India’s Debt Dilemma

Barry Eichengreen
University of California, Berkeley and NCAER

Poonam Gupta
NCAER

Ayesha Ahmed
NCAER

India Policy Forum

July 6–7, 2023

NCAER | National Council of Applied Economic Research

NCAER India Centre, 11 Indraprastha Estate, New Delhi 110002
Tel: +91-11-23452698, [Link]
NCAER | Quality . Relevance . Impact
 2 | India Policy Forum 2023

The findings, interpretations, and conclusions expressed are those of the authors and do not
necessarily reflect the views of the Governing Body or Management of NCAER.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 3

India’s Debt Dilemma*

Barry Eichengreen
University of California, Berkeley and NCAER

Poonam Gupta
NCAER

Ayesha Ahmed
NCAER

India Policy Forum

July 6–7, 2023

Abstract

India was an outlier on fiscal outcomes pre-pandemic, before drifting further in the high debt direction
during COVID. Its debt-to-GDP ratio is not likely to fall anytime soon. Current high levels of debt limit the
resources available for other priorities such as health, education and climate change abatement. At the same
time, there is no immediate crisis of debt sustainability: institutional factors limit rollover risk, and interest rates
have not risen with additional debt issuance. But financial stability and sustainability risks may arise in the
future, and lack of resources to meet pressing needs is a drag on growth. Consolidation would require lower
primary deficits achieved through greater tax revenue generation and privatization, all while protecting and
prospectively increasing capital spending. Contingent liabilities have posed risks to the public finances of the
States and should be minimized by fiscal-management reforms. As their debt manager, the RBI should ensure
that States face the market interest rates warranted by their current and projected debt levels. Fiscal
Commissions should provide stronger incentives for prudence.

JEL Classification: H6, H7, H61, H63

Keywords: Debt Management, Debt Sustainability, Finance Commission, Fiscal Deficit,

Public Debt

* Preliminary draft. Please do not circulate beyond the NCAER India Policy Forum 2023, for which this
paper has been prepared.
Prepared for the India Policy Forum. We thank Navya Srivastava and S. Priyadarshini for research
assistance.
4 | India Policy Forum 2023

1. Introduction
India’s public finances paint a mixed picture. The country was an outlier in fiscal
outcomes before the pandemic. Its fiscal deficits and public debts were among the
highest in the developing world; its interest payment/GDP ratio and primary deficits
were large. The pandemic reinforced these trends. At their peak in 2020-21, the debt
and deficit stood at 89 and 13 per cent of GDP respectively. (Contingent liabilities – the
present value of the prospective stock – are estimated at an additional 5 per cent of
GDP). A large part of the additional public debt accumulated during COVID was
incurred by the Central Government, while a majority of the debt accumulated in pre-
COVID years was owed by the States. With the recovery of nominal GDP, the country’s
debt and deficit ratios have fallen from these multi-decade highs. But at 84 and 9 per
cent, they are still high relative to other emerging market and middle-income countries,
where they average 60 and 5 percent.1
In this paper we assess the sustainability of the public finances, with a focus on
the next five years.
A first criterion for sustainability is whether the debt ratio will remain stable. We
find, under reasonable assumptions, that the debt ratio will remain broadly stable. This
stability rests on the assumption of a largely unchanged primary budget deficit and a
favorable growth-rate-interest-rate differential, the latter reflecting India’s positive
growth prospects and also institutional factors limiting upward pressure on interest
rates. The institutional factors in question include a captive market for public debt
among state banks, private banks, insurance companies and provident funds. Together
with household savings, these have enabled the government to fund its deficits without
undue pressure on borrowing costs.
A second for sustainability is whether there is significant rollover risk. We find
that these same factors, together with the currency composition and maturity of the
debt, also limit rollover risk. In this respect our conclusions differ from those of
Blanchard, Felman and Subramanian (2021).
Counterbalancing these happy conclusions is the unhappy fact that India is
unlikely to significantly reduce its debt ratio absent extensive and politically-fraught
reforms. Smaller primary budget deficits will be difficult to achieve given pressure for
social and infrastructure spending, including on climate-change abatement and
adaptation and the green transition, and the difficulty of boosting tax revenues. Faster
growth rates or lower interest rates are pleasant to imagine but difficult to achieve.
What are the costs of living with high public debt? First, interest payments will
continue to absorb a significant share of the government’s resources, limiting their
availability for other economic and social priorities. Second, available fiscal resources
leave no room for meeting emerging priorities, including health, education, and climate
change adaptation. Third, debt dynamics leave little scope for responding to negative
shocks, such as declining rates of domestic or global growth. Fourth, having banks hold
large amounts of government debt creates financial stability risks; this is the "diabolic
loop” seen a decade ago in Europe and in the case of the Silicon Valley Bank. Fifth, and

1 These numbers and the categorization of countries, 95 in number, as “emerging-market and middle
income” are from the IMF’s Fiscal Monitor, April 2023. The fiscal year runs from April to March. For
example, fiscal year 2023-24 refers to April 1, 2023-March 31, 2024.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 5

relatedly, with further financial liberalization and reform, the government when
marketing its debt may be forced to rely less on captive domestic investors and more on
foreign institutional investors. Though rollover risk may be limited now, it may rise in
the future with this change in investor composition.
Section 2 summarizes trends in India’s public finances, while Section 3 considers
salient features of debt composition. Section 4 presents a debt sustainability analysis,
after which Section 5 analyzes past episodes of debt consolidation. In Section 6 we
assess the implications and risks of the current levels of debt. Section 7 summarizes the
implications.

2. Debt and Deficits in India

Public debt has not only been high in India, but it has increased markedly over
the past four decades (Figure 1). Having averaged 60 percent of GDP in the 1980s, it
rose to 70 percent in the 1990s and 80 percent in the 2000s. From these highs, it
declined to 69 percent of GDP the following decade, before increasing to nearly 90
percent of GDP in 2020-21 in the wake of COVID and hovering at 85-87 percent for the
last two years.
The budget deficit has fluctuated around 7-8 percent of GDP, as shown in Figure
1. It rose to an unprecedented 13.1 percent of GDP in 2020-21. This increase was due
mainly to higher expenditure, and to a lesser extent due to slower revenue growth and
contraction of nominal GDP. This unprecedented deficit resulted in a commensurately
large increase in public debt to nearly 90 percent of GDP, surpassing the previous peak
of about 83 percent in the early 2000s.
Interest payments have averaged 5 percent of GDP for three decades. They rose
from 11.5 percent of total revenue in 1980-81 to fully a quarter of total revenue in
2022-23. Government spends more on interest than on education and health combined.
Interest payments exceed total capital expenditure. The General Government’s primary
deficit (deficit net of the aforementioned interest payments) averaged a bit over 2
percent of GDP in the two decades preceding COVID. The General Government has in
fact run a primary surplus only once in the past 40 years, in 2007-08. Since then, there
have been two sharp increases in the primary deficit, to 4.6 percent of GDP in 2009-10
and 7.8 percent of GDP in 2020-21.2

2 Subsequently, the primary deficit decline to 3.7 percent in 2022-23.

 6 | India Policy Forum 2023

Figure 1: General Government (Federal and States) Debt and Fiscal Indicators
Figure 1A: Total Public Debt3 Figure 1B: Fiscal Deficit
Percent of GDP Percent of GDP

Total Public Debt 14 Fiscal Deficit

100 12
80 70.4 68.7 86.5 10 7.8
78.2 7.6
60.0 8 6.8
60 8.9
6 7.9
40
4
20 2
0 0
1980-81
1983-84
1986-87
1989-90
1992-93
1995-96
1998-99
2001-02
2004-05
2007-08
2010-11
2013-14
2016-17
2019-20
2022-23

1980-81
1983-84
1986-87
1989-90
1992-93
1995-96
1998-99
2001-02
2004-05
2007-08
2010-11
2013-14
2016-17
2019-20
2022-23
Figure 1C: Primary Deficit Figure 1D: Interest Payments
Percent of GDP Percent of GDP
Interest Payments
Primary Deficit
10 7
5.7
6 5.0
8 4.7
5
6 4 5.2
4.9 3.0
4 3
2.7 3.7
2.1 2.1 2
2
1
0 0
1980-81
1983-84
1986-87
1989-90
1992-93
1995-96
1998-99
2001-02
2004-05
2007-08
2010-11
2013-14
2016-17
2019-20
2022-23

1980-81
1982-83
1984-85
1986-87
1988-89
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2022-23
-2

Source: CEIC (Compiled from Reserve Bank of India). Dashed horizontal lines are decadal averages from
1980-81 to 1989-90, 1990-91 to 1999-2000, 2000-01 to 2009-10, and 2010-11 to 2019-20, respectively.

Revenues have increased only slowly, compared to the increase in other large
emerging markets (Figure 2). Between 1980-81 and 2022-23, tax revenue rose by 3.3
percentage points of GDP, reflecting tax buoyancy (elasticity of revenues with respect to
income) only slightly above 1. Non-tax revenue, which includes interest and dividends,
has similarly remained stagnant as a proportion to GDP. The elasticity of revenues with
respect to income is higher in other large middle-income economies, with the sole
exception of Indonesia.4 In comparison, the expenditure-to-GDP ratio has been close to

3 Total Public Debt in India includes debt issued and other liabilities in Public Account consisting of
National Small Saving Fund (NSSF), Provident Fund, Deposit and Reserve funds, securities issued to
finance subsidies on oil, food, and fertilizers, etc.
4Whereas direct tax collection has increased in proportion to GDP, indirect taxes as a proportion of GDP
have declined, indicating a tax buoyancy of more than 1 for direct taxes, and less than one for indirect
taxes (Appendix A).
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 7

the median of other emerging countries. This gap has resulted in a perennially large,
and even increasing, budget deficit compared to other emerging markets.
Expenditure overall has remained broadly stable for two decades, the only large
increase occurring during COVID. Nearly 85 percent has been revenue or committed
expenditures.5 Capital spending has been low, rising modestly from 2.3 percent of GDP
in 1994-95 to 3.6 percent of GDP by 2011-12, and hovering close to that level over the
period 2020-21. It then rose by 1.4 percentage points to 5.0 percent of GDP in the past
two years, reflecting the government’s infrastructure push.
Figure 2: General Government (Federal and States) Revenue and Expenditure

Figure 2A: Total Revenue Figure 2B: Tax Revenue and Non-Tax Revenue
Percent of GDP Percent of GDP
24
Total Revenue Tax Revenue and Non-Tax Revenue
22 20
19.8
20 14.7 14.5 15.5
17.9 20.2 15 16.7 16.7
18 19.2
18.6 10 Tax Revenue
16
Non-Tax Revenue
14 5 3.9 3.5 3.6 3.1 3.5
12 0
1980-81
1983-84
1986-87
1989-90
1992-93
1995-96
1998-99
2001-02
2004-05
2007-08
2010-11
2013-14
2016-17
2019-20
2022-23

1980-81
1983-84
1986-87
1989-90
1992-93
1995-96
1998-99
2001-02
2004-05
2007-08
2010-11
2013-14
2016-17
2019-20
2022-23
Figure 2C: Total Expenditure Figure 2D: Revenue and Capital Expenditure
Percent of GDP Percent of GDP

Total Expenditure Revenue Expenditure and Capital

35 30
26.7 26.1
Expenditure 23.8
30 24.4 25 23.3 22.6
21.8
25 28.7 20
20 Revenue
15 Capital
15
10
10 3.4 3.6
5 2.6 5.0
5
0
0
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2022-23
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2022-23

Source: CEIC (Compiled from Reserve Bank of India). Dashed horizontal lines are decadal averages from
1980-81 to 1989-90, 1990-91 to 1999-2000, 2000-01 to 2009-10, and 2010-11 to 2019-20, respectively.
Interest payments are high by global and emerging market standards (Figure 3).
The IMF (2023) projects a further rise in the interest-payments-to-GDP ratio over the
2023-27 period as global rates trend upward.

5 Revenue expenditures are expenditures incurred for purposes other than the creation of physical or
financial assets. They are incurred for the normal functioning of the government departments, interest
payments, and grants to State governments and other parties.
 8 | India Policy Forum 2023

While India’s debt ratio is comparable or lower than in the advanced economies,
this is scant comfort. Advanced-country governments enjoy lower interest rates and
consequently have lower interest-payment-to-GDP ratios. Debt-to-GDP ratios of
advanced economies averaged 112 percent in 2022, whereas interest payments
averaged 1.5 percent of GDP. In contrast, India pays as much as 5 percent of GDP in
interest on debt.
Figure 3: Comparing India’s Fiscal Indicators with Other Country Averages
(General Government)

3A) General Government Debt, Global 3B) General Government Debt, Emerging Markets
100 Public Debt Global, % of GDP Public Debt Emerging Markets, %
100 of GDP

50
50

0 0
2000 2010 2019 2022 2027 2000 2010 2019 2022 2027
Median Q1 level Median Q1 level
Q3 level India Q3 level India

3C) Fiscal Deficit, Global 3D) Fiscal Deficit, Emerging Markets

15 Fiscal Deficit Emerging Markets, %
Fiscal Deficit Global, % of GDP 20
of GDP
10
10
5

0 0
2000 2010 2019 2022 2027 2000 2010 2019 2022 2027
-5
Median Q1 level -10 Median Q1 level
Q3 level India
Q3 level India

3E) Interest Payments, Global 3F) Interest Payments, Emerging Markets

Interest Payments Global, % of GDP Interest Payments Emerging
6 10 Markets, % of GDP
4

0 0
2000 2010 2019 2022 2027 2000 2010 2019 2022 2027
Median Q1 level
Median Q1 level
Q3 level India
Q3 level India

Source: Fiscal Monitor Database, IMF April 2023. Figures show median and interquartile range of the
respective variables and respective country or country groups.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 9

Figure 4 shows that the revenue-to-GDP ratio is below that of most other
emerging markets (see also Rao 2018). Not only is the level below that in other
countries, but India has one of the slowest rates of increase over the last 20 years. In
contrast, the public-expenditure-to-GDP ratio is not atypical and, if anything, has
increased more slowly. India’s deficit is evidently more a problem of low revenues than
one of high expenditure.6
Figure 4: Comparing India’s Fiscal Indicators with Emerging Market (EM)
Averages, General Government
4A) Total Revenue to GDP (EM Median, Interquartile 4B) Total Expenditure to GDP (EM Median,
Range and India) Interquartile Range and India)

Total Revenue (% of GDP) Total Expenditure (% of GDP)

40 50

20
0

0
2000
1990
1992
1994
1996
1998

2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022

India EM Q1 level
India EM Q1 level EM Median EM Q3 level
EM Median EM Q3 level
4E) Total Revenue to GDP as of 2022 (select EMs) 4F) Increase in Total Revenue to GDP between 2000
and 2022
Total Revenue (% of GDP), 2022 Increase in Total Revenue (% of GDP)
27.7 27.6 25.8 25.5
30
21.6 15 12.1 between 2000 and 20022
25 20.1 19.2 18.3
20 15.2 10 8.3
15 6.4
4.0 3.9
10 5 2.5 2.1 1.8 1.8
5
0 0

4G) Total expenditure to GDP as of 2022 (select EMs) 4H) Increase in Total Expenditure to GDP between
2000 and 2022
40 34 33 32 20 16.8
35 30 29
30 27 26 15
25 21 9.9 9.5
18
20 10 7.7
15 6.2 6.0
10 5 3.2 3.0 2.2
5
0 0

Source: Fiscal Monitor, IMF April 2023. Figures 4A and 4B show median and interquartile range of Emerging
Market and Middle-Income Economies (83 countries) and India. Data for India is for fiscal years.

6 We return to this point in Section 5 below.

10 | India Policy Forum 2023

3. Debt Composition
Nearly 90 percent of General Government debt is long-term, as measured by
residual maturity (Figure 5).7 There has been a concerted effort to reduce rollover risk
by issuing long-tenor securities. As a result, the weighted average maturity periods for
both Central and State government loans have been increasing (Figure 6).
Figure 5: Duration of Debt (General Government)

Duration of Debt (% of Total Debt)

100

80

60

40

20

0
2006-07

2010-11
2000-01
2001-02
2002-03
2003-04
2004-05
2005-06

2007-08
2008-09
2009-10

2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
Short Term (% of Debt) Long Term (% of Total Debt)

Source: Data for short-term debt for 2000-01 to 2009-10 are from Status Paper, Ministry of Finance,
September 2016; and then from Status Paper on Government Debt, Ministry of Finance, April 2022. Long-
term debt is calculated as total minus short-term debt.

Tenors vary. The share of Central Government debt with a maturity greater than
20 years rose from 13 to 20 percent between 2012 and 2021. In the two most recent
years, a majority of debt issued by the Central Government has had a maturity of 14
years or longer, and 30 percent has had a 30- or 40-year maturity.
State debt has a lower average maturity. As of March 2022, about 5 percent of
the outstanding State Development Loans (SDLs) had a maturity of less than a year.
Maturity periods for 30 percent of SDLs were 1-5 years, for 45 percent 5-10 years. The
remaining 20 percent had a maturity of 10 years or longer (of which a small proportion
had a maturity of more than 20 years).8 The market for long term debt is thin, and the
term premium for all but the highest quality borrowers (insurance companies and the
like) can be significant. The States seek to minimize interest costs; they therefore issue
shorter-term debt while waiting for the market in longer-term debt to develop further.

7 Short-term debt of the Center includes 14-day intermediary treasury bills, 91-day, 182-day, and 364-day
treasury bills, dated securities maturing in the ensuing year, and external debt with residual maturity of
less than one year. For the states, short-term debt includes market loans maturing within the next year,
loans to the Center due in the ensuing year, and short-term borrowings from the RBI through Ways and
Means Advances (WMA).
8According to the RBI’s Report on State Finances (January 2023), “Though 63.3 per cent of the
outstanding State government securities is in the residual maturity bucket of five years and above,
redemption pressure is expected to remain high till 2030-31.”
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 11

Figure 6: Weighted Average Maturity of Outstanding Debt

Weighted Average Maturity of Weighted Average Maturity of

Centre and State Government General Government Securities
14 Securities 14
12
12
10
10
Years

Years
8
6
4 6
2 4
0 2
0

Central Govt Weighted Average Maturity

State Govt Weighted Average Maturity
Weighted Average Maturity on
Dated Securities

Source: For the Center, Status Papers (April 2022 and September 2016), Ministry of Finance, from 2003-
04 to 2020-21, Public Debt Management Quarterly Report, RBI (March 2023) for 2021-22 and Q3 2022-
23; For state governments: Monthly Reviews of the Economy, Clearing Corporation of India (CCIL). GDP
Deflator is from the Economic & Political Weekly Research Foundation (EPWRF).
Note: We use the shares of the Center and states in total debt as weights to calculate weighted average
maturity on the General Government outstanding stock, for Q3 2022-23, the shares are assumed to be the
same as those for 2021-22. For Q3 2022-23, the weighted average maturity is the average of the weighted
average maturities for the period Q1-Q3 2022-23.

While the average maturity of public debt has risen, yields have declined, if
slightly. The General Government weighted average coupon fell from 8 percent in 2011-
12 to 7.3 percent in 2022-23 (Figure 7). The average yield on Central Government debt
has been slightly lower than that on State debt.
Strikingly, bond yields in India have not moved with the level of indebtedness or
inflation. This is true at both the Central and State Government levels. Moreover, the
interest rate at which different States raise their debts does not vary significantly by the
level of indebtedness, primary deficit, or the rate of economic growth.9 Mishra and Patel
(2018) suggest that this reflects an implicit guarantee from the central government, and
the fact that the largest investors in government bond markets (public sector banks,
insurance companies and provident funds) are owned by the Central Government, and
as such are not purely profit maximizing entities. These institutional investors are all
required to hold government bonds as a statutory requirement (see Appendix C below).
In addition the Reserve Bank of India (RBI), by carefully scheduling the calendar
of borrowing and coaxing government-owned investors to hold the bonds of the States,
ensures that interest rates on State debt remain in a tight range. Evidently, it does not
want perceptions of debt distress or unsustainability of the debts of some States to

9 The calculations are based on the average nominal weighted average yield on new issues.
 12 | India Policy Forum 2023

infect others. We are not convinced of the advisability of this policy; we will have more
to say about it below.
Figure 7: Cost of Debt (Outstanding Debt)

Weighted Average Coupon on Center and Weighted Average Coupon of General

State Securities Government Securities
10 10
Weighted Average Coupon

Weighted Average Coupon

9 9
8 8
7 7
6 6

5 5
4
4

Q3 2022-23
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22
Q3 2022-23
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22

Central Govt Weighted Average Coupon Weighted Average Coupon on Dated

State Govt Weighted Average Coupon Securities

Source: For the Center Nominal WAC: Status Papers (April 2022 and September 2016), Ministry of
Finance, from 2003-04 to 2020-21, Public Debt Management Quarterly Report (March 2023) for 2021-22
and Q3 2022-23. For State Nominal WAC: Monthly Review of the Economy, Clearing Corporation of India
(CCIL).
Note: The shares of the Center and States in total debt are used as weights to calculate the weighted
average coupon on General Government outstanding stock.
The average yield on new issuances has also declined over time, from about 11
percent in 2000-01 to about 7.5 percent currently (Figure 8).
Figure 8: Cost of Debt (new issues in the year)

Yield on Dated Securities Weighted Average Yield General

12.0 Government
12.0
Weighted Average Yield

10.0
10.0
Weighted Average Yield

8.0
8.0
6.0
6.0
4.0
4.0
2.0
2.0
-
Q3 2022-23
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21

-
Q3 2022-23
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21

Central Govt Weighted Average Yield

State Govt Weighted Average Yield
Weighted Average Nominal Yield on Dated…

Source: For the Center and State WAY, RBI (till 2020-21). For the Center, Public Debt Management
Quarterly Report (March 2023) for 2021-22 and Q3 2022-23; and for the State: State Finances Report
(2023), RBI, for 2021-22 and 2022-23. Yield is for primary issues in the year indicated.
Note: Shares of the Center and States in total debt are used as weights to calculate the Weighted Average
Yield (WAY) on General Government primary issues (new issues in the year).
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 13

As Figure 9 shows, less than 4 percent of General Government debt in 2020-21 is

offered at floating rates. (Only the Central Government offers floating debt.10) Thus, the
country’s debt portfolio is largely insulated from short-run interest rate volatility.

Figure 9: Fixed and Floating Rate Debt

Fixed/Floating Rate Debt (% of Total Debt)

100
90
80
% of Total Debt

70
60
50
40
30
20
10
0
2001-02
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
Fixed Rate (% of Total Debt) Floating Rate (% of Total Debt)

Source: Data for years 2000-01 to 2009-10 is from the Status Paper (September 2016); data for 2010-11
to 2020-21 are from the Status Paper (April 2022), Ministry of Finance.

Figure 10 shows the proportion of General Government debt securities in

different hands. In 2015-16, about 42 percent of General Government debt was owned
by commercial banks. The bank share then dropped to 37 percent in 2021-22, as
various regulatory requirements mandating their holding government bonds were
relaxed (see below, including Appendix D). Foreign portfolio investors owned about 3
percent of public debt securities in 2015-16; their share similarly dropped to 1 percent
in 2021-22. Correspondingly, the shares of insurance companies, provident funds and
the RBI increased over time.

10A floating rate bond is based on a benchmark rate, such as the repo rate, reverse repo rate, treasury bill
yield, or saving schemes interest rates, plus a fixed spread that is determined at the time of first issuance.
14 | India Policy Forum 2023

Figure 10: Ownership of Debt

Ownership Pattern of General Government Dated Securities

100
90
80
70
60
50
40
30
20
10
0
2015-16 2016-17 2017-18 2018-19 2019-20 2020-21 2021-22

Commercial Banks Insurance Companies Provident Funds

RBI Foreign Portfolio Investors Others

Source: Public Debt Statistics, RBI. Note: Provident funds are retirement funds run by the government. Others
include Co-operative Banks, Non-Bank PDs, Mutual Funds, Corporates, Financial Institutions, and Others.

In 2000-1, about 13.5 percent of Central Government debt was issued externally.
Since then there has since been a steady decline in the share of external debt, which
stood at just 3.7 percent in 2021-22 (Figure 11). The remainder is long-term
instruments, concessional, and owed to multilateral and bilateral investors (amounting
to 3 percent of the total debt) 11 Holdings of foreign institutional investors are just 1
percent of the total debt. Foreign banks hold negligible quantities of Indian government
debt.
As is to be expected, most of this externally-held debt is denominated in foreign
currency. Debt denominated in foreign currency dropped from about 10 percent of the
total in 2002-03 to 4.3 percent in 2020-21 (Figure 11). Consequently, the debt portfolio
is largely insulated from currency risk.

11In 2003-04, IDA was the largest source of multilateral external debt. Since then, its share has dropped
by half (from 54 percent of the external debt to 26 percent in 2021), with a corresponding increase in
debt from IBRD and ADB, which contributed to 16 percent and 19 percent of the external debt,
respectively, as of 2020-21. Among the bilateral sources, Japan has consistently been the largest
contributor, accounting for 24 percent of the external debt in 2020-21, followed by Germany and Russia.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 15

Figure 11: External Debt (% of Total Debt), General Government

External Debt (% of Total Debt) Foreign Currency Denominated Debt (% of

12 Total Debt)
12 10.9 10.9
10 10
8.5 8.4
7.5 7.8 8 7.5 7.8
8
6.5 6.5
5.8 5.8
6 5.1 6 5.1
4.5 4.2 4.3 4.5 4.2 4.3
4 4

2 2

0 0

2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21

Source: Status Paper on Government Debt (September 2016), Ministry of Finance, for the data for 2002-
03, and Status Paper on Government Debt (April 2022), Ministry of Finance, for the data from 2003-04 to
2020-21. Note: External debt is debt to foreign lenders: banks, nonbank financial institutions,
international organisations and foreign governments, among others.

4. Debt Sustainability
We now use extrapolations of the debt/GDP ratio as a way of thinking about debt
sustainability. We use Equation 1 to project the trajectory of public debt.12

𝑏𝑡−1 (𝑟𝑡 −𝑔𝑡 )

∆ 𝑏𝑡 = + 𝑝𝑑𝑡 (1)
1+𝑔𝑡

Here bt is the debt-to-GDP ratio, pdt is the primary-deficit-to-GDP ratio (deficit net
of interest payment), gt is growth of real GDP, and rt is the real interest rate on public
debt; all in year t. ∆ 𝑏𝑡 is the change in debt-to-GDP ratio between t and t-1.

General Government
We consider a baseline scenario and several additional scenarios. As the
baseline, real GDP growth, the real interest rate, and the primary deficit will be at the
same levels for the next five years as their respective averages from 2013-14 to 2022-
23 (Table 2) – that is, 5.7 percent, 2.8 percent and 2.9 percent, respectively.13 This yields
an annual increment to the debt/GDP ratio of 0.5 percentage point a year, implying a
cumulative increment of 2.2 percentage points over five years. General Government
debt is projected to reach 88.7 percent of GDP in 2027-28 (Table 3).

12The exercise is based on the assumption that g, r, and pd are exogenous, that is, they are not impacted
by the level of debt.
13For comparison, in 2022-23 growth was 7.0 percent, the real interest rate was -1.0 percent, and the
primary deficit was 3.7 percent.
16 | India Policy Forum 2023

The second scenario assumes faster GDP growth. Our third scenario then adds a
favorable change of half a standard deviation in the primary deficit from the average
level over the past decade for each variable (Table 3). Thus, we assume GDP growth of
7.9 percent a year, or a primary deficit of 1.9 percent, respectively.
In this second scenario, the debt/GDP ratio declines by 1.2 percentage points a
year, reaching 81.0 percent in 2027-28. In the third scenario, it declines by 0.5
percentage points a year, reaching 83.9 percent in 2027-28.14 Thus, even under
optimistic assumptions, the debt/GDP ratio will remain high relative to comparator
countries.
The debt ratio will also remain high relative to India’s Fiscal Responsibility and
Budget Management (FRBM) targets, which foresee a debt-to-GDP ratio of no more than
60 percent.15 But adherence to these targets is not mandatory. There is no formal
mechanism to monitor compliance, and there are no penalties for breaching the targets.
It follows that governments have not been able to adhere to these limits on deficits and
debts.
Table 1: Average Values and Standard Deviations of the Key Parameters for
General Government
Ten-year average Five-year average
(2013-14 to 2022-23) (2018-19 to 2022-23)
Mean Std dev Mean Std dev
Nominal GDP growth (γ) 10.7 4.3 4.1 5.9

Deflator growth (π) 4.7 2.3 5.6 2.7

Real GDP growth (g) 5.7 5.3 10.0 7.9

Nominal interest rate (i) 7.5 0.8 7.0 0.7

Real interest rate (r) 2.8 2.6 1.4 2.9

Primary deficit (pd) 2.9 2.1 4.0 2.6
Growth-interest 3.0 4.8 2.7 7.2
differential (g-r)

14 We obtain similar pathways for public debt under most other reasonable scenarios.
15 The Sarma Committee on Fiscal Responsibility Legislation was set up in 2000 to recommend fiscal
reforms. After several rounds of reviews and modifications, its deliberations led to the formulation of the
Fiscal Responsibility and Budget Management Act. In 2016, a committee under N.K. Singh was then tasked
with suggesting changes in the Act. It suggested using General Government debt as the primary target for
fiscal policy, with a General Government debt-GDP target of 60 percent to be achieved by 2023 (40
percent for the Centre and percent for the States). Accordingly, the Finance Act of 2018 included the
following amendments to the FRBM Act. First, the fiscal deficit should be reduced to 3 percent of GDP by
2020-21. Second, the revenue deficit (the difference between recurrent expenditure and recurrent
earnings) and effective revenue deficit (revenue deficit minus any grants that the states received from the
Center for capital expenditure) were no longer targeted. Third, General Government debt again was not
to exceed 60 per cent of GDP, while Central Government debt was not to exceed 40 per cent of GDP, but
now by the end of 2024-2025.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 17

Table 2: Evolution of General Government Debt-to-GDP Ratio

Scenarios Scenario Debt level Primary Real Real Change in Cumulative
Description in 2022- Deficit GDP Interest Debt in Change in
23 (pd) growth Rate first year Debt in next
(bt-1) (g) (r) (∆ 𝒃𝒕 ) five years

Baseline Baseline: Past 10- 86.5 2.9 5.7 2.8 0.5 2.2
(S1) year averages
S2 Higher real GDP 86.5 2.9 7.9 2.8 -1.2 -5.5
growth rate
S3 Lower Primary 86.5 1.9 5.7 2.8 -0.5 -2.6
Deficit
S4 S1 plus contingent 86.5 2.9 5.7 2.8 1.5 6.9
liabilities
absorbed (1
percentage point
of GDP) each year
S5 S1 with Higher 86.5 1.9 7.9 2.8 -2.2 -10.1
real GDP growth
rate and Lower
Primary Deficit
Note: Projections start from 2023-24. For 2022-23, estimates of the level of debt are from the Economic
Survey.

Figure 12: Evolution of General Government Debt-to-GDP Ratio under Different

Scenarios

100
93.4

90 88.7
83.9
80 81.0

70 76.4
Baseline
S2
60
S3
S4
50
S5
40

Note: The estimate of debt for 2022-23 has been taken from Chapter 3 of Economic Survey 2022-23.

The RBI in its reports on state finances and the IMF in its Article IV Reports warn
of the impact of contingent liabilities on debt sustainability. RBI (2023) observes that
“State government guarantees increased sharply by end-March 2021, which has
implications for their debt sustainability.” IMF (2022a) reports that “[f]iscal risks reflect
higher macroeconomic uncertainty, particularly from the external sector, and
contingent liabilities from public sector banks and electricity generation corporations.”
Past contingent liabilities have been on account of Air India, public sector banks,
electricity distribution companies, public-private partnerships (PPPs) in infrastructure
18 | India Policy Forum 2023

provision, and other state-owned enterprises (SOEs). They materialized when

governments assumed the debts of companies, rescuing and recapitalizing them.
Blanchard et al. (2021) apply Equation 1 to historical data for India and take the
difference between actual and implied changes as the realization of contingent
liabilities. They find these to have been substantial. Alternatively, the Office of the
Comptroller and Auditor General and RBI have attempted to estimate contingent
liabilities directly; as of March 2021 they put these at 2.5 percent of GDP for the Central
Government and 3.7 percent of GDP for the States (Figure 15).
We assume that contingent liabilities will be taken onto the budget at a rate of
one percentage point of GDP each year for the next five years. Unsurprisingly, this will
add another 6.9 percentage points of GDP to the debt, taking it above 93 percent of GDP
under baseline assumptions.16
The bottom line is that even under an exceptionally favorable scenarios, General
Government debt to GDP is unlikely to decline below 80 percent on current policies.
And less benign scenarios are possible.
Figure 13: Contingent Liabilities
Contingent Liabilities of the Center Contingent Liabilities of the States

Center's Contingent Liabilities, % of GDP Contingent Liabilities of the States, % of

GDP
4.0 4.0
3.5 3.5
Per cent of GDP

Per cent of GDP

3.0 3.0
2.5 2.5
2.0 2.0
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0

Source: Financial Audit Reports on Account of Union Government, CAG, Union Budget Statements, and
RBI.

Table 3: GDP-Growth-Rate-Interest-Rate Differential and Accumulation of Public

Debt
Average Average Debt Level Debt Level Change in
g-r Primary in 1981-82 in 2019-20 Debt-to-GDP
Deficit
1981-82 to 1.9 2.9 48.8 75.7 26.9
2019-20

16Were such liabilities instead taken onto the budget at a rate of two percentage points of GDP, this would
straightforwardly add 13 percentage points of GDP to the debt, and so forth.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 19

Central Government
We use Equation 1 to project public debt for the central government in scenarios
similar to those for General Government. In the baseline, for the next five years GDP
growth, the real interest rate, and the primary deficit will be at the same levels as their
respective averages from 2013-14 to 2022-23 (5.7 percent, 2.6 percent and 1.7 percent;
see Table 4). This yields a stable debt-to-GDP ratio (actually, a small reduction of about
0.3 percentage points over the period, as shown in Table 5). In the second scenario
where we assume faster GDP growth, debt to GDP declines by a cumulative 5.6
percentage points. A similar reduction is projected in the third scenario of a lower
primary deficit. The only scenario in which the debt of the central government is
projected as rising relative to GDP is when contingent liabilities materialize.
Table 4: Average Values and Standard Deviations of the Key Parameters for
Central Government
Ten-year average Five-year average
(2013-14 to 2022-23) (2018-19 to 2022-23)
Mean Std dev Mean Std dev
Nominal GDP growth (γ) 10.7 5.3 10.0 7.9
Deflator growth (π) 4.7 2.3 5.6 2.7
Real GDP growth (g) 5.7 4.3 4.1 5.9
Nominal interest rate (i) 7.3 0.9 6.8 0.8
Real interest rate (r) 2.6 2.6 1.2 3.0
Primary deficit (pd) 1.7 1.8 2.8 2.0
Growth-interest differential 3.2 4.8 2.9 7.2
(g-r)

Table 5: Evolution of Debt-to-GDP Ratios

Scenarios Scenario Debt Primary Real Real Change in Cumulative
Description level in Deficit GDP Interest Debt in change in
2022-23 (pd) growth Rate first year Debt in next
(bt-1) (g) (r) (∆ 𝒃𝒕 ) five years
Baseline Baseline: Past 10- 60.5 1.7 5.7 2.6 -0.1 -0.3
(S1) year averages
S2 Higher real GDP 60.5 1.7 7.9 2.6 -1.2 -5.6
growth rate
S3 Lower Primary 60.5 0.9 5.7 2.6 -0.9 -4.4
Deficit
S4 B1 plus 60.5 1.7 5.7 2.6 0.4 2.1
contingent
liabilities
absorbed (0.5
percentage point
of GDP) each year
S5 Higher real GDP 60.5 0.9 7.9 2.6 -2.1 -9.6
growth rate and
Lower Primary
Deficit
20 | India Policy Forum 2023

Figure 14: Evolution of Central Government Debt-to-GDP Ratio

Projetions of Gross Central Government Debt/GDP

70
62.6
60 60.2
56.1
54.9
50 Baseline
S2 50.9
40 S3
S4
30
S5
20

Source: CEIC and RBI State Finances Reports (multiple years). Projections are for 2023-24 onwards.

State Governments
For purposes of projection, we take the debt-to-GDP ratio, growth of nominal
GDP, rate of inflation, and growth of real GDP as identical for the Center and the States.
However, primary deficits and interest rates differ (Table 6). In most scenarios
including in the baseline, the debt-to-GDP ratio of the States is projected to increase
(Table 7). By implication, the projected increase in General Government debt can be
primarily (even entirely) attributed to the increase in debt to GDP ratio of the States.
The contrast reflects higher interest rates. States pay about 0.5 percent higher
interest than the Center. As a result, g-r is less favorable. This is why the States’ debt has
accumulated faster than the Center’s despite lower primary deficits.

Table 6: Average Values and Standard Deviations of the Key Parameters for State
Government
Ten-year average Five-year average
(2013-14 to 2022-23) (2018-19 to 2022-23)
Mean Std dev Mean Std dev
Nominal GDP growth (γ) 10.7 5.3 10.0 7.9
Deflator growth (π) 4.7 2.3 5.6 2.7

Real GDP growth (g) 5.7 4.3 4.1 5.9

Nominal interest rate (i) 7.8 0.8 7.4 0.7
Real interest rate (r) 3.1 2.5 1.8 2.9
Primary deficit (pd) 1.3 0.5 1.4 0.6
Growth-interest 2.6 4.8 2.3 7.2
differential (g-r)
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 21

Table 7: Evolution of the State Government’s Debt-to-GDP Ratios

Scenarios Scenario Debt Primary Real Real Change in Cumulative
Description Level in Deficit GDP Interest Debt in Change in
2022-23 (pd) Growth Rate First Year Debt in the
(bt-1) (g) (r) (∆ 𝒃𝒕 ) Next Five
Years
Baseline Baseline: Past 28.0 1.3 5.7 3.1 0.6 2.9
(S1) 10-year
averages
S2 Higher real GDP 28.0 1.3 7.9 3.1 0.1 0.3
growth rate
S3 Lower Primary 28.0 1.0 5.7 3.1 0.3 1.6
Deficit
S4 S1 plus 28.0 1.3 5.7 3.1 1.3 6.4
contingent
liabilities
absorbed (0.5
percentage
point of GDP)
each year
S5 S1 with higher 28.0 1.0 7.9 3.1 -0.2 -1.0
real GDP growth
rate and Lower
Primary Deficit

Figure 15: Evolution of the State Government’s Debt-to-GDP Ratio

Projections of Gross State Government Debt/GDP

40
Baseline
35 34.4
S2
S3 30.8
30 S4 29.5
S5 28.2
25 27.0

20

15

10

Source: State Finances Report, RBI (2013-14 up till 2022-23). State debt refers to Total Outstanding
Liabilities of States including loans and advances from the Central Government. Projections are for 2023-
24 onwards.
 22 | India Policy Forum 2023

A few states such as Gujarat and Maharashtra have managed their public
finances well.17 Their debts have increased least since 2014-15, remaining below 25
percent of state GDP (Table 8). At the other end of the spectrum are Punjab, Rajasthan,
and Kerala, whose debts have increased on average by 12 percentage points of GDP
since 2014-15 and exceeded 40 percent of state GDP at the end of 2020-21.
Table 8: Levels and Changes in Debt Levels across the Indian States
(Percent of Their Respective GDP)
Change in Change in Debt
Debt-to-
Debt-to-GDP in Debt-to-GDP Debt to GDP to GDP b/w
GDP in
States 2019-20 in 2020-21 b/w 2019-20 2020-21 and
2014-15
(2) (3) and 2014-15 2014-15
(1)
(4=2-1) (5=3-1)
Punjab 31.7 42.8 48.9 11.1 17.2
West Bengal 38.6 37.8 43 -0.8 4.4
Rajasthan 24.1 35.4 40.3 11.3 16.3
Kerala 28 32.9 40.3 4.9 12.3
Bihar 29 33.2 38.7 4.2 9.8
Andhra Pradesh 23.4 33.2 36.9 9.9 13.5
Uttar Pradesh 31 32.3 36.6 1.3 5.5
Jharkhand 20 30.5 36.3 10.4 16.3
Goa 29.5 30.2 35.2 0.7 5.7
Haryana 21.2 29.9 33.2 8.7 12
Tamil Nadu 17.3 26.5 31.5 9.2 14.2
Madhya Pradesh 22.7 22.8 30.2 0.1 7.6
Chhattisgarh 14.1 25 28.8 10.9 14.7
Telangana 14.4 23.7 28.8 9.4 14.4
Odisha 16.2 26.7 26.4 10.5 10.2
Karnataka 17.3 21 25.9 3.7 8.6
Gujarat 22 20.4 22.2 -1.6 0.2
Maharashtra 18.1 18.1 20.9 0 2.8
Note: GDP refers to the GDP figures of the respective States.

We compare some key variables across these two sets of States in Table 9. We
define a dummy variable that equals 1 for States with an above-median increase in debt
to GDP, and 0 for those below the median.18 The results show that States with large
increases in debt ratios had primary deficits and contingent liabilities more than twice
those of States with small increases. Although they also had slightly slower GDP growth,
this differential was not significant. Inflation and interest rates did not differ across the

17In this section, we focus on the 18 largest Indian states. Erstwhile Special Category States and the Union
Territories are not included.
Similar results are obtained if instead of comparing the States which are below and above the median
18

we compare the values of these variables for the top one-third of the States for the increase in debt-to-
GDP ratio with the bottom one-third of the States.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 23

two classes of States. Recall how we noted the absence of an interest rate differential
above.
Table 9: Comparing States with a Large Increase in the Debt-to-GDP Ratios with
Those with a Smaller Increase in the Debt-to-GDP Ratio

(1) (2) (3) (4) (5) (6) (7)

Real Inflation Nominal Primary Capital Contingent Debt to
GDP Interest Deficit Expenditu Liabilities GDP
Growth Rate GDP re to GDP to GDP

Dummy =1 -0.73 0.01 0.05 1.23*** -0.34 2.57* 1.53

for above (1.08) (0.04) (1.04) (5.13) (0.60) (2.19) (0.46)
median
increase in
debt
Constant 6.01*** 3.40*** 7.67*** 0.91*** 4.17*** 1.94* 26.79**
(12.67) (14.13) (225.9) (5.38) (10.27) (2.34) *(11.4)
No. of 18 18 18 18 18 18 18
observation
s
Note: Median change in the debt to GDP ratio in 2014-15 and 2020-21 is 11.1. The dummy takes a value 1
for states where the increase in the debt-to-GDP ratio between 2014-15 and 2020-21 exceeded 11.1, and
a value 0 for the states with a below median increase. Variables are averaged over the period. Inflation is
the rate of growth of the state-specific GDP deflator. t statistics are in parentheses *, **, *** refer to
significance at 10, 5, and 1 percent levels, respectively.

In Tables 10-12 and Figures 16-17, we present debt sustainability analyses for
Punjab and Gujarat, states representative of those with high and low debt/GDP ratios.
Debt is unlikely to stabilize and may even increase further in Punjab. In Gujarat, on the
other hand, debt is projected to decline as a share of state GDP in all scenarios.
In sum, States in a less favorable position are likely to face graver problems of
debt sustainability, owing to slower economic growth, larger contingent liabilities, and
higher interest rates for States overall. Given projections of a stable debt-to-GDP ratio
for the Central Government, the behavior of these problem States constitutes the main
threat to debt sustainability.
A question is why these problem states have had so much room to run. One
answer is that, as we have already noted, borrowing costs do not vary across States.
Despite different debt levels (and different projected primary deficits and contingent
liabilities), Gujarat and Punjab issue at equivalent interest rates. This reflects the RBI’s
efforts to equalize interest rates across States.19 De facto, this results in States in better
fiscal health subsidizing those whose health is worse. It relaxes market discipline on
errant States.

19 We have not found much scholarly literature on this issue. Practitioners have pointed us to the
following: (1) SDLs of different states are all eligible for the RBI’s repo facility subject to the same haircut;
(2) banks are allowed to mark to market different states’ SDLs identically; (3) all SDLs held by banks
carry zero risk weights; (4) the RBI provides states with short term loans up to a specified percentage of
its borrowing needs; (5) at the end of the day SDLs are covered by a broader central bank and
government guarantee. Verifying these hypotheses and identifying their relative importance is an
important topic for future research.
24 | India Policy Forum 2023

The horizontal devolution of taxes among States, awarded by the Finance

Commission (FC) every five years, also does not provide incentives for fiscal rectitude.20
FCs are mandated to allocate more resources to States with larger revenue deficits,
which is an obvious source of moral hazard. The 15th FC included tax effort (the ratio of
per capita own tax revenue to per capita state GDP in the previous three years) as one
criterion in its larger devolution matrix, but this did not solve problems on the
expenditure side. Some States keep significant expenses and liabilities off budget. FCs
do not have data, mechanisms, or a clear mandate to estimate contingent liabilities. The
15th FC was asked to recommend performance incentives for States in areas like the
power sector and solid waste management. But FCs have not been asked to consider
overall fiscal prudence or contingent liabilities (except indirectly through reforms of the
power sector) when recommending allocations.
FCs are dissolved immediately after they report to the President. There is no
parallel institution or body to monitor States’ finances and assess whether they have
departed from the course projected by the FC. Thus, it would be desirable to establish a
permanent Fiscal or Expenditure Council to monitor state finances, assess the quality of
data and forecasts, and inform the public of the fiscal stance and debt sustainability of
different States.21

Table 10: Average Values and Standard Deviations of the Key Parameters for
Punjab
Ten-year Average Five-year Average
(2013-14 to 2022-23) (2018-19 to 2022-23)
Mean Std dev Mean Std dev
Nominal GDP growth (γ) 8.0 3.6 6.3 4.5
Deflator growth (π) 3.0 1.1 2.5 1.2
Real GDP growth (g) 4.8 3.0 3.7 4.0
Nominal interest rate (i) 7.8 0.7 7.4 0.6
Real interest rate (r) 4.8 0.9 4.9 1.1
Primary deficit (pd) 1.5 3.0 0.7 1.0
Growth-interest differential 0.0 3.2 -1.2 4.0
(g-r)
Contingent Liabilities as % of 8.9 7.0 3.6 2.4
GDP

20The Finance Commission (FC) is a constitutional body formed by the President of India every five years
to recommend the devolution of revenue to the States and its horizontal distribution. The sixteenth FC is
slated to be announced later in 2023-24. An earlier literature (von Hagen and Eichengreen 1996)
suggests that vertical fiscal imbalances (where the center raises taxes but states are responsible for
spending programs) provide states with incentives to run larger deficits, in the expectation of
consequently receiving larger transfers from the center. To the extent that tax reforms have located more
revenue-raising capacity at the center, this vertical fiscal imbalance and associated deficit bias may have
grown more acute.
21 See Rao (2018) on Fiscal Council and Debroy and Sinha (2023) on Expenditure Council.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 25

Table 11: Evolution of Debt-to-GDP Ratio for Punjab

Scenarios Scenario Debt Primary Real Real Change in Cumulative
Description Level in Deficit GDP Interest Debt in Change in
2022-23 (pd) Growth Rate the First Debt in the
(bt-1) (g) (r) Year Next Five
(∆ 𝒃𝒕 ) Years

Baseline Baseline: Past 10- 47.8 1.5 4.8 4.8 1.5 7.6
(S1) year averages
S2 Higher real GDP 47.8 1.5 6.4 4.8 0.8 4.1
growth rate
S3 Lower Primary 47.8 0.04 4.8 4.8 0.0 0.2
Deficit
S4 Contingent 47.8 1.5 4.8 4.8 1.5 13.2
liabilities absorbed
(1.12 percentage
point of GDP) each
year
S5 Higher real GDP 47.8 0.04 6.4 4.8 -0.6 -3.1
growth rate and
Lower Primary
Deficit

Figure 16: Evolution of Debt to GDP Ratio for Punjab

Projections of Punjab Debt/GDP

70

60

50

40

30

20

10

Baseline (S1) S2 S3 S4 S5

Source: CEIC, EPWRF, and RBI State Finances Reports (multiple years). Projections are from 2023-24
onwards.
 26 | India Policy Forum 2023

Table 12: Average Values and Standard Deviations of the Key Parameters for
Gujarat
Ten-year Average Five-year Average
(2012-13 to 2021-22) (2017-18 to 2021-
22)
Mean Std dev Mean Std dev
Nominal GDP growth (γ) 12.4 5.0 11.1 6.8
Deflator growth (π) 3.6 2.1 3.6 2.6
Real GDP growth (g) 8.5 3.7 7.2 5.0
Nominal interest rate (i) 7.9 0.9 7.3 0.7
Real interest rate (r) 4.3 2.3 3.7 2.9
Primary deficit (pd) 0.4 0.3 0.4 0.4
Growth-interest differential (g-r) 4.2 4.6 3.5 6.4
Contingent Liabilities as % of GDP 0.4 0.2 0.2 0.1
Contingent Liabilities as % of GDP 0.2
(as of 2021-22)

Table 13: Evolution of Debt to GDP Ratio for Gujarat

Scenarios Scenario Debt Primary Real Real Change in Cumulative
Description level in Deficit GDP Interest Debt in Change in
2021-22 (pd) Growth Rate the First Debt in the
(bt-1) (g) (r) Year Next Five
(∆ 𝒃𝒕 ) Years

Baseline Baseline: Past 10- 19.9 0.4 8.5 4.3 -0.4 -1.7
(S1) year averages
S2 Higher real GDP 19.9 0.4 10.3 4.3 -0.7 -3.1
growth rate
S3 Lower Primary 19.9 0.2 8.5 4.3 -0.5 -2.4
Deficit
S4 Contingent 19.9 0.4 8.5 4.3 -0.4 -1.7
liabilities absorbed
(0.04 percentage
point of GDP) each
year
S5 Higher real GDP 19.9 0.2 10.3 4.3 -0.8 -3.8
growth rate and
Lower Primary
Deficit
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 27

Figure 17: Evolution of Debt to GDP Ratio for Gujarat

25

20
18.2
17.5
Baseline (S1) S2 16.8
16.1
15
S3 S4

S5
10

Source: CEIC, EPWRF, and RBI State Finances Reports (multiple years). Projections are from 2022-23
onwards.

5. Sharp Reductions in Debt

We focus now on past episodes of debt consolidation. We define consolidations
as episodes when the general government debt ratio fell consistently for at least five
consecutive years. Using data starting in 1990, this yields two consolidation episodes:
1991-92 to 1997-98; and 2004-05 to 2012-13. Debt reduction was 6.7 percent of GDP in
the first episode, 16.9 percent in the second.22
The first episode followed a balance-of-payments crisis during which India
signed up for an IMF program.23 The IMF loan was conditioned on fiscal consolidation
designed to reduce the Central Government’s deficit from 8.5 percent of GDP in 1990-91
to 5 percent in 1992-93 (Chopra and Collyns 1995). This decline was premised on lower
recurrent and capital expenditure. Inflation accelerated (the average annual rate of GDP
inflation was about 10 percent), reflecting exchange rate depreciation in 1991-92.24
Consolidation proceeded despite the fact that growth was slower than in control years,
and despite the fact that tax revenues also grew more slowly.
2004-05 to 2012-13, in contrast, was marked by faster growth, especially
between 2004-05 and the Global Financial Crisis. Tax and administrative reforms
yielded dividends in the form of higher revenues. In this second episode unlike the first,
the decline in the primary deficit was underpinned by higher tax revenue rather than by
lower expenditure; capital expenditure, in particular, was protected.

22This is in contrast to Eichengreen and Esteves (2022), who also required the debt ratio to fall by at least
10 percentage points in order for it to qualify as a consolidation episode.
23 India signed the IMF program agreement in October 1991 and exited it in June 1993.
24 The exchange rate was first sharply devalued from its artificially appreciated levels, and was later
floated (managed float).
 28 | India Policy Forum 2023

Of the reduction of 17 percentage points in the debt to GDP ratio, nearly 10

points were accounted for by the States. These initiatives by State Governments were
supported by a Debt Swap Scheme (DSS) in 2002-03/2004-05 and a Debt Consolidation
and Relief Facility (DCRF) in 2005-06/2009-10. Under DSS, States could prepay
expensive loans from the Central Government and instead raise cheaper loans from the
market. Under DCRF, debt relief was provided to the states through debt reduction,
rescheduling of debt and lower interest rates, conditional on enacting and
implementing Fiscal Responsibility and Budget Management legislation. Debt relief was
linked to the improvement in fiscal performance (assessed in terms of the reduction in
revenue deficits). This experience is a reminder that debt consolidation at the State and
Central Government levels is not independent; the Central Government can play an
important role in providing incentives to the States.

Table 14: Episodes of Debt Consolidation

Consolidation Duration Initial Terminal Change in debt to
(Years) Debt Debt GDP ratio (∆ 𝒃𝒕 )
1991-92 to 7 74.2 67.5 -6.7
1997-98
2004-05 to 9 83.6 66.7 -16.9
2012-13

In Table 15 we regress a set of outcome indicators on a dummy variables equaling

1 in years of debt consolidation, defined as above, and 0 otherwise. The results show that
inflation is more than twice as high during consolidation episodes, while the primary
deficit is about 1 percentage point lower. Higher inflation might be thought to make for a
lower real interest rate, but the real interest rate is significantly lower than in the control-
group years only in the second consolidation episode, when its low level was largely
attributable to the low level of nominal rates post-Global Financial Crisis. On average,
growth is not significantly different than in normal (non-consolidation) years.

Table 15: A Comparison of Key Variables during the Consolidation Episodes and
Normal Years
(1) (2) (3) (4) (5) (6)
Inflation Growth Real Nominal Real Growth Primary
Interest Interest – Real Deficit
Rate Rate Interest
Rate
Dummy =1 for 4.49*** -0.48 -0.20 4.29*** -0.28 -1.14*
1991-92 to 1997-98 (4.44) (0.37) (0.18) (6.17) (0.18) (1.63)
Dummy=1 for 2004- 2.90*** 1.08 -3.54*** -0.63 4.61*** -1.29**
05 to 2012-13 (3.12) (0.91) (3.46) (0.99) (3.30) (2.02)
Constant 4.89*** 5.78*** 3.54*** 8.43*** 2.24*** 3.19***
(8.93) (8.3) (5.89) (22.44) (2.72) (8.44)
No. of Observations 33 33 33 33 33 33
Note: Data are from 1990-1991 to 2022-2023. t statistics in parentheses. *, **, *** refer to significance at
10, 5, and 1 percent levels, respectively.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 29

Table 16 compares total revenue, tax revenue, total expenditure, and capital
expenditure across consolidation episodes and normal years. While the primary deficit
was lower in both episodes compared with control years, its reduction was achieved in
different ways. In 1991-92/1997-98, a lower primary deficit was attained by
compressing expenditure, including capital expenditure. The consequences were not
growth friendly. In the second episode, in contrast, the decline in primary deficit was
obtained mainly through higher revenue collection, including by raising tax revenue. The
result was at least growth neutral.

Table 16: A Comparison of Key Variables during the Consolidation Episodes and
Normal Years

(1) (2) (3) (4)

Total Tax Total Capital
Revenue Revenue Expenditure Expenditure
Dummy =1 for -0.75 -1.13** -1.64** -0.74**
1991-92 to 1997- (1.54) (2.18) (2.22) (2.01)
98
Dummy=1 for 1.08** 0.79* -0.33 0.30
2004-05 to 2012- (2.42) (1.67) (0.49) (1.09)
13
Constant 18.91*** 15.67*** 26.95*** 3.42***
(71.98) (56.24) (67.57) (20.83)
No. of Observations 33 33 33 29
Note: Data are from 1990-1991 to 2022-2023. Capital expenditure data is available from 1994-1995
onwards. t statistics are in parentheses. *, **, *** refer to significance at 10, 5, and 1 percent levels,
respectively.

In sum, consolidation is easier when debt is reduced by both the Center and the
States. Contrary to assertions in the literature (Alesina, Favero and Giavazzi 2019), it is
not obvious that consolidations achieved by cutting spending has worked better than
consolidations achieved by raising revenues. The Alesina et al. result is obtained from
data for advanced countries, where spending and taxes are arguably too high. As we
showed above, international comparisons suggest that tax revenues are too low in India
(not that spending is too high). This is a reminder of the need to tailor advice to context.
In neither case was it possible to maintain the lower levels of debt achieved in
the consolidation episode. In both cases more than half the reduction was reversed
subsequently. After the 7 percentage point reduction in debt/GDP from 1991-92 to
1997-98, debt rose from 68 percent of GDP in 1998-99 to 85 percent in 2003-04, more
than fully reversing its preceding fall. Debt rose despite an acceleration in GDP growth
from 5.3 percent to 5.9 percent per annum. This rise was attributable to an increase in
primary deficit from 2 percent to 3.3 percent and to some decline in inflation that
translated into higher real interest rates.
The reduction of debt achieved from 2004-05 to 2012-13 was partially reversed
in 2013-14/2019-20, when the debt ratio rose from 67 percent to 75 percent. The
period was marked by the same primary deficit ratio as in the preceding consolidation
30 | India Policy Forum 2023

period, a slight deceleration in growth (from 6.9 percent to 6.7 percent), and once more
a fall in inflation that translated into higher real rates.
These post-consolidation experiences speak to the limited role of inflation in
debt consolidation. It is tempting to think that a country whose debt is at long tenors
can inflate away a significant portion. Inflation was higher during both consolidation
episodes than in other periods, consistent with this presumption. Inflation worked to
reduce debt, especially in the second of the two episodes, by helping to depress the real
interest rate. But once the burst of inflation passed and inflation came down, this effect
was reversed. Arslanalp and Eichengreen (2023) analyze annual data on inflation and
debt for a panel of countries stretching to 1800. They estimate the relationship using
local projections and simulate the effect of an inflation shock. Consistent with what we
find here for India, they show that the impact of an inflation shock on the debt ratio is
temporary. That impact effect is reversed over time as interest rates, maturities and
spending begin to adjust.

6. Costs and Risks

What are the costs and risks of India’s high debt and deficits? In the introduction
to our paper we identified five.
First, interest payments absorb resources, limiting their availability for other
economic and social purposes. Interest payments exceed 25 percent of general
government revenues. This share is roughly twice the emerging market and developing-
country average. At 5 percent of GDP, they are again twice the emerging market
average (Figure 3 above). This difference reflects not high borrowing costs but rather
high levels of debt. In contrast, government expenditure as a share of GDP is in line with
other emerging markets. With interest payments absorbing a larger share of revenues,
less is left for other needs. As noted above, the government spends more on interest
than on education and health combined. Its interest payments exceed its capital
expenditure.
Second, and relatedly, available fiscal resources leave no room for meeting
emerging priorities, notably climate change abatement and adaptation and the green
transition. McKinsey (2022) estimates that, owing to its exposed geography, India will
have to invest half as much again as advanced economies as a share of GDP to maintain
its economic development in the face of climate change and in order to build low-carbon
infrastructure. According to its Net Zero 2050 scenario, India will have to spend 11
percent of GDP between now and 2050 on decarbonization and low-carbon growth,
compared to the global average of 7.5 percent. This reflects elevated heat exposure of
urban residents in particular, as well as the need for extensive spending on low-
emissions assets and enabling infrastructure. Not all of this investment must be
financed by government revenues and borrowings, of course. Global investment funds,
oil and gas majors, foreign utilities, Indian conglomerates, government companies, and
pension funds are all taking equity stakes in Indian renewable energy projects (Jaiswal
and Gadre 2022). Wind and solar power companies issue debt to finance their
investments, borrowing from domestic and international banks and development
finance institutions. In 2019-21 some 50 percent of their debt financing is sourced
overseas, a growing share in the form of green bonds. This said, regulatory risk
(changes in tariffs and rates), planning risk (mis-estimation of power generation
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 31

capacity) and extreme weather risk (including from climate change) make a significant
public-finance contribution unavoidable.
Third, debt dynamics leave little room for responding to shocks, such as
declining rates of domestic and global growth. India was not strongly constrained in
responding to COVID-19; it reacted with a fiscal stimulus of 20 trillion rupees, or
roughly 9 percent of GDP. About a third of this was above-the-line measures (spending
on social protection and health care and foregone revenues); the remaining two thirds
of below-the-line measures involved various forms of business support (IMF 2022b).
The combined response, while smaller than in the advanced economies (Hudson et al
2022), was nonetheless substantial. Mishra and Patel (2022) argue that the resulting
increase in debt has put upward pressure on interest rates, although our own analysis
fails to find much evidence of this (see Appendix C). Be this as it may, at some point
responding in this way to shocks will begin to show up in interest rates, especially as
regulations encouraging investments in bonds by insurance companies, provident funds
and banks are progressively relaxed. At some point, this will begin to throw debt
sustainability into doubt. Conversely, maintaining debt sustainability in the face of such
shocks will leave the government countercyclically constrained, amplifying cycles.
Fourth, high government debt creates the potential for financial stability risks.
For the moment, such risks remain limited.25 Banks are required to hold government
securities in order to satisfy their Statutory Liquidity Ratios (SLR); they are required to
hold liquid assets, including government bonds, of a specified minimum percentage of
deposits.26 Risks to their balance sheets can therefore develop with the repricing of
these assets when interest rates rise. But SLR has been cut from 38 percent in the early
1990s to 18 percent in recent years (see Appendix D). As a result, banks now hold a
smaller share of their assets in government securities. The RBI has also mandated that
banks hold highly liquid assets as Investment Fluctuation Reserves (IFR), intended as a
buffer against fluctuations in their investment portfolios. As of December 2022, banks
held more than the mandated level of reserves. Moreover, public sector banks are no
longer more exposed to interest rate risk than private banks or foreign banks. All this
has limited portfolio repricing risk, though by no means eliminating it.
But this reduction in mandated bank holdings means, in turn, that government is
relying on a more diverse set of investors to hold its debt. The share of insurance
companies, provident funds and other non-bank investors in Central Government
securities has increased from 20 percent in 1990-91 to 46.6 percent in 2021-22 (as
noted earlier).27 For the moment, insurance companies and provident funds are
required by regulation to hold roughly 50 percent of their investable funds in
government securities. But if regulations compelling the insurance companies and
provident funds to hold government bonds are further relaxed and/or domestic savings
decline, Central and State governments will be forced to place additional debt, including

25Such is our judgment. In addition, the RBI in its biannual Financial Stability Reports does not flag the
holding of government securities, or changes in the interest rate, as significant risks to Indian banks.
26Business Standard, Aug 16, 2022, notes some of the developments which have led to a decline in the
interest rate risk for the Indian banks: [Link]
in-better-position-to-manage-rbi-s-interest-rate-risks-122081601071_1.html
27The RBI also holds a larger fraction of public debt than in the past. The RBI’s share has increased from
7.8 percent of the Central Government debt in 2007-08 to 16.6 percent in 2021-22. The corresponding
numbers are smaller for the General Government.
32 | India Policy Forum 2023

short-term debt, with foreign investors, in the manner of other emerging markets. If
externally-held debt is denominated in foreign currency, as in other countries, this will
increase the sector’s currency mismatch, creating debt-servicing and financial
difficulties when exchange rates move. Even if India succeeds in placing rupee-
denominated debt with foreign investors, this nevertheless raises the risk of a capital-
flow reversal, an investor strike and a bond-price collapse, since the currency mismatch
will now be on foreign balance sheets, encouraging foreign investors to flee at the first
sign of trouble (Carstens and Shin 2019).
For the moment, India may be able to place most of its debt with “patient”
domestic investors. But if this becomes less true going forward, run risk – and volatility
– will rise.

7. Conclusion
Our central conclusion is that India’s general-government-debt-to-GDP ratio,
which is high by emerging market standards, is unlikely to decline significantly in the
next five years. In the best-case scenario, it might fall from its current level of some 90
percent of GDP, which is half again as high as the emerging market average, to 80
percent of GDP, where it would be 30 percent again as high. But less rosy scenarios are
also possible.
What might be achieved with more ambition? In purely mathematical terms,
India could bring down its debt to 70 percent of GDP through a combination of lower
primary deficits, higher inflation, and faster GDP growth. A percentage point increase
each in growth and inflation and a percentage point reduction in the primary deficit
would reduce public debt to 70 percent of GDP in five years. The requisite changes
could be achieved through an amalgam of the following factors:
• Raising additional revenue through higher tax, non-tax, and privatization
receipts. Along with better tax administration and digitalization, recent tax
reforms have succeeded in boosting revenue growth. Yet in a fast-growing
economy, where nominal GDP has been growing on average at 11-12 percent, the
rate of tax-revenue growth still has not exceeded that of GDP growth, in contrast
to other fast-growing emerging markets.
• Continuing to re-orient spending toward capacity- and infrastructure-enhancing
investment that promises to further boost GDP and revenues.
• Limiting contingent liabilities, which have been a chronic problem at the state
level.
But imagining sharp changes along these lines borders on wishful thinking.
Meanwhile, economic and social development will require additional spending on
health and education. Government will have to contribute significantly to the country’s
decarbonization and climate-change-adaptation investments, which are large by
international standards. Eventually, interest rates will adjust upward in response to
inflation, eliminating any favorable debt-consolidation effects. As a result of these
factors, India will almost certainly be living with high public debt for years to come.
All this said, the country faces no immediate crisis of debt sustainability. Our
baseline scenario does not point to exploding debt ratios. For the moment, rollover risk
is limited. Most public debt securities are held by banks, insurance companies and
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 33

other patient domestic investors. It is denominated in rupees. Little is at short

maturities or floating rates.
But the preceding does not mean that the country’s relatively high public debt is
without costs. Devoting a large share of financial resources to servicing debts leaves the
Central Government and States with fewer resources for other investments. At some
point, it will leave less room for responding to shocks. Even if volatility is limited now,
this could change with financial liberalization and deregulation. The bottom line is that
India’s high public debt leaves no room for missteps.
34 | India Policy Forum 2023

References
Acharya, Viral V. 2020. “Understanding and Managing Interest Rate Risk at Banks,”
Chapter 3 in Quest for Restoring Financial Stability in India, Sage Publications, pp. 197-
215.
Alesina, Alberto, Carlo Favero and Francesco Giavazzi. 2019. Austerity: When it Works
and When it Doesn’t Work, Princeton: Princeton University Press.
[Link]
ely%20controversial,times%20and%20at%20all%20costs.
Arslanalp, Serkan and Barry Eichengreen (2023), “Living with High Public Debt,” paper
prepared for the Federal Reserve Bank of Kansas City Jackson Hole Symposium
(August).
Arslanalp, Serkan, and Takahiro Tsuda. 2014. “Tracking Global Demand for Emerging
Market Sovereign Debt,” IMF Working Paper WP/14/39, Washington, D.C.: IMF.
Blanchard, Olivier, Josh Felman, and Arvind Subramanian, 2021, “Does the New Fiscal
Consensus in Advanced Economies Travel to Emerging Markets?” Policy Brief 21-7,
Washington, D.C.: Peterson Institute of International Economics.
Brunnermeier, Markus., Luis Garicano, Philip R. Lane, Marco Pagano, Ricardo Reis, Tano
Santos, David Thesmar, Stijn Van Nieuwerburgh, and Dimitri Vayanos, 2016. “The
Sovereign-Bank Diabolic Loop and ESBies,” American Economic Association Papers and
Proceedings 106: pp.508–512.
Carstens, Agustin and Hyun Song Shin. 2019. “Emerging Markets Aren’t Out of the
Woods Yet,” Foreign Affairs (15 March),
[Link]
yet?gpp=vbUtssg/sGSAFl22Ax97dDpZckdNM3FZT0d0WU5LS2VQUTRuNEszOHNDQVd
KRlJqTFNoR2h6TmVMSmRUQWR0b044R0Q0WnNWUDUyQkppNjFnOjVjYzA1YmFiOW
Q4OTEzZThkYzQ0ZGM3YTdkNGUwOTM0YzA2YTA1NzZhNjNkMjA4YjIyZDk2ODMyZDF
mZWM3ZTc%3D
Chopra, Ajai and Charles Collyns, 1995, “The Adjustment Program of 1991/92,”
Washington, D.C.: IMF.
Clearing Corporation of India Ltd. (CCIL). Various Editions. Monthly Reviews of the
Economy, published by CCIL on their website: CCIL Publications - Rakshitra
([Link]).
Debroy, Bibek and Aaditya Sinha. 2023. “An Expenditure Council Could Help Promote
Fiscal Prudence,” LiveMint (6 June).
Eichengreen, Barry and Rui Esteves. 2022. “Up and Away? Inflation and Debt
Consolidation in Historical Perspective,” Oxford Open Economics,
[Link]
Hudson, Callum, Benjamin Watson, Alexandra Baker and Ivalio Arsov. 2022. “The Global
Fiscal Response to COVID19,” Quarterly Bulletin Reserve Bank of Australia (June),
pp.100-109.
International Monetary Fund. 2022a. Article IV Review of India, Washington, D.C.: IMF.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 35

International Monetary Fund. 2022b. “Policy Responses to COVID-19,” Washington, D.C.:

IMF, last updated 2 July.
International Monetary Fund. 2023. IMF Fiscal Monitor, Washington, D.C.: IMF.
Jaiswal, Shantanu and Rohit Gadre. 2022. “Financing India’s 2030 Renewables
Ambition,” BloombergNEF (June).
McKinsey & Company. 2022. “The Net-Zero Transition: What It Would Cost, What It
Could Bring,” Washington, D.C.: McKinsey.
Ministry of Finance. 2016. “Status Paper on Government Debt,” Department of Economic
Affairs, Government of India, September.
——— . 2022. “Status Paper on Government Debt for 2020-21,” Department of
Economic Affairs, Government of India, April.
Mishra, Neelkanth, and Prateek Singh, 2018. “A Deep Dive into State Budgets in India,”
India Policy Forum, Sage Publications, Delhi.
Mishra, Prachi and Nikhil Patel. 2022.,“The Anatomy of Public Debt Reductions in India,”
Center for the Advanced Study of India, University of Pennsylvania (August).
Mishra, Sangita, Kirti Gupta, and Pushpa Trivedi, 2020. “Subnational Government Debt
sustainability in India: An Empirical Analysis,” RBI Working Paper 8/2020.
Patnaik, Ila and Radhika Pande. 2019. “Savings and Capital Formation in India,”
Working paper No. 271, National Institute of Public Finance and Policy, New Delhi.
Rao, Govinda M. 2018. “Public Finance in India: Some Reflections, Decision 45, pp.113-
127.
Reserve Bank of India, Various Years. “Public Debt Statistics,” Mumbai: Reserve Bank of
India.
——. Various Editions. “Public Debt Management Quarterly Reports,” Mumbai: Reserve
Bank of India.
——. Various Editions. “Public Debt Management Quarterly Reports,” Mumbai: Reserve
Bank of India.
——. 2022. RBI Bulletin, “State Finances, A risk Analysis,” Mumbai: Reserve Bank of
India.
——. 2023. “State Finances: A Study of Budgets of 2022-23,” Mumbai: Reserve Bank of
India.
Von Hagen, Juergen and Barry Eichengreen (1996), “Federalism, Fiscal Restraints and
European Monetary Union,” American Economic Review 86, pp.134-138.
36 | India Policy Forum 2023

Appendix A
Table A.1: Data

Indicator Source
Total Liabilities of General 1. CEIC. Estimate for 2022-23 has been taken from Chapter
Government 3 of the Economic Survey 2022-23. We have used the
words debt and liabilities interchangeably in the paper.
General Government Primary Calculated as the difference between Fiscal Deficit and Interest
Deficit Payments
General Government Interest CEIC
Payments, Total Revenue, Tax
Revenue, Non-Tax Revenue,
Total Expenditure, Revenue
Expenditure, Fiscal Deficit
Data for the center and the state For key fiscal variables, we considered data from RBI (Handbook
governments of Statistics on Indian Economy), State Finances Report, RBI,
CEIC Database, Economic Survey, and India Series in the
Economic and Political Weekly Research Foundation’s (EPWRF)
Database. While largely data for Center and State Government
match across these sources (with EPWRF’s estimates slightly
different than those provided by RBI and CEIC), the data for debt
does not add up to General Government Data. We calculated
Centre’s debt as the difference between General Government
Outstanding Liabilities and State’s Outstanding Liabilities net of
loans and advances from the Centre. For the other variables, we
used the RBI’s Database on Indian Economy and its State
Finances Report.
General Government Debt for Fiscal Monitor, IMF April 2023
Global and Emerging Markets
General Government Fiscal Overall balance of General Government, Fiscal Monitor, IMF
Deficit for Global and Emerging April 2023
Markets
Interest Payments on General Calculated as the difference between fiscal deficit and primary
Government Debt for Global and deficit, compiled from the Fiscal Monitor, IMF April 2023
Emerging Markets
Contingent Liabilities The data for contingent liabilities is available from 2008-09 till
2021-22 for the Central government and has been compiled
from various annual financial audits of the union government
conducted by the CAG. The outstanding guarantees data for
states is available from 1991-92 onwards and is published by
the State Finances Report. The data for Centre’s outstanding
guarantees and state’s outstanding guarantees are available for
2008-09 to 2021-22 to get the general government contingent
liabilities.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 37

Appendix A: Tax Buoyancy

Figure A.1: Tax Buoyancy

Tax Buoyancy (ratio)

2.0

1.5

1.0 1.2
1.1 1.0 0.8
0.9
0.5

0.0
1981-82
1982-83
1983-84
1984-85
1985-86
1986-87
1987-88
1988-89
1989-90
1990-91
1991-92
1992-93
1993-94
1994-95
1995-96
1996-97
1997-98
1998-99
1999-00
2000-01
2001-02
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22
2022-23
-0.5

Source: Author’s calculations. Note: Tax buoyancy is measured as the ratio of tax revenue growth relative
to nominal GDP growth for each of the years shown in the chart. If gross tax receipts increase more than
proportionally to an increase in nominal growth (that is, the ratio is greater than 1), then we say that the
tax system is buoyant. Horizontal dashed lines are for decadal averages from 1981-82 to 1989-90, 1990-
91 to 1999-2000, 2000-01 to 2009-10, and 2011-12 to 2019-20.

Figure A.2: Direct and Indirect Taxes

Direct Tax Revenue and Indirect Tax Revenue (% of GDP)

14
12.2
12 11.1
10.1 10.3
10
8
Direct 6.4
6 Indirect
4 5.3

2 3.2
2.5
0
1980-81
1981-82
1982-83
1983-84
1984-85
1985-86
1986-87
1987-88
1988-89
1989-90
1990-91
1991-92
1992-93
1994-95
1995-96
1996-97
1997-98
1998-99
1999-00
2000-01
2001-02
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22
1993-94

2009-10

Source: RBI. Direct taxes refer to taxes levied on property or income such as income tax and personal
property tax. Indirect taxes refer to. Indirect taxes are levied on goods and services such as GST and
customs and excise duties. The data for 2020-21 is a Revised Estimate and for 2021-22 is a Budget
Estimate. Horizontal dashed lines are for decadal averages from 1981-82 to 1989-90, 1990-91 to 1999-
2000, 2000-01 to 2009-10, and 2011-12 to 2019-20.
 38 | India Policy Forum 2023

Appendix B: Debt and Deficit of the Center and the States

Figure B.1: Public Debt of the Central and State Governments
Debt of the Central Government (% of GDP) Debt of the State Government (% of GDP)

70 35

59.9 60.5 30
60
29.4 28.0
54.7
25 23.8
22.0
50
46.1 20

40 15

10

1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13

2016-17
2018-19
2020-21
2022-23
2014-15
30
1990…
1992…
1994…
1996…
1998…
2000…
2002…
2004…
2006…
2008…
2010…
2012…
2014…
2016…
2018…
2020…
2022…

Source: State Finances Report, RBI (States). For 2021-22: Revised Estimates and for 2022-23 the Budget
Estimates for State’s Total Debt. The charts show the total outstanding liabilities of Central Government
and State Government as % of GDP. States total liabilities include the debt it owes to the Center. Centre’s
Total Debt has been calculated as the difference between general government total outstanding liabilities
and state government liabilities net of loans and advances from the Centre. Horizontal dashed lines are
the respective decadal averages.

Figure B.2: Debt Owed to the Central Government by the States and Net Public
Debt of the Central Government
Debt of the State Government (% of GDP) Net Debt of the Center and Gross Debt of the
States

States Total Debt and Debt owed 70

Public Debt, % of GDP
to Centre, % of GDP 60
58.5
48.4 48.8
50 44.9
35
30 40
25 30 22.0 23.8
20 20 29.4 28.0
15
10 10
5 0
0
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17

2020-21
2022-23
2018-19
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05

2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2022-23
2006-07

Centre (Net Debt)

Total Outstanding Liabilities of States
States (Total Debt)
Loans and advances from Centre

Source: State Finances Report, RBI (States). For 2021-22 we have used the Revised Estimates and for
2022-23 the Budget Estimates for State’s Total Debt. The chart in the left of the panel shows the total
outstanding liabilities of states as % of GDP as well as the component for loans and advances from Centre
as % of GDP. The chart in the right of the panel shows the states total debt and the central government net
debt calculated as the difference between general government total debt and states total debt. Horizontal
dashed lines are the respective decadal averages.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 39

Figure B.3: Share of Centre (Net Debt) and States (Total Debt) in Total Public Debt

Public Debt, share of Centre and States (%)

120

100

80

60

40

20

Centre (Net Debt) States (Total Debt)

Source: State Finances Report, RBI (States). Net public debt for the Centre has been calculated as the
difference between general government total outstanding liabilities and state’s total outstanding
liabilities. For 2021-22 we have used the Revised Estimates and for 2022-23 the Budget Estimates for
State’s Total Debt. In this chart, we are showing the share of Centre’s debt and State’s debt in the total
general government debt (following the specification in Figure B2)

Figure B.4: Total Debt of States and Total Debt excluding Debt on account of UDAY

Public Debt, % of GDP

35
30
29.4 23.8
25 22.0
20
15
10
5
0

States (Total Debt) States (Total Debt excluding UDAY) UDAY

Source: State Finances Report, RBI (States) and CEIC (compiled from Clearing Corporation of India for Ujjwal
DISCOM Assurance Yojana, UDAY). For 2021-22 we have used the Revised Estimates and for 2022-23 the Budget
Estimates. The chart shows the total outstanding liabilities of State Government as % of GDP, total outstanding
liabilities of States excluding UDAY as % of GDP, and debt incurred on account of UDAY as % of GDP. Under the UDAY
scheme, State Governments assumed contingent liabilities on account of the loss-making electricity distribution
companies (governments issued bonds in lieu of the debt owed by these companies to the banks).
40 | India Policy Forum 2023

Figure B.5: Deficit of Center and States

Fiscal Deficit, % of GDP

10
9 Centre States

8
7
5.8 6.5
6 4.8
5 4.3
4
3 3.2
2 3.1 3.1
2.5
1
0
1990-91

1992-93
1993-94
1994-95
1995-96
1996-97
1997-98
1998-99
1999-00
2000-01
2001-02
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08

2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22
2022-23
1991-92

2008-09
Source: Handbook of Statistics on Indian Economy, RBI (Centre) and State Finances Report, RBI (States).
For Centre, data for 2021-22 is Actual and 2022-23 is a Revised Estimate from CEIC. For States, data for
2021-22 is a Revised Estimate and for 2022-23 is a Budget Estimate. The chart shows the fiscal deficit as
% of GDP for both Centre and State governments. Horizontal dashed lines are the respective decadal
averages.

Figure B.6: Primary Deficit of Centre and States

Primary Deficit, % of GDP

6

5
Centre States
4
3.0
3

2 1.6
1.2
0.9 1.5
1
1.2
0.8 0.9
0

-1

-2

Source: Handbook of Statistics on Indian Economy, RBI (Centre) and State Finances Report, RBI (States).
For Centre, data for 2021-22 is Actual and 2022-23 is a Revised Estimate from CEIC. For States, data for
2021-22 is a Revised Estimate and for 2022-23 is a Budget Estimate. The chart shows the primary deficit
as % of GDP for both Centre and State government. Horizontal dashed lines are the respective decadal
averages.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 41

Figure B.7: Interest rate Paid by Centre and State governments

Weighted average yields on primary issues

in the year for Centre and State Government
10
9
8
7
6
5
4
3
2
1
0

Centre States

Source: RBI (Centre and States), CEIC (individual state governments). Yields refer to weighted average
yields on new issues of securities during the year.
42 | India Policy Forum 2023

Appendix C: Debt-to-GDP Ratio and Interest Rates on Government

Debt (General Government)

We regress nominal interest rates on debt-to-GDP ratio of the General Government, for
the entire period 1990-91 to 2022-23, and different subperiods. Interest rates have
been calculated as weighted average yields on Centre and State government securities
using the shares of Centre and States debt in total debt as weights.
Results indicate that the interest rates do not react positively to the level of debt
(Table C1). In other words, the government does not pay a premium to raise debt when
its debt levels are already high. One would have expected this to be perhaps more true
in the earlier years, when financial repression through high SLR and CRR, and even
through the automatic monetization of deficit by the RBI was much higher. But it also
remains the case for the period starting in 2010-11.
As we show in the next appendix, over time the financial repression (at least
through commercial banks) has declined, and the investor base has become more
diversified. Yet the non-relationship (or the reverse relationship) between interest rates
and debt levels has persisted.28 This could be attributed to three factors: (i) There are
adequate savings and lack of alternative safe assets. (ii) Financial repression has
continued but has just shifted from commercial banks to other investors, notably
insurance companies and provident funds. (iii) The RBI, with a strong balance sheet
(along with other large players in the market, such as the State Bank of India), ensures
that yields remain low.
We repeat the exercise with real interest rates, and find that the real interest
rates do not react positively to the level of debt (Table C2).

Table C1: Results from Regressing Nominal Interest Rate on Debt- to-GDP Ratio of
the General Government
(1) (2) (3)
Nominal Nominal Nominal
interest rate interest rate Interest rate
(yields) (yields) (yields)
Debt to GDP ratio -0.18*** -0.08** -0.08***
(3.23) (2.77) (4.71)
22.42*** 13.65*** 13.72***
Constant (5.44) (6.37) (10.62)
No. of Observations 33 23 13
Years included 1990-91 to 2000-01 to 2010-11 to
2022-23 2022-23 2022-23
Note: t statistics in parentheses. *, **, *** refer to significance at 10, 5, and 1 percent levels, respectively.
Nominal interest rate for general government has been calculated as the weighted average yield on
Centre and State government securities (using the shares of Centre and States debt in total debt as
weights).

28Lack of positive relationship prevails when we do a similar exercise separately for Central and State
governments.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 43

Figure C 1: Co-movement of Debt-to-GDP Ratio and Nominal Interest Rate

Debt to GDP ratio and Nominal Interest Rate Debt to GDP ratio and Nominal Interest Rate
(Scatter plot) (Time Series)
Debt to GDP (LHS)
14

100 Nominal interest rate 16

90 14
12

80 12
10
70
10

8
60
6
50
8

4
40 2
30 0
6

1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07

2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2022-23
2008-09
65 70 75 80 85 90
Debt to GDP

Source: CEIC (General government debt to GDP ratio). Nominal interest rate for general government has
been calculated as the weighted average yield on Centre and State government securities (using the shares
of Centre and States debt in total debt as weights).

Table C2: Results from Regressing Real Interest Rate on Debt- to-GDP Ratio of the
General Government

(1) (2) (3)

Real interest Real interest Real Interest
rate rate rate
Debt to GDP ratio -0.11 -0.06 -0.12
(1.52) (0.72) (1.32)
Constant 10.80* 6.31 10.96
(1.98) (1.05) (1.60)
No. of Observations 33 23 13
Years included 1990-91 to 2000-01 to 2010-11 to
2022-23 2022-23 2022-23

Note: Data are from 1990-1991 to 2022-2023. t statistics in parentheses. *, **, *** refer to significance at
10, 5, and 1 percent levels, respectively. Real interest rate has been calculated as the difference between
nominal interest rate and growth rate of GDP deflator (inflation rate).
 44 | India Policy Forum 2023

Figure C2: Co-movement of Debt-to-GDP Ratio and Real Interest Rate

Debt to GDP ratio and Real Interest Rate Debt to GDP ratio and Real Interest Rate (Time
(Scatter plot) Series)
Debt to GDP (LHS)
100 Real interest rate 10
8

90
8
80
6

70 6
60 4
4

50
40 2
30
2

0
20
-2
10
0

0 -4

1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07

2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2022-23
2008-09
-2

65 70 75 80 85 90
Debt to GDP

Source: CEIC (General government debt to GDP ratio). Nominal interest rate for general government has
been calculated as the weighted average yield on Centre and State government securities (using the shares
of Centre and States debt in total debt as weights).
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 45

Appendix D: Bank-Sovereign Nexus

The following five kind of investors hold government securities: banks, insurance
companies, provident funds, the RBI, and a residual category, which includes retail
investors, cooperative banks, and mutual funds, among others. Their relative shares
have changed in the last decade, over which the share of banks has declined, whereas
those of the other four investors have increased (Figure D2).
The government owns a large part of each segment. For instance, it owns 12
banks (21 banks are private) and 7 of the largest insurance companies (50 insurance
companies are private). Government banks accounted for 60 percent of total bank
assets), while government insurance companies accounted for about 80 percent of the
industry’s total assets (as of 2020-21).
Banks, insurance companies, and provident funds have statutory requirements
to invest in government securities (for the banks, for example, see Figure D1). But
public banks have traditionally held more than the mandated share of their assets in
government securities (Gupta, Kochhar and Panth, 2015). Their SLR ratio has declined
from about 40 percent in the early 1990s to 18 percent currently, while public sector
banks have reduced their excessive shares in these securities. They now hold only a
slightly larger share of their assets in government securities compared to private
banks.
Both insurance companies and provident funds face statutory requirements to
invest about 50 percent of their respective investable funds in government securities. In
recent years, the provident fund has requested the government to allow it to increase
the share of its investments in government securities, from 50 percent to 65 percent in
2016, and again to 75 percent in 2022. This request indicates a lack of options as far as
other safe, long-term, and liquid assets are concerned (the corporate bond market is
thin, and its secondary market has very little volume and liquidity).
Figure D1: Statutory Liquidity Ratio

Statutory Liquidity Ratio (India)

45
40
35
30
25
20
15
10
5
0
1990-91
1991-92
1992-93
1993-94
1994-95
1995-96
1996-97
1997-98
1998-99
1999-00
2000-01
2001-02
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22
2022-23

Source: CEIC (Compiled from Reserve Bank of India). Daily SLR has been averaged during the fiscal year
to get annual average SLR for the respective fiscal years.
 46 | India Policy Forum 2023

Figure D2: Share of Public Sector Banks, Private Banks and the RBI in Central
Government Securities
Shares of RBI and Scheduled Commercial Banks Shares of Private and Foreign Banks and Public
Sector Banks

Scheduled Commercial Banks Private and Foreign Banks

75 RBI 30.0 70 Public Sector Banks

70 60
25.0
65 50
60 20.0
40
55
15.0 30
50
45 10.0 20
40 10
5.0
35 0
30 0.0

1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05
2006-07
2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
1990-91
1992-93
1994-95
1996-97
1998-99
2000-01
2002-03
2004-05

2008-09
2010-11
2012-13
2014-15
2016-17
2018-19
2020-21
2006-07

Source: Handbook of Statistics on Indian Economy, RBI.

Table D1: Shares of Institutions in Holdings of Central Government Securities

Insurance Public Private RBI Provident Foreign Others
Companies Sector and Funds Institutional
Banks Foreign Investors
Banks
2007-08 19.2 32.3 17.3 7.8 2.9 20.4
2008-09 17.0 31.6 16.1 8.5 3.0 23.7
2009-10 17.4 33.7 15.4 11.0 3.4 19.0
2010-11 19.5 34.7 15.6 10.9 3.7 15.6
2011-12 18.7 36.8 17.3 13.3 3.8 1.6 8.4
2012-13 18.6 24.3 19.0 17.0 7.4 1.6 12.1
2013-14 19.5 23.8 19.8 16.1 7.2 1.7 12.0
2014-15 20.9 22.7 19.9 13.5 7.6 3.7 11.9
2015-16 22.2 20.5 20.5 13.5 6.0 3.6 13.7
2016-17 22.9 22.7 16.6 14.7 6.3 3.5 13.3
2017-18 23.5 20.9 20.7 11.6 5.9 4.4 13.0
2018-19 24.3 17.8 21.5 15.3 5.5 3.2 12.3
2019-20 25.1 19.6 19.4 15.1 4.7 2.4 13.6
2020-21 25.3 20.0 16.9 16.2 4.4 1.9 15.3
2021-22 25.9 18.6 18.2 16.6 4.6 1.6 14.5
Source: Handbook of Statistics on Indian Economy, RBI. Others include Mutual Funds, Co-operative Banks,
Primary Dealers, Financial Institutions, Corporates, and State Governments. Besides RBI and Scheduled
Commercial Banks, the data for other institutions is only available since 2007-08.
 India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 47

Figure D3: Shares of Institutions in Ownership of General Government Securities

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
2007-082008-092009-102010-112011-122012-132013-142014-152015-162016-172017-182018-192019-202020-212021-22

Insurance Companies Public Sector Banks Private and Foreign Banks

RBI Provident Funds Foreign Institutional Investors
Others

Source: Handbook of Statistics on Indian Economy, RBI. Others include Mutual Funds, Co-operative Banks,
Primary Dealers, Financial Institutions, Corporates, and State Governments. Besides the RBI and
Scheduled Commercial Banks, the data for other institutions is only available from 2007-08.

Table D2: Shares of Institutions in Ownership of Total General Government

Securities
Insurance Public Private and RBI Provident Foreign Others
Companies Sector Foreign Funds Institutional
Banks Banks Investors
2007-08 19.7 35.0 14.8 6.6 4.0 19.9
2008-09 17.6 35.6 13.7 7.1 4.0 22.0
2009-10 18.3 37.5 13.1 8.9 4.3 17.9
2010-11 20.6 37.1 13.1 8.6 4.6 16.0
2011-12 20.3 38.9 14.4 10.4 4.8 1.3 10.0
2012-13 20.7 27.7 16.9 13.3 9.2 1.3 10.8
2013-14 22.0 27.0 17.5 12.5 8.9 1.3 10.7
2014-15 23.8 25.7 16.6 10.3 9.5 2.8 11.4
2015-16 24.9 23.6 17.1 9.9 8.6 2.8 13.1
2016-17 25.8 25.5 12.9 10.3 9.5 2.5 13.5
2017-18 26.8 23.2 15.7 8.0 10.2 3.1 13.1
2018-19 27.1 18.9 17.8 10.4 10.8 2.2 12.8
2019-20 27.3 23.1 13.7 10.0 10.6 1.6 13.7
2020-21 26.9 22.2 12.7 11.0 10.4 1.3 15.6
2021-22 26.8 21.1 13.7 11.2 10.1 1.0 16.0
48 | India Policy Forum 2023

Figure D4: Concentration of Ownership in Central Government and General

Government Securities
Herfindahl Index for Central Government Herfindahl Index for General Government
0.38 0.38
0.36 0.36
0.34 0.34
0.32 0.32
0.30 0.30
0.28 0.28

0.26 0.26

0.24 0.24

0.22 0.22
0.20
0.20

2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14

2015-16
2016-17
2017-18

2019-20
2020-21
2021-22
2014-15

2018-19
2007-08
2008-09
2009-10
2010-11
2011-12
2012-13
2013-14
2014-15
2015-16
2016-17
2017-18
2018-19
2019-20
2020-21
2021-22

Source: Author’s calculations. The index is calculated by adding the squared shares of RBI, Scheduled
Commercial Banks, Provident Funds, Insurance Companies, Foreign Portfolio Investors, and Others
(which include Mutual Funds, Co-operative Banks, Primary Dealers, Financial Institutions, Corporates,
and State Governments) in total Central Government or General Government securities.
India’s Debt Dilemma | Barry Eichengreen, Poonam Gupta and Ayesha Ahmed 49

NATIONAL COUNCIL OF APPLIED ECONOMIC RESEARCH

NCAER India Centre, 11 Indraprastha Estate, New Delhi 110002. INDIA.
Tel: +91-11-2345 2698, 6120 2698 [Link]

NCAER | Quality . Relevance . Impact