Structure of government debt
Data extracted in June 2021
Planned update: June 2022
This article presents recent statistics on the structure of general government debt and its relationship to gross domestic product (GDP) in the European Union (EU). In the context of the Stability and Growth Pact's Excessive deficit procedure notification process, Eurostat publishes government debt data twice a year, in April and October, as well as quarterly government debt data transmitted in line with Regulation (EU) No 549/2013 (ESA 2010 transmission programme).
In order to analyse the debt structure in European countries, Eurostat collects additionally data in an annual survey providing information on Member States' government gross debt by sector of debt holder, by instrument, by initial and remaining maturity and by currency of issuance. The survey also collects information on (one-off and standardised) guarantees granted by the general government to non-government units as well as the market value of the Maastricht debt instruments and the apparent cost of government debt. One-off guarantees are contingent liabilities which are not included in general government gross debt.
This article examines the main results of the latest questionnaire, fully or partly completed by the EU Member States and Norway.
Maastricht debt as a percentage of GDP
Maastricht government debt has followed an upward trend following the financial crisis. From a high point at the end of 2014 (86.6 % of GDP), at EU level, a decrease in the debt to GDP ratio have been noted up to the end of 2019 (77.5 % of GDP).
During 2020, the downward trend was sharply reversed. Up to the end of 2020, EU general government gross debt increased to 90.7 % of GDP), reflecting governments' increased financing needs as a result of the impact of the economic downturn resulting from the COVID-19 pandemic as well as due to policy measures undertaken to mitigate the impact of the pandemic.
Between the end of 2019 and the end of 2020, the general government gross debt to GDP ratio increased in the EU (from 77.5 % of GDP at the end of 2019 to 90.7 % of GDP at the end of 2020 corresponding to an increase of 13.2 percentage points) and the euro area (from 83.9 % of GDP at the end of 2019 to 98.0 % of GDP at the end of 2020 or by 14.1 percentage points). At the level of the EU and euro area, the increases represent the sharpest increases observed in the time series since 1995, as well as the highest level of general government gross debt as a percentage of GDP recorded.
Between the end of 2019 and the end of 2020, the debt to GDP ratio increased in all 27 EU Member States and Norway. The largest increases were recorded in Greece (+25.1 pp.), Spain (+24.5 pp.), Cyprus (+24.2 pp.), Italy (+21.2 pp.), France (+18.1 pp.), Portugal (+16.8 pp.), Belgium (+16.1 pp.), Croatia (+15.9 pp.), Slovenia (+15.2 pp.) and Hungary (+15.0 pp.).
At the end of 2020, 14 out of 27 EU Member States reported debt to GDP ratios higher than the reference value of 60.0 %, while seven EU Member States recorded debt to GDP ratios of more than 100.0 %: Greece recorded the highest debt to GDP ratio at 205.6 %, followed by Italy (155.8 %), Portugal (133.6 %), Spain (120.0 %), Cyprus (118.2 %), France (115.7 %) and Belgium (114.1 %).
The lowest debt to GDP ratio was registered by Estonia at 18.2 % of GDP at the end of 2020, followed by Luxembourg (24.9 %), Bulgaria (25.0 %), Czechia (38.1 %), Sweden (39.9 %), Denmark (42.2 %), Latvia (43.5 %), Lithuania and Romania (both 47.3 %), as well as Norway (46.0 %).
Breakdown by subsector of general government
- Central government (S.1311);
- State government (S.1312);
- Local government (S.1313);
- Social security funds (S.1314).
Figure 2 gives an overview of the subsector breakdown, as a percentage of total debt for all subsectors, i.e. not consolidated between different levels of government.
For 24 of the 27 EU Member States, the central government represented more than 75.0 % of the general government debt (not consolidated between subsectors) at the end of 2020, while other subsectors of general government had a comparatively large share in Germany, Spain and Sweden as well as Norway.
The share of local government debt (non-consolidated between subsectors) was at 40.7 % in Norway. In Sweden, the share of local government debt was 31.6 %. Local government debt also played an important role in Denmark (16.7 %), Finland (15.8 %) and Estonia (15.2 %).
In Germany (28.2 %) and Spain (18.8 %), state governments accounted for a significant share in the total gross debt (not consolidated between subsectors). State government as a subsector of general government exists only in four Member States, namely Belgium, Germany, Spain and Austria. The debt share of state government was 14.6 % in Belgium and 6.8 % in Austria.
The impact of social security funds on the general government debt continues to be relatively small: contributions of less than 5 % were recorded in 23 countries (out of the 25 reporting countries with a social security funds' subsector). Only two countries had somewhat higher ratios of debt for social security funds: France (9.8 %) and Spain (5.3 %).
Impact of consolidation
The general government debt has to be consolidated within each subsector and between subsectors at the level of general government. This implies that the debt issued by one subsector and held by another one should be excluded from the general government debt. When debt of one subsector of general government is held by another subsector, general government gross debt is then lower than the sum of subsector gross debt. Table 1 illustrates this effect in percentage of total non-consolidated debt (consolidated within subsectors but not between).
The majority of countries show a limited consolidation impact between the subsectors of general government, where the general government debt was reduced by less than 5 % in 16 of the 27 EU Member States as well as in Norway. A significant consolidation effect was observed in Cyprus (36.6 %), Latvia (23.1 %), Spain (20.2 %), Estonia (16.8%) and the Netherlands (16.0 %). Significant consolidation effects are often due to central government liabilities in deposits held by local government and social security funds, for example in Estonia, Latvia and Cyprus. The level of consolidation tends to be influenced by the growing prevalence of cash pooling arrangements.
On the other hand, eight countries showed no consolidating amounts between subsectors of general government or consolidating amounts below 1 %: Denmark, Germany, Croatia, Lithuania, Malta, Slovenia, Finland and Norway.
Breakdown by financial instrument
The Maastricht debt is divided into the following categories according to the ESA 2010 classification:
- currency and deposits (AF.2);
- debt securities (AF.3); and
- loans (AF.4).
The breakdown of debt by financial instrument is presented in Figure 3.
For the EU, 81.8 % of the general government debt was made up by debt securities, 15.1 % was made up by loans and 3.2 % by currency and deposits. For the EA-19, 82.1 % of the general government debt was made up by debt securities, 14.6 % was made up by loans and 3.3 % by currency and deposits.
For 25 of the 27 EU Member States, the most used debt instrument remained debt securities at the end of 2020; ranging from a share of 19.7 % of Maastricht debt in Greece to 93.4 % of the general government debt in Czechia.
Greece and Estonia as well as Norway registered high shares of loans, with loans making up a share of 78.4 %, 57.9 % and 58.1 % respectively. Significant loan to total debt ratios were also recorded for Cyprus (32.9 %), Sweden (32.7 %) and Croatia (29.6 %). The countries reporting a higher share of loans tended to be those either having a relatively low level of general government gross debt (e.g. Estonia), a relatively high share of debt of subcentral sectors of government (e.g. Sweden) or having benefited in recent years from loans of EFSF, ESM, IMF and other international assistance (e.g. Greece and Cyprus).
At the end of 2020, currency and deposits represented less than 5 % of total debt for 22 countries. In contrast, currency and deposits accounted for 12.2 % of total general government gross debt in Portugal (due to saving certificates), 10.9 % in Ireland (due to defeasance structures) and 8.9 % in Italy (due to saving certificates).
Breakdown by sector of debt holder
Figure 4 presents general government gross debt by sector of the debt holder: non-financial residents (non-financial corporations, households and non-profit institutions serving households), financial residents (financial corporations) and non-residents (rest of the world).
At the end of 2020, government debt was mainly held by resident financial corporations sector in thirteen EU Member States for which data is available, as well as in Norway. Its share was the highest in Sweden (72.7 %), followed by Croatia (67.2 %) and Denmark (66.6 %). At the other end of the scale, the smallest proportion of debt held by resident financial corporations was recorded in Cyprus (15.3 %), ahead of Estonia (27.5 %), Latvia (29.6 %), Lithuania (29.9 %) and Ireland (33.8 %).
The debt share held by non-residents (rest of the world sector) was also significant in a large number of countries but highly variable between countries for which data is available. 'Rest of the world' (non-residents) was the largest debt holder in thirteen EU countries, with percentages starting from 49 % and higher: Cyprus (81.9 %), Estonia (70.0 %), Lithuania (69.5 %), Latvia (66.8 %), Austria (63.7 %), Finland (60.8 %), Slovenia (58.9 %), Belgium (55.9 %), Ireland (55.5 %), Slovakia (53.6 %), Romania (50.9 %), Luxembourg (50.0 %) and Portugal (49.0 %). In contrast, this proportion was only 18.2 % in Malta and 20.0 % in Sweden.
The resident non-financial sectors (non-financial corporations, households and non-profit institutions serving households) played a major role as debt holder in Hungary (25.0 %), Malta (21.4 %), Portugal (13.8 %) and Ireland (10.7 %).
Breakdown by initial maturity
The debt questionnaire aims to provide detailed information on the time structure of government debt based on its initial maturity. The maturity is subdivided into several maturity brackets: less than one year, one to five years, five to seven years, seven to ten years, ten to fifteen years, fifteen to thirty, and more than thirty years, as well as the summary category of more than one year. For some countries, which did not provide the complete breakdown, only two categories are shown: less than one year (short-term) and more than one year (long-term). For the other seventeen countries, a detailed debt maturity breakdown is available. The share of short-term debt to total debt is illustrated in Figure 5.
General government gross debt classified by maturity reveals a common pattern: between 70.1 % (in Sweden) and nearly 100 % of the outstanding debt was issued on a long-term basis. Short-term debt levels of less than or equal to 1 % were recorded in Lithuania (0.0 %) and Bulgaria (0.1 %).
The short-term debt ratio was significant in Sweden (29.9 %), Denmark (21.6 %), Portugal (16.7 %) and Finland (15.6 %) as well as Norway (21.3 %), while the short-term debt ratio also exceeded 10 % in the Netherlands, Italy, France, Germany, Malta and Ireland.
The countries providing a detailed long-term debt breakdown showed very different structures. This is shown in Figure 6.
Breakdown by remaining maturity
While the initial or original maturity of debt measures the time between issuance date and redemption date, the remaining maturity of debt measures the time left until the redemption date.
Figure 7 shows the share of central government gross debt with a remaining maturity of less than one year, at end 2019 and at end 2020, i.e. the share of the central government debt which was to be redeemed in 2020 and is due to be redeemed in 2021.
Data is available for 21 out of the 28 countries having completed the survey.
At the end of 2020, the highest shares of short-term remaining maturity debt in total central government debt were reported by Sweden (39.7 %), followed by Estonia (28.4 %), Portugal (26.4 %), Latvia (23.8 %) and Italy (23.4 %), while the lowest shares were observed for Bulgaria (3.1 %) and Romania (3.6 %).
Between the end of 2019 and the end of 2020, large reductions in the share of short-term remaining maturity debt were observed for Estonia (49.2 % at end 2019, 28.4 % at end 2020), Hungary (22.5 % at end 2019, 16.7 % at end 2020) and Romania (18.4 % at end 2019, 3.6 % at end 2020).
Breakdown by currency of issuance
As shown in Figure 8, of the 27 countries for which 2020 data is available, seventeen Member States issued all or almost all (>99 %) of their central government gross debt in national currency. In particular, Belgium, Estonia, Ireland, Cyprus, Lithuania, Luxembourg, Portugal and Finland issued 100 % of their debt in national currency (all in euro), while this ratio was above 99 % in Spain, France, Italy, Latvia, Malta, the Netherlands, Austria, Slovenia and Slovakia. (Please see the notes below.)
Only three countries issued more than 50 % of their debt in foreign currency: Bulgaria (82.0 %, note that Bulgaria has a currency board arrangement vis a vis the euro), Croatia (72.1 %) and Romania (51.1 %). All three countries are not part of the euro area and the major share of their foreign currency issuances are denominated in euro.
As concerns changes in the ratio of foreign currency debt to total debt between the end of 2019 and the end of 2020: Denmark and Romania increased the share of foreign currency debt, while the share of foreign currency debt decreased in Poland.
Figure 9 presents the share of outstanding central government debt issued in euro at the end of 2020. The debt denominated in euro is equal to the debt issued in national currency for the 19 euro area member countries. 100 % of the stock of government debt was denominated in euro for Belgium, Estonia, Ireland, Cyprus, Lithuania, Luxembourg, Portugal and Finland. A share higher than 99 % of total central government gross debt was denominated in euro in Latvia, France, Malta, Spain, Slovakia, the Netherlands, Slovenia, Italy and Austria. Greece and Germany reported shares exceeding 95 % (in descending order).
In contrast, the major issuing currency in the non-euro countries Czechia (92.0 % issued in Czech koruna), Denmark (89.9 % issued in Danish krone), Sweden (80.0 % issued in Swedish Krona), Hungary (78.3 % issued in Hungarian forint) and Poland (77.1 % issued in Polish zloty) was their respective national currency.
Apparent average cost of government debt
The apparent average cost of central government debt (accrued interest payable over the period as a percentage of the average outstanding debt) shows the differences between countries in terms of their cumulated past conditions for accessing financial markets. Based on 27 replies from EU Member States and Norway, the analysis of apparent average cost of central government debt is shown in Figure 10.
The apparent average cost of central government gross debt varied between 0.4 % in Estonia, 0.7 % in Finland, 0.8 % in Germany and 3.2 % in Hungary and 3.8 % in Romania in 2020.
Comparing the 2020 data with 2019, decreases in implicit rates were observed for all reporting countries. The largest decreases were observed for Lithuania (3.1 % in 2019 versus 1.9 % in 2020), as well as for Denmark (2.6 % in 2019 versus 1.4 % in 2020), Croatia (3.5 % in 2019 versus 2.8 % in 2020), Poland (3.0 % in 2019 versus 2.4 % in 2020), Luxembourg (1.7 % in 2019 versus 1.1 % in 2020) and Malta (3.2 % in 2019 versus 2.7 % in 2020). As this measure of the cost of debt depends on interest rates prevailing at the moment of issuance in the past, it is normally not very sensitive to the most recent market trends, provided that the composition of debt is mainly long-term. The large decreases observed between 2019 and 2020 are a reflection of the very low interest rates prevailing during 2020 as well as the large volume of new issuances in 2020 as a result of the COVID-19 pandemic.
Government guarantees as a percentage of GDP
Countries were additionally asked about the amount of government guarantees. These guarantees (both "one-off" and "standardised") are not part of government gross debt, as they are contingent liabilities, being contingent on the actual call of the guarantee. They should not be added to the Maastricht debt.
Based on 27 replies from EU Member States the ratio of government guarantees provided by central government on debt of non-government units, as a percentage of GDP, is shown in Figure 10. At the end of 2020, the amount of government guarantees as a percentage of GDP exceeded 10 % for seven countries (Finland, Austria, Denmark, Italy, France, Luxembourg and Spain) out of 24 countries for which data is available. At the end of 2019, the amount of guarantees in Finland, Austria, Denmark, Germany (latest year available) and Luxembourg exceeded 10% of GDP.
At the end of 2020, a ratio to GDP of less than 5 % was recorded in Slovakia (0.1 %), Bulgaria (0.2 %), Ireland (0.6 %), Czechia (0.9 %), Lithuania (1.1 %), Latvia (1.5 %), Croatia (1.7 %), Estonia (1.9 %), Poland (2.0 %), and Romania (3.4 %).
Between 2019 and 2020, the stock of guarantees granted by central governments increase for all countries for which data is available. This is due to guarantees granted in the context of EU schemes (SURE, PEGF) as well as national schemes newly formed or extended as a policy response to the COVID-19 pandemic.
Market vs. face value
The market value is the price as determined dynamically by buyers and sellers in an open market.
In Council Regulation (EC) No 479/2009, as amended, the face value is used. This is equal to the undiscounted amount of the principal that the government will have to pay to creditors at maturity.
Debt statistics cover data for general government as well as its subsectors: central government (S.1311), when applicable state government (S.1312), local government (S.1313) and when applicable social security funds (S.1314).
Maastricht debt comprises only the following instruments:
- AF.2: The category 'currency and deposits' consists of currency in circulation and all types of deposits in national and in foreign currency.
- AF.3: The category 'debt securities' consists of negotiable financial instruments that are bearer instruments and are usually traded on secondary markets or can be offset on the market, and do not grant the holder any ownership rights in the institutional unit issuing them.
- AF.4: The category 'loans' consists of financial instruments created when creditors lend funds to debtors, either directly or through brokers, which are either evidenced by non-negotiable documents or not evidenced by documents.
Debt figures on general government statistics and each of its subsectors are reported consolidated.
Consolidation is a method of presenting statistics for a grouping of units, such as institutional sectors or sub-sectors, as if it constituted a single unit. Consolidation thus involves a special kind of cancelling out of flows and stocks: eliminating those transactions or debtor/ creditor relationships that occur between two transactors belonging to the same grouping. Usually the sum of subsectors should exceed the value of the general government sector. Subsector data should be consolidated within each subsector, but not between them. ESA 2010 recommends compiling both consolidated and non-consolidated financial accounts. For macro-financial analysis, the focus is on consolidated figures. The Maastricht debt is also consolidated.
The Eurostat 2021 government debt structure survey
The survey launched by Eurostat on government debt structure contains the following tables: a set of five tables (central government debt, state government, local government debt, social security funds debt and general government debt) for 2017, 2018 and 2019, and the same set of tables for 2020 detailing gross debt by financial instrument, sector of debt holder and initial maturity, plus a table with additional classifications of government debt (remaining maturity, currency of issuance, apparent cost of debt, market value of gross debt and guarantees (contingent liabilities)).
The survey presents breakdowns for the general government and its subsectors for the two previous calendar years, categorising the debt by sector of the debt holder, by instrument, by detailed initial and remaining maturity, by currency of issuance, as well as guarantees granted by the government to non-government units. The latter are contingent liabilities. In addition, the survey contains information on the market value of Maastricht debt instruments as well as on the apparent cost of government debt.
For the EU-27 and EA-19, data from ESA table 28 (quarterly government debt) and EDP notification supplied in April 2021 is used to complement the analysis. For all countries, the data reported in the structure of government debt survey corresponds to the general government gross debt totals submitted in the context of the April 2021 Excessive Deficit Procedure notification.
For all 27 EU Member States and Norway, the data used in this article was reported in the structure of government debt survey. The survey is not fully completed by all countries. Hence the number of countries shown for each breakdown of gross debt as well as guarantees varies.
The analysis on breakdown by currency, apparent cost of the debt and state guarantees is based on central government data.
The GDP supplied to Eurostat in the context of the April 2021 EDP notification is used in the graphs and analysis.
Foreign currency breakdown: The reporting is not currently fully harmonised between reporting countries. Countries are encouraged to report hedged foreign currency amounts under domestic currency. A small number of countries may report the stock of foreign currency debt at nominal value, implying a small distortion in the shares of domestic and foreign currency.
The coverage of guarantees may still vary across countries and have an expanding coverage over time.
Some breakdowns of gross debt by detailed maturity, by sector of debt holder and by currency may not sum to the total debt instrument / gross debt, in case detailed information was not available for some items.
Portugal: Information on the detailed split of long-term loans by initial maturity is partially available. For this reason, total long-term loans exceed the detailed breakdown.
Greece: The coverage of data relating to the remaining maturity of debt, currency of issuance, apparent cost of debt excludes extra-budgetary units of central government. Thus the coverage of central government data for these indicators is limited to S.1311.1, the budgetary central government. For guarantees: the amounts for scheme of standardised guarantees are included (taking into consideration COVID-19 related measures).
Denmark: The coverage of data relating to the remaining maturity of central government gross debt is limited to debt securities and deposits.
Germany: Only asset guarantees provided to publicly controlled monetary financial institutions are included under guarantees given by government units on non-government borrowing to financial corporations (S.12) sector, see also https://www.destatis.de/EN/Themes/Government/Public-Finance/EU-Directive-Budgetary-Frameworks/Tables/guarantees.html.
Monitoring and keeping government debt in check is a crucial part of maintaining budgetary discipline which is essential as Europe undergoes dramatic demographic changes. Its ageing population, in particular, is expected to pose major economic, budgetary and social challenges.
The Protocol on the excessive deficit procedure (EDP) annexed to the Maastricht Treaty specifies that the ratio of gross government debt to GDP must not exceed 60 % at the end of the preceding fiscal year. The application of the Protocol is made operational by Council Regulation (EC) No 479/2009, as amended.
Fiscal data are compiled in accordance with national accounts rules, as laid down in the European System of Accounts (ESA 2010) adopted in the form of a Regulation (EU) No 549/2013 of the European Parliament and of the Council of 21 May 2013. The full text of compilation of General government debt data complies with ESA 2010 rules concerning the sector classification of institutional units, the consolidation rules, the classification of financial transactions and of financial assets and liabilities and the time of recording. The valuation is however different. Debt liabilities in ESA 2010 are valued at market value, whereas Maastricht debt is valued at nominal (face) value. Most data in the publication on the structure of government debt refer to general government gross debt (and various breakdowns) expressed at face value.
- Annual government finance statistics (t_gov_10a)
- Government deficit and debt (t_gov_10dd)
- Quarterly government finance statistics (t_gov_10q)
- Government statistics (gov)<noprint>, see:
- Annual government finance statistics (gov_10a)
- Government deficit and debt (gov_10dd)
- Structure of government debt (gov_10dd_sgd)
- Quarterly government finance statistics (gov_10q)
- Structure of government debt (ESMS metadata file — gov_10dd_sgd_esms)
- Government deficit and debt (ESMS metadata file — gov_10dd_esms)
- Quarterly financial accounts for general government (ESMS metadata file — gov_10q_ggfa_esms)
- Quarterly government debt (ESMS metadata file — gov_10q_ggdebt_esms)