High public debt in India: 9 stylised facts

The Covid-induced surge in public debt in India was unique compared to its own history, but also bigger and driven by different factors relative to the average emerging market economy. In this post, Mishra and Patel document nine stylised facts on the recent evolution of sovereign debt and fiscal deficits in India – examining issues such as the cost of high debt levels, whether there are silver linings, and the path ahead.

India’s sovereign debt reached unprecedented levels in 2020, partly driven by the policy response to Covid-19, but also by low growth and high interest rates (Figure 1). Some have argued that high levels of debt may be less concerning in an environment of low interest rates. However, there is also a significant body of evidence that points to several mechanisms through which high levels of sovereign debt can have negative effects on the economy (see for instance Koh et al. 2020).

Figure 1. India’s sovereign debt, 1990-2020

Source: Global Debt Database, International Monetary Fund (IMF) World Economic Outlook (WEO) Database. Actual data till FY2020, WEO projection (for deficit) for 2021.

Against this background, we document stylised facts on the recent evolution of sovereign debt and fiscal deficits in India and ask the following questions: What are the costs of high debt levels in India? Are there any silver linings? And what lies ahead? We analyse macroeconomic outcomes during and after debt “surges”, “stabilisation”, and “reduction” periods in India and other countries and ask whether past experiences with surges and reductions shed light on different policy options and the trade-offs for India during the post-pandemic recovery.

The Covid-induced surge in debt in India was unique compared to its own history, but also bigger than that for the average emerging market (EM) economy. The drivers of the debt surge were different too. Both fiscal expansion and the collapse in growth played a proportionately larger role in India compared with the average EM, even as higher inflation played a greater role in reducing debt in India.

Notwithstanding the high level of sovereign debt, there are a few silver linings for India. The share of sovereign debt held by foreigners – an important predictor of crises in the literature (for example, in Reinhart and Rogoff (2011)) – is low. Moreover, although global waves of debt surges have been followed by restructuring or default, India has not had any such episode so far. Furthermore, long-term real rates remain low in India, comparable to the median EM. That said, we find substantial heterogeneity across countries. While India was closer to the 25^th percentile during the last decade, it has now caught up with the median.

High interest payments, less space for countercyclical policies, crowding out of social spending

We document substantial costs of high debt. A major one is foregone resources on account of strikingly high interest payments, which at almost 30% of overall revenues during Covid-19, are close to three times higher for India than the typical EM (Figure 2). Unlike debt to GDP (gross domestic product) ratios, which entail a stock variable in the numerator and a flow one in the denominator, these interest expenses to government revenues ratios comparing two flow variables tend to provide a cleaner measure of the burden of high public debt on a yearly basis.

Figure 2a. PPP-GDP weighted averages across country groups

Source: IMF WEO Database.

Figure 2b. Fraction of variation in debt explained by business cycle fluctuations

Source: World Bank World Development Indicators, 2019.

Limited fiscal space: Business cycle shocks account for much lower fraction of debt fluctuations

High expenditures on interest payments reduce the resources available for countercyclical fiscal policies in the event of negative shocks such as Covid-19, as well as for social spending in critical areas such as health and education, where India’s public spending remains markedly below peers.

Indeed, our analysis suggests that business cycle fluctuations explain a smaller fraction of the variation in debt in India compared with peers, reaffirming the limits to countercyclical fiscal policy on account of high debt levels (Figure 3).

Simple calculations suggest that reducing India’s interest payments to revenue to the EM average of 10% would release resources of close to Rs. 6-8 trillion, a figure comparable to India’s pre-Covid general government education expenditure, and about three times its health expenditure.

Figure 3. Contribution of shocks to the historical decomposition of public debt

Notes: (i) Historical decomposition based on a vector autoregression identified using narrative sign restrictions. (ii) The primary balance shock is orthogonal to business cycle (demand and supply) shocks. (iii) Sample comprises of 33 EMs from 1990-2020.

High spreads, especially since the pandemic; and high elasticity of borrowing costs to debt

Another cost of high public debt in India is its impact on borrowing costs. Although real rates in India are low and in line with the median EM, we find that they have increased over time, and that the elasticity of borrowing costs to a unit increase in debt is higher for India than the typical EM (Figure 4).

For example, on average, an increase in debt to GDP by 1 percentage point (pp) increases long-term borrowing costs by 0.19 pp in India, while for a median EM, it increases by only 0.01 pp.

Figure 4a. Debt and spread pre-pandemic, and 2022 values

Note: Each arrow represents a country – the beginning of the arrow represents the debt and spread pre-pandemic, and the end of the arrow their values in 2022.

Source: IMF WEO Database, Bloomberg (for CDS spreads).

Figure 4b. Sensitivity of long-term real rates to debt levels in India

Notes: (i) Based on an unbalanced panel of 19 EMs. (ii) Simple scatter plot not controlling for fixed effects. (iii) Numbers reported in the text are based on regressions with country fixed effects.

Sources: IMF WEO Database, IMF International Financial Statistics (IFS) database, Bloomberg, Haver, Jorda-Schularick-Taylor Macro history database, OECD, Mauro et al. (2015), Global Debt Database.

Public debt and deficits in India are much larger than similarly rated emerging market peers

Finally, public debt exemplifies an important factor in the assessments of rating agencies too, where India’s debt and deficits stand out as being markedly higher than similarly rated peers (Figure 5).

Figure 5a. Central government fiscal deficit (% of GDP)

Note: Sample includes seven countries with comparable ratings to India – Colombia, Croatia, Indonesia, Morocco, Philippines, Romania and Uruguay. For 2021, sample excludes Indonesia and Croatia due to data availability.

Source: Haver and IMF staff calculations.

Figure 5b. General government debt (% of GDP)

Note: Sample includes 7 countries with comparable ratings to India – Bulgaria, Colombia, Croatia, Indonesia, Kazakhstan, Philippines and Romania.

Source: Government Debt Database and IMF staff calculations.

Sovereign debt episodes and transitions from surges

In order to understand where to go from here, we look at India’s own history and also draw on cross-country experiences. Since 1913, India has had nine episodes of debt surges, five episodes of reduction, and six episodes of debt stabilisations.

Surges have typically ended in stabilisations in India, whereas in an average EM, 75% of surges end in reductions (Figure 6). In other words, India has been able to sustain debt at high levels without default or restructuring. Across reduction episodes, India reduced debt ratios by 2 pp per year, compared to more than double the figure for the average EM.

Figure 6a. How surges typically end

Source: IMF WEO, Chapter 3, April 2023.

Figure 6b. Episodes of debt surges and reductions in India

Larger and longer lasting public debt surges are followed by lower growth

We also find that debt surge episodes are associated with worse macroeconomic outcomes – low economic growth and public investment – compared with debt reduction episodes.

Moreover, cross-country evidence suggests that the greater the magnitude of the rise in debt, and longer lasting the episode, the greater the associated reduction in growth around the surge (Figure 7).

Figure 7a. Debt surges and GDP growth

Note: The binscatter controls for real per capita GDP (logs) and country groups and year dummies.

Figure 7b. Effects of longer debt surges on growth

Source: Global Debt Database, IMF WEO Database; author calculations (see WEO, April 2023, Chapter 3 for details).

More India grows out of debt, lesser the required adjustment: Possible scenarios to reduce debt by 20 percentage points in 10 years

How much debt could India reduce? One way to approach this question is to look at interest payments and additional budgetary resources that could be generated by lower sovereign borrowing.

For example, getting interest payments down to 22% (still much higher than the EM average of 10%) would require reducing the debt ratio to 70%, bringing it closer to the median for similarly rated peers.

What is a possible path and how long would it take to get there? The higher the growth rate and the lower the borrowing costs, the lower the need for fiscal adjustment. Simulation exercises suggest that if we assume constant values for real GDP growth rate at 7% and real rate at 2% in line with the IMF World Economic Outlook (WEO) assumptions, a general government primary and fiscal deficit of lower than 1.7% and 5.9% of GDP, respectively, would be needed every year to reduce debt ratios to 70% in the next 10 years (and interest payments to 22% of revenues).

This would require a sharp adjustment when compared with the FY 2022-23 primary and fiscal deficit at projection of 4.5% and 9.9%, respectively, according to WEO. Importantly, the higher the growth rate and the lower the interest rate, the less the required adjustment. For example, a growth rate of 9% or a real rate of 0% would open up more space with a primary deficit of more than 3% of GDP instead of 1.7%, still ensuring the same debt reduction (Figure 8).

Figure 8a. Primary deficit (% of GDP), growth rate (%), and real interest rate (%)

Figure 8b. Real interest rate and growth rates

Source: Global Debt Database, IMF WEO Database, Actual data till FY 2020, WEO projection (for deficit) for 2021.

Possible trajectories for deficits and debt

While the calculations above assume constant primary and fiscal deficits, allowing for some transitional dynamics and smoothing the adjustment path, we report in Figure 9, illustrative scenarios for debt and fiscal consolidation for India over the next five years.

Figure 9. Primary fiscal deficit (left), total fiscal deficit (middle) and debt (right)

Notes: (i) The alternate scenario with added consolidation assumes a primary deficit 0.5 pp below the WEO projection for years 2022-2027. (ii) The alternate scenario with higher growth and added consolidation assumes, in addition to the above consolidation, a 9% growth rate for years 2022-2027. (iii) The alternate scenario with only consolidation reduces debt to 80% by 2030, whereas the scenario with consolidation and high growth reduces debt to 68%.

Fiscal consolidations in emerging markets reduce debt without hurting growth much

Indeed, evidence across emerging economies suggests that primary balance consolidations outside of recessions could, in fact, be successful in reducing debt, and do not tend to be detrimental to growth as multiplier effects roughly balance the positive impulse from other channels such as higher confidence (Figure 10).

The composition of revenues and expenditures during consolidations also has a significant bearing, and there is evidence suggesting that consolidations that are more geared towards cutting government consumption rather than government investment tend to have lower output costs, or even positive effects on output.

Figure 10. Impact of discretionary fiscal consolidations on debt and growth

Notes: (i) Impulse response to a primary balance consolidation shock identified in a structural vector autoregression with narrative sign restrictions. (ii) Sample includes 33 emerging economies from 1990-2019. (iii) The model includes five variables: GDP growth, primary balance to GDP, debt to GDP, inflation, and effective interest rate on debt.

The views expressed here are those of the author and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.

A version of this post was first published as a Policy Brief of the Isaac Centre for Public Policy, Ashoka University.