Courtesy of Ugo Panizza.
Italian bond issued in France to finance a railway in Italy

CORE Insights Public debt:
threat or opportunity?

The authors of this Insight are:
Barry Eichengreen: University of California, Berkeley
Ugo Panizza: Graduate Institute of International and Development Studies

28 April 2022

Highlights

  • For hundreds of years, governments have issued debt to finance their expenditure as a substitute for raising taxes, especially during challenging times.
  • Public debt in the U.S. has been increasing over the past four decades. Political polarization of parties has been offered as a potential explanation.
  • We can understand the dynamics of debt by looking at four variables: primary deficit, real interest rate, growth rate, and the stock of debt in the previous period.
  • Public debt can be repaid by running a primary surplus, or by promoting economic growth. However, in some circumstances, governments also have an incentive to use inflation to reduce debt.
  • State sovereignty limits the extent to which governments can be sued if they default on their debt, especially internationally.
  • Default may have negative consequences in terms of reputation and access to debt on favorable terms in the future.
  • Statutes and non-partisan institutions known as autonomous fiscal councils have been implemented in many countries to keep the government accountable for its use of debt.

CORE Insights

Concepts in this Insight are related to material in:

Recommended reading before starting this Insight:

  • Section 14.8 of The Economy for an overview of the government’s finances.

1 Introduction

COVID-19 was designated as a pandemic in March of 2020 and recognized by governments as an all-hands-on-deck emergency. They responded by ramping up health spending, commissioning vaccine research, extending subsidies to employers seeking to meet payroll, and providing relief payments to households. They did this despite the fact that tax revenues were in sharp decline with businesses closed and residents locked down.

bond
A type of financial asset for which the issuer promises to pay a given amount over time to the holder. Also known as: corporate bonds.

How did governments accomplish this feat? Answer: they did so by borrowing. When governments spend more than they raise in taxes, they finance the difference by selling bonds to the public. Investors purchase those bonds with cash, which the government then uses to extend payroll subsidies, make relief payments, and finance other spending. Worldwide, on average, governments borrowed 10.2% of GDP in 2020 and an additional 7.9% of GDP in 2021. As they borrowed they became more heavily indebted. General government debt as a share of global GDP rose from 83.6% in 2019, on the eve of the pandemic, to an estimated 97.8% in 2021.1 The extent of borrowing and increase in debt were unprecedented in peacetime. We would need to go back to the Second World War—an all-hands-on-deck emergency if there ever was one—to see anything similar.

interest rate
The price of bringing some buying power forward in time.
disposable income
Income available after paying taxes and receiving transfers from the government.

Few begrudge this borrowing. Governments that do not meet this kind of crisis by mobilizing additional resources, including those that can only be mobilized by incurring debt, face losing their legitimacy and public support. But now there is a financial legacy. There is a bill coming due. How should we think about this? When governments issue debt, they commit to paying interest to the bondholders and to buying back or replacing bonds when they mature. The spending happens today, benefiting the currently living, but interest and principal payments continue for as long as the bonds remain in circulation, potentially burdening successive generations. Thomas Jefferson, the third U.S. president, warned that ‘the principle of spending money to be paid by posterity, under the name of funding, is but swindling futurity on a large scale.’2 Future generations must pay taxes to finance the state’s interest payments, leaving them less disposable income with which to put food on the table and invest in commercial and industrial ventures. Writing shortly before Jefferson, Adam Smith, in The Wealth of Nations, warned that these ‘enormous debts presently oppress, and will in the long run probably ruin, all the great nations of Europe.’3

countercyclical
Tending to move in the opposite direction to aggregate output and employment over the business cycle.

Smith and Jefferson are eminent figures, but not everyone shares their pessimistic assessment. Governments borrow for good reasons. They borrow to invest in roads, bridges, and broadband, and in so doing enhance the growth of the economy. They borrow to invest in early childhood education and technical training that make for a more productive labor force. Borrowing to finance these productive public investments bequeaths more debt and thus obliges households, now or in the future, to pay more taxes. But it also aims to deliver more robust growth and higher incomes, both topping up the government’s coffers and leaving those households better off even after paying taxes. In addition, governments, by borrowing, are able to boost public spending in periods when private spending is weak. They are able to engage in countercyclical fiscal policy, in other words, smoothing economic fluctuations.

To read a book-length discussion of these issues, see In Defense of Public Debt by Barry Eichengreen, Asmaa El-Ganainy, Rui Esteves, and Kris James Mitchener, or Sovereign Debt: A Guide for Economists and Practitioners by S. Ali Abbas, Alex Pienkowski, and Kenneth Rogoff. To read recent surveys that focus on debt crises in high-income and emerging economies, see ‘Sovereign Debt in the 21st Century: Looking Backward, Looking Forward’ by Kris James Mitchener and Christoph Trebesch, or ‘Enough Potential Repudiation: Economic and Legal Aspects of Sovereign Debt in the Pandemic Era’ by Anna Gelpern and Ugo Panizza.

In the Insight, we shall see that public debt is both a threat and an opportunity. It depends what it is used for and the circumstances in which it is taken on. The view from history tells us how and why governments have borrowed. Economic reasoning shows that the relationship between the economy’s growth rate and the interest rate explains the dynamics of the debt and enables us to understand the sustainability of the debt. From a political standpoint, we want to understand why governments repay what they borrow, in light of the fact that, as sovereign states, they are mostly immune from litigation in the courts. And we want to understand what happens when they choose otherwise.

2 Concepts and distinctions

Debt vs deficit

government debt
The total amount of money owed by the government at a specific point in time.
stock
A quantity measured at a point in time. Its units do not depend on time. See also: flow.
flow
A quantity measured per unit of time, such as annual income or hourly wage.
government budget deficit
When the government budget balance is negative.

Debts and deficits are not the same. Government debt is a stock that measures the total amount of money owed by the government at a point in time. For example, the debt of France at the end of 2020 was €2.65 trillion. The budget deficit is a flow. It measures the difference between government expenditure and revenues over a period of time; when expenditure exceeds revenues, it is a deficit and in the opposite case, it is a surplus. (For more details on the government’s finances, see Section 14.8 of The Economy.) For example, the budget deficit of the French government during 2020 was €212 billion. These two variables are related, because the difference between the stock at the end of this year and the stock at the end of last year should equal the deficit during the year.

Debt-to-GDP ratio

gross domestic product (GDP)
A measure of the market value of the output of final goods and services in the economy in a given period. Output of intermediate goods that are inputs to final production is excluded to prevent double counting.
insolvent
When the value of an entity’s assets is less than the value of its liabilities.

In analyzing public debt, it is useful to scale that debt by GDP. To learn more about how GDP is calculated, see Unit 13 of The Economy. The same nominal amount of debt will be less burdensome for a larger economy, and economies grow over time. Whereas U.S. public debt stood at $233 billion in 1949, it was $5.6 trillion in 1999. At first glance, this 25-fold increase might look worrisome. But it appears very different when one observes that U.S. GDP grew from $270 billion to nearly $10 trillion, a 37-fold increase, over the period. The debt-to-GDP ratio fell from 86% to 56%. The debt-to-GDP ratio is also convenient for international comparisons. For instance, a debt of $88 billion (170% of GDP) brought the Lebanese government to the brink of insolvency in 2020. But $88 billion of debt would have been mere rounding error for a country like Germany, with a GDP of nearly $4 trillion.

Some authors suggest adjusting interest payments for the erosion of debt by inflation, replacing
[(Debt + (Interest rate × Debt)) ÷ GDP]
by
[(Debt + ((Interest rate − inflation rate) × Debt)) ÷ GDP],
where Debt is nominal public debt.

The debt-to-GDP ratio is the most widely used metric of public debt, because it is useful for analyzing the factors influencing the evolution of the debt burden over time (as we show starting in Section 5). But it is not the only possible way of scaling the debt. Alternatively, one can compare interest payments with GDP. This measure focuses on the current financing situation. It has the advantage that we are comparing a flow (interest payments this year) with another flow (GDP this year). Its corresponding disadvantage is that interest payments and GDP this year tell us nothing about the likely future evolution of the debt burden going forward.

In addition, it is important to note that the share of taxes in GDP varies substantially across countries. According to the International Monetary Fund (IMF), in 2021, tax revenues were 14% of GDP in Uganda, for example, but 52% in France. (The average for high-income economies was 42%, while the average for emerging and developing economies was 27%.) Typically, government revenues are dominated by tax receipts. However, some governments such as those of Norway and Saudi Arabia also derive large revenues from the extraction and sale of natural resources. If debt is serviced from government revenues, then these large differences suggest that the typical high-income economy can sustain a higher debt-to-GDP ratio than the typical emerging economy. For some purposes, therefore, debt relative to government revenues can be a more useful metric.

Different ways to classify debt

For a brief explanation of the diabolic loop, see Section 6.1 of A crash course on the euro crisis by Markus K. Brunnermeier and Ricardo Reis.

debt securities
Another, equivalent way to refer to bonds. See also: bonds.

For some purposes it is important to disaggregate public debt according to who holds it. Debt securities issued by the government can be purchased by individual investors, mutual funds, pension funds and banks. Accumulation of large amounts of government bonds by banks in particular can create macroeconomic risks, since any problems of servicing the government’s debt will then become problems for the banks and, equally, any problems for the banks that force them to liquidate their holdings of government bonds can become problems for the government. In Europe, where these linkages were prominent starting in 2008, they became known as ‘the diabolic loop’.

inflation
An increase in the general price level in the economy. Usually measured over a year.

For other purposes it is important to distinguish whether debt is denominated in domestic or foreign currency, whether its term to maturity is long or short, and whether the interest rate is indexed to inflation. (For more about inflation, see Section 15.1 of The Economy.) When debt is denominated and issued in a foreign currency, as foreign investors have historically required of most governments, changes in the exchange rate can have implications for its sustainability. When debt is short in term, there will be little scope for using inflation to reduce its real value, since as soon as their debt securities mature, investors will be able to demand a higher interest rate in compensation for inflation, limiting any erosion in the real value of their investments. Thus, the maturity of the debt may influence a government’s policy choices. When the interest rate is indexed to inflation, there will be absolutely no scope for inflating away the debt. But there can be unanticipated consequences for debt-servicing costs. If the government, unlike investors, is confident that there will be no inflation, indexed debt can be cheaper than debt with a fixed interest rate. However, if inflation turns out to be higher than expected by investors, indexed debt will end up being more expensive for the government.

A further complication is the existence of different levels of government. One could focus on debt issued by the central government alone, or on general government debt, including that of state and municipal governments. The difference between central and general government debt can be substantial in federal countries. In 2020, when the central government debt of Switzerland was 21% of GDP, general government debt was 42% of GDP (the average debt of cantons, the 26 states of the Swiss Confederacy, was 12% of GDP and municipal debt 9% of GDP). Although statistics on general government debt are more encompassing, some countries do not keep good track of local government debt. Hence general government debt statistics are not always comparable across countries.

Gross debt vs net debt

To obtain a measure of net debt, one might subtract the government’s assets from the gross debt. Large differences between gross and net debt can arise when government agencies hold a large fraction of the government’s debt. For instance, about one-quarter of U.S. federal government debt is held in the Social Security Trust Fund, which funds future payments to retirees. In this case, one agency of government (the Treasury) is making payments to another agency of government (the Trust Fund). Therefore, U.S. statistical sources normally report ‘debt held by the public’, which nets out these cross-holdings.

From Figure 1, we can see that the difference between net and gross debt is large in Japan, where the government owns land and other assets valued at about 100% of GDP. At the other extreme, there are countries where the government holds few assets and for which gross and net debt are almost identical, as is the case of Barbados and Nigeria, for example.

liquidity
Ease of buying or selling a financial asset at a predictable price.

But calculating net debt can be problematic, since some government assets are not traded, and so can be difficult to value. In terms of liquidity, government assets may not easily be sold if the government needs to raise cash in a debt crisis when it needs to buy back debt. This renders gross debt a better measure of vulnerability.

Further, standard public debt measures do not include payments to future pensioners for which trust fund resources have not yet been put aside, or other implicit liabilities of the public sector (for example, the need in the future to inject resources into loss-making state-owned enterprises). These unfunded liabilities can be important, but measuring them is not easy, conceptually or practically.

  Net debt (% of GDP) Gross debt (% of GDP)
Australia 38.4 63.1
Barbados 147.6 149.0
Belgium 101.4 115.0
Brazil 62.7 98.9
Chile 8.7 32.5
Czech Republic 25.8 37.6
France 104.3 113.5
Germany 50.0 68.9
Ghana 72.8 78.0
Indonesia 33.0 36.6
Italy 142.0 155.6
Japan 169.2 256.2
Nigeria 34.6 35.1
South Africa 70.2 77.1
United Kingdom 93.8 103.7
United States 103.2 127.1

Figure 1 Gross and net debt as a share of GDP, selected countries, 2020.

IMF, 2021, World Economic Outlook April 2021; U.S. Department of the Treasury, 2022, Debt to the Penny; Federal Reserve Bank of St. Louis, 2022, FRED; Bank of Japan, 2022, Statistics.

In recent years, central banks have bought considerable amounts of public debt. The Bank of Japan holds 43% of Japanese government debt, and the U.S. Federal Reserve System holds 22% of U.S. federal government debt. If we remove these amounts from the net debt figure in Figure 1, Japan’s debt-to-GDP ratio would fall to 50% of GDP, and the U.S. debt ratio to 75%. It is tempting to make this adjustment, since the central bank is an arm of the public sector. The Federal Reserve System, like the Social Security Trust Fund, is part of the U.S. public sector. However, to purchase public debt, the Federal Reserve must issue liabilities of corresponding value. These take the form of reserves held by member banks at the central bank, on which the Federal Reserve pays interest. Because those reserves are not counted as public debt, it is conventional not to net out government securities held by the central bank.

Question 1 Choose the correct answer(s)

Which of the following statements about government debt are correct?

  • The Netherlands (GDP in 2004 of €601 billion) had government debt of approximately €204 billion at the end of 2003, and €214 billion at the end of 2004, so its deficit for the year 2004 was approximately €10 billion.
  • If the Argentinian government issues bonds denominated in U.S. dollars, it will become easier for the government to repay its debt if the dollar appreciates.
  • The difference between the gross and net measures of government debt is that unfunded pension liabilities are included in the gross but not in the net measure.
  • Different levels of government enable a country to monitor its public debt more precisely.
  • Deficit can be regarded as the difference in the stock of debt between 2003 and 2004.
  • If the dollar appreciates, it will become harder for Argentina to repay its debt, as the Argentine peso will be weaker compared to the dollar. For more details on the exchange rate, see Section 15.9 of The Economy.
  • Unfunded pension liabilities are included in neither definition; the difference is that government assets are deducted from the gross measure to calculate the net measure.
  • The existence of local governments makes keeping track of a country’s debt more difficult, with consequences for the precision of the statistics collected.

3 Public debt in history

States and sovereigns have long borrowed to mobilize resources to meet emergencies. Aristotle described how, in the fifth century BCE, Dionysius of Syracuse borrowed to finance military campaigns against the Carthaginians.4 The English monarch Edward III (1312–1377) borrowed to continue the Hundred Years War with France. In the 1340s, with the arrival of the Black Death, the Italian city-states of Venice and Siena borrowed to mobilize public health resources.

For more details on the tax smoothing logic, you can read Federal deficit policy and the effects of public debt shocks by Robert J. Barro, or On the determination of public debt by Robert J. Barro, or The motives to borrow by Ugo Panizza, Andrea F. Presbitero, Antonio Fatás, and Atish R. Gosh.

tax smoothing logic
A large increase in taxation to fund an unanticipated increase in government spending (such as for a war or pandemic) leads to tax avoidance and other behavioral responses that reduce the economy’s output and distort the allocation of resources. It is preferable to spread out the tax increases and borrow to fund the expenditure.

Until the late 20th century, debt finance by governments was predominantly war finance. War is the prototypical emergency, prompting governments to mobilize all available resources, including those that can only be mobilized by borrowing. What economists refer to as ‘tax smoothing logic’ suggests that the costs associated with mobilizing those resources should be spread over time rather than paid up front, all at once. High tax rates are especially distortionary, meaning that they cause inefficient behavior. For example, taxing investment income near or at a rate of 100% weakens or eliminates all incentive to invest. Rather than raising taxes sharply in times of war and lowering them back down when peace is restored, it therefore makes more sense to raise taxes moderately both in the present and in the future, while issuing debt to finance the immediate gap between revenues and essential spending.

Figure 2 illustrates this for the case of the United States. It shows the evolution of debt relative to GDP (that is, relative to the size of the economy, and how that history has been punctuated by war).

In this line chart, the horizontal axis shows years, from 1790 to 2020, and the vertical axis shows the debt-to-GDP ratio in the U.S., expressed as a percentage and ranging from 0 to 140. From 1790 to 1980, the debt-to-GDP ratio rose sharply during wartime, for example, from 10% to 18% during the War of 1812, from near zero to around 35% during the U.S. Civil War in 1865, and from 10% before the First World War to nearly 120% during the Second World War. After the wars, the debt-to-GDP ratio decreased gradually over time. However, since 1980, the debt-to-GDP ratio has followed an increasing trend, from around 35% in 1980 to nearly 130% in 2020.

Figure 2 Evolution of the debt-to-GDP ratio in the U.S. (expressed in percentage terms), 1790–2020.

Federal Reserve Bank of St. Louis, 2022, FRED; TreasuryDirect, 2022, ‘Monthly Statement of the Public Debt (MSPD) and Downloadable Files’; TreasuryDirect, 2022, ‘Historical Debt Outstanding: Annual 1790–1849’.

Figure 2 also hints that the purposes for which governments have borrowed broadened in the 20th and 21st centuries. The U.S. federal government went deeper into debt, for example, in the Great Depression of the 1930s. The Depression, when the unemployment rate soared to 25%, was seen as a social crisis tantamount to war. It highlighted the government’s role in providing social insurance and other essential services in hard economic times. Because tax receipts fell off in the economic slump, the government had to borrow to finance their provision. The U.S. government, along with others, similarly borrowed to meet the emergencies created by the global financial crisis in 2008–2009 and the COVID-19 recession and health crisis.

Up through the Second World War, the U.S. government, having issued additional debt in response to geopolitical and economic emergencies, regularly reduced the debt-to-GDP ratio when the crisis passed. (Section 7 describes how in more detail.) Prudent governments that utilize their borrowing capacity in an emergency appreciate the need to restore that capacity subsequently, since there is no telling when it will have to be utilized again.

Since the 1980s, however, the picture looks different. U.S. federal government debt as a share of GDP has trended steadily upward, excepting only the 1990s, when strong economic growth and fiscal restraint under President Bill Clinton reduced the debt-to-GDP ratio. But taking the last four decades as a whole, the U.S. debt-to-GDP ratio has soared from 30% of GDP to more than 120%. What went up showed little tendency to come back down subsequently.

Question 2 Choose the correct answer(s)

Based on the information in this section, which of the following statements are true?

  • The logic of tax smoothing implies that raising debt can be seen as an alternative to raising taxes when the government needs to finance its expenditure.
  • Figure 2 shows that as GDP in the U.S. increased, so did the U.S. government debt-to-GDP ratio.
  • In the past, governments have only raised debt to finance war.
  • In recent decades, the trend of the debt-to-GDP ratio has changed in the U.S.
  • Tax smoothing logic presents the idea that raising debt can be seen as an alternative to raising taxes when the government needs to finance its expenditure.
  • Over time, the U.S. debt-to-GDP ratio increased, but Figure 2 does not plot U.S. GDP, so we cannot draw any conclusions about the correlation between these two variables.
  • In the past, governments have also raised debt in response to other problems, for example meeting the health emergency of pandemics.
  • Whereas in the past, governments tended to reduce the debt-to-GDP ratio once a crisis passed, this has not been the case since the 1980s.

Exercise 1 Debt-to-GDP ratio in other countries

Collect data on the debt-to-GDP ratio of a country of your choice (other than the U.S.), for the longest time span available. For example, you can use the IMF Global Debt Database. Plot the data on a graph similar to Figure 2. Compare your graph with Figure 2 and describe the similarities and differences you see. (Hint: Doing Economics offers guidance on how to plot a line chart in Excel).

4 The political economy of public debt

The recent upward trend of government debt as a share of GDP in the U.S. is part of a global trend evident in advanced (high-income), emerging, and low-income countries alike (as shown in Figure 3). Economists do not agree on what explains this trend.

There are three bar charts. Chart 1 shows data for advanced economies. The horizontal axis shows years ranging from 2007 to 2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 40 to 140, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 70% of GDP in 2007 to over 120% in 2021, while interest expenses have fallen from around 2.5 to 1.8% in the same time interval. Chart 2 shows data for emerging economies. The horizontal axis shows years ranging from 2007–2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 20 to 70, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 35% of GDP in 2007 to around 65% in 2021, while interest expenses have remained at roughly 2% over the same time interval. Chart 3 shows data for low-income developing countries. The horizontal axis shows years ranging from 2007–2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 20 to 70, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 30% of GDP in 2007 to around 50% in 2021, while interest expenses have increased from 1% to nearly 2% in the same time interval.
There are three bar charts. Chart 1 shows data for advanced economies. The horizontal axis shows years ranging from 2007 to 2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 40 to 140, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 70% of GDP in 2007 to over 120% in 2021, while interest expenses have fallen from around 2.5 to 1.8% in the same time interval. Chart 2 shows data for emerging economies. The horizontal axis shows years ranging from 2007–2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 20 to 70, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 35% of GDP in 2007 to around 65% in 2021, while interest expenses have remained at roughly 2% over the same time interval. Chart 3 shows data for low-income developing countries. The horizontal axis shows years ranging from 2007–2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 20 to 70, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 30% of GDP in 2007 to around 50% in 2021, while interest expenses have increased from 1% to nearly 2% in the same time interval.
There are three bar charts. Chart 1 shows data for advanced economies. The horizontal axis shows years ranging from 2007 to 2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 40 to 140, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 70% of GDP in 2007 to over 120% in 2021, while interest expenses have fallen from around 2.5 to 1.8% in the same time interval. Chart 2 shows data for emerging economies. The horizontal axis shows years ranging from 2007–2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 20 to 70, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 35% of GDP in 2007 to around 65% in 2021, while interest expenses have remained at roughly 2% over the same time interval. Chart 3 shows data for low-income developing countries. The horizontal axis shows years ranging from 2007–2021, the primary vertical axis shows government debt as a percentage of GDP, ranging from 20 to 70, and the secondary vertical axis shows the interest expenses as a percentage of GDP, ranging from 0 to 5. Government debt has increased from around 30% of GDP in 2007 to around 50% in 2021, while interest expenses have increased from 1% to nearly 2% in the same time interval.

Figure 3 Interest expense and government debt, 2007–2021.

IMF Fiscal Monitor, April 2021. The blue bars plot government debt as percent of GDP (left scale) and the orange line plots interest expenses as percent of GDP (right scale).

Some see this increase in debt as a symptom of political dysfunction (for more details on how political interests can constrain governments and policymakers see Section 22.12 of The Economy). Casual observation and data suggest that political polarization, defined as the distance between the policy preferences of different political parties, is on the rise.5 6 When their spending preferences are very different—when polarization along this dimension is severe—the party in power will want to maximize spending on its preferred programs while in office, since it knows that the opposition will not continue those programs upon supplanting it. The party in power will be tempted to spend on its preferred programs whether or not the revenues are there.

For more details on ‘starving the beast’, you can read Do tax cuts starve the beast? The effect of tax changes on government spending by Christina D. Romer and David H. Romer.

In addition, a political party that prefers low levels of government spending in general may rely on debt finance when in office in order to tie the hands of its successor. By requiring that successor to devote substantial resources to paying interest on an inherited debt, it can prevent it from devoting them to social programs. (This strategy, which increases reliance on debt finance by cutting taxes, is sometimes referred to as ‘starve the beast’.) Or the party in power may increase the deficit just prior to elections as a way of temporarily stimulating the economy and increasing its political chances.

Alternatively, what we see in Figure 3 might be efficient and rational. If interest rates on government debt decline, as they in fact have over the last four decades (see Figure 4), then it makes sense for governments engaged in tax smoothing to issue additional debt today, since the future taxes required to service it will be correspondingly lower.

In this line chart, the horizontal axis shows the years from 1980 to 2019, and the vertical axis shows the real interest rate, ranging from −2% to 7%. In high income economies, the real interest rate has fallen from over 6% in 1982 to −0.3% in 2019. In the U.S., the real interest rate has fallen from 2% in 2003 to −0.5% in 2012, then fluctuated at around 0–1% from 2013 onwards.

Figure 4 Real interest rate estimated from inflation linked bonds in high-income economies and the U.S.

Rachel and Summers, 2019. The U.S. rate is based on treasury inflation-protected securities (TIPS). These are U.S. Treasury bonds that are indexed to inflation.

5 Analyzing debt dynamics

A simple equation can help us understand how the debt-to-GDP ratio changes over time as a result of economic conditions and policies. We start with the numerator of the ratio, debt. The change in debt is the difference between government expenditure and government revenues, as explained in Section 2. When a government spends more than it earns, it finances the difference by issuing debt. Conversely, when a government spends less than it accrues in revenue, it uses the surplus to extinguish debt or accumulate assets. Thus, debt today () is equal to debt in the previous period plus government expenditure () minus government revenues (). The letter stands for bonds, which is the most frequent form of government debt (although here includes all types of government debt, not only bonds). Debt today is calculated as:

Note that variables without a subscript refer to the current period. While formal notation uses a subscript of t for all variables in the current period, this can look cluttered so we have chosen to exclude it.

primary deficit
The government deficit (its revenue minus its expenditure) excluding interest payments on its debt. See also: government debt.

It is useful to distinguish interest payments from other expenditures such as public employee salaries, social transfers, and military procurement. The non-interest part is referred to as ‘primary’ expenditure. Correspondingly, the difference between primary expenditure and total revenues is the primary budget deficit (also known as the primary deficit). As total expenditure is equal to primary expenditure () plus interest payments (), we can write that debt is equal to:

Since the primary deficit (denoted by ) is equal to , the equation can be rewritten as:

We distinguish the primary deficit because policymakers have the capacity to adjust it by changing either taxes or primary expenditure. Interest payments, in contrast, are largely determined by the size of the debt (which is due to past decisions) and by the interest rates demanded by investors in order to hold government debt securities.

real interest rate
The interest rate corrected for inflation (that is, the nominal interest rate minus the rate of inflation). It represents how many goods in the future one gets for the goods not consumed now.

We now define as the debt-to-GDP ratio , where refers to GDP), and as the primary deficit-to-GDP ratio , where we use lowercase letters to denote variables that have been divided by GDP.

Real (net of inflation) interest expenditure as a share of GDP is equal to the real interest rate (the interest rate net of inflation), denoted , multiplied by the debt-to-GDP ratio . The overall deficit (as a share of GDP) is then the primary deficit plus interest payments: . Note the use here of instead of . This is because the government pays interest on the debt contracted in the previous period.

So far, we have seen that the debt-to-GDP ratio will change when the overall deficit (as a share of GDP) changes. If, for the moment, we assume that the economy is not growing then:

However, the equation above ignores the rate at which the economy grows, which also affects how the debt-to-GDP ratio changes. A given size of debt is a smaller proportion of GDP if the economy grows. Equivalently, the government will have greater resources available to pay back its debt.

If is the rate of growth of GDP, where is defined like this: , then, as set out in more detail in ‘Find out more: Debt restructuring’, when GDP changes, we can write the change in the debt-to-GDP ratio as:

The steps to derive Equation 5 are set out in ‘Find out more: Debt restructuring’. If the initial debt-to-GDP ratio is, say, 100%, is 4%, is 3% and is 1%, the debt-to-GDP ratio will increase by 2 per cent:

If, however, is 3% and is 4%, the debt-to-GDP ratio will remain constant, even if the country runs a deficit equal to 1% of GDP. Use Equation 5 to verify this for yourself.

Find out more Derivation of the basic equation for debt dynamics

Here we use some basic arithmetic manipulations to derive Equation 5. We want to show that , where is the stock of debt, is the primary deficit, is GDP, and are the debt-to-GDP ratio and the primary deficit-to-GDP ratio (that is, and ), and and are the real interest rate and real growth rate respectively.

Start by writing:

The stock of debt today is equal to the stock of debt in the previous year plus primary expenditure (), plus interest payments, minus government revenues ().

Analogously, real GDP is equal to real GDP in the previous year plus real GDP growth: (1 + ). Substituting A2 into A1:

Using as the denominator:

Or, simplifying:

And, using the fact that ,

Since tends to be small, and .

Note that the standard formula described above in the text, which is Equation 5, is an approximation. We should have written .

We can now put Equation 5 to work. Say government officials want to calculate the primary deficit or surplus necessary to keep the country’s debt-to-GDP ratio constant. They then set  = 0, and solve Equation 5 to obtain:

Thus, the primary deficit compatible with keeping the debt-to-GDP ratio stable will rise with the GDP growth rate , but fall with the interest rate on government securities . To understand this, notice that faster GDP growth reduces the debt burden by raising the denominator of over time, and a higher interest rate means that there are fewer resources to devote to debt retirement (or the need to issue additional debt to meet interest payments). Thus, is critically important for debt dynamics. ‘Find out more: The importance of r − g in the history of debt dynamics’ summarizes the global historical record.

Find out more The importance of r − g in the history of debt dynamics

As we’ve seen, the dynamics of public debt depend importantly on the difference between the real (inflation-adjusted) interest rate and the economy-wide rate of economic growth . The financial historian Paul Schmelzing has heroically assembled information on government bond yields for the now-advanced (high-income) economies (which include Europe and, in the subperiod from when it existed as a sovereign nation, the U.S.) and estimated growth rates over the last eight centuries (Figure 5). There’s lots of volatility, but shows a tendency to trend downward over the long sweep of historical time.

In this line chart, the horizontal axis shows the years from 1317 to 2017 and the vertical axis shows the difference between the real interest rate and the GDP growth rate, denoted as r minus g and ranging from −10% to 25%. r minus g has decreased over time, from around 10–15% in the 1300s and early 1400s, to 2–12% in the 1500s, 1600s and 1700s, to −10–0% in the 20th century.

Figure 5 700 years of global for the now-advanced (high-income) economies.

Own elaborations based on data from Schmelzing, 2020.

Several factors contributed to this development. The cost of borrowing declined as sovereign debt came to be recognized as an obligation of the state rather than of the individual occupying the throne, with the establishment of parliaments and legislatures in which creditors were represented, and with the creation of markets in which debt securities could be bought and sold. Meanwhile, growth rates went up with the industrial revolution and the transition to modern economic growth in the nineteenth and twentieth centuries.

There has been a further fall in in high-income economies in the last four decades, prompting debate about whether prevailing levels of debt are becoming less worrisome. To some extent, current low interest rates on government debt reflect high savings rates in emerging economies such as China and the high savings propensities of the wealthy, who have enjoyed disproportionate income gains in high-income economies. In addition, they reflect a safe-asset shortage due to the value investors attach to safe and liquid government bonds in a volatile economic and financial environment. Whether these favorable conditions for public-debt management will continue to prevail going forward is anyone’s guess.

Imagine now that the growth rate of the economy is larger than the interest rate. The country then can run a primary deficit (a positive value of ) and still keep its debt-to-GDP ratio from rising. For instance, if is 3%, is 1% and the initial debt-to-GDP ratio is 50%, the country can run a primary deficit of 1% of GDP without its debt-to-GDP ratio rising.

The opposite is true, however, if the interest rate is higher than the growth rate. If  = 1%,  = 3%, and is again 50%, then keeping the debt ratio stable will require a primary (budget) surplus of 1% of GDP ( will have to be −1%). The surplus required to stabilize the debt will increase with the size of the initial debt ratio. For instance, with  = 1%,  = 3%, and  = 100%, debt stabilization now requires a primary surplus of 2% of GDP. A primary surplus smaller than that will cause the debt ratio to rise without limit.

Another special case is when  = . Then, as Equation 5 shows, the increase in the debt-to-GDP ratio is equal to the primary deficit ratio . In other words, the debt ratio remains constant when primary expenditure (non-interest spending) is limited to tax revenues.

Using the public debt simulator

In the public debt simulator, in SIM1, you can choose your own values of , and to compute the debt stabilizing primary deficit.

Follow the steps in Figure 6 to understand what happens to government debt in four different scenarios.

Figure 6 Geometric representation of the debt dynamic equation.

Panel A: r  > g and d > 0
: In this diagram, the horizontal axis shows the initial level of debt, denoted b_{t−1}, ranging from negative to positive values, and the vertical axis shows the change in debt, denoted delta b, ranging from −2 to 6%. The evolution of debt is an upward-sloping line that intersects the line at a positive value, denoted , and has a slope of r minus g.

Panel A: r  > g and d > 0

This diagram shows the initial level of debt on the horizontal axis and the change in debt on the vertical axis. We can use a line to show how debt will change, given the interest rate and the growth rate. The vertical intercept of the line is the primary deficit, and the slope of the line shows the difference between the interest rate and the growth rate. In this case, the interest rate is higher than the growth rate, so the debt line slopes upwards. When the change in debt (vertical axis value) is negative, debt is decreasing, and when the change in debt is positive, debt is growing. The horizontal axis intercept (point A) shows the initial level of debt at which the debt stays the same over time (change in debt is zero). In this case, the initial level of debt required is negative, so for any initial level of debt greater than 0, the change in debt is positive and debt will grow. We move north-east along the debt line, as shown by the arrows.

Panel B: r > g and d < 0
: In this diagram, the horizontal axis shows the initial level of debt, denoted b_{t−1}, ranging from negative to positive values, and the vertical axis shows the change in debt, denoted delta b, ranging from −2 to 6%. The evolution of debt is an upward-sloping line that intersects the line at a negative value, denoted , and has a slope of r minus g.

Panel B: r > g and d < 0

As in the previous case, the interest rate is higher than the growth rate, so the debt line slopes upwards. However, debt is stable at point C. This is because the government is now running a primary surplus ( < ) which compensates for the difference between interest rate and growth rate. Notice that the vertical intercept of the line is negative. The direction of the arrows shows that for any change in debt , debt will grow (north-east) or reduce in size (south-west).

Panel C: r < g and d > 0
: In this diagram, the horizontal axis shows the initial level of debt, denoted b_{t−1}, ranging from negative to positive values, and the vertical axis shows the change in debt, denoted delta b, ranging from −2 to 6%. The evolution of debt is a downward-sloping line that intersects the line at a positive value, denoted , and has a slope of r minus g.

Panel C: r < g and d > 0

The interest rate is now lower than the growth rate, so the debt line slopes downwards. In this case, debt is only stable at point F. Note that debt can be stable even if the government is running a primary deficit (). The vertical intercept of the line is positive. The direction of the arrows shows that for any level of initial debt, this will converge to the level shown by point F.

Panel D: r < g and d < 0
: In this diagram, the horizontal axis shows the initial level of debt, denoted b_{t−1}, ranging from negative to positive values, and the vertical axis shows the change in debt, denoted delta b, ranging from −2 to 6%. The evolution of debt is a downward-sloping line that intersects the line at a negative value, denoted , and has a slope of r minus g.

Panel D: r < g and d < 0

As in the previous case, the interest rate is lower than the growth rate, so the debt line slopes downwards. Notice that for any initial level of debt, , debt will always contract and converge to point G. This is because the government is running a primary surplus ( < ), and the rate at which the economy is growing is faster than the rate at which debt is being paid back. Like in the previous case, the direction of the arrows shows that for any level of initial debt, this will converge to the level shown by point G.

Exercise 2 Plotting the dynamics of debt

For this exercise, you will need sheet SIM1 of the public debt simulator.

  1. Choose two countries from the list provided, such that the debt-to-GDP ratio continually increases for one and stabilizes for the other. Calculate the debt-stabilizing primary deficit for each country using the spreadsheet.
  2. Draw by hand the geometric representation of the debt dynamics for each country, as in Figure 6.
  3. Use your diagram to explain the dynamics of debt in each country. (Hint. You can check your working by using sheet SIM2 of the public debt simulator.)

Figure 7 uses this setup to analyze the evolution of the debt of a country that starts the year 2020 with a 50% debt-to-GDP ratio (SIM3 of the public debt simulator).

The purple line plots a situation in which  =  = 0% and  = 2%. The economy is not growing and the government is not running a primary deficit, but the interest rate on the debt is positive. Given that , and  = 0%, the debt-to-GDP ratio will rise to 51% in 2021, 52.02% in 2022, and 80.4% in 2045. The debt ratio rises exponentially (it rises by larger amounts in later years) because in every successive year the difference between and is applied to a higher level of debt.

The orange line in Figure 7 depicts the situation where and the country runs a deficit ( = 1%,  = 0%,  = 2%). With primary spending now exceeding revenue, the debt ratio rises even faster than when  = 0%.

The red line ( = 1%,  = 2%,  = 1%) is when . Since the country is growing and , the denominator of the debt-to-GDP ratio grows faster than it accumulates interest obligations, so the government can run a primary deficit and still maintain a constant debt ratio. When  = 1% and is 50%, as here, the debt stabilizing primary deficit is 0.5%. Since the country runs a primary deficit of 1% of GDP in the case shown, debt will increase until it reaches 100% of GDP. At this point it will stabilize, since the condition  = () is finally met. But when , the debt-to-GDP ratio never grows to infinity (see also Panel C of Figure 6).

The blue line ( = 1%,  = 3%,  = 1%) is when  = (). In this case, the debt remains constant at its initial value of 50%.

The green line ( = −2%,  = 2%,  = 3%) is when , but the country runs a surplus that more than compensates for this difference, causing the debt-to-GDP ratio to decrease rapidly.

SIM3 allows you to enter your chosen values for , and and simulate the path of the debt-to-GDP ratio over a 50-year period. The same simulation sheet also provides values for these variables for a sample of emerging and high-income economies.

In this line chart, the horizontal axis shows years from 2020 to 2044, and the vertical axis shows the debt-to-GDP ratio as a percentage, ranging from 0 to 120. Five scenarios are shown, all with an initial debt-to-GDP ratio of 50%. In scenario 1,  = 1%,  = 0%, and r = 2%, and the debt-to-GDP ratio increases over time to 110% in 2044. In scenario 2,  = 0%,  = 0%, and r = 2%, and the debt-to-GDP ratio increases over time to around 80% in 2044. In scenario 3,  = 1%,  = 2%, and r = 1%, and the debt-to-GDP ratio increases slightly to around 60% in 2044. In scenario 4,  = 1%,  = 3%, and r = 1%, so the debt-to-GDP ratio remains constant over time. In scenario 5,  = −2%,  = 2%, and r = 3%, and the debt-to-GDP ratio decreases over time to around 10% in 2044.

Figure 7 Evolution of the debt-to-GDP ratio under different assumptions for , , and .

Factoring in inflation

So far, we worked with real variables and did not consider the role of inflation. However, in some circumstances, inflation can play a key role in the evolution of the debt-to-GDP ratio. Fortunately, it is straightforward to rewrite Equation 5 in terms of inflation () and the nominal interest rate () instead of the real interest rate:

The relationship between the nominal interest rate and expected inflation is described by the Fisher equation, which states that , where is expected inflation. For more details, see Section 15.1 of The Economy.

Fisher equation
The relation that gives the real interest rate as the difference between the nominal interest rate and expected inflation: real interest rate = nominal interest rate – expected inflation.

Equation 7 shows that, other things being equal, an increase in inflation will reduce the debt-to-GDP ratio. This is because when we compute the debt-to-GDP ratio, we use nominal GDP which is affected by both real growth () and inflation (). However, other things are rarely equal. Higher than expected inflation will lead to an increase in the nominal interest rate. This increase in the nominal rate will compensate for the debt reduction effect of higher inflation. In fact, if we define unexpected inflation as u as the difference between expected inflation (e), and actual inflation (), we can use the Fisher equation to rewrite Equation 7 as + (u) . This clarifies that, in the absence of financial repression (see Section 7), only unexpected inflation can reduce the debt-to-GDP ratio.

A higher deficit can be driven by higher expenditure or lower taxes (or both). The belief that tax cuts can pay by themselves is one of the basic tenets of supply side economics, as you can read in Foundations of Supply-Side Economics by Victor Canto, Douglas Joines, and Arthur Laffer. For a critical evaluation, see ‘Evidence on the High-Income Laffer Curve from Six Decades of Tax Reform’ by Austan Goolsbee, or ‘The Elasticity of Taxable Income with Respect to Marginal Tax Rates: A Critical Review’ by Emmanuel Saez, Joel Slemrod, and Seth Giertz.

An important complication concerns the link between deficit spending and GDP growth. Issuing debt to finance productive investment expenditure can have a positive effect on long-run GDP growth. Similarly, countercyclical deficit spending can have a positive effect on GDP growth in the short run. In Equation 7, a larger deficit will lead to a proportional increase in the debt-to-GDP ratio, other things being equal. But when debt is issued to undertake actions that affect the evolution of GDP, other things are no longer equal. Those actions will raise the denominator of the debt-to-GDP ratio. Faster GDP growth will lead to higher tax revenues. In theory, there are even conditions in which a higher deficit could reduce the debt-to-GDP ratio.

SIM4 of the public debt simulator allows you to enter values for , , , and and simulate the debt-to-GDP ratio over a 50-year period.

Question 3 Choose the correct answer(s)

Which of the following statements about debt dynamics are true?

  • If , the debt-to-GDP ratio will always increase.
  • The nominal value of the debt the government has to repay is reduced by a higher rate of inflation.
  • It is not possible for a government to run a primary deficit and keep its debt-to-GDP ratio constant.
  • If the rate at which the economy grows is lower than the interest rate on debt, it is still possible for debt-to-GDP ratio not to increase.
  • Looking at Figure 6 Panel B, for levels of initial debt below C, the debt-to-GDP ratio will decrease. The reason is that the primary surplus is sufficient to offset the degree to which is increasing the interest burden.
  • Inflation does not affect the nominal value of the debt since it is fixed in nominal terms. Inflation affects the real burden of the debt.
  • The blue line in Figure 7 shows that when , a government can run a primary deficit and keep its debt-to-GDP ratio constant.
  • Figure 6 Panel B shows that the debt-to-GDP ratio can decrease even when because the country runs a surplus that more than compensates for this difference. This is also shown by the green line in Figure 7 ( = −2%,  = 2%,  = 3%).

Question 4 Choose the correct answer(s)

Suppose that a country’s debt-to-GDP ratio is 100%, the economy is growing at a rate of 1%, and the government is intending to run a 1% deficit. Which of the following statements are true?

  • The debt-to-GDP ratio can be kept constant if inflation is at the same rate as the interest rate on debt.
  • In an economy that is expanding, when  = 0%, a small change in debt will cause a debt-to-GDP ratio that increases without limit.
  • Even if the current real interest rate is above the economy’s growth rate, a forward-looking government could justify an increase in its borrowing on economic grounds.
  • A spike in inflation will make it more difficult for the government to reduce its debt.
  • From Equation 5 combined with the Fisher equation, we know that when . Using the numbers provided: , which is only true when and have the same value.
  • When   = 0, we know that . From Figure 6, we can see that a slight change in debt will not cause it to increase without limit.
  • The government may anticipate a lower real interest rate in the future; its borrowing may finance investment that is expected to raise the economy’s growth rate.
  • Unexpected inflation, i.e. inflation that is not reflected in the nominal interest rate, reduces the real burden of debt. The nominal value of the debt will not change, and it will become easier to pay back in real terms. Recall that by the Fisher equation, .

Exercise 3 Using the debt dynamics equation

Go to the OECD Data website. For this exercise, make sure you download the data in yearly frequency and as a percentage of GDP when necessary. For this exercise, you will need to download data for the following variables:

  • nominal long-term interest rates
  • inflation rates
  • real GDP growth (hint: look for quarterly GDP and select a yearly frequency)
  • primary deficit and debt-to-GDP ratios (hint: look for general government deficit and general government debt, respectively); note that on the OECD Data website, a positive value of the primary deficit ratio means that the government is running a surplus.
  1. Find and download data for the long-term interest rate and inflation rate of a country of your choice, over the longest time span for which data is available. Using the Fisher equation, calculate the real interest rate for each year observed.
  2. Plot the real interest rate on a line chart. How does the trend you observe in the real interest rate compare with that described in Figure 4?
  3. Now add the growth rate of real GDP to your chart. From a quick look at your chart, in which periods would you expect debt to expand? In which periods would you expect it to contract?
  4. Finally, add to your plot the primary deficit and the debt-to-GDP ratio. Does the debt-to-GDP ratio follow your predictions from Question 3? (Hint: For clarity, you can plot the debt-to-GDP ratio adding a secondary axis.)
  5. Take the average values of all the variables you plotted over the most recent available 10 years in the data. Use these values to predict how debt for your chosen country will behave in the future. You can use Figure 6 to help you comment on its likely trend over time. (Hint: You can check your prediction by using the public debt simulator, sheet SIM3.)
  6. Discuss whether you think that your prediction is reasonable.

6 The unexplained part of public debt

When working with real-world data, sometimes the left- and right-hand sides of Equation 7 don’t match. The statistics published by the government for may be larger or smaller than one would expect from the values of , , , and . This can happen because the deficit and interest payments are flows (they are measured over a period of time), whereas debt is a stock (it is measured at a point in time).

Public-debt analysts accommodate this fact by appending another term to the equation:

The term ‘off-budget resources’ refers to expenditures that are not included in the government budget. For instance, the government may set up a special entity to borrow money for a particular task (such as to finance bank recapitalization or a specific investment project). Since borrowing is not done directly by the government, it is often not included in the government budget. There are cases, however, in which the government needs to borrow directly to recapitalize such off-budget entities and this borrowing can lead to sudden jumps in public debt which are not reflected in budgeted public expenditures. For details, see ‘Underground Government: The Off-Budget Public Sector’ by James Bennett (2004).

The stock-flow reconciliation is represented by sf. This cumbersome name comes from the fact this residual item exactly reconciles the deficit, interest payments, and the change in GDP with the stock of debt. This term is also referred to as ‘the unexplained part of public debt’.

Equation 7 works well enough, obviously, when such discrepancies are small. At times, however, they can be large. This will be the case, for example, when a banking crisis forces the government to use off-budget resources to inject funds into the banking system, or when it recapitalizes a large state-owned corporation. It can be the case when public debt is denominated in foreign currency and the exchange rate depreciates, causing the value of the debt to shoot up relative to GDP (which is denominated in domestic currency). ‘Find out more: The case of COVID-19’, where we use Equation 8 to decompose the increase in debt ratios during the COVID-19 pandemic, shows that this reconciliation can be important in practice.

Find out more The case of COVID-19

Governments around the world increased public spending and ran substantial budget deficits in response to the COVID pandemic, leading to sharp increases in debt-to-GDP ratios. In the advanced economies, debt-to-GDP ratios rose from about 100% in 2019 to more than 120% in 2020. In emerging and developing economies, ratios went from an average of 54% to 63% (see Figure 8).

Figure 9 uses Equation 8 to decompose the growth of the debt ratio into the contributions of the primary deficit, interest payments, inflation, real GDP growth (in this case, negative growth), and the stock-flow reconciliation. The variables above the zero line increased the debt-to-GDP ratio, while the variables below the line reduced it. The figure shows that the two main drivers of the increase in the debt-to-GDP ratio were the increase in deficits and contraction of GDP. The red bars show that the primary deficit added 10 percentage points to the debt ratio in the advanced economies and 7.5 percentage points in emerging and developing economies.

Negative growth (the blue bars) also contributed to the increase in the debt-to-GDP ratio, where the effect was again larger in the advanced economies. Note that if growth had been positive instead of negative, the blue bars would have been in the bottom part of the graph. This difference reflects two elements. Firstly, the GDP contraction was smaller in emerging and developing economies (2 percentage points versus nearly 5 percentage points). Secondly, the initial level of debt was smaller in emerging and developing economies (remember that, in Equation 8, GDP growth is interacted with the initial debt/GDP ratio). In particular, the contribution of negative growth to the increase in the debt ratio is high in Latin America both because the recession was deeper than in other emerging economies and because the initial debt level was higher.

Finally, in all regions inflation played some role in mitigating the growth of debt.

In this bar chart, the horizontal axis shows seven regions or types of economies, and the vertical axis shows the debt-to-GDP ratio as a percentage, ranging from 0 to 140. In 2019, the debt-to-GDP ratio was around 105% in advanced economies, 56% in emerging and developing economies, 58% in developing Asia, 30% in developing Europe, 70% in Latin America, 50% in the Middle East, and 52% in sub-Saharan Africa. In 2020, the debt-to-GDP ratio was 122% in advanced economies, 62% in emerging and developing economies, 64% in developing Asia, 38% in developing Europe, 78% in Latin America, 56% in the Middle East, and 58% in sub-Saharan Africa.

Figure 8 Debt-to-GDP ratio in advanced and emerging economies.

Authors’ elaborations based on IMF Data.

In this bar chart, the horizontal axis shows seven regions or types of economies (advanced economies, emerging and developing economies, developing Asia, developing Europe, Latin America, Middle East, and Sub-Saharan Africa). The vertical axis shows five measures: negative real GDP growth, inflation, interest payments, the primary deficit, and the stock-flow reconciliation, expressed as contribution to the debt-to-GDP ratio in percentage points, ranging from negative 10 to 25. For all regions and types of economies, negative GDP growth, primary deficits, and interest payments contributed to the debt-to-GDP ratio, while inflation helped reduce the debt-to-GDP ratio. In all regions except for Developing Asia, the stock-flow reconciliation reduced the debt-to-GDP ratio.

Figure 9 Drivers of the increase of the debt-to-GDP ratio over 2019 to 2020.

Authors’ elaborations based on IMF Data.

7 How governments repay

Recall the debt dynamics equation, which reads:

This highlights three ways governments can prevent their debts from becoming unsustainable (that is, that they can prevent the debt-to-GDP ratio from rising explosively and spiraling out of control). Firstly, they can run primary surpluses and use the excess revenues to retire (that is, repay) outstanding debt. In other words, they can ensure that the deficit ratio, , is negative.

However, there are relatively few recent instances where governments have succeeded in running substantial surpluses for extended periods. One study found just three cases of countries that ran surpluses of 5% of GDP for 10 consecutive years.7 When revenues rise, governments face pressure from voters to increase spending and not just retire debt. Economic shocks, from recessions to financial crises and pandemics, can throw the effort to maintain primary surpluses off course.

austerity
A policy where a government tries to improve its budgetary position in a recession by increasing its saving.

In addition, a sudden shift from deficit to surplus (for example, by undergoing a period of austerity) risks precipitating a recession by depressing aggregate demand. Efforts to reduce the debt-to-GDP ratio that end up only reducing GDP can be counterproductive even from the narrow standpoint of fiscal consolidation. Thus, euro area countries, having borrowed in response to the Global Financial Crisis of 2008–2009, sought to reduce those deficits starting in 2010 before recovery from recession was secure. The result was that the euro area slid back into recession starting in June 2011, causing the debt-to-GDP ratio to rise rather than fall.

For more details on the effects of austerity policies, see Section 14.6 of The Economy.

This points to a second way of stabilizing or reducing debt relative to GDP, through economic growth () that raises the denominator of the ratio. The sharp reduction in the U.S. debt-to-GDP ratio after the Second World War shown in Figure 2, like similar reductions in other high-income economies, was aided by fast economic growth from 1945 to 1975 (known in France as Les Trente Glorieuses, or the thirty glorious years).

Replicating that growth success today, unfortunately, is unlikely. In many countries, labor supply is growing more slowly. Productivity growth has slowed since the mid-1970s.

A third way of stabilizing or reducing the debt ratio is by keeping the interest rate on government debt securities () low. Governments do this by cultivating a reputation for repaying, so that investors will have no reason to demand a risk premium on top of the risk-free interest rate. Central banks can act as lenders and liquidity providers of last resort, limiting bond price volatility for which investors require additional compensation.

After the Second World War, policy makers also used statutory regulation to depress . In the U.S., they placed ceilings on the interest rates that banks were permitted to offer on time deposits. This encouraged depositors to shift into treasury securities, driving their prices up and yields down. Economists refer to these regulations and policies depressing as ‘financial repression’.

Financial markets today are more lightly regulated, however, limiting scope for financial repression. International capital mobility is higher than after the Second World War, enabling investors faced with artificially low interest rates in one national market to shift their funds to another.

Light-touch regulation and international capital mobility also limit the scope for using inflation to reduce the value of the debt relative to GDP. Inflation might seem like a foolproof way of raising the denominator of the debt-to-GDP ratio. But if it leads to a commensurate increase in the cost of servicing that debt (in the nominal interest rate), then the government saves nothing. In countries such as Brazil, where the vast majority of public debt is either short term (maturing in a year or less) or indexed (where the contract specifies that payments rise one-to-one with inflation), there is little scope for inflating away the debt.

Question 5 Choose the correct answer(s)

Based on the information in this section, which of the following statements are true?

  • Primary (budget) surpluses are not a viable option to reduce debt.
  • If the economy grows faster, the debt-to-GDP ratio will be reduced solely because of an increase in GDP in the denominator.
  • A trustworthy government will likely face a lower real interest rate.
  • Governments can easily turn to inflation to reduce the real interest rate it faces.
  • They are theoretically a viable option to reduce debt (look at Equation 5), although they may be difficult politically to implement.
  • GDP in the denominator plays a role. However, increases too, decreasing the difference between and in Equation 5.
  • If a government establishes a good reputation for duly paying back debt, investors will be less inclined to demand a higher interest rate on debt.
  • Increases in inflation may lead to increases in the nominal interest rate, thus keeping unchanged.

8 And what happens when they don’t?

If governments fail to take the necessary measures to reduce debt, they may eventually renege on their debts. They can announce an inability to meet their contractual obligations and just stop paying. They will be declared to be in default by the International Swaps and Derivatives Association, the organization that oversees trading of credit default swaps (derivative securities that provide insurance against defaults). But because national governments enjoy a degree of immunity from being sued in foreign courts, foreign creditors have relatively little legal recourse.

Governments tend to be reluctant to default on domestic creditors, since the latter do, in fact, possess recourse: they can eject the defaulting government from office. Hence, governments are more likely to use indirect means to reduce the value of domestic debts. Specifically, they may attempt to inflate them away. A classic example is Germany after the First World War, when the government had a debt-to-GDP ratio of 100%. By running a very high rate of inflation, that debt was liquidated. As shown in Figure 10, its real, purchasing-power value was reduced to zero.

In this line chart, the horizontal axis shows the years from 1910 to 1938 and the vertical axis shows the debt-to-GDP ratio of Germany as a percentage, ranging from 0 to 120. The debt-to-GDP ratio remained steady at around 10% from 1910–1914, sharply increased to 100% in 1919, and rapidly decreased to 0% in 1923. It then increased gradually to 20% by 1938.

Figure 10 Public debt ratio in Germany, 1910–1938.

Eichengreen, El-Ganainy, Esteves & Mitchener, 2021. In Defense of Public Debt. New York: Oxford University Press.

If there is an easy way out, then why don’t governments always default or inflate away their debts? The obvious explanation is potential for retaliation by the creditors. Investors may refuse to buy treasury bonds in the future, or they may demand a substantial risk premium. Germany again illustrates this point: the Weimar government had to pay very high interest rates when seeking to resume borrowing after 1923.

But adverse financial consequences don’t always deter opportunistic behavior by indebted governments, which is why sovereign defaults are common in history. Some economists suggest that in many cases, a government’s immediate savings on interest and principal payments are likely to dominate any costs it incurs as a result of being penalized by investors at some future date.

That we see governments continuing to pay and investors continuing to lend therefore implies that there are other, indirect costs of default. In the nineteenth century, those costs took the form of gunboat diplomacy. Creditor-country governments would send in troops to seize the resources needed to make payments. In the 1930s, countries that defaulted were subject to retaliatory tariffs by creditor-country governments. After the Second World War, the doctrine of sovereign immunity was weakened, allowing the creditors to sometimes seize planes and vessels belonging to the national airline or government when these landed or docked outside the country. (See ‘Find out more: Debt restructuring’ for more details.) Foreign banks, seeing the government as unreliable, can stop providing trade credit, so exports and economic growth may suffer.

Find out more Debt restructuring

When a household or firm is unable to pay a debt, its finances are reorganized and the creditors receive compensation through a legal proceeding—that is, through the deliberations of a bankruptcy court. Not so when a sovereign government is unable to pay. If the debt contract is issued under domestic law, then the government can simply adopt new legislation changing its terms or repudiating it entirely—end of story. Even if it is subject to a foreign law, the creditors still have limited legal redress, since under the prevailing doctrine of sovereign immunity, sovereign states can be sued only with their permission, which they rarely grant.

free ride
Benefiting from the contributions of others to some cooperative project without contributing oneself.

The creditors’ only option is to negotiate. But a number of factors can make it hard to reach agreement. Firstly, sovereign debtors are better able to judge how much adjustment effort is economically and politically possible and therefore what they can offer their creditors. The creditors may therefore reject the government’s proposal as an attempt to underpay them even when that offer is realistic. Secondly, there can be conflicts of interest among the creditors. Opportunistic investors, sometimes referred to as ‘vulture funds’, may buy up bonds in default for a few cents on the dollar and demand full repayment, in effect ‘free riding’ on the debt relief agreed by others. They may try to convince a compliant court that the doctrine of sovereign immunity doesn’t apply. For example, following a 2001 default, the government of Argentina famously waged a 15-year battle with its holdout creditors. A defining moment was when a subsidiary of the U.S. hedge fund Elliot Capital Management managed to convince a Ghanaian court to seize a three-masted tall ship visiting the country as part of a goodwill mission by the Argentine government.

A number of mechanisms have been developed for addressing these problems. The International Monetary Fund may use its data-gathering capacity to help overcome information asymmetries. It may extend bridge loans to tide over countries acting in good faith but facing delays in negotiations. Legislators and regulators in the principal financial centers may pass laws and rules weakening the position of holdout creditors. Issuers may add renegotiation-friendly contractual provisions to their loan agreements, such as the majority action clauses and pari pasu clauses requiring all creditors to be treated equally. Creditors may form committees in an effort to solve coordination problems.

There have also been proposals for the creation of an international bankruptcy court with the power to ‘cram down’ settlement terms on both sides. Given the absence of a global government to oversee it, however, these proposals have come to nothing.

Exercise 4 Causes and consequences of sovereign default

Use the following sources to answer the questions below. You may also want to do your own research to find more background readings on this topic.

  1. Provide one example of a debt default that was preceded by a recession in the domestic economy, and one example where the default resulted from political events. In each case, discuss the economic consequences of the default.
  2. Are governments punished for defaulting on public debt? Should they be?

9 Rules and institutions for public debt management

Maintaining sustainable debt levels can be challenging for political as well as economic reasons. Governments have taken two approaches to this problem: budgetary rules on the one hand, and institutions and procedures on the other. Rules may be embodied in a legal statute or a country’s constitution. Germany’s ‘debt brake’, for example, was written into the Federal Republic’s constitution in 2009. This limits annual federal government borrowing to no more than 0.35% of GDP. Articles 121 and 126 of the ‘Treaty on the Functioning of the European Union’ (originally the Treaty of Rome) and associated protocols commit EU member states to limiting their debt and deficits to 60% and 3% of GDP, respectively. In other jurisdictions, rules set separate targets or limits for the growth of government spending and/or revenues.

For further details on the Greek debt crisis in 2009, you can read Greece’s Debt by the Council on Foreign Relations; or Breaking the Greek debt impasse by Barry Eichengreen, Peter Allen, and Gary Evans; or Putting the Greek debt problem to rest by Barry Eichengreen, Emilios Avgouleas, Miguel Poiares Maduro, Ugo Panizza, Richard Portes, Beatrice Weder di Mauro, Charles Wyplosz, and Jeromin Zettelmeyer.

The strength of fiscal rules is their simplicity. Everyone knows what is meant by 60% of GDP, and monitoring compliance is relatively straightforward. Or is it? The Greek debt crisis in 2009 was triggered when a new government announced that its predecessor had cooked the books and that the public-sector debt and deficit were significantly higher than acknowledged previously. And as we saw earlier, there are many different ways of calculating the debt.

The weakness of rules is their arbitrariness. No single limit on the deficit or debt is right for all countries at all times. There is no single threshold where debt is too much—where it suddenly becomes a drag on growth. (See ‘Find Out More: Public debt and economic growth’ for further discussion.) The EU’s 60% ceiling on the debt ratio just happened to be the average in Europe when the Maastricht Treaty was negotiated, and a 3% deficit just happened to be the level needed to keep that debt ratio constant, given then prevailing interest rates and growth rates. Interest rates, growth rates and debt ratios are all very different today, of course. An arbitrary rule not tailored to circumstances is unlikely to garner respect and compliance. Thus, both Germany and France violated the Maastricht Treaty’s 3% ceiling on deficits within five years of the euro’s creation.

Find out more Public debt and economic growth

Does too much public debt hamper economic growth? High levels of debt could put upward pressure on interest rates and crowd out private investment. They could increase uncertainty about future taxation and inflation. If heavy debts make it difficult for the government to issue additional debt, they could limit the pursuit of countercyclical fiscal policies.

For surveys of the literature on these questions, see ‘Public Debt and Economic Growth in Advanced Economies: A Survey’ by Ugo Panizza and Andrea Presbitero; ‘The 90% Public Debt Threshold: The Rise and Fall of a Stylized Fact’ by Balázs Égert; or ‘Do Higher Public Debt Levels Reduce Economic Growth?’ by Philipp Heimberger.

In an influential 2010 article, the economists Carmen Reinhart and Kenneth Rogoff concluded that high levels of public debt are negatively correlated with economic growth once debt exceeds 90% of GDP. Subsequent research reanalyzed their data and criticized their findings. Those subsequent studies did not find consistent evidence of threshold effects at 90% or other levels. It still makes sense, of course, that at some point, public debts become so heavy that they constitute a burden for society and for growth. But the debt-to-GDP ratio at which any negative effects begin to become evident will vary with country conditions. Equation 5 in Section 5 shows that the level of debt is less likely to be a concern when the growth rate of the economy is high and when the interest rate is low. The level of debt at which negative effects begin to become evident will be higher when a country has the economic, financial, and political wherewithal to mobilize a higher share of GDP in the form of taxes and run primary surpluses. But analyzing the growth-debt nexus requires careful study of specific country cases, not a search for magic numbers.

causality
A direction from cause to effect, establishing that a change in one variable produces a change in another. While a correlation is simply an assessment that two things have moved together, causation implies a mechanism accounting for the association, and is therefore a more restrictive concept.

An important issue is causality. Even if there is evidence that public debt is, on average, negatively correlated with economic growth, correlation does not necessarily imply causation. Low economic growth could simply be leading to high levels of debt. Alternatively, the observed correlation between debt and growth could be due to a third factor that affects both debt and growth. Again, this points to the need for caution before making general statements about the effect of public debt levels on economic growth.

Another weakness of fiscal rules is that if rules are too rigid, they will not accommodate the special circumstances of a recession, financial crisis, or pandemic. But if they are too flexible and easily relaxed, they will lack credibility. Observers expecting exceptions to be invoked will doubt that the rules will be enforced. The EU has tried to strike a balance between credibility and flexibility, enhancing the first by adding to its own agreement a version of the German debt brake, while enhancing the second by making provision for the business cycle when gauging countries’ compliance with its rules. This balancing act remains a work in progress.

autonomous fiscal council
A nonpartisan, fiscal watchdog providing the government with independent estimates of the public finances and their sustainability.

The alternative is strengthening institutions and procedures in order to increase transparency and avoid the kind of problems experienced by Greece in 2009. Countries can create a non-partisan fiscal watchdog such as the U.S. Congressional Budget Office, the Netherlands Bureau for Economic Policy Analysis, or the UK Office for Budget Responsibility to provide independent estimates of the public finances and their sustainability. They can create an autonomous fiscal council with its own budget, and staffed by experts serving long terms in office, to provide realistic forecasts of tax revenues and economic growth. They can require budgeting to incorporate those forecasts, thereby preventing governments from basing spending plans on unrealistic assumptions. Chile, for example, has an Autonomous Fiscal Council that calculates cyclical adjustments to observed budgetary outcomes, comments on possible deviations from announced structural balance targets, and proposes mitigation measures. A growing number of countries are moving in this direction.

Exercise 5 The role of fiscal councils

Look for the website of the autonomous fiscal councils in two countries of your choice. You can use the IMF Fiscal Council Dataset, for example, if you want to find out which countries have one.

  1. Briefly summarize the purposes of your chosen councils. How do they provide incentives for the government to manage its debt appropriately?
  2. The economist Beetsma and his colleagues suggest that countries with an autonomous fiscal council tend to do better at fiscal forecasts and compliance with fiscal rules. Why might a government not want to establish such a council? (Hint: You may find the article ‘What are fiscal councils, and what do they do?’ helpful.)
  3. Find a country that does not have an autonomous fiscal council. Might this have impacted the management of public finances?

Question 6 Choose the correct answer(s)

Based on the information in this section, which of the following statements are true?

  • The EU set a 60% debt-to-GDP and 3% deficit-to-GDP limit because these are the optimal amounts of debt and deficit relative to GDP.
  • Relatively high levels of debt can harm growth, but only if the level exceeds 90% of GDP.
  • Bending fiscal rules in extreme circumstances can give them credibility.
  • Autonomous fiscal councils decide the government’s budget.
  • Those values happened to be the prevailing values when the Maastricht Treaty was negotiated.
  • Relatively high levels of debt can harm growth, but there is no consensus on how ‘high’ they have to be to hamper growth.
  • Fiscal rules that are too strict risk not being accommodated when exceptional circumstances happen, for example during a recession. A certain degree of flexibility for these occasions can increase their credibility.
  • They are watchdogs, that is, bodies which provide independent estimates of tax receipts, growth, and public finances—so their role is a monitoring one.

10 Conclusion

Public debt allows governments to mobilize resources to meet emergencies, from wars, pandemics, and natural disasters to financial crises and recessions. Costs can be spread over time, avoiding the need to raise taxes to highly distortionary levels. Public debt issuance enables the government to finance productive investments that enhance the economy’s capacity to grow and that generate a stream of returns, enabling the Treasury to pay interest and repay the principal. A market in public debt allows the authorities to finance budget deficits in economic downturns. It allows them to use fiscal policy as a macroeconomic stabilization and insurance tool.

But public debt can also be issued imprudently to advance the private ends of those in power. Unproductive borrowing that does nothing to enhance the economy’s capacity to grow can be a burden on future generations required to pay additional taxes to service inherited debts. Incumbent politicians may borrow in order to boost spending just prior to elections and increase their political chances. A political party when in office may borrow to boost spending on its preferred programs, knowing that if it is supplanted by the political opposition no such spending will occur. Managing public debt poses a ‘common pool problem’. Each special interest favors issuing a little more to finance increases in its preferred programs without internalizing its contribution to the total debt burden.

Countries devise political and institutional mechanisms to restrain such behavior. They impose numeral limits for debts and deficits, incorporating these into statutory laws and constitutional amendments. They create independent government agencies to assemble and publish accurate data on the public finances, informing policy decisions. They constitute independent fiscal commissions to advise on fiscal choices. Such arrangements increase the likelihood that the government’s borrowing capacity will be used prudently. But they do not work perfectly, which is why public debt, unavoidably, retains its mixed reputation.

11 Acknowledgements

The authors would like to thank the CORE editors who invited them to write this Insight and provided helpful suggestions and feedback.

CORE is grateful to Sam Glendenning, Simran Krishna-Rogers, and Elena Xu for their valuable feedback.

12 References

  1. Fiscal Monitor. 2021. ‘Strengthening the Credibility of Public Finances’ (October). Washington, DC: International Monetary Fund. 

  2. Thomas Jefferson. 1816. Letter to John Taylor Monticello May 28, 1816, para. 7. 

  3. Adam Smith. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. Volume 5, Book 3. 

  4. Eichengreen, Barry, Asmaa El-Ganainy, Rui Esteves, and Kris James Mitchener. 2021. In Defense of Public Debt. New York: Oxford University Press. 

  5. Marina Azzimonti. 2013. ‘The Political Polarization Index’. Working Paper 13–41. Federal Reserve Bank of Philadelphia. 

  6. Robert B. Talisse. 2019. ‘Political Polarization is about Feelings, not Facts’. The Conversation. Updated 31 July 2019. 

  7. Barry Eichengreen and Ugo Panizza. 2016. ‘A Surplus of Ambition: Can Europe Rely on Large Primary Surpluses to Solve its Debt Problem?’. Economic Policy 31(85): pp. 5–49.