CONTENTS

December 30, 2005

Dr. Rakesh Mohan
Deputy Governor
Reserve Bank of India
Mumbai

Dear Sir,

Submission of the Report of the Working Group on Compilation of State Government Liabilities

We are pleased to submit herewith the Report of the Working Group on Compilation of State Government Liabilities.

(Narendra Jadhav)

(K. R. Lakhanpal)                                             (Yogesh Khanna)                                                           (Kavita Gupta)

(V. S. Senthil)                                                      (K.G. Mahalingam)                                                         (M. J. Joseph)

   Member                                                                   Member                                                                     Member

(R. H. Dholakia)

ACKNOWLEDGEMENT

The Working Group gratefully acknowledges the invaluable contributions and support of erstwhile members, Shri S.C. Garg, then Joint Secretary (PF-I), Ministry of Finance, Government of India and Shri Rahul Sarin, Principal Finance Secretary, Government of Jharkhand in the discussions of the Group and in the preparation of the Report.

The Working Group places on record its appreciation for the invaluable contributions of the special invitees, Shri Prabal Sen, Chief General Manager-in-Charge, Department of Government and Bank Accounts of the Reserve Bank of India (RBI) and Shri B. Mahapatra, then Chief General Manager-in-Charge, Internal Debt Management Department (IDMD), RBI, in the deliberations of the Group and in the preparation of the Report.

The Working Group acknowledges the outstanding professional contributions of Dr. R.K. Pattnaik, Adviser, Department of Economic Analysis and Policy (DEAP), and Shri Somnath Chatterjee, Director, IDMD, RBI, who brought to bear their experience and expertise in the fiscal area, in the discussions of the Group and in the preparation of the Report.

The Working Group would like to place on record its appreciation for the arrangements made for its meetings and the hospitality extended by Shri Ramesh Chander, then Regional Director, RBI, New Delhi. The Working Group would also like to thank Shri Rajan Goyal, Director and Smt. Sangita Misra, then Research Officer, DEAP, RBI, New Delhi for logistical support in organizing the meetings of the Group at New Delhi.

The Working Group appreciates the support of Shri P.D. Jeromi, Assistant Adviser, DEAP, Kochi and Shri Jai Chander, Research Officer, Division of Central Finances, DEAP, of the RBI, in the preparation of the Report. Helpful suggestions and analytical inputs provided by Shri A.K. Mitra, Assistant Adviser, IDMD are gratefully acknowledged. The contributions of the officers and staff of the Division of State and Local Finances, DEAP, RBI, Central Office, in providing data and logistical support, and in particular, of Shri Rajmal and Shri J.B. Singh, Research Officers, are also gratefully acknowledged.

• At present, there appears to be no unanimity about the exact level, composition, and the methodology for compiling the liabilities of State Governments in India. In fact, there is a great deal of ad hocism in the compilation of debt statistics.

• The budget documents of the State Governments do not provide data on their outstanding liabilities. Such data are, however, provided in the ‘Finance Accounts’ of the State Governments, but these reports are not available in respect of all State Governments on a timely basis. The Combined Finance and Revenue Accounts of the Central and State Governments are also released with a great deal of time lag. Moreover, some issues relating to coverage of liabilities need to be addressed.

• Debt and liabilities should be considered as synonymous. Accordingly, all borrowings which are repayable and/or on which interest accrues, are to be considered as debt.

• Compilation of data on debt should be consistent with the Gross Fiscal Deficit (GFD).

• Total budgetary liabilities of State Governments may be decomposed in four categories viz. (i) Public Debt; (ii) Ways and Means Advances and Overdrafts from the RBI or any other bank; (iii) Public Accounts; and (iv) Contingency Fund. This treatment would be in conformity with the international best practices.

• Public Debt would include open market borrowings, borrowings from banks and financial institutions, Special Securities issued to the National Small Savings Fund, Bonds/Debentures which are issued by the State Government and loans from the Central Government.

• Public Accounts would include State Provident Funds, Small Savings, Insurance and Pension Funds, Reserve Funds and Deposits and Advances.

• The following items (under the Capital Account of the State Government budget) should not be included in debt or GFD: (i) Remittances; (ii) Suspense and Miscellaneous; (iii) Appropriation to Contingency Fund (since all these are largely adjusting heads which eventually get cleared within or across accounts); and (iv) Decrease in Cash Balances (since it does not induce additional liability). These may, however, be shown as memo items.

• The implicit liabilities of State Governments including guarantees, off-Budget Borrowings, Pension Fund and State Public Sector Liabilities should be excluded from the definition of State Government budgetary liabilities or debt.

• The information on the State Government liabilities may be published in the State Government budget documents under the following Statements: (1) Budgetary Liabilities of State Government (outstanding at end-March) and their break-up; (2) Details of Guarantees given by the State Governments (GASAB Format); (3) Assessed Fiscal Risk of State Government Guarantees; (4) Off-Budget Borrowings of State Government; (5) Liabilities of State Government Public Sector Undertakings; and (6) Other Implicit Liabilities of the State Government (including pension liabilities). An additional Statement (7) on Subsidies provided by the State Government may also be published.

• Four Annexes to Statement 1 may also be published which would provide details (outstanding amount, rate of interest, date of maturity, etc) of open market borrowings, loans from Centre, borrowings from banks/financial institutions and special securities issued to NSSF, respectively.

• The Group recognizes that there are widely differing views on the inclusion of various implicit liabilities in order to obtain the aggregate liabilities of the State Governments, on which a consensus may emerge over a period time. Priority may, however, need to be accorded to the task of collating and publishing data on the different parameters, as set out in Statements 1 to 7, which should be available with the State Governments.

• Statements (1) alongwith the Annexes and (2) above may be published in the budget documents of the State Governments with effect from the fiscal year 2006-07. The remaining Statements should be published by the State Governments as soon as possible. In case it is not possible to bring out all the remaining Statements at the same time, a graduated approach for publication could be adopted.

• The Government of India, RBI and other institutions could help the States in creating necessary capacity, systems and processes and acquiring technology to compile data on liabilities.

• The following arrangements may be made for compiling the data: (a) RBI will provide the data on outstanding market borrowings to the State Governments; (b) The Central Government may provide the details regarding the loans from Centre to the State Governments and Special securities issued by the States to NSSF; and (c) The data on borrowings from banks and financial institutions and any other such transactions, may be provided by the State Governments.

• For ensuring the consistency in the data, it is desirable that a single agency compiles and disseminates the information on outstanding liabilities of all the States. Although the State Governments are the most reliable sources of such information, the task cannot be fully entrusted to them unless it becomes an obligatory part of the State Budget documents. RBI can then act as a single agency putting estimates of liabilities of all States together in a single publication, as it does for the State Budgets.

• Till such time that the State Governments are not in a position to publish the requisite data on their outstanding liabilities in their budget documents, all the above data may be furnished by the concerned institutions to the RBI, as a transitional measure, to enable consolidation and publication.

• The RBI should compile the (latest available) ‘accounts’ (un-audited) and (revised and budget) estimates of liabilities of all the State Governments and publish the same in its regular annual publication on State Budgets, from the viewpoint of data dissemination and to facilitate academic and policy research. The CAG may provide the latest available audited data on liabilities and the same could also be reported by the RBI in its annual study on State budgets, alongwith the ‘accounts’ (un-audited), revised estimates and budget estimates of more recent years.

Report of the Working Group on Compilation of State Government Liabilities

1.1 Recognising the various problems associated with the database of the State Government liabilities, the 14^th Conference of State Finance Secretaries held at RBI, Mumbai in August 2004, considered and approved the proposal of constituting a working group on the methodology and compilation of data on the liabilities of the State Governments. As a follow up to this proposal, a Working Group was constituted with the following members:

1The Report reflects the views of the members and not necessarily of the institutions to which they belong.
2Shri Rahul Sarin was succeeded by Shri M. Singh in November 2005.
3Shri S.C. Garg, predecessor of Shri V.S. Senthil, was associated with the Working Group till May 2005.

1.2 In the first meeting of the Group, held on November 20, 2004, in Mumbai, the terms of reference were set as under:

1.3 The remainder of the Report is organised as follows. Section II while discussing the present status regarding the compilation of the liabilities of the State Governments also highlights the limitations of the existing estimates. Section III discusses the international practices and definitions of the pubic debt. Section IV sets out the analytical framework for defining the outstanding public debt. On the basis of the foregoing discussion, Section V offers recommendations of the Group for compiling the data on the liabilities of State Governments.

2.1 Transparency, reliability and consistency in the fiscal data are crucial for successful management of government finances. The time series data on the liabilities of government is of particular importance as it affects future state of affairs of government finances. At the policy making level as well as at the academic level, a need to streamline the reporting of data on State Government liabilities is increasingly felt. The number of studies making an assessment of the financial health, and the mixed results generated by them, highlight, among other things, the need for having reliable and credible statistics on public debt which are comparable across States, countries and time period. In this regard, the methodology for compiling the data on debt/liabilities of State Governments assumes considerable importance.

2.2 At present, there appears to be no unanimity about the exact level, composition, and the methodology for compiling the liabilities of State Governments in India. The budget documents of the State Governments do not provide data on their outstanding liabilities. Such data are, however, provided in the ‘Finance Accounts’ of the State Governments, but these reports are not available in respect of all State Governments on a timely basis.

2.3 A survey of the existing important publications presenting data on fiscal liabilities shows substantial differences in definition and coverage of liabilities among these sources. This essentially reflects that debt statistics are compiled in an ad hoc manner. This issue needs urgent attention and action.

2.4 The State Governments in their budget documents currently do not publish the data on outstanding liabilities. The accounting information and other details relating to public debt and other liabilities of the State Governments, as published in the Finance Accounts, are compiled and audited by respective Accountants General under authority of the CAG. Finance Accounts are generally made available in public domain with a time lag of one year. The definition and coverage adopted by CAG is clear from Statement No. 4 of Finance Accounts, which contains summary of totals of public debt and other liabilities of the State Governments, as derived from Statements No. 16 and 17, classified under the category of internal debt, loans and advances from the Central Government, small savings and provident funds and obligations like reserve funds and deposits, both interest and non-interest bearing. On the other hand, Statement No.16 contains details of receipts, disbursements and balances under heads of account relating to Debt, Contingency Fund and Public Account. Similarly, Statement No.17 relates to detailed Statement of Debt and Other Interest Bearing Obligations of Government. A close scrutiny of Statements No. 16 and 17 would reveal that detailed information and statistics about public debt and other liabilities of the State Governments are made available, in parts, in both the Statements and the reader is expected to assimilate complete position after studying them. The CAG also publishes consolidated details of public debt and other liabilities of all State Governments as part of "Combined Finance and Revenue Accounts of Union and State Governments in India". At present, this document is being published after a time lag of two to three years, which is expected to be brought down in line with Finance Accounts. A Staff Paper published by CAG on the subject of ‘Fiscal Liabilities (Union and States) – Growth and Sustainability’ has also used same definition as outlined above. As far as Combined Finance Accounts are concerned, it is, however, observed that the latest available issue pertains to the year 1999-2000 and does not cover liabilities under Small Savings/National Small Savings Fund (NSSF). Since the NSSF was constituted in 1999-2000, it is expected that future issues of the Combined Finance Accounts would incorporate the same under the liabilities of the State Governments. Furthermore, there are negative entries in respect of the outstanding liabilities under negotiated loans of some of the State Governments, as reported in the Combined Finance Accounts, which need to be elucidated.

2.5 In order to provide the consolidated position of the State Government liabilities, RBI publishes the estimated data based on the State Government budget documents on a yearly basis, but again with a lag of approximately a year largely due to the delay in preparation of budget documents by the State Governments.

2.6 At present, data on outstanding liabilities of State Governments for various years as published by the RBI are compiled from (i) the outstanding debt (stock) data (under various categories) reported in the CAG’s "Combined Finance and Revenue Accounts of the Union and State Governments in India" for the year 1986-87 and (ii) ‘flows’ data (net of repayments) on the corresponding items reported in the budget documents of the State Governments for the subsequent years. The estimates of outstanding liabilities are obtained by progressively adding the ‘flow’ data for each year to the stock data for 1986-87.

2.7 The items that are included in the liabilities of State Governments are (i) Internal Debt (including Special Securities issued to the National Small Savings Fund (NSSF) and WMA from the Reserve Bank); (ii) Loans from the Central Government; (iii) Small Savings, Provident Funds, etc (including State Provident Funds, Insurance and Pension Funds, Trusts and Endowments, and Small Savings). The following items are not included: Reserve Funds, Deposits and Advances, Suspense and Miscellaneous, Contingency Fund and Remittances. The (item-wise) consolidated and State-wise data on liabilities, as published in the latest Study of State Government Budgets, 2004-05, are shown in Appendix 1 and 2, respectively.

2.8 As stated in the Report of the Twelfth Finance Commission (2004), "Previous finance commissions had followed the practice of excluding the short-term components of debt viz., ways and means advances and reserve funds and deposits." Thus, the broad definition followed by Finance Commissions includes the same heads as that by the Office of CAG. Its narrow definition differs only in respect that it excludes WMA and Overdrafts from RBI.

2.9 There are several studies by individual scholars attempting to define and measure debt in India.  Buiter and Patel (1992) construct debt series aggregated across Centre (including external liabilities), States and Central Public Sector Undertakings (PSUs) after excluding monetized deficit from the total.  Rajaraman & Mukhopadhyay (1999) construct an aggregated debt series (Centre and States) for the period 1951 to 1997. While they follow the Buiter and Patel study in terms of excluding monetization (although their definition of monetization differs from Buiter & Patel), external and PSU liabilities are, however, excluded by them, because their objective was to obtain domestic non-monetized deficit. The items included by them, relevant for the States, are: Internal debt (excluding WMA and Overdrafts from RBI or other banks), Provident Funds and Reserves and Deposits. Loans from the Centre, being inter-government loans, are excluded from State Government debt because the objective was to obtain an aggregate series. This study, however, does not attempt to link the non-monetized component of Gross Fiscal Deficit to the increment in debt in any given year. Gurumurthy (2002) breaks Gross Fiscal Deficit (GFD) in terms of the following flow components: Loans from the Central Government (net), Market Borrowings (net), Loans from Financial Institutions, Provident funds, Reserve funds, Deposits and Advances, etc. However, while computing the stock of debt, he includes only Internal Debt including Market Loans and WMA, loans from banks and other financial institutions, loans from the Central government, Provident funds and Small Saving in his definition of State government debt and excludes Reserve funds and Deposits & Advances. Thus, in this study also, the linkage between GFD in a given year and increment in liability for that year is not formally established. Rangarajan and Srivastava (2003), while recognizing this linkage, compute ‘Derived Fiscal Deficit’ (DeFD) as the difference between the total liabilities of the Central government for any two consecutive years. They also note the growing disparity between DeFD and reported GFD in the series constructed by them. They follow the same definition as followed in Finance Accounts for the Centre by the CAG Report on Union Government (No 1 of 2003, p 109) for calculating the fiscal deficit. This includes External debt (at historical rates), Internal debt, Small savings, Provident funds etc. and Reserve funds and Deposits.

2.10 Dholakia (2003) and Dholakia and Karan (2004, 2005) are some of the comprehensive individual research studies that have attempted to define and measure State Government liabilities in recent years. The second study is more comprehensive and covers 25 States over the period 1989-2003. Compared to the CAG study, Dholakia and Karan (2004, 2005) exclude WMA and OD from RBI, but include Suspense & Miscellaneous and Contingency Fund in their definition of liabilities. Even more recently, Rangarajan and Srivastava (2005) have published data on the combined (Centre and States) debt-GDP ratio over the period 1951-52 to 2002-03. According to the authors, "The debt-GDP ratio has risen from 61.7 per cent in 1990-91 to about 76 per cent in 2002-03, when external debt is considered at historical exchange rates and liabilities of states on account of reserve funds and deposits are not included. When these are included and external debt is evaluated at current exchange rates an upward adjustment of about 9 percentage points of GDP is called for, consisting of 3 and 6 percentage points for the two factors, respectively, taking government liabilities to about 85 per cent of GDP at the end of 2002-03."

2.11 Figure 1 shows the components included in the definition of liabilities by the above publications.

2.12 When looked at the component-wise details, the comparison is not without complications. This is apparently because of different data sources. CAG, which prepares the Finance Accounts of the State Governments and that of the Centre, has direct access to government data. RBI draws on the receipts and expenditure statements in State Government budget documents for its publications on State finances. However, a closer look makes it clear that the difference arises on account of the classification of sub-heads. Thus, the broad-based head of ‘Public Debt’ consisting of (i) Internal Liabilities; and (ii) Loans & Advances from the Centre, provides exactly the same estimates by RBI and CAG when considered together, but gives substantially different estimates for its two components when considered separately. In its estimates for accumulation in a given year, CAG considers WMA & OD as a part of Internal Debt and provides a figure net of repayment (Statement 15, Finance Accounts). RBI, on the other hand, considers Internal Debt gross of discharge of Internal Debt and excludes WMA & OD from it (Annexure III, Handbook of Statistics on State Government Finances, 2004). Similarly, CAG considers Loans & Advances from the Centre net of repayments, whereas RBI considers it gross. When we make these adjustments, the ‘Public Debt’ (Net) as well as the two sub-components turn out to be identical before 1999-2000 for most of the States. For the years 1999-2000 and thereafter, while the aggregate estimates of ‘Public Debt’ are identical, individual components sharply differ for almost all the States. This is because of the creation of National Small Savings Fund (NSSF) on April 1, 1999, consequent of which the small savings receipts from the Centre were shifted from the head Loans & Advances from the Centre to the head Internal Debt.

3.2 In its Manual on Government Finance Statistics (GFS), 1986, IMF describes government debt as ‘a stock of liabilities with different time dimensions accumulated by government operations in the past and scheduled to be extinguished by the government operations in futures’. The Manual has listed out the characteristics pertaining to government debt which are important for the debt management. These are: (1) the schedule of maturing debt indicates when domestic or foreign funds will be needed for its repayments. (2) Because the stock of outstanding is not homogeneous, plans for sale of additional debt require additional information: on whether the existing debt is negotiable or nonnegotiable, in bearer form or registered, placed by compulsion or voluntary, sold by auction or by subscription, with interest payable by coupon or through discount, indexed or unindexed, redeemable at maturity only or at penalty before maturity, and finally, on what interest it carries and by whom the debt is held. (3) Plans for servicing outstanding debt through interest payment must also be based on the size and coupon rate of issues being serviced. It may be noted that the information currently being provided by many countries is limited to only a few of the above characteristics.

3.3 The IMF’s GFS Manual 1986 specifies that only the recognised direct financial obligations of the government which are serviced through interest payments and/or redemption should be included in debt. Government guarantees of the debts of others should be excluded, along with any other contingent liabilities, unless and until the government is called upon to take over and service that debt. The contingent or actuarial liabilities of government insurance schemes or social security systems are also excluded from the totals of recognised, fixed-term direct government debt. The monetary authorities’ obligations – for currency issues, for example are excluded from government debt, as monetary authorities’ functions are considered to be not a part of government but of financial institutions sector. The floating debt of unpaid government obligations is also excluded from government debt figures unless recognised and converted into fixed-term contractual obligations. Floating debt statistics may usefully supplement government debt data, however, and are covered by memorandum items in expenditure.

3.4 On measurement side the Manual specifies that the basis for measurement of outstanding debt should be the amount that will have to be repaid by government and not the amount of money received when the debt arose. The amount to be repaid may differ from the actual amount of government borrowing receipts by discounts or premia.

3.5 The International Accounting Standards Committee (IASC) Framework for the Preparation and Presentation of Financial Statements defines Liability as:".. a present obligation of the enterprise arising from past events, the settlement of which is expected to result in an outflow from the enterprise of resources embodying economic benefits."

3.6 Based on the IASC definition, the International Federation of Accountants (IFAC) identifies certain fundamental characteristics of liabilities:

• the existence of a present obligation arising from past events. That is, a transaction or other event in the past has given rise to a duty or responsibility to a third party, which has not yet been satisfied; and

• that liabilities have adverse financial consequences for the reporting entity. That is, the entity has to incur additional liabilities, or dispose of cash or other assets to settle the obligation.

3.7 IFAC finds these characteristics present in the definition of Liability used by most countries.

3.8 The Public Debt Committee of the International Organisation of Supreme Auditing Institution (INTOSAI) published the guidelines and other information for by Supreme Audit Institutions (SAIs) to encourage the proper reporting and sound management of the public debt.

3.9 The Committee in its guidelines on definition of public debt has specified that the preparers of reports on public debt need to ensure that any definitions used are precise, clear, consistent, appropriate and comprehensive. The elements of liabilities and other commitments incurred by public bodies or by corporations sponsored by such bodies can be shown as lying on a spectrum that extends from direct borrowing through a range of other financial obligations from trade accounts payable to various contingencies and commitments. These commitments may or may not be recorded as liabilities in financial statements. However, they may have a significant effect on future borrowing needs and, therefore, future demands on the country's economic resources.

3.10 The public debt is characterised as an obligation on a public body to make payments to third party at some future date, subject to the occurrence of one or more uncertain future events if it constitutes a contingent liability. Measuring public debt means to assign a value, in monetary terms to the total amount due.

3.11 The Public Debt Committee of INTOSAI underscores the importance of the regular disclosure of country's public debt as it will reveal whether debt levels have been kept within the country's ability to support them and can help ensure that potential problems visible. The disclosure requirements in the United Kingdom ensure that the public sector makes following disclosure in connection with public debt. The explanation of risk profile and risk management policies should give information on the nature and purpose for which financial instruments are held/issued, interest rate policies, accounting policies for derivatives, and hedging policies. The numerical disclosures requirements, inter alia, include interest rate profile, debt maturity analysis and liquidity, currency risk, and fair values of debt instruments.

3.12 Public Debt Committee of INTOSAI recommends that the contingent liabilities may be included in the total liabilities of the government by assigning some value based on estimates. The degree of uncertainty in respect of these items of debt implies that different items need to be assigned values on the basis of different estimates. However, IMF’s GFS Manual (1986) recommends exclusion of such contingent liabilities from public debt. There is, therefore, a difference in the practice followed across countries for disclosing the extent of contingent debt as opposed to actual debt. In Canada, for instance, a note to the statements concerning public debt is appended. This note is audited, but the value of individual contingent liabilities is not included within the overall public debt totals. In the United Kingdom, contingent liabilities are also excluded from public debt totals but, where possible, individual liabilities are identified and disclosed. In Portugal, public debt (actual and some contingent amounts) is disclosed in financial statements and is audited. However, only contingent liabilities relating to guarantees are disclosed and with less information than for actual debt.

3.13 Financial information about public debt may be reported in a wide variety of documents, for instance, budgets, central bank bulletins and a variety of other reports to legislatures. Several countries disclose their planned and actual public debt periodically as part of the ongoing budget decision-making and accountability process.

4.1 From the above discussion, it is clear that the concepts of public debt and liabilities need clearer definition to avoid confusion and erroneous use. For policy purposes, we can consider them as synonymous. The concept must have grounding in the analytical framework for its effective use in policy making. The most important use of debt or liabilities of a State Government is for assessing its fiscal sustainability. All the available alternative measures of fiscal sustainability require correct measurement of debt and liability of a State Government. The famous Domar Equation and its derivation from first principles provide the basic analytical framework where the difference between growth of income and effective interest rate on the State’s debt plays the determining role.

4.2 Now total interest payment is the same irrespective of the definition of debt we choose to follow. However, the weighted average interest rate, which is also calculated as the ratio of total interest payment to debt, will depend on the value of debt. Broader definitions of debt would be expected to yield smaller average interest rates, thus raising the differential between income growth and interest rate. A definitional change might reverse the conclusions about sustainability if the differential changes sign (see, Dholakia, 2003 for an illustration with Gujarat data).

4.3 Another critical element in the analytical framework is the relationship between fiscal deficit and change in debt. Fiscal deficit in any given year has to be financed either by additional borrowings or by creating new liquidity. Since States’ access to money finance is severely limited in India, there is a strong link between fiscal deficits and the stock of debt for a State. Thus, the analytically correct definition of debt would be one which satisfies the following:

D_t+1 = D_t + (Debt Increase)_t+1

D_t+1 = D_t + DFD_t+1 ; Where DFD: Deficit financed by borrowings

Now, GFD_t+1 = DFD_t+1 + MFD_t+1; Where MFD: Monetized Deficit

Therefore, D_t+1 = D_t + (GFD- MFD)_t+1

4.4 The monetised deficit is that part of deficit which is financed through WMA & OD and reduction in cash balances of the State government. The remaining part of the deficit could, thus, be referred to as deficit financed through borrowings (DFD).

4.5 None of the compilations of State liabilities discussed in section II above satisfy this basic criterion of analytically sound definition of debt. Moreover, since the correct measurement of debt critically hinges on fiscal deficit, the latter should be measured correctly.

4.6 RBI defines fiscal deficit as per the well-accepted prevalent concept in the country:–

GFD = Total Expenditure (TE) (including repayment of debt) – Revenue Receipts (RR) – Non-Debt Creating Capital Receipts (NDCR) – Recovery of Loans and Advances – Repayment of debt

4.7 The above definition of the gross fiscal deficit is followed with some variation by various agencies including State Governments. The main source of the variance of actual calculation of GFD from the above definition has its roots in the classification of some of the capital receipts items. Thereby, the practice followed by various agencies presently is normally characterised by the following discrepancies:

5The Introductory on page V of Finance Accounts 2001-02 for Kerala (common to all State Finance Accounts) says 'In the Public Account, the transactions relating to 'Debt' (other than those included in Part I), 'Deposits', 'Advances', 'Remittances' and ' Suspense' are recorded. The transactions under 'Debt', 'Deposits' and 'Advances', in this part are those in respect of which Government incurs a liability to repay the moneys received or has a claim to recover the amounts paid, together with the repayments of the former ('Debt', and 'Deposits') and the recoveries of the latter ('Advances'). The transactions relating to 'Remittances' and 'Suspense' in this part embrace merely adjusting heads under which appear such transactions as remittances of cash between treasuries and currency chests, transfers between different accounting circles, etc. The initial debts or credits to these heads will be cleared eventually by corresponding receipts or payments either within the same circle of account or in another account circle.'

4.8 Fiscal deficit should, therefore, be computed as follows:

GFD = Total Expenditure (TE) including Repayment of Debt + Civil Advances (net) [CAG data] – Revenue Receipts (RR) – MCR – Inter-State Settlement – Recovery of Loans and Advances – Repayment of Debt.

4.9 This GFD will be financed by additional borrowings and monetization. As explained above, GFD = DFD + MFD. Since the extent of monetization is reasonably limited in case of State Governments, we replace the term MFD and call the items included therein as Memo Items. The following items would fall under this head:

4.10 On the other hand, DFD would include those heads in respect of which a State Government incurs a liability to repay the money received. Thus, within our framework, there is no difference between Debt and Liabilities. The following heads are included in DFD, grouped under the Consolidated Fund, Public Accounts and Contingency Fund:

6Please see para 4.11.

4. Contingency Fund.

4.11 The treatment to Ways & Means Advances and Overdrafts from RBI warrants attention. It is a form of money finance and should ideally be included among memo items. However, in keeping with the international practice, as outlined above (Section III), that the sources of finance on which interest is payable should be taken as debt (or liabilities), WMA & OD have been included among liabilities of the State Government.

4.12 Figure 2 shows the sub-components of DFD and Memo Items, and how these together go to meet the fiscal deficit.

4.13 The present report mainly deals with compilation of debt statistics regarding explicit liabilities of the State Governments. For assessing the sustainability of finances of State Governments, however, the implicit commitments and contingent liabilities are equally important. The contingent liabilities are those commitments where government faces liabilities if a specific event occurs. These liabilities should include the guarantees given by a State government, provision of relief in the event of natural disaster and the financial commitments of institutions involved in quasi-fiscal activities which the State Government may have to honour in specific events. It may be noted that at present, only guarantees given by State Governments are published by some of the State Governments in their budget documents and the same is published by the Reserve Bank in its reports/documents on the relevant subject. These, however, are excluded from the purview of liabilities of the State Governments. In order to assess the sustainability of the State Government finances, the fiscal risk involved in the guarantees should be identified and quantified. The quantification of the fiscal risks of guarantees implies that a part of them will enter into the total liabilities of the State Governments.

4.14 The State Governments in India have been assigned the larger responsibilities of health services which implies that the States Governments have implicit liabilities in terms of health expenditure. Pensions are another important form of liabilities of the State Governments. This is particularly important as some of the State Governments are having pay-as-you-go systems and also the Old Age Pensions schemes. The assessment of liabilities on account of the pensions could be made from the total amount of funds an employee should contribute so that his pension is completely financed from the funds he has contributed and income earned from the investment of the same. Thus, the total estimates of pensions liabilities could be made in the form of funds that would be sufficient to finance the current pensions which are paid on pay-as-you-go system.

7Please refer to the Report of the Group to Assess the Fiscal Risk of State Governments Guarantees (published in the RBI Bulletin, June 2003) and the Report of the Group on Model Fiscal Responsibility Legislation at State Level (published in the RBI Bulletin, March 2005), for further details. It may be also mentioned that the RBI has so far organized three workshops on the subject of fiscal risk of guarantees, for the benefit of State Government officials.

4.15 Apart from the confirmed liabilities of the State Governments discussed so far, there is a rising phenomenon of the Off-Budget Borrowing by some State Governments through their Special Purpose Vehicles (SPVs) where the respective State Governments give guarantee. Thus, in principle, such liabilities are contingent liabilities. However, in practice, since SPVs do not have any independent source of own revenue generation, such liabilities are actual liabilities of the respective State Governments. In particular, in case the liability is incurred for a purpose or project without a revenue stream, leaving no surplus to service the liability (indicating complete absence of due diligence), the fiscal risk of the State Government guarantee would have to be very high and in that sense, such liabilities would be more ‘explicit’ than ‘contingent’. The Group also notes that liabilities could arise in the future on account of externalities from developmental projects. The budgetary provision in this regard would be restricted only to the outlays for the projects and not incorporate future liabilities arising from externalities. These are technical aspects of the measurement of such liabilities, which are beyond the scope of examination of the present Group. At this stage, it may be noted that all such liabilities are not created through any State budget process and hence do not appear as a part of the fiscal deficit of the State during the year they were incurred. Under the existing practice, the State either redeems the liability through injecting equity to SPVs and repaying the loans on behalf of the SPVs or contributes to the reserve fund specially created for meeting all such contingent liabilities in future at the time of their redemption. Under both these cases, the fiscal deficit of the current year will increase, leading to a corresponding increase in the debt or liabilities of the State in the next year. It has also been felt by some that certain subsidies that are provided under existing policies and that result in committed liabilities to the State Government (as for example power provided at concessional rates) also need to be reported in a separate Statement detailing their nature and amounts, from the viewpoint of enhancing fiscal transparency.

4.16 A Group of State Finance Secretaries to assess the fiscal risk of State Government Guarantees had examined the problem and submitted its report in July 2002. The Group recommended exercise of financial prudence by financial institutions like NABARD, HUDCO, PFC etc. while lending for project-based financing to SPVs with prior concurrence of Government of India (GoI) since all borrowings by the State Governments on the strength of guarantees have been brought under Article 293 (3) of the Constitution. Moreover, the Group also recommended State-specific caps on the levels of borrowings as prescribed under the (erstwhile) MTFRP and the creation of a sinking fund with a suitable contribution from the annual net small savings collections realizable to States. Such a fund would be managed by RBI and States contributing to the fund would be eligible for assistance therefrom. The creation of the Guarantee Redemption Fund is a step in the right direction because it formally incorporates future contingent liabilities in the present budgets. The Twelfth Finance Commission has, in fact, recommended that States should set up guarantee redemption funds through earmarked guarantee fees, which should be preceded by risk weighting of guarantees.

4.17 In line with the practice recommended and followed by IMF (Section III), the Group recommends that these contingent liabilities may be excluded from the formal definition of liabilities. In order to incorporate this dimension, however, the Group recommends the reporting of these liabilities on an annual basis under a separate head. In some cases, it is indicated a priori that the liability for repayment of principal and/or interest payment of the borrowings of SPVs would be met by the State Government from the provisions in its Consolidated Fund. Such borrowings of SPVs should be included under the explicit liabilities of the State Government.

4.18 Similar issues can be raised regarding the liabilities of a State Government regarding the pension to its employees in future and the State PSUs liabilities. It may be, however, noted that there is no clear-cut policy to calculate the pension liabilities for employees who entered in service till end-December 2003. Moreover, under the National Pension Scheme, which came into effect from January 2004, the Central Government has not extended any guarantee on the returns to the new entrants. If the same logic were to be extended, then there would not be any liability with the State Governments on this account. As far as PSU liabilities are concerned, there is a suggestion that an index on the rating of each State PSU needs to be worked out, on the basis of which the PSU liabilities could be computed and published. These are again real problems and appropriate provisions for them need to be made in the State budget. These will invariably increase the current and future fiscal deficits of the State. However, if account is taken of these liabilities as existing now and merged with the other liabilities of the State Government, the same problem of consistency with past fiscal deficits of the State Government would arise. Thus, in line with the practice followed by IMF (Section III), we may exclude the Pension Fund and Public Sector Liabilities from our formal definition of Liabilities. In order to incorporate this dimension, however, we recommend the reporting of these liabilities on an annual basis under a separate head. These estimates can be prepared over time when the required data are properly compiled.

5.1 It is proposed that debt and liabilities be considered synonymous. Accordingly, all borrowings which are repayable and on which interest accrues are recommended to be considered as debt. Alternatively, the practice followed in the Government of India budget may be adopted by the States as well. Thus, total budgetary liabilities are recommended to be decomposed in four categories viz. (i) Public Debt; (ii) Ways and Means Advances and Overdrafts from the RBI; (iii) Public Accounts; and (iv) Contingency Fund. The above treatment would be in conformity with the international best practice. As alluded to earlier, the IMF Manual on Government Finance Statistics, 1986, IMF describes government debt ‘a stock of liabilities with different time dimensions accumulated by government operations in the past and scheduled to be extinguished by the government operations in future.’

5.2 It may be noted that the above definition of liabilities brings about a closer alignment with that of GFD, as discussed in Chapter IV. The remaining discrepancy between computation of GFD and liabilities is essentially on account of certain ‘Memo items’ (i.e. withdrawal of cash balances, Appropriation to Contingency Fund, Suspense and Miscellaneous and Remittances) which are treated as financing items of the GFD, but excluded from liabilities. The Group recommends that withdrawal of cash balances may continue to be taken as a financing item of GFD. Cash balances, in any case, do not form part of liabilities. The remaining memo items viz., ‘Appropriation to Contingency Fund’, ‘Suspense and Miscellaneous’ and ‘Remittances’ may, however, be treated as non-debt items and excluded from the computation of GFD and liabilities, with a view to maintaining consistency between the definitions of GFD and liabilities.

5.3 The implicit liabilities that include Off-Budget Borrowings, Contingent Liabilities, Pensions and State Public Sector Liabilities should be excluded from the definition of State Liabilities or Debt. However, it is recommended that they may be disclosed in the financial statement along with estimates of debt. In those cases, however, where it is indicated a priori that the liability for repayment of principal and/or interest payment of the borrowings of SPVs would be met by the State Government from the provisions in its budget, the same should be included under the explicit/budgetary liabilities of the State Government.

5.4 As discussed earlier, the coverage of debt should be consistent with GFD. The break-up of GFD is then used to construct a measure of debt in annual flow i.e. Increase in Debt = Increase in [Internal debt (including WMA & OD) (net) + Loans & Advances from the Centre (net) + Small Savings & PF etc. + Deposits (net) + Reserve Funds (net) + Contingency Fund (net) While both stock and flow components are available from finance accounts- Statements 15 & 16, albeit with a delay of two years, the flow components can be obtained from budget documents or RBI reports on State Finances as shown in the following item exhibit.

5.5 It is also felt that the demarcation of Public Debt into Internal Debt and Loans & Advances from the Centre can be done away with. It is instead proposed to report a single major head viz. Public Debt and include all items currently reported under Internal Debt and Loans & Advances from the Centre as minor heads within it. Figure 3 gives the final definition of Liabilities, as proposed by this working group.

5.6 There could be many sources of the data as many agencies are involved in the transaction and reporting of all the items included in the State Government Liabilities. For ensuring the consistency in the data, it is desirable that a single agency compiles and disseminates the information on outstanding liabilities of all the States. Although the State Governments are the most reliable sources of such information, the task cannot be fully entrusted to them unless it becomes an obligatory part of the State Budget documents. The Government of India, RBI and other institutions could help the States in creating necessary capacity, systems and processes and acquiring technology to compile data on liabilities. RBI can then act as a single agency putting estimates of liabilities of all States together in a single publication, as it does for the State Budgets. The proposed scheme for the publication of liabilities is set out below.

5.7 It is proposed to publish in the State Government Budget documents the information of the State Government liabilities under the following statements:

An additional Statement on Subsidies provided by the State Government may also be provided as indicated in para 4.15.

5.8 The format of Statements 1 to 7 is given below. It is recommended that Statements (1) (alongwith the Annexes) and (2) may be published in the budget documents of the State Governments with effect from the fiscal year 2006-07. The remaining Statements should be published by the State Governments as soon as possible. In case it is not possible to bring out all the remaining Statements at the same time, a graduated approach for publication could be adopted. Information provided in the format of Statements 1 to 7 would ensure uniformity of practices of compiling data on liabilities across State Governments as well as enhance fiscal transparency. The Group recognizes that there are widely differing views on the inclusion of various implicit liabilities in order to obtain the aggregate liabilities of the State Governments, on which a consensus may emerge over a period time. Priority may, however, need to be accorded to the task of collating and publishing data on the different parameters, as set out in Statements 1 to 7, which should be available with the State Governments. Once the modalities for collating and publishing information on the various liabilities (both explicit/budgetary and implicit) of the State Governments as well as on the associated fiscal risks of the implicit liabilities get firmly entrenched, policy makers and researchers would be provided the option of progressively including various implicit liabilities (guarantees, off-budget borrowings, pensions, etc), or their risk-weighted component, in order to obtain a more comprehensive picture of the overall liabilities of State Governments. Efforts to build up a consensus on the inclusion of such implicit liabilities could then be expedited.

* Rate of interest guaranteed in case of loans, debentures, etc. is to be given.

5.9 It is proposed that all the data relating to the liabilities of State Governments may be published in the budget documents of the respective State Governments. The State Governments should also bring out quarterly, if not monthly reports on their accounts and liabilities. It may be noted that data to be provided in the Statements 1 to 7 relate to three years viz., the budget estimates of the current year, revised estimates of the previous year and ‘accounts’ data of the year before. It may also be noted that these ‘accounts’ data provided in the budget documents are un-audited figures and that ideally, audited data should be used which would ensure comprehensive accuracy of the relevant magnitudes. As indicated in para 2.4, timely availability of audited data on State Government budgetary transactions, however, continues to be beset with some difficulties, which need to be addressed by the concerned entities at the earliest. The CAG may also compile and publish the audited data on liabilities in addition to the Finance Accounts of the States. The Group, however, suggests that the non-availability of audited data should not delay the reporting of data on liabilities as per the ‘accounts’ (un-audited), revised estimates and budget estimates of the latest years.

5.10 In order to facilitate the process of data compilation, the Group recommends the following:

5.11 Till such time that the State Governments are not in a position to publish the requisite data on their outstanding liabilities in their budget documents, all the above data may be furnished by the concerned institutions to the RBI, as a transitional measure, to enable consolidation and publication. As alluded to earlier, the RBI’s annual Study on State Government provides, inter-alia, State-wise detailed data on the budget estimates of the current fiscal year, the revised estimates of the previous year and the ‘accounts’ (un-audited) data of the year before. This is the only publication to provide such updated data in respect of all the State Governments on a regular and timely basis. Keeping this in view, the Group recommends that the RBI should compile the (latest available) ‘accounts’ (un-audited) and (revised and budget) estimates of liabilities of all the State Governments and publish the same in its regular annual publication on State Budgets, from the viewpoint of data dissemination and to facilitate academic and policy research. The CAG may provide the latest available audited data on liabilities and the same could also be reported by the RBI in its annual study on State budgets, alongwith the ‘accounts’ (un-audited) and the revised estimates and budget estimates of more recent years.

1. Buiter, Willem H. and Urjit R. Patel (1992): "Debt, deficits, and inflation: An application to the public finances of India", Journal of Public Economics, Volume 47, Issue 2, March 1992, Pages 171-205.

2. CAG (2002): State Finances - A Critical Appraisal (1997-98 - 2001-02), International Centre for Information Systems and Audit, Office of the Comptroller & Auditor General of India.

3. CAG (2004): State Finances – A Critical Appraisal, International Centre for Information System and Audit (ICISA), Office of the Comptroller & Auditor General of India, October.

4. CAG (199 ): Combined Finance and Revenue Accounts of the Union and State Governments in India for the Year 1991-92, CAG, New Delhi.

5. CAG (2004): Combined Finance and Revenue Accounts of the Union and State Governments in India for the Year 1996-97, CAG, Delhi.

6. Dholakia, Ravindra H. (2003): "Measurement Issues in Comparing Fiscal Performance of States," Economic & Political Weekly, Vol.38, No.10, March 8

7. Dholakia, Ravindra H. and Karan Navendu (2004): Consistent Measurement of Fiscal Deficit and Debt of States in India, Working Paper, IIM, Ahmedabad.

8. --------------- (2005): "Consistent Measurement of Fiscal Deficit and Debt of States in India", Economic and Political Weekly, June 18, 2005.

9. Eleventh Finance Commission (2000): Report of the Eleventh Finance Commission.

10. Twelfth Finance Commission (2004): Report of the Twelfth Finance Commission

11. Government of India: Major and Minor Heads of Accounts of Union and State Governments, Ministry of Finance, New Delhi.

12. Gurumurthi, S (2002): "Twelfth Finance Commission and States' Debt Burden", Economic & Political Weekly, October 5, 2002

13. IMF (2001): Government Finance Statistics Manual, International Monetary Fund, Washington.

14. Rajaraman, Indira and Abhiroop Mukhopadhyay (1998): "Sustainability of Public Domestic Debt in India", NIPFP, mimeo.

15. Rangarajan C. and D.K. Srivastava (2003): "Dynamics of Debt Accumulation in India: Impact of Primary Deficit, Growth and Interest Rate", Economic and Political Weekly, November 15, 2003.

16. -------------------------------------------- (2005): "Fiscal Deficits and Government Debt – Implications for Growth and Stabilisation", Economic and Political Weekly, July 2, 2005.

17. RBI (2004): State Finances: A Study of Budgets of 2004-05.