees of these authorities. Non-financial statutory corporations include Air India, Indian Airlines and other statutory corporations. Figures of PF relating to Air India and Indian Airlines are collected from these institutions. For other statutory corporations information is culled out from their respective annual reports. The PF contribution by the employers as available in the annual reports is doubled to include the employees’ contributions. For want of requisite data no adjustments are made for withdrawals, interest accrued and the employees’ contributions made in excess of the minimum prescribed limit. Information on PF of RBI and Commercial Banks is obtained from the RBI while for others the same is obtained from respective institutions or their annual reports. No data are available on PF contributions of the employees of various non­Government educational institutions. This has been estimated up to 1981-82 on the basis of wages and salaries to the employees in all recognised non-Government educational institutions. Employees’ contribution to the PF is estimated assuming that it forms 6.25 per cent of the wages and salaries. Since 1982-83, the PF of this category is included under Employees PF Scheme, 1952 and as such no separate estimates are prepared.

VI.133 Data relating to the PF contributed by the employers and the employees of various establishments covered under the Employees’ Provident Fund Act, 1952 are made available by the Central Provident Fund Commissioner (CPFC) through CSO. Data on PF contributions by the employees of Dock Labour Boards and Port Trusts are also collected from each Board/ Trust. The data on New Pension Scheme is obtained through a special return from MoF. This item has been included from 2004-05.

Methodology Used

VI.134 The methodology used is the same as has been prescribed except for two items, viz., non-Government provident fund and pension fund. These two items are estimated on the basis of past trends in growth.

Validating the Data as well as Result

VI.135 Most of the data items are obtained from actual data. For the remaining two items the data are reconciled at the time of preparation of flow of funds accounts.

Deviations from Prescribed Procedures

VI.136 The prescribed methodology is partially followed for this item, due to non­availability of data for two items, viz., non-Government provident fund and pension fund.

Discussions in the HLC Meetings

VI.137 The HLC on discussion with the CSO felt that the existing system may be continued.

Recommendations

VI.s As the present system of estimation of household savings in provident and pension funds is reasonably satisfactory, the HLC recommends that the existing system may continue.

(Action: CSO)

VI.t Since there is a need to capture household savings in the form of pension funds, the relevant apex body, namely, the Provident Fund Regulatory and Development Authority (PFRDA) should collect and disseminate such information on a regular basis.

(Action: PFRDA)

Vi.u The share of 6.0 per cent of wages and salaries paid to the employees of local authorities as their savings in PF is based on an old exercise conducted by the CSO. Therefore, the HLC recommends that the CSO should conduct a fresh study to update this ratio and keep it updated from time to time.

(Action: CSO)

Claims on Government

VI.138 Certain categories of ‘Claims on Government ’, namely ‘small savings’, are a major financial instrument for household savings in India.

Instruments

VI.139 The instruments are: (i) Small Savings, (ii) Investment in Government Securities, (iii) National Rural Development Bonds, (iv) National Deposit Scheme, (v) Capital Investment Bonds, (vi) Deposit scheme for retiring Government employees, (vii) relief bonds and savings bonds, (viii) compulsory deposits, and (ix) special bearer bonds.
Identification of Data Source

Identification of Data Source

• Budget documents and Division of Central Finance and Division of State and Local Finance of DEAP, RBI,

• Department of Government and Budget Accounts (DGBA), RBI, and

• Information from CSO.

Methodology Prescribed

VI.140 It includes investment in Government securities, small savings, bearer bonds, capital investment bonds, national rural development bonds, national deposit scheme, compulsory deposits and other such schemes brought out by the Government from time to time, less household ’s net borrowing from the Government. The household investment in Government securities is estimated on the basis of data on sale of total securities available in the budget document of the Central and State Governments using the proportion of securities purchased by the households to total securities obtained from the RBI survey on ‘Ownership of Central and State Government Securities’. In the absence of data, no corresponding estimates relating to local authorities can be prepared. The small savings comprise national savings certificates, national plan savings certificates, post office savings bank deposits, post office cash certificates and defence certificates, cumulative time deposits, national defence certificates, treasury savings deposit certificates and annuity deposits, Indira Vikas Patras and Rahat Patras etc. The bearer bonds, capital investment bonds, national rural development bonds, national deposit schemes and compulsory savings are taken from the budget documents of the Central Government. The households’ share is derived by adjusting the investment made in small savings out of provident fund contributions and deposit linked insurance funds. The investment in small savings out of provident fund contributions by the employees is worked out on the basis of information obtained from provident fund institutions on the pattern of investment of provident fund contributions. Information on capital investment bonds, national rural development bonds, national deposit schemes, etc., is obtained from the Department of Government and Budget Accounts (DGBA), RBI.

Methodology Used

VI.141 The RBI survey on ownership of Government securities has been discontinued since 1992. Therefore, historical ratios of 0.13 and 0.06 are taken as household shares in Government securities.

Validating the Data as well as Result

VI.142 In the absence of survey on ownership pattern of Government securities, there is no scope to validate the data on households’ investment on Government securities. The data on other items are validated by CSO’s estimates.

Deviations from Prescribed Procedures and Reasons Thereof

VI.143 Apart from the investment in Government securities, there is no deviation from the prescribed procedure. The reason for deviation in the former case is the absence of the survey of ownership pattern of Government securities. Therefore, the estimate for this item is being carried out with historical ratios.

Discussion in the HLC Meetings

VI.144 It was discussed in the HLC meetings that estimate of household investment in Government securities (both Centre and States) needs to be firmed up. Presently, the share of households in the total sale of Government securities is obtained from the RBI survey on ownership of Central and State Government Securities published in the December 1994 issue of the RBI Bulletin. In this connection, the HLC noted that the work on the last survey on ownership of Central and State Government Securities was initiated in 1991 and it was published as ‘Ownership of Government Rupee Securities: Outstanding as on March 31, 1990 and Preliminary Trends in 1991 and 1992’ in December 1994 issue of RBI Bulletin. Subsequently, the survey was discontinued and in its place a Table titled ‘Ownership Pattern of Central and State Government Securities’ has been published in the Handbook of Statistics on Indian Economy since 1999-2000. This table, however, covers only a part of the household sector such as ‘employees provident fund scheme’ and ‘coal mines provident fund scheme’. Therefore, the HLC felt that a revised format on the ownership pattern of Government securities comprising six categories, viz., 1) Government, 2) Banks, 3) Financial Sector Other than Banks, 4) Non-Service Sector (Corporates, Firms and Societies), 5) External Sector and 6) Household Sector, may be prepared and sent to IDMD, RBI to provide the data based on the revised format as at end-March of every year.

VI.145 It may be mentioned that at the instance of the HLC, IDMD, RBI has put in place a data generating mechanism to provide information along the above lines for estimation of household savings under Government securities.

Recommendation

VI.v The HLC recommends that an appropriate format as prescribed in this Report on the ownership pattern of Government securities (consolidated for all levels of Government), comprising six categories, viz., 1) Government, 2) Banks, 3) Financial Sector other than Banks, 4) Non-Service Sector (Corporates, Firms and Societies), 5) External Sector and 6) Household Sector, may be prepared by the RBI to provide the data based on the revised format (Annex A6.4) as at end-March every year, so that the residual status can be accorded to household (as defined in national accounts and FoF framework) in this instrument. The concerned Department within the RBI should ensure the providing of such information for estimation of household financial savings estimates.

(Action: RBI)

Shares, Debentures, Mutual Funds and Commercial Bonds

VI.146 Households route part of their financial surplus in the private corporate and the Government sectors, through the capital market by investing in shares and debentures of non-Government companies, cooperative institutions, Government companies, including banks and financial institutions such as UTIMF and other mutual funds.

Instruments
VI.147 The instruments are household investment in: (i) Banks, (ii) UTI mutual fund schemes (iii) mutual funds other than UTIMF, (iv) Private corporate business, and (v) Public sector bonds.

Raj Committee
VI.148 The relevant observations/recommendations of the Raj Committee with respect of household savings in the form of shares, debentures and bonds are as follows.

“In the estimation of household savings in the form of shares, care should be taken to ensure that the estimates in regard to bonus shares are not duplicated at any stage.”

Identification of Data Source

• NABARD’s publication ‘Statistical Statements Relating to Co-operative Movement in India’,

• Special return from Unit Trust of India Mutual Fund (UTIMF),

• Special returns from private mutual funds,

• RBI studies on joint stock companies,

• Paid-up capital figures from Ministry of Company Affairs,

• Information from Ministry of Finance,

• Information from Capital Market Division (CMD), DEAP, RBI, and

• SEBI data on Ownership distribution of mutual funds.

Methodology Prescribed

VI.149 The household investment in shares and debentures includes investment in shares and debentures issued by non-Government non-financial and financial companies and co ­operative banks and societies, bonds issued by public sector enterprises, units of UTI, other mutual funds and shares and debentures issued by financial corporations other than UTI. Information on household investment in shares and debentures of non-Government (financial and non-financial) companies is not available separately. This is derived as a residual by subtracting the investment of the public sector and the private corporate sector from the total investment in this instrument. First, a global estimates of shares and debentures is arrived at by blowing up the sample estimates of shares and debentures of public and private limited companies, based on the ratio of global paid up capital (PUC) available from the Ministry of Company Affairs (MCA) to sample PUC of public and private limited companies as obtained from the RBI studies on Finances of Public Limited Companies and Finances of Private Limited Companies. The estimates so arrived at are cross-checked with the primary market data obtained from the Capital Market Division (CMD) of RBI. For estimating household share in investment in shares and debentures, the results of the survey of Ownership of Capital Of Joint Stock Companies are made use of. This survey has been discontinued after 1995. However, the ratios from this survey are being used. Details of ownership of share capital for each type of co-operative institutions are available in the NABARD’s publication ‘Statistical Statements Relating to Co-operative Movement in India’. The share capital contributed by individuals and others is assumed as household investment in the co­operative share capital. The household investments in the bonds issued by Public Sector Undertakings are either obtained from the Ministry of Finance or from respective institutions. Similarly, the households’ investment in the units of UTIMF is estimated on the basis of information obtained from the UTIMF. Household investment in this case consists of adults / individuals, minors, Hindu Undivided families (HUF), trusts and dividend accrued to households on account of investment in the units of UTIMF. In the new series with base 1999-2000, income accrued to household but not paid on account of investment to the units of UTIMF has also been included in the estimates of households’ savings in the form of shares and debentures. Household investment in mutual funds other than UTIMF is estimated on the basis of information collected directly from various mutual funds. The data on household investment in shares and debentures of financial corporations are available from their balance sheets.

Methodology Used

VI.150 The total mobilisation in the primary market in the form of ‘shares and debentures and bonds’ by non- Government public limited companies are available from Capital Market Division (CMD), DEAP, RBI. The share of public non-financial companies is estimated as 83.09 per cent of the total shares. The share of private non-financial companies is estimated as 35.84 per cent of public non-financial companies. Of the total debentures figures 82.87 per cent is taken as debentures issued by non-financial public limited companies. The debentures of non-financial private limited companies are estimated at 1.14 per cent of public non-financial limited companies. For netting out sectoral contribution, 23.0 per cent for shares and 28.0 per cent for debentures are taken as households’ investment. These historical ratios are taken from Survey of Ownership of Capital of Joint Stock Companies as on end march 1995. This survey has been discontinued since then. The financial and investment companies’ data are estimated on the basis of the past trends in growth. The same applies for the public sector bonds. The investment in shares and debentures of commercial banks is estimated on the basis of past trends in growth. For investment in UTIMF units the figures are taken from UTIMF directly through special correspondence. For mutual funds (other than UTIMF) the total mobilisation data is available from SEBI. The information about household’s share is obtained from SEBI.

Validating the Data as well as Result

VI.151 Since the global paid-up capital figures are not available from MCA, there is no scope to validate the data on investment in non-financial companies and financial and investment companies. In the absence of NABARD publications, the data on co-operative sector cannot be validated.

Deviations from Prescribed Procedures

VI.152 In the absence of global paid-up capital figures from Ministry of Corporate Affairs, RBI has been using the IPO data maintained by it in its Capital Markets Division (CMD). This is a major deviation from the prescribed procedure. For the mutual funds (other than UTIMF), use is made of SEBI’s information. In the absence of NABARD publication, the data for co-operative sector is estimated.

Discussion in the HLC Meetings

VI.153 It is well recognised that household financial savings is increasingly invested in the form of shares, debentures, mutual funds and bonds. Presently, these instruments are worked out as per the results of the survey of ownership of capital of joint stock companies as at end-March 1995. In view of the fact that the estimation is based on results from a back-dated survey and that significant change has taken place in India’s capital market over the years, the HLC considered it appropriate to have an ownership survey so that the data on subscription for the recent years on each of the items namely, new issues of shares, debentures, mutual funds and commercial bonds can be obtained for the relevant ownership categories. It may be noted that apart from shares and debentures, mutual funds and commercial bonds have now become a fast emerging avenue of investment for retail investors in the recent years.

VI.154 In the absence of an up-to-date survey, the HLC thought it appropriate to have the involvement of SEBI in providing the information on a regular basis to the RBI for household financial savings estimation. The SEBI representative agreed that providing information on household ownership of shares, debentures, mutual funds and commercial bonds (each separately) would be feasible given the information base available with the top two registrars to issue, who among themselves have a market share of approximately 70 per cent to 80 per cent of public issues. The application form for the public issue of shares, debentures and bonds, contains the data fields for the following type of investors viz. (i) Individuals, (ii) Hindu Undivided Families, (iii) Bodies Corporate, (iv) Banks and Financial Institutions, (v) Mutual Funds, (vi) Non-Resident Indians, (vii) Insurance, Companies (viii) Venture Capital Funds (ix) Foreign Institutional Investor (x) Foreign Venture Capital Investor, (xi) Multilateral Agencies, (xii) Bilateral Development Financial Institution. It is learnt that the contents relating to the aforesaid data fields are captured by the registrars to the issue and they would be in a position to furnish the same.

VI.155 Among the above mentioned investor categories, category (i) and (ii) fall under the household sector. Category (iii) is the private corporate sector. Categories (iv) (v), (vii), (viii) are covered under banks and other financial institutions. Categories (vi), (ix), (x), (xi) and (xii) fall under the rest of world sector.

VI.156 It may be mentioned that the HLC’s interaction with the SEBI in this regard has been encouraging. SEBI has agreed to initiate the process of collecting the above information on a regular basis.

VI.157 Significant data lag with respect to household investment in shares and debentures from the co-operative banks and credit as well as non-credit societies persists due to considerable delay in the NABARD publication on ‘Statistical Statements relating to the Co­operative Movements in India’. It is felt that involvement of the NABARD is essential to provide projection of household sector’s holding in the above mentioned instruments, for the period for which actual data are not available.

Recommendations

VI.w The HLC recommends the involvement of the SEBI in providing information on a regular basis to the RBI on areas like annual subscription to public issues of shares, debentures, mutual funds and commercial bonds (each separately and consolidated for all depositories) as per the ownership categories (categories of subscribers are Qualified Institutional Investors, FIIs, NRIs, corporate, trusts and other categories) from which the RBI will extract household investment in shares, debentures, mutual funds and commercial bonds for household financial savings estimation (Annex A6.5.1). Registrars/Depositories/ AMFI need to be involved for regular data support. Henceforth, the SEBI would arrange to collect the required data from these entities, which are the primary sources of those data, by sensitising them and explaining the requirement of their data for estimation of household financial savings.

(Action: SEBI and RBI)

VI.x The SEBI may collect the consolidated balance sheet position for all registrars and broker houses and provide the data to the RBI as per the flow-of-funds format (Annex A6.5.2) to be forwarded by the RBI, so that the same can be incorporated in the FoF accounts.

(Action: SEBI)

VI.y The HLC recommends that the consolidated statements of assets/liabilities of all capital market institutions (namely, merchant bankers and registered brokers) should be compiled by the SEBI or have them compiled and forwarded to RBI for compilation of flow-of-funds account of the Indian economy.

(Action: SEBI and RBI)

VI.z The NABARD should be involved in providing projections of household investment in shares and debentures of co-operative banks and credit as well as non-credit societies to get around the problem of considerable time lag in the publication such as “Statistical Statements Relating to the Co-operative Movement in India”

.(Action: NABARD)

VI.aa. In view of the back dated Survey of Ownership of Capital of Joint Stock Companies as on end-March 1995, there is a need to update this at least once in five years. The SEBI may look into the modalities of conducting such a survey.

(Action: SEBI)

VI.bb Till a reliable flow of data from the SEBI on various parameters is available on a regular basis, the need for a blow-up factor for computing household investment in the shares, debentures, mutual funds and bonds of all companies is there. Accordingly, global paid-up capital series needs to be provided to the RBI by the Ministry of Corporate Affairs.

(Action: MCA, Government of India and RBI)

VI.cc The CSO should forward requisite formats to SEBI for estimation of savings, capital formation and GDP for the bodies under the regulatory purview of SEBI.

(Action: CSO and SEBI)

Trade Debt

Instruments

VI.158 The instruments are household’s trade debt (net) in (i) non-financial companies and (ii) financial and investment companies.

Identification of Data Source

VI.159 The sample studies on company finances by Company Finance Division, DSIM.

Methodology Prescribed

VI.160 The trade debt (net) has been estimated as change in trade dues in respect of sundry creditors minus change in loans and advances to sundry debtors. This information is available from the RBI studies on company finances published in RBI Bulletin from time to time. The sample results are blown up on the basis of coverage of sample PUC to total PUC of all companies.

Methodology Used

VI.161 Since the global paid-up capital figure is not available, it is not possible to apply the prescribed method. The past trend in growth rate is applied.

Validating the Data as well as Result

VI.162 Without the global paid-up capital figures from companies, it is not possible to validate the figures.

Deviations from Prescribed Procedures

VI.163 Due to non-availability of company finance studies for the year of compilation of savings, the past trends in growth rate is applied.

Discussion in the HLC Meetings

VI.164 The desirability of improvement in the statistics of household’s trade debt (net) in (i) non-financial companies and (ii) financial and investment companies was discussed at length in the HLC meetings. It was generally agreed that such improvement hinges on overall firming up of corporate sector statistics based on company finance studies of DSIM, RBI.

Recommendation

VI.dd Till the MCA 21 data is available and can be used for the purpose of financial savings estimation under trade debt, the existing methodology of computing trade debt may be continued. For estimating trade debt figure for all companies from the sample studies of the RBI, there is a need to have an appropriate blow-up factor. As recommended earlier, global paid-up capital series needs to be provided to the RBI by the Ministry of Corporate Affairs.

Loans and Advances by Households

VI.165 Household financial liabilities are a major subtraction item in the savings estimation, since it is used for acquiring a financial asset and/or physical asset. The household sector’s financial savings is estimated as the change in the sector’s financial assets net of their financial liabilities. The household sector receives loans from the banking sector, Government sector, financial corporations, and non-banking companies and insurance companies, electricity boards, and cooperative societies.

Raj Committee

VI.166 The relevant observations/recommendations of the Raj Committee with respect to the financial liabilities of the household sector are as under.

“In the present estimation of net claims on Government, the loans taken by the households from the local authorities are not taken into account due to non availability of data. However, with the acceptance of the recommendation of the working group suggesting that the State Statistical Bureaux should consolidate statistics of local bodies on a complete enumeration basis, it should be possible to obtain the data on loans also.

There is considerable time lag in the consolidation of data on loans and advances to households extended by the primary credit and non-credit co-operative societies. The RBI may be requested to expedite such consolidation so that appropriate data for the latest years are included in the CSO’s quick estimates.”

Chelliah Committee

The Chelliah Committee has made the following observations/recommendations concerning financial liabilities of the households.
“As the data on loans and advances of non-financial non-Government companies to its employees are not collected, arrangements may be made to collect such information ad include it in the financial liabilities while working out household sector’s savings in the form of financial assets.”

Instruments

VI.167 The instruments are households ’ loans and advances from (i) banks, consisting in turn of RBI, commercial banks and cooperative banks and credit societies, (ii) other financial institutions (consisting of financial corporations, insurance companies and financial and investment companies), (iii) cooperative non-credit societies, and (iv) Government (both central and State and SEBs).

Identification of Data Source

• Information from Department of Government and Budget Accounts (DGBA), RBI,

• Information of aggregate credit of commercial banks available from Division of Money and Banking (DMB), DEAP, RBI,

• BSR survey on credit by DSIM, RBI,

• ‘Statistical Statements Relating to the Co-operative Movement in India’ (NABARD), and

• Information through special returns and annual reports of various institutions.

Methodology Prescribed

(a)  Loans and Advances from Banks

VI.168 The bank credit to households comprises (i) bills purchased and discounted and (ii) loans/advances, cash credit and over-drafts to institutions like partnerships, proprietary concerns, joint families etc., and individuals. Loans and advances by the RBI to its staff are also included separately on the basis of information obtained from the Department of Government and Budget Accounts (DGBA), RBI. Data on the bank credit of the scheduled commercial banks including the Regional Rural Banks (excluding inter-bank advances), are obtained from statutory returns filed under 42(2) of the RBI Act, 1934. Data on total scheduled commercial banks credit are published in the RBI Bulletin. But institution-wise details are not available. In order to estimate the proportion of credit given to the households by commercial banks, the information collected by the RBI through their survey on ‘Organisation-wise Classification of Outstanding Credit of Scheduled Commercial Banks according to Occupation’ is utilised. The household proportion is estimated after excluding the food credit, which is assumed to be extended to the Government sector. The proportion is then applied to total bank credit (excluding food credit) to arrive at the estimates of bank credit to household.

(b) Loans and Advances from other financial institutions (OFI)

VI.169 The information on loans extended to the households, as available from the following sources, has been made use of:

(i) Loans extended by State Financial Corporations (SFCs) and State Industrial Development Corporations (SIDCs), etc. to staff members as well as to proprietor and partnership concerns, Hindu Undivided Families and Trusts;

(ii) Loans and advances granted by other Government financial corporations, namely, Industrial Financial Corporation of India (IFCI), Rural Electrification Corporation (REC), to their staff members and loans by HDFC, EXIM Bank, SIDBI and Housing and Urban Development Corporation (HUDCO) to individuals; and

(iii) Information on loans and advances by the chit fund and mutual savings fund companies is specially obtained from the RBI and loans and advances given by SEBs to their employees is obtained from their annual report. However, the figures for OFI are available from National Accounts Statistics, CSO.

(c) Loans and Advances from Co-operative Societies

VI.170 Data on loans and advances given to the households by co-operative societies (credit and non-credit) and banks available annually from the publication ‘Statistical Statements Relating to the Co-operative Movement in India’ (NABARD) is made use of.

(d)  Loans and Advances from Government

VI.171 The figures for loans and advances from Central Government are available from CSO. The figures for the State Governments are available from Division of State and Local Finances (DSLF), DEAP, RBI.

Methodology Used

VI.172 The data on loans and advances from banks are estimated on the basis of prescribed methodology. The loans and advances from OFI and Government are obtained from the CSO and DSLF, DEAP. The data for co-operative sector are estimated on the basis of the past trends in growth.

Validating the Data as well as Result

VI.173 The data are generally validated at the time of compilation of flow of funds.

Deviations from Prescribed Procedures

VI.174 Due to non-availability of Survey on Credit from DSIM for the year in question the latest available survey was used. Due to availability of data in respect of other instruments with a lag of one year, these are estimated.

Discussions in the HLC Meetings

VI.175 The HLC discussed at length the need for firming up the database on financial liabilities of the household sector to get a clear picture of household financial savings in any given year. The HLC reaffirmed that the detailed information on bank credit according to organisation and occupation (industrial activity) as available in BSR survey on outstanding credit of scheduled commercial banks may be used for this purpose. It may be mentioned that the occupational groups classified for the purpose are: (i) agriculture, (ii) medium and large industry, (iii) SSIs, (iv) wholesale trade – food procurement and other purposes, (v) other priority sectors, and (vi) all other sectors (including personal loans). Under these occupational groups, the household sector covers institutions like partnership, propriety concerns, joint families and individuals. Besides, the amount shown under credit limits up to Rs.2 lakh is shown as loans to the household sector.

VI.176 It was thought desirable to take the help of apex financial bodies namely, IRDA and NABARD in the estimation of loans and advances to household sector emanating from the insurance and co-operative sectors, respectively. The involvement of apex financial bodies is thought to be an urgent need in view of the requirement of dealing with a very diverse kind of database for estimating household financial savings.

VI.177 The issue of the present practice of subtraction of household financial liabilities from the gross financial assets to arrive at the household net financial savings was discussed in the HLC meetings on April 18, 2008 and November 20, 2008. It was agreed that while such a practice is conceptually appropriate, the problem with this practice, however, is that the household financial liabilities may have been incurred for the purpose of acquisition of not only the financial assets but also physical assets. In the context of the deployment of savings, the concept of ‘transferable savings’ is more relevant and it is, thus, useful to look at ‘gross financial savings’ and not ‘net financial savings’. Hence, it may be more appropriate to present the total household savings as:

[Gross Financial Assets + Physical Assets – Financial Liabilities]

Instead of the present practice of reporting as:[Net financial savings (i.e., Gross Financial Assets – Financial Liabilities + Physical Assets].

Recommendations

VI.ee The HLC recommends that as BSR data on bank credit is available with a lag of one year, proportion of the previous year may be applied wherever necessary. Care should be taken to ensure that the database on loans and advances by sectors/institutions/categories are comparable.

(Action: RBI)

VI.ff The HLC recommends incorporation of an estimate of loans to households and employees from the non-financial companies. This will entail requisite modifications/ inclusion in the questionnaire forwarded to companies by the RBI for conducting the Company Finance Studies.

(Action: RBI)

VI.gg The NABARD should provide a projection of loans and advances to the household sector from the co-operative credit and non-credit societies to get around the problem of considerable time lag in the publication such as ‘Statistical Statements Relating to the Co­operative Movement in India’.

(Action: NABARD)

VI.hh The IRDA should provide an estimate of loans and advances to households from the insurance companies – both public and private – by way of providing regular information to the RBI.

(Action: IRDA)

VI.ii While the present practice of subtraction of household financial liabilities from the gross financial assets to arrive at the household net financial savings is conceptually appropriate, the problem with this practice, however, is that the household financial liabilities may have been incurred for the purpose of acquisition of not only the financial assets but also physical assets. In the context of deployment of savings, what is important is the concept of ‘transferrable savings’. Hence, the CSO should present the total household savings as:

[Gross Financial Assets + Physical Assets – Financial Liabilities]

Instead of the present practice of reporting as:

[Net financial savings (i.e., Gross Financial Assets – Financial Liabilities + Physical Assets].

It would be appropriate for the CSO to present household financial savings by instruments in gross terms and not net of respective liabilities as is the present practice. In addition, for presenting household savings in form of financial assets, the estimate would consist of gross financial assets less financial liabilities.

(Action: RBI and CSO)

New Data Bases

VI.178 With reference to the ToR to suggest new data bases, if any, to be devised/built up for improving the reliability or checking validity of the estimates, it was agreed by the HLC that presently all the financial instruments available for deployment of household savings are taken into account while estimating the household financial savings. However, there is a need to continuously monitor the emergence of new instruments for incorporating them for in the savings estimates. Accordingly, new data base for such instruments need to be developed.

Recommendations

VI.jj On examination of the instruments of financial savings for households, it was agreed by the HLC that presently all the financial instruments available for deployment of household savings are taken into account while estimating the household financial savings. However, there is need to continuously monitor the emergence of new instruments for incorporating them in the savings estimates. Accordingly, new data base for such instruments need to be developed.

(Action: RBI)

VI.kk New databases should be devised/built-up for improving the reliability or checking validity of the estimates of household savings. The HLC examined the present data collection procedure adopted for estimating household financial savings and arrived at a consensus that in the existing scheme of things, although some apex bodies (namely, NABARD and NHB) are already involved, there is a possibility of involving the other apex bodies such as IRDA and SEBI for the purpose of building up an alternative database in respect of financial instruments under their regulatory purview and in respect of assets and liabilities of capital market institutions for the purpose of compiling the flow-of-funds accounts. Towards this pursuit, SEBI’s involvement was envisaged in respect of household investment in shares, debentures, mutual funds and commercial bonds, IRDA’s in respect of ‘life funds’ of insurance companies, NABARD’s in respect of deposit, credit and investment data pertaining to co-operative banks, credit and non-credit societies and NHB’s in respect of household deposits with the housing finance companies. The HLC recommends creation of a regular data supply mechanism from the apex bodies to the RBI for which specific forms will be supplied. Going forward, it is also envisioned that all the apex bodies are engaged in the ongoing review of the data on household financial savings from the following standpoints:

i. Identification of the database used;
ii. Methodology prescribed and in practice used;
iii. Validating the data as well as results; and
iv. Comment on changes required in the procedure.

(Action: NABARD, NHB, IRDA and SEBI)

VI.ll The HLC agreed that there is a need for field studies by different institutions like CSO and NSSO for developing new databases. The appropriate case studies and type studies to be undertaken may be funded by CSO and RBI.

(Action: CSO and NSSO)

Institutional Support for Statistics

VI.mm The issue of institutional support for generating reliable statistics as public good was discussed. The HLC was convinced that constraints by way of inadequate staffing of the desks involved with timely compilation of massive statistical exercises have led to difficulties in completion of such exercises in time. The HLC recommends that the National Statistical Commission (NSC) should identify such units in NSSO, CSO and the Research Departments in the RBI for strengthening them through appropriate staffing so that in future statistical exercises like income-expenditure survey, Company Finance Studies, BSR, FoF and similar exercises can be completed expeditiously at these organisations.

Chapter VII : Estimation of the Savings of the Private Corporate Sector

VII.1 The private corporate sector is classified into two segments, one comprising joint stock companies and the other comprising co-operative banks & societies and non-profit corporate institutions (in the previous Committees’ Reports, these were referred to as quasi-corporate bodies). Joint stock companies are further classified into non-Government non- financial companies and non-Government financial companies. While non­Government non-financial enterprises comprise public and private limited companies that also includes foreign direct investment companies, non-Government financial companies consist of all private sector commercial banks (scheduled and non-scheduled), other financial and investment companies engaged in hire-purchase, share transactions, investments in securities, housing finance, insurance, etc. Cooperative institutions include both cooperative banks (excluding State cooperative banks) and societies (credit and non-credit).

VII.2 Within private corporate sector, major share of savings is generated by the joint stock companies (Table 7.1). These contributed 91-96 per cent of private corporate sector’s savings during 2001-07. The lion’s share of about 90 per cent or more of the savings of joint stock companies, in turn, emanates from non-Government non-financial companies (NGNFCs). The non-Government non-financial corporate sector is dominated by public limited companies in terms of paid up capital (82 to 85 per cent) and by private limited companies in terms of number of companies (88 to 90 per cent). Therefore, the financial performance of private corporate sector can reasonably be gauged by focusing on large public limited companies.

VII.3 A main feature in case of private limited corporate sector is the existence of large number of small sized PUC companies and as such unlike in the case of public limited companies, it becomes difficult to secure a good coverage in terms of PUC by studying a limited number of companies. Non-banking financial companies also exhibit a similar limitation with the substantial growth in non-banking financial companies in recent years.

VII.4 It may be noted that the size of the private corporate sector in terms of its contribution to gross domestic savings of the country is rising in recent years (Table 7.2). The percentage share of this sector in gross domestic savings was 18 per cent in 1999-00 which declined to 14.4 per cent in 2001-02 and since then it followed an increasing trend to reach 22.4 per cent level in 2006-07.

VII.5 The improved productivity, lowering of the tax and interest rates as part of conducive policy environment, more efficient entrepreneurial activity and financial restructuring leading to low debt- equity ratio have collectively resulted in improving financial performance in the private corporate sector in recent years. Improved profitability as reflected in the progressive increase in retained profits resulted in more than doubling of the private corporate sector saving rates.

VII.6 The Annual Report of Ministry of Corporate Affairs (MCA) 2006-07 states that there are as many as 7,32,169 companies, limited by shares, which were at work in the country as on March 31, 2006. These comprised 7,30,817 non-Government companies and 1,352 Government companies. Out of 7,32,169 companies limited by shares at work, 80,141 companies were public limited and 6,52,028 were private limited companies. Besides, there were 3,605 companies with limited liability by guarantee and associations not for profit and 498 companies with unlimited liability as on March 31, 2006. As per the MCA, its MCA21 repository at present has reasonably good quality annual accounts data of around 1,80,000 companies.

System in Place for Estimating Corporate Savings

VII.7 Under an arrangement between the Reserve Bank (RBI) and the Central Statistical Organisation (CSO), RBI prepares the estimates of savings for the non-Government non-financial, financial and investment companies (excluding commercial banks) in India. Such estimates are compiled at all-industries level and also for specific industry-groups. These estimates are based on the results of the analysis of balance sheets and profit and loss accounts of select companies. The population estimates are obtained by blowing up the sample results on the basis of PUC of the sample companies to the PUC of all companies. Ministry of Corporate Affairs (MCA) supplies the information on PUC of all companies. These estimates are forwarded to CSO. The CSO estimates savings of other segments of the sector, viz., private commercial banks, insurance companies, co-operative institutions and non-profit corporate institutions.

VII.8 The estimates of savings of life insurance companies in private sector are prepared by analysing their annual reports. For non-life insurance companies, the profit and loss account as published in annual report of the Insurance Regulatory & Development Authority (IRDA) is used. There are at present 17 general insurance companies which have been granted registration for doing non-life insurance business in the country. Of these, 6 are in public sector and the rest in private sector. Of the 11 private sector companies, two have been granted license during 2007-08. The analysis presented in the IRDA report pertain to the private insurance companies like Bajaj Allianz, Tata AIG, Reliance, IFFCO TOKIO, ICICI Lombard, Cholamandalam, HDFC CHUBB and Royal Sundaram.

VII.9 In case of private commercial banks, data published in the RBI’s ‘Statistical Tables Relating to Banks in India’ are used.
VII.10 As regards co-operative institutions, the publication entitled ‘Statistical Statements Relating to Cooperative Movements in India’ published by NABARD is used.

VII.11 Few annual accounts are analysed for estimating savings of non-profit corporate institutions.

VII.12 As per recommendation of Advisory Committee on National Accounts, reinvested earnings of FDI companies are excluded from the savings of non-Government corporate enterprises and are included instead in the net factor income and net capital inflow from abroad. The CSO is making this adjustment in the corporate savings estimates since 1999-2000 onwards. No such adjustment to corporate savings estimate is made for earlier years due to non-availability of data on reinvested earnings of FDI companies. The addition of this adjusted figure with those estimated by the CSO and the RBI for other segments gives the corporate savings estimate for the private corporate sector.

Present Methodology of Estimating Corporate Savings

Estimation for Non-Government Companies

VII.13 Data: RBI is conducting regular annual studies on finances of joint stock companies in the private sector since 1950-51 which form the primary source for analysing the financial performance of the private corporate sector. These studies cover selected companies classified according to size of PUC and also of different industry groups. The studies provide comprehensive financial statistics on income and expenditure, assets and liabilities, sources and uses of funds, etc and relate to the financial year ended in March. The annual accounts for the financial year become available only after September. The estimates of savings and capital formation of the private corporate business sector are based on a larger set of companies compared to the number of companies covered in the published studies. At present, the RBI publishes the studies on the financial performance of the corporate sector in respect of (i) non-Government non-financial public limited companies, (ii) non­Government non-financial private limited companies, and (iii) non-Government financial and investment companies based on the audited annual accounts of the selected companies.

VII.14 The data items required for estimating savings from the studies include profits retained, non-operating surplus (+)/deficit (-) and depreciation provision from the income and expenditure Statement. The other data used in the estimation of savings and capital formation are the type-wise population PUC of all non-Government companies received from the MCA every year. Data on sample PUC are collected from balance sheets of selected companies to obtain the coverage of the studies in the population.
VII.15 Time frame: A company under Section 210 (3) (b) of the Companies Act 1956, is required to present within six months of the date of the closing of accounts, its balance sheet and profit and loss account at an annual general body meeting and simultaneously file copies of these with the concerned Registrar of Companies. Thus, the accounts of companies generally become available six months after the account closing date and as such the company finance studies of RBI would be available after some lag.

VII.16 Industrial Classification: The company finance studies in RBI adopt the National Industrial Classification of All Economic Activities. With the available information on sales of different products, a company is classified into an industry group based on the sales of that product which account for more than one half of its total sales or from which the company derives more than half of its main income. This classification is revised every year in the RBI studies. MCA, on the other hand, classifies a company on the basis of the main activity of the company indicated in its memorandum of association filed at the time of its registration, and as such changes that subsequently occur in the product diversification of the companies do not seem to have always been taken into account for all companies.

VII.17 Methodology: In order to estimate savings, the relevant data items of the sample companies are blown-up or inflated to cover entire population based on ratio of population paid-up capital to that of sample companies. The blow-up factors are obtained separately for three types of companies mentioned above. Depending upon the availability of global PUC data and the number of select companies, RBI estimates obtained for the latest financial year are generally ‘tentative’, the estimates for the preceding financial year are ‘revised’ and the estimates for the second-last preceding financial year are made ‘firm or final’. The tentative estimates are based on a small sample of 1,700 – 1,800 companies. The final estimates are, however, prepared with sufficiently larger sample of about 6,000 companies. The estimates are prepared separately for the ‘non-financial’ and the ‘financial’ sector.

VII.18 Data Gaps / Limitations: The selection of companies in the RBI studies is largely governed by the availability of balance sheets and is not drawn using any statistical sampling technique. While other parameters like sales, total net assets, etc., may have high correlation with various financial parameters, use of PUC as the characteristic for blow up procedure is chosen mainly on consideration of the data availability at population level and partly on account of its relative robustness. Another limitation is due to the fact that while the selection of companies for the studies excludes non-operating, defunct, and under-construction companies, the population PUC includes data on such companies. Thirdly, in the absence of the population PUC figures at the time of preparing tentative estimates, it is estimated based on average growth of population PUC in last 3 to 5 years. Whenever the population PUC figures are made available subsequently by MCA, the population estimates of savings and capital formation undergo revision because of change in the multiplier used for arriving at the population estimates.

Estimation for Private Commercial Banks

VII.19 The savings of private commercial banks is estimated as addition to their reserve funds. The transfers to reserve funds include net amount carried to reserves, depreciation provision, amount allocated for other special purposes and amount carried forward to next year’s account net of surplus/deficit of the previous year. Details of appropriation of profit of private scheduled commercial banks as transfer to (a) statutory reserve, (b) capital reserve, (c) investment reserve, proposed dividend, dividend tax and balance carried forward to balance sheet are available from the “Appropriation of Profit of Scheduled Commercial Banks” published in the RBI’s ‘Statistical Tables Relating to Banks in India’.

Estimation for Insurance Companies

VII.20 The savings in respect of life insurance companies in private sector represents change in life fund. This savings, however, is attributed to the household sector, since it is the households that make investment in the life funds of private insurance companies. The relevant information in this regard is collected from the annual reports of such companies. For non-life insurance companies, estimates are worked out by the CSO from the profit and loss accounts of private sector non-life insurers published in the IRDA Annual Report. Net Savings is estimated as sum of net transfer to reserves and net transfers to balance sheet.

Estimation for Co-operatives

VII.21 In case of co-operative societies, data on statutory reserves/funds, bad debt reserves and other reserves are available in the ‘Statistical Statements Relating to Co-operative Movement in India’, published by NABARD. The gross savings is taken as the increase in statutory funds and other reserves/funds. In the absence of data on time, the estimates of savings have been prepared on the basis of trends observed in the value added of banks and trade sector.

Estimation for Non-profit Corporate Institutions

VII.22 For estimating the gross savings of Non-profit Corporate Institutions, a few annual reports are analysed and their savings is estimated as the excess of current receipts over current expenditure. Also compiled for the base year were the GDP estimates for these institutions. A ratio of savings to GDP for these institutions was then compiled. Applying this ratio on the overall GDP of the Non-profit Corporate Institutions (which in turn is estimated by multiplying the total workforce in Non-profit Corporate Institutions and the average value added per worker in these institutions), the annual estimates of savings of Non-profit Corporate Institutions are compiled. Currently, the coverage under this segment relates to non-profit organisations serving to private corporate sector.

Some Issues, Recommendations of Past Committees and Present Status

VII.23 Concepts and present methodology used for estimating savings and capital formation of private corporate sector are largely based on the recommendations of the Raj Committee (1982) and subsequently by Chelliah Committee (1996) and National Statistical Commission (2001). Most of the methodological issues discussed in these reports are fairly common, as there has not been any substantive improvement in coverage of non-financial/financial companies in corporate savings estimate during last three decades. Some of the major issues are listed below:

VII.24 Statistical sampling scheme is not used for preparing population estimates -Population frame consisting of all companies is not available. Every year a large number of companies register with the Registrar of Companies (ROC) and get liquidated before they file their first annual report. As a corrective step, Chelliah Committee (1996) recommended that top 1500 companies in terms of net fixed assets or sales must be included in the sample. Estimates could then be made on the basis of the information of the top companies and for the remaining companies, estimates may be built up on a sample basis. DCA (now MCA) should make available this information. In case the information does not flow or takes time to come, it would be necessary to apply on the companies the legal provisions for timely supply of information. Registrars of various States are expected to send a copy of annual reports to RBI. Due to non-availability of sufficient and reliable information from MCA on top companies, this recommendation could not be implemented. However, efforts are made to include all large BSE listed companies in the RBI annual studies.

VII.25 Blow-up factor - DCA (now MCA) should be part of the system in rendering help in the estimation of savings and capital formation of the private corporate sector. It is necessary to strengthen the Statistical and Research Unit of MCA to enable estimation of population PUC. MCA has recently started MCA21 data reporting system.

VII.26 Substitution of PUC by other characteristics such as Gross Formation of Assets (GFA) or sales – It was not possible to implement this suggestion due to non-availability of information on these variables at a population level.

VII.27 Coverage: Undertaking census studies on all segments of private corporate sector on a quinquennial frequency - RBI should continue its company finance studies with greater regularity placing the studies on better footing through regular censuses covering companies of all sizes. The census studies have not been found possible by the RBI due to resource constraints.

VII.28 Resumption of the studies on the finances of foreign controlled rupee companies and branches of foreign companies – As stated earlier, following the recommendation of Advisory Committee to National Account Statistics, CSO has started excluding reinvestment earnings of FDI companies in India from the savings estimates for non-Government financial/non-financial companies made by RBI from the years 1999-2000 onwards.

VII.29 Data gaps in respect to Cooperative institutions – Noting the inordinate time lag in the publication of NABARD’s annual studies on Statistical Statements Relating to the Cooperative Movement in India, earlier committees emphasised the importance of speeding up the publications. However, the lags in these publications have increased to four to five years which have adversely affected the estimation.

VII.30 The issue of data gap and inordinate delay in publication of NABARD’s publications was discussed by the present HLC. It was felt that gaps in these data could not be filled without the direct involvement of NABARD. It was thought that apart from the speeding up of publications, it was imperative to involve NABARD in making provisional estimates of relevant data for the years for which the full-fledged publications were not available. In such cases, NABARD could gather information on major institutions and undertake ‘guess estimates’ for the rest.

Corporate savings – Existing and Suggested Definition