Blog post

The Brexit bill: uncertainties in the estimate of EU pension and sickness insurance liabilities

Pension and sickness insurance liabilities for EU staff could be an especially contentious part of negotiations on an EU-UK financial settlement: the

Publishing date
17 February 2017

The ‘Brexit bill’, the financial settlement between the UK and remaining 27 EU member states, is already a subject of intense discussion. Commentators are trying to ascertain exactly which items (both assets and liabilities) will be included in the settlement. One key issue is whether the UK will be obliged to pay for the pensions of European officials. In this post we do not discuss the legal aspects, but instead highlight the fact that estimations of the EU’s pension liability are burdened with considerable uncertainty.

Key points

  • The balance sheet of the European Union includes an estimate for the present value of pension and sickness insurance liabilities related to EU officials, which remained broadly stable at about €35 billion in 2005-11, but has almost doubled to €63.8 billion in 2011-15.
  • The number of EU officials contributing to the pension scheme increased by only 8% from 2011-15, so increased employment cannot be the main reason for the increased pension/sickness liability.
  • The main reason for the doubling of the pension/sickness liability is a large fall in the discount rate which is used to calculate the present value. The discount rate may increase in the coming years, which will lead to a fall in the present value of pension/sickness liabilities.
  • Parallel with the doubling of pension/sickness liabilities in the EU balance sheet, the contribution rate payed by EU officials from their salaries has actually fallen. While the same actuarial method is used to calculate the balance sheet liability and the employee contribution rate, the underlying assumptions differ enormously.
  • It should be a common interest to find a “reasonable” way of calculating the pension/sickness liability in the Brexit financial settlement.
  • Only two thirds of the “reasonable” value should be considered as a liability to which the UK should contribute, because one third of pension costs are paid by EU officials. If the assumptions for EU employee contributions are used for Brexit bill calculations, then the UK should pay a share of a liability of approximately €29bn.


Employees of EU institutions and agencies participate in unique pension and sickness insurance schemes. Officials contribute one-third to the financing of the pension scheme via a compulsory contribution from their salaries, while two-thirds comes from the member states (via the general EU budget), who jointly guarantee the payment of these benefits.

There is no fund behind the pension and sickness schemes: they operate on a pay-as-you-go basis. When employee contributions in a year exceed actual annual payments (which was the case for several years after the inception of the pension scheme), this surplus goes to the general EU budget and thereby reduces the national contributions to the EU budget. When annual pension payments exceed the contributions by employees, the difference is paid out of the EU budget.

In each year, the Consolidated Annual Accounts of the EU present an estimate for the present value of future pension and sickness insurance liabilities of current and former EU officials. To calculate this present value, the Commission projects expected pension and sickness spending several years ahead. But one euro now is worth more than one euro in the future (because if you invest one euro now, its value will increase to more than one euro in the future).

So using an interest rate, which is called ‘discount factor’, the current euro equivalent of all future euro obligations is calculated, which is called ‘present value’. The larger the discount factor, the smaller the present value.

The present value of these liabilities was €63.8 billion at the end of 2015, which consists of three main items:

  • The Pension Scheme of European Officials (PSEO) – €55 billion;
  • The pension scheme for top officials (current and former members of the Commission, European Parliament, Court of Justice, Court of Auditors, Secretaries General of the Council, the Ombudsman, the European Data Protection Supervisor, the European Union Civil Service Tribunal) – €1.5 billion;
  • The Joint Sickness Insurance Scheme – €7.4 billion.

Doubling of pension and sickness insurance liabilities from 2011-15

As Figure 1 below shows, this liability was fairly stable at around €35 billion in 2005-11, but it has almost doubled since 2011.

Employment growth explains only a small part of the increased liabilities

The number of EU employees who contribute to the main pension scheme increased by 8 percent from 2011 to 2015 (Figure 2). Therefore, the main reason for the doubling of the present value of benefits was not an increase in employment. (In the annex we provide further information about the number of EU officials.)

A fall in the discount rate is the main explanation

The main reason for the increase in the pension/sickness liability of the EU balance sheet is a big decline in the discount rate from 2011-15, as indicated on Figure 3. Indeed, the 2001-05 fall in the discount rate was also mirrored by the increase in pension/sickness liabilities in these years.

However, interest rates may not remain as low as they are now. The reports on the EU consolidated accounts acknowledge that the present value of future pension and sickness insurance liabilities will likely fall when the discount rate increases.

Staff contribution rates fell while liability doubled

According to the EU’s Staff Regulations, contributions made by the staff are meant to cover one third of the cost of pensions rights acquired every year. Given that final pension rights are an inherently uncertain amount, the contribution rate applied on the staff’s salaries in order to yield the required revenue is subject to adjustment on the basis of a report by Eurostat every year.

Based on this Eurostat calculation, up to 2013 a Council decision, and from 2014 a Commission decision set the contribution rate, which applies from 1 July of the year in question until 30 June of the following year.

While the main principles to assess the actuarial cost to adjust the staff contribution rate are the same as the pension liability calculation for the EU’s balance sheet, there are three main differences in the assumptions.

  • First, EU balance sheet calculations use the most recent values of the nominal discount and inflation rates, but the employee contribution rate calculations use a historical average: a 12-year average was used till 2012, it is currently 18 years, and this will be gradually increased to a 30-year average by 2021.
  • Second, EU balance sheet calculations use the expected inflation rate for the euro area (derived from inflation-linked government bonds of euro countries), but the employee contribution rate calculations use a historical average of past inflation for the EU as a whole.
  • Third, EU balance sheet calculations use a 19-year maturity discount rate, but the employee contribution rate calculations consider a 21-year maturity discount rate; both consider only the euro area.

These differences in assumptions lead to major differences in the parameters which are used for the actuarial calculations (Table 1).

Interest rates are currently low, so the EU balance sheet calculations use a relatively low nominal discount rate (2.0%), while the average nominal discount rate of the past 18 years, which is used for the employee contribution rate calculations, is much higher (4.8%).

The different consideration of inflation also leads to differences, so the real discount rate, which is actually used in the calculations, is much lower for the EU balance sheet calculations (0.6%) then for the employee contribution rate calculations (2.7%).

Therefore, the present value of the same liability (future pension and sickness insurance payments related to EU officials) looks much larger in the EU balance sheet calculations (because of the lower discount rate) than in the EU employee contribution rate calculations (because of the higher discount rate).

The implication is that using the 2.7% discount rate (instead of the actually used 0.6%) for the EU balance sheet calculations, the present value of the pension/sickness liability would be much smaller than the €63.8 billion number indicated in the 2015 EU balance sheet.

Unfortunately we cannot make a precise calculation, because we do not have sufficient data. But we can do a simple calculation, to illustrate the magnitudes, if we assume a one-time “big” liability in 19 years from now instead of a stream of liabilities in each year in the future. We consider 19 years because this is the maturity of the discount rate used and therefore the average duration of future pension liabilities could be around 19 years.

By taking the €63.8 billion value and the 0.6% discount rate, we can calculate the hypothetical value of this “big” liability in 19 years from now and then discount this hypothetical value with the 2.7% discount rate to get an alternative estimate of the present value. This simple calculation shows that a present value of €43.1 billion would be consistent with a 2.7% discount rate when the €63.8 billion present value is consistent with the 0.6% discount rate. This calculation illustrates well the large sensitivity of the present value calculations to the discount rate.

The differences in the assumptions result in opposite dynamics: while Figure 1 showed that the EU balance sheet liability has almost doubled from 2011-15, Figure 4 shows that the contribution rate that EU officials pay has actually been falling in recent years.

Most likely, the main reason for using historical average interest rates in the EU employee contribution calculations is to avoid significant jumps in the annual payments required of employees. For example, if EU balance sheet calculation assumptions were used for the employee contribution rate calculations, then the contribution rates of EU employees would have increased from about 11% of their salaries in 2011 to a much higher value by 2015.

Need for a political deal on “reasonable” valuation of pension/sickness liability in Brexit bill negotiations

It is understandable that the Staff Regulation adopted a system which aims to avoid big jumps in contribution rates for EU officials. However, in light of that, one may wonder whether the end-2015 €63.8 billion pension/sickness liability in the EU balance sheet is the right number to consider in the Brexit financial settlement.

The method used to calculate this pension/sickness liability in the EU balance sheet is the Projected Unit Cost Method, in line with International Public Sector Accounting Standard (IPSAS) 25. This is correct, the EU should follow international standards.

But given that the current low interest rates were the main reason for the doubling of the EU’s pension/sickness liability from 2011-15 and interest rates may increase in the coming years, and given that exactly the same liability is treated very differently in the calculation of contributions by EU officials, common sense suggests that the assumptions used in the actuarial calculations can fall within the scope of political negotiations. After all, the Brexit financial settlement should be prepared “reasonably”. A reasonable option seems to be the use of the assumptions made for the EU official pension contribution rate calculations.

Only two thirds of the “reasonable” present value should be considered in the Brexit financial settlement

While the EU budget provides a guarantee for the full amount of the accumulated pension rights, one third of contributions will be paid by EU officials from their salaries. There are two possible options to consider for this factor, in terms of its place in the ‘Brexit bill’.

One option is that the UK pays its share of the total “reasonable” present value now, but is gradually reimbursed for this one third as future EU officials pay their contributions over the next half a century. But such a long reimbursement period looks cumbersome.

The other option is to consider only two thirds of the “reasonable” present value when calculating the financial settlement. Above we used the Commission’s assumptions about EU official contributions to calculate the present value. This suggested that the “reasonable” value of the pension/sickness liabilities is €43.1bn instead of €63.8bn. Two thirds of this liability is approximately €29bn. If this calculation is correct, then it is a share of this €29bn that should be assigned to the UK in the financial settlement.

Of course, there is then also the question of what share of this pension/sickness liability should be charged to the UK in the Brexit bill settlement. We are working on this issue too and plan to publish our comprehensive calculations for the Brexit bill considering all EU assets and liabilities soon.

And a few technical issues

In the meantime, we invite the Commission to provide more transparency about certain assumptions within the calculations, such as:

  1. The method used to project (real) wage growth of EU officials;
  2. The selection of the nominal discount rate (e.g. why euro-area average, when it is not a risk-free rate, and why a 19-year maturity rate, when the whole term structure of interest rates should be available, which may lead to a more accurate estimate);
  3. The derivation of inflationary expectations for each country, including those countries for which inflation indexed bonds are not available, or available only for certain maturities;
  4. The reasons for the geographical discrepancy in calculations: for the balance sheet calculations euro area interest rate and inflation are considered, but for the EU employee contribution rates euro area interest rate and EU inflation is used.

Annex: the number of EU officials

Figure 2 above shows the number of EU officials who contributed to the main PESO pension scheme between 2003 and 2015. The increase form 2003-08 could be explained (at least partly) by the enlargement of the EU from 15 to 27 members.

After 2008 employment at most EU institutions (including the European Commission) remained broadly stable, but there was a large expansion of employment at ‘Other institutions and agencies’ and at the European Parliament.

  • ‘Other institutions and agencies’ account for the largest increase in EU employment, from 4,787 people in 2008 to 11,579 people in 2015. The number of such institutions increased from 19 in 2005 to 52 in 2013, due to the creation of new agencies, such as the European External Action Service, Europol, Frontex, European Banking Authority and European Asylum Support Office. The full list of the 52 institutions and agencies are included in Table 2.
  • The other main contributor to the increase in employment of EU officials from 2008-15 was the European Parliament, where employment increased from 6,320 people in 2008 to 9,582 people in 2015.


About the authors

  • Zsolt Darvas

    Zsolt Darvas is a Senior Fellow at Bruegel and part-time Senior Research Fellow at the Corvinus University of Budapest. He joined Bruegel in 2008 as a Visiting Fellow, and became a Research Fellow in 2009 and a Senior Fellow in 2013.

    From 2005 to 2008, he was the Research Advisor of the Argenta Financial Research Group in Budapest. Before that, he worked at the research unit of the Central Bank of Hungary (1994-2005) where he served as Deputy Head.

    Zsolt holds a Ph.D. in Economics from Corvinus University of Budapest where he teaches courses in Econometrics but also at other institutions since 1994. His research interests include macroeconomics, international economics, central banking and time series analysis.

  • Konstantinos Efstathiou

    Konstantinos, a Greek citizen, works in Bruegel as an Affiliate Fellow in the Macroeconomics and Governance area. Before joining Bruegel, Konstantinos was at the European Commission, as a Blue Book Trainee in the Cabinet of President Juncker. He also interned at the Central Bank of Luxembourg (BCL) as a research assistant, involved in projects related to the Wage Dynamics Network (WDN) research group.

    He holds a Master in International Economics and International Relations from the Johns Hopkins University School of Advanced International Studies (SAIS), where he specialized in Quantitative Methods and European Studies.

    Konstantinos’ research interests include macroeconomics, European economic governance and international economics.

    Konstantinos is a native speaker of Greek, speaks fluent English and has good knowledge of French.

    Declaration of interests 2017

    Declaration of interests 2018

  • Inês Goncalves Raposo

    Inês Gonçalves Raposo is an Affiliate Fellow at Bruegel in the areas of European macroeconomics, governance, finance and financial regulation. Previously she worked for the Financial Stability Department of the Bank of Portugal. Inês holds a MSc in economics from Nova SBE with a major in Macroeconomics and Financial Markets and a BA in applied mathematics from the University of Lisbon.

    Her research interests include political economy, monetary and fiscal policy and applied macroeconomics. She is a native Portuguese speaker and is fluent in English and French.

    Declaration of interests 2018

Related content