The costs of managing public pension schemes

The primary purpose of this report was to compare the management costs of the main public sector pension schemes with those of the private sector pension scheme. In particular, this study covers the civil and military retirement pension scheme (PCMR), as well as the various schemes managed by the social policy directorate (DPS) of the savings and deposit bank (CDC), including the French national insurance fund for local government workers (CNRACL), the supplementary pension institution for contractual employees working under a public law contract (IRCANTEC), and the civil servants’ supplementary pension fund (RAFP).

The report acknowledges the methodological limitations inherent to this type of exercise, which have already been established by previous reports of the General Inspectorates for Finance and Social Affairs, due to differences in the organization and nature of the schemes that are difficult to overcome. By making comparisons between schemes rather than between the organizations that manage them, by integrating the cost of certain services that are not recharged and by focusing on the cost components of the production function of each scheme, it is still possible for us to make comparisons. In 2021, the managing costs for the abovementioned schemes were as follows: €96 million for PCMRs, €122 million for CNRACL, €102 million for IRCANTEC, €33 million for RAFP, approximately €1.09 billion for the general pension scheme and €1.28 billion for AGIRC-ARRCO. The analyses performed show that the management cost per member of the CNRACL is 60% higher than that of the PCMR, and twice as high per euro of benefit paid, even though these two schemes are very similar in terms of how they work. This can mainly be explained by the fact that salary costs per staff member are 30% higher than for the CNRACL with social benefits included, and 17% higher without them. Staff costs are in fact the main component of costs (60% of management costs on average), ahead of information systems (on average around 25% but 12% for the PCMR) and property (around 5%, with costs per staff member four times higher than the DPS). Going into more detail requires further work by managers to quantify the complexity factors specific to each scheme. Finally, the cost differentials observed cannot be linked to differentials in the quality of services offered due to the lack of a reliable means of comparison.

Determining the full costs of all schemes required adjustments using a repeatable approach for the future. In particular, the budget documents show that the management costs of the PCMRs are underestimated by 20% due to a failure to account for certain support functions performed by the Directorate General of Public Finances (DGFiP). However, the general scheme’s costs are overstretched and to an undetermined extent by the “inter-scheme” duties undertaken by the French national old age pension fund (CNAV). With regard to the CNRACL, IRCANTEC and RAFP, the mission encountered the complexity – perhaps even a lack of transparency – of the DPS cost model. Since 2018, the DPS has progressively diversified the prices charged to the schemes from the actual cost incurred for their administrative management as determined by the cost accounting. This diversification, which mainly results in an underestimation of the CNRACL’s management costs by 20%, is specifically intended to appear more competitive and to obtain a mandate to manage new pension schemes – primarily that of the PCMR. Furthermore, a quarter of the costs incurred by the DPS are related to multi-disciplinary and support services for headquarters (known as “federal” costs), based on key aspects that are unfavorable to the DPS and have a questionable variability over time. For example, the decision of the CDC’s Director General in November 2022 means that €24 million of the €75 million in federal costs incurred by the DPS will no longer be charged to it in the future in the interests of competitiveness.

As a result, neither the costs charged nor the analytical result of the DPS reflect the economic reality, but rather a political and administrative balance between principals, custodians and the CDC, and within the latter. Management by the State and by the administrators of the schemes entrusted to the DPS is therefore based on management data that are inherently not representative of the reality of the costs incurred, which has a detrimental impact on how they operate. For example, the trajectory for reducing management costs set by Article 14 of the 2018-2022 French National Public Finance Programming Act cannot be verified without full access to the actual underlying costs; the mission estimates that the CNRACL deviates from it by 10 to 20%. It is therefore critical to standardize the DPS’ communication of the budgets, accounts and invoices of the schemes by aligning them with the standards applicable to the State, and the CDC must demonstrate transparency on its actual costs and how they are determined, which implies establishing an analytical accounting system linked to the general accounting system for this purpose. In any case, increased mutualization of pension schemes, or even merging some of them, would be an effective lever for controlling management costs. In fact, empirical data highlights a negative correlation between the total sum of pensions paid and the proportion of operating costs. The current reform of the PCMRs and the merger of the AGIRC and ARRCO have also created synergies that demonstrate the benefit of these mergers.

However, the mission identifies a major weakness in the way public pension management is organized related to the illegality of the terms for granting management mandates to the CDC for six schemes. According to analyses it has performed, shared by the legal affairs department of the Ministry of Economics and Finance, this administrative management constitutes an economic activity and the management mandates are therefore public contracts with a cumulative value of €230 million per year, yet these contracts are entered into without being advertised or put to tender and are open-ended. However, neither the non-economic nature of pension schemes in themselves, nor the theory of them being economic services for public benefit, are sufficient to justify an exclusive right for the CDC to manage these schemes without any competition; and all the more since the CDC positions itself as an economic actor in a competitive way and in search of growth.

There is therefore a high risk that, in the event of litigation before the Council of State or the Court of Justice of the European Union, the mandates will be terminated and the
State shall be held liable. It would therefore be appropriate for the public authorities to prevent this situation by ensuring that the organization of public pensions be made compliant.

This can only be achieved by increasing the control level of the principals and the State on the DPS’ pension activity by placing it under the “in-house providing” scheme. Therefore, the proposed target scenario in the mid-term involves setting up a new public institution in two stages, starting with transforming the State Pensions Service into an institution in 2025, and then transferring to it the DPS resources for management of the CNRACL, the RAFP and other smaller integrated schemes. This institution would be controlled by its principals, i.e. all the pension schemes for civil servants and servants with equivalent status, who would sit on its Board of Directors, thus facilitating the convergence of these schemes with each other and then towards a universal system.

However, the IRCANTEC would not be integrated into this group. In fact, the efficiency
gains that would be achieved by sharing the management of this scheme with the AGIRC-ARRCO,  in view of the proximity of the rules for the acquisition of rights and affiliates, argue in favor of a merger of these two institutions. The resulting decrease in IRCANTEC management costs would be 20% to 40% depending on the selected terms.

In addition, the deteriorating situation of the IRCANTEC will to a certain extent require there to be shared parameters between the schemes in the near future. Lastly, the better efficiency shown by the IRCANTEC can only be explained by temporarily favorable demographics.
In the interests of simplification, rationalization and fairness, the mission therefore recommends starting to phase out the special IRCANTEC scheme in the medium term, which will lead to the alignment of contractual employees working under a public law contract with the AGIRC-ARRCO regulations.

ABBREVIATIONS & ACRONYMS