Who will pay the COVID debt?

From the perspective of public finances, the COVID pandemic is reminiscent of a period of war. The sharp and temporary increase in government spending to combat the pandemic was financed in developed countries not from collected taxes, but from debt and money creation. Does this then mean that we will be passing on the “COVID bill” to the coming generations? This article shows that most of the new debt has already been offset by inflation and bond revaluations. COVID expenditure has thus already been partially “paid for” from the investment losses of government bond holders and need not become a burden we are leaving for future taxpayers. However, historical experience shows that the “post-war” years are also of key importance, with high interest and deflation compensating investors for their wartime capital losses. This effect is also apparent in the current high interest rates, yet it is far from targeted deflation. In the long run, it is therefore likely that, like after the Second World War, COVID debt will be reduced by a combination of inflation and growth, while the contribution of surplus budgets (i.e. future taxpayers) will not be large.

Published in Global Economic Outlook – October 2023 (pdf, 1.1 MB)

Who pays the state

“Our kids are never gonna forgive us for this.” In the  first debate of the candidates for the Republican Party’s nomination for President of the United States in August 2023, politician and former US ambassador to the United Nations Nikki Haley was referring to the dramatic increase in the country’s government debt since the start of the COVID pandemic.[1] Yet is Haley right that we are leaving the COVID debt burden to the coming generations?  “It’s time to put an accountant in the White House,” Haley concluded her speech. So let’s start from an accounting perspective.

Every state needs resources to perform its functions, and they acquire them in three ways: by taxing residents, by issuing bonds and by creating money. In all these cases, the state receives real resources – only the countervalue it offers for them is different. A taxpayer simply hands over part of their wealth to the state. By contrast, a creditor receives a bond for their money, i.e. receives a financial claim of the nominal value of the principal and additionally a promise of interest income. In the third case, the banknote is the countervalue. This is similar to a bond, the difference being that money does not bear interest but can be used to pay for things. Virtually every modern state relies on a combination of all three sources – taxation, bonds and money. However, their relative importance differs significantly, both across states and over time.

Chart 1 – The COVID pandemic brings short-term but dramatic growth in government expenditure
(general government expenditure as a share of GDP, %)

Chart 1 – The COVID pandemic brings short-term but dramatic growth in government expenditure
Source: Eurostat, ONS (UK), FRED (US), author’s calculation

Is debt or monetary financing of public budgets merely deferred taxation? In normal times, when inflation is low, interest rates do not change much and governments meet their obligations to creditors, bond yields are predictable and money retains relatively stable purchasing power. In such a situation, the government’s debt rises and falls depending on whether the budget is in deficit or surplus, and public finances can only be healed by increasing tax revenues or cutting expenditures. Public debt can indeed be seen as a “burden” that today’s taxpayers create for future taxpayers.[2]

However, in turbulent times, such as wars in the past in particular, this may not be the case. This is because the taxpayers share the burden with holders of government bonds. A clear example is the situation in which the state fails to honour its obligations to creditors (e.g. by declaring partial or total insolvency). Yet investors may also suffer losses even in less extreme situations, such as when the market value of the government bonds they hold falls or when unexpected inflation erodes the real value of both principal and interest.

The COVID bill has already arrived

The COVID-19 pandemic led to a sharp increase in government spending (Chart 1). This short-term but dramatic increase in government expenditure was similar to what happens during a war. Hall and Sargent (2022), for example, show that the increase in federal spending in the USA in 2020 and 2021 was similar to that in the two world wars. Likewise, the proportion of people who did not work during the COVID shutdowns and received contributions from the state was comparable to the proportion of the population that served in the army during the Second World War.

As government revenues did not increase, the COVID “war” expenditure led to a significant increase in the already high debt levels in developed countries. The way the pandemic expenditure was financed thus also has a close parallel in past wars. There has been practically no increase in taxes to cover the COVID expenditure. All developed countries financed the expenditure shock by issuing new bonds and money.[3]

It may therefore seem strange that financial markets and rating agencies have shrugged off the risks of public debt. The price of CDS credit derivatives, which are de facto insurance against sovereign default, actually decreased for the median developed country in 2020 and remained below pre-pandemic levels in 2021 and 2022 (Chart 2). A similar picture is painted by rating agencies, which improved the ratings of developed countries more often than downgrading them over this period.[4] Part of the explanation may lie in the record easing of monetary policies during COVID, pushing the interest costs of public debt to historic lows. Yet why did nothing change even in 2022, when central banks changed course and initiated the sharpest increase in the cost of borrowing since the 1980s?

Chart 2 – In spite of the sharp rise in public debt, insurance against sovereign defaults has not become more expensive
(CDS spread in bp; deviation from the 2015–2019 average; developed countries, 2020–2023)

Chart 2 – In spite of the sharp rise in public debt, insurance against sovereign defaults has not become more expensive
Source: Refinitiv, author’s calculation

This is because most of the “COVID bill” has already been repaid. So it will not necessarily require future increases in taxation or cuts. One useful example is provided by the USA. Chart 3 shows the evolution of the federal debt from the beginning of 2019 to the first quarter of 2023. The blue line shows that if neither the size of the economy nor prices had grown since 2019, the volume of debt would have risen from 100% of GDP at the end of 2019 to almost 135% of GDP at the end of March 2023. The red line shows the ratio of debt to actual GDP, which first rose sharply but then fell again to 110% of GDP due to economic growth and (especially) inflation. The yellow line shows the same but uses its market value instead of the nominal valuation of the debt. This can be thought of as the amount the US government would need if it wanted to repay all of its debts.[5] Expressed in this way, the ratio of federal debt to the size of the economy is at roughly the same level as before COVID. In other words, if the USA wanted to repay its entire public debt, it would now need the same proportion of the country’s economic output as before it increased its debt during the “war” against COVID.

Chart 3 – Repayment of the US federal debt would require the same proportion of the country’s economic output as before the pandemic
(development of the US federal debt, % of GDP)
Chart 3 – Repayment of the US federal debt would require the same proportion of the country’s economic output as before the pandemic
Source: BIS, author’s calculation

How is this possible? COVID expenditure has been repaid by holders of government bonds through investment losses and inflation (Hall and Sargent, 2023a). During the COVID years, government bond investors lent to developed countries at extraordinarily low – and even negative – interest rates. The subsequent sharp rise in rates caused a sharp fall in the value of such securities, i.e. a loss for investors. The loss in real terms was exacerbated by inflation, which eroded the purchasing power of the currency in which the investments were denominated.

However, this effect was felt differently across countries depending on the size and structure of the government debt, interest rate developments and inflation. Chart 4 breaks down the changes in the debt-to-GDP ratio (market value) between the end of 2019 and the first quarter of 2023 for ten selected economies. The UK’s long average bond maturity, for example, has caused more severe revaluation than in other economies (because it will take investors much longer to swap the old, low-interest bonds for new, high-interest bonds). In Italy, France, Spain and Belgium, higher public debt contributed to greater revaluation. In Poland and Hungary, the real debt burden was pushed down mainly by inflation. In addition to inflation and revaluation, which shift the “burden” of COVID debt to investors, economic growth has also contributed to partial deleveraging in some countries. This does not shift the burden, but reduces it – making it easier for society as a whole to pay for the COVID expenditure.

Chart 4 – COVID expenditure has (so far) been paid by investors in government bonds
(breakdown of changes in public debt in market valuation to GDP, 2019 Q4 – 2023 Q1, pp)

Chart 4 – COVID expenditure has (so far) been paid by investors in government bonds
Source: BIS, ONS, FRED, Eurostat, author’s calculation

Investor participation in financing wars has been common in history. Hall and Sargent (2021 and 2023b) use examples of American and British wars since the 19th century to document the development of investment yields on government bonds during and after these wars. According to their research, up to the First World War it was common for wartime inflation to  wipe out much of the public debt (in real terms) and for investors to suffer large losses. In the post-war period, however, economic policy sought a return to the gold standard (i.e. pre-war prices). The combination of deflation and high interest rates enabled investors to achieve high returns. Such compensation of investors for their wartime capital losses was intentional. It was designed to ensure that investors were willing to finance the next war as well. In the end, the final bill for the war (after taking into account the post-war years) fell to the taxpayers. It was not until the Second World War that this historical pattern was disrupted, after which countries no longer sought to return to pre-war prices. By contrast, high inflation continued in the post-war years and contributed significantly to the reduction of the debt burden. Real growth was the second, comparably important, effect. By contrast, the contribution of surplus budgets – i.e. taxpayers – was significantly smaller both in the USA (Hall and Sargent, 2021) and in Britain (Wickens, 2022).

The “post-war” compensation of investors for their COVID losses is already under way – at least in part. The fierce fight by the major central banks to tame inflation from 2022 onwards, which has also driven long-term government bond yields high, will lead to relatively generous (real) investment returns in the decade after the pandemic.[6] Yet these efforts are far from managed deflation. As in the aftermath of the Second World War, investors will most likely not be fully compensated and the COVID debt will remain “amortised” primarily by a combination of inflation and growth, while the contribution of surplus budgets (i.e. taxpayers) will be less important.

What if the state borrows from itself?

However, the current episode is still unusual in historical context. Central banks today are important creditors of many countries. In the aftermath of the global financial crisis and again in response to the outbreak of the pandemic, the central banks of many developed countries launched extensive quantitative easing programmes, during which they purchased securities, most often their own government bonds, using the newly created liquidity. The public sector as a whole thus de facto exchanged the government’s long-term liabilities for the central bank’s short-term liabilities.

Then, when the value of long-term government bonds fell in 2022, the central banks bore part of the investment losses. The distinction whether a loss occurs in government or central bank accounts can, for many accounting and legal reasons, play a role.[7] From a fundamental economic point of view, however, the central bank is part of the public sector, so its financial results ultimately affect public finances – directly or indirectly.[8] Therefore, if we want to estimate the part of the public COVID debt paid by private investors through their asset losses, the original estimate needs to be adjusted for the investment losses incurred by taxpayers through the central bank. Chart 5 shows the original breakdown of the debt-to-GDP changes for the UK and the USA, adjusted for the accounting loss of the central bank arising from the revaluation of the portfolio of securities holdings.´

Chart 5 – However, central banks also incurred part of the investment losses
(breakdown of changes in net government and CB debt in market valuation to GDP, 2019 Q4 – 2023 Q1, pp)

Chart 5 – However, central banks also incurred part of the investment losses

Source: BIS, ONS, FRED, Eurostat, HM Treasury, FED, CNB, author’s calculations
Note: The ECB does not disclose the market value of its QE portfolio, so the size of the bank’s (unrealised) losses cannot be estimated.

In the case of the United States and Britain, while the inclusion of the central bank reduces the contribution of private investors to government “deleveraging”, it remains clear. Even when the central bank’s loss is taken into account, the conclusion that the COVID debt has already been largely “repaid” through the investment losses of private holders of government bods remains true for these countries.


To avoid any misunderstanding, this article does not describe the long-term evolution of public debt, which is likely to continue to grow in the coming years (Gaspar et al., 2023). It focuses solely on the COVID episode that led to a sharp one-off rise in debt at record-low interest rates. In the USA, the United Kingdom, the euro area and the other countries under comparison, it has already been significantly reduced through revaluation and inflation – i.e. by the investment losses of the debt creditors – and need not be a burden for future generations.

However, the final distribution of the COVID burden will also depend on the post-pandemic period. Not all investment losses have been realised, and future developments in interest rates and inflation may redistribute wealth back to investors at the expense of taxpayers. Indications so far are that part of the investment losses will return to holders of government bonds over the next decade, which, according to the financial markets, will be characterised by low inflation and fairly high real yields. However, these will not be able to fully reverse the inflation episode and the majority of the COVID debt (in real terms) is likely to be amortised by inflation and real growth (as in the aftermath of the Second World War).

Who will ultimately pay the COVID debt plays a role. The average bond holder is wealthier than the average taxpayer, which is why the current inflation and revaluation of bonds has led to a redistribution of wealth in favour of poorer households that benefited from government support programmes during the COVID period. Further redistribution may take place between residents and non-residents. This is because investors in government bonds are often non-resident, while taxes are almost exclusively paid by residents.

Written by Martin Kábrt. The opinions expressed in this article are his own and do not necessarily reflect the official position of the Czech National Bank. The author would like to thank Anna Drahozalová for her research assistance, and colleagues at the CNB for their helpful insights.


Barro, R.  J. (1979), “On the determination of the public debt“, Journal of Political Economy 87(5): 940-971.

Arslanalp, S. and Eichengreen, B. (2023), “Living with High Public Debt”, Working Paper presented at the Jackson Hole Economic Policy Symposium, August 2023.

Frait, J., Komárková, Z. and Szabo, M. (2021), “Rostoucí zadlužení státu, provázanost mezi vládním a finančním sektorem a rizika pro finanční stabilitu”, cnblog, https://www.cnb.cz/cs/o_cnb/cnblog/Rostouci-zadluzeni-statu-provazanost-mezi-vladnim-a-financnim-sektorem-a-rizika-pro-financni-stabilitu/.

Gaspar, V., Poplawski-Ribeiro, M. and Yoo, J. (2023), “Global Debt Is Returning to its Rising Trend“, IMF blog, 13 September 2023, https://www.imf.org/en/Blogs/Articles/2023/09/13/global-debt-is-returning-to-its-rising-trend.

Hall, G. J. and Sargent, T. (2021), “Debt and Taxes in Eight U.S. Wars and Two Insurrections”, chapter 27 of The Handbook of Historical Economics (editors Alberto Bison and Giovanni Federico), Academic Press, 2021.

Hall, G. J. and Sargent, T. (2022), “Three World Wars: Fiscal-Monetary Consequences”, Proceedings of the National Academy of Sciences, Vol. 119, No 18, 3 May 2022.

Hall, G. J. and Sargent, T. (2023a), “Fiscal Consequences of the US War on COVID”, paper presented at the 2023 Bank of Korea International Conference, 15 May 2023.

Hall, G. J. and Sargent, T. (2023b), “Financing Big U.S. Federal expenditures Surges: COVID-19 and Earlier U.S. Wars”, Chapter 10 in How Monetary Policy Got behind the Curve – And How to Get Back (editors Michael D. Bordo, John H. Cochrane, and John B. Taylor) Stanford, CA: Hoover, 2023.,

Checherita-Westphal, C. (2019), “Interest rate-growth differential and government debt dynamics”,  ECB Economic Bulletin, Issue 2/2019, https://www.ecb.europa.eu/pub/economic-bulletin/focus/2019/html/ecb.ebbox201902_06~0c96ee6f7c.en.html.

Komárek, L. (2019), “How heavy a fiscal burden are we carrying to interest rate base camp? The fiscal and monetary space in OECD countries”, Global Economic Outlook 3/2019, Czech National Bank.

Wickens, M. R. (2022), “How may the UK’s Debt-GDP ratio be reduced? Evidence from the last 120 years”, CEPR Discussion Paper DP17172, ISSN 0265-8003.


public debt, inflation, central banks

JEL Classification

E58, F34, F62

[1] Experts, too, are hinting at gloomy prospects. For example, B. Eichengreen and S. Arslanalp (2023) prepared their audience for the “new reality” of high debt in the most discussed contribution to the Jackson Hole central bankers’ symposium.

[2] Albeit with three reservations. First, residents are not only the debtors, but also the creditors of public debt, as the vast majority of government bonds are held directly or indirectly by domestic investors (see, e.g., Frait, Komárková and Szabo, 2021). Thus, both debts and financial claims are left to “future generations”. Second, the growing prosperity of the economy reduces this future “burden”. If GDP grows faster than interest accumulates, it will be easier for future generations to repay the debt than it would be for today’s generation. However, this condition often does not hold in developed countries (Checherita-Westphal, 2019). Third, unlike households, the state is “immortal”. It can refinance its debts forever and never has to repay them – meaning each future generation can pass on the debts to the one coming after them. Beyond a certain level, however, debt can limit the government’s fiscal space or even enter an unsustainable trajectory. These considerations are discussed in more detail in an earlier issue of Global Economic Outlook (Komárek, 2019).

[3] Such a response is also recommended in the literature, because a sharp increase in taxes would create distortions in the economy and deepen an already serious economic crisis (e.g. Barro, 1979).

[4] For example, Moody's upgraded its ratings for seven countries and downgraded only one (Britain). Fitch made four upgrades and one downgrade (Slovakia). 

[5] By contrast, the nominal value is the amount (principal) borrowed by the US government. On the one hand, the nominal value of the debt underestimates the actual debt burden, as it does not take into account the interest (coupon rate) that the government has to pay regularly. On the other hand, it overestimates the nominal value of the debt burden, as it does not in any way discount the principal that may be payable decades into the future. The market value of the debt takes into account coupon rates and discounting.

[6] At the end of September 2023, ten-year US bond yields are close to 4.6%, although the financial markets expect average inflation to be just above the Fed’s 2% target in the same period. Similar real yields are implied by the prices of US inflation-linked bonds.

[7] A central bank cannot, for example, become insolvent – it can always meet its obligations by creating new money. Unlike the government, it also determines how it will remunerate its obligations. In practice, however, the use of these privileges to repay its own debt is incompatible with the price stability mandate.

[8] Central banks usually distribute their profits to public budgets. Losses are treated differently in different jurisdictions – either they are paid by the government (and affect public budgets immediately) or remain on the central bank’s balance sheet and are covered by its future profits (thus affecting future public budgets through lower central bank profit levies).