7/26/2019 SSRN-id2606547

1/37

1

Accounting for Pension Flows and Funds:

A case study for accounting, economics and public finances

15 Mai 2015

Yuri Biondi (Cnrs ESCP Europe), yuri.biondi@gmail.com

http://yuri.biondi.free.fr/

Research Director and Member of the Executive Committee of the Labex ReFi (Financial Regulation Research Lab), ESCP Europe, 79 avenue dela Rpublique, Paris 75011, France. http://www.labex-refi.com/en/

Marion Sierra (University of Paris Dauphine), marion.sierra-torre@dauphine.fr

https://www.linkedin.com/in/marionsierra

EGPA XII Permanent Study Group Public Sector Financial Management

Workshop in Zurich-Winterthur (Switzerland), May 7-8, 2015The Accountant, the Economist and the Politician and other stories in Public Sector Financial Management

Abstract:

Accounting for pension obligations has been coevolving with political and financial economic strategiesaimed to prompt and promote active financial markets and institutional investors, as well astransnational harmonisation and convergence of accounting standards between private and publicsectors. In this context, our article provides a theoretical analysis of accounting for pension obligations,drawing upon a comprehensive review of existing practice and regulation. The latter are stillinconsistent with the actuarial representation that has been adopted by the IPSAS 25 (EmployeeBenefits) and the IAS 19 (Employee Benefits). According to our frame of analysis, a variety of viable

modes of pension management exists and shall be acknowledged by accounting and financialregulations. Accounting (and financial economic) concepts and regulatory recommendations are thenelaborated in view to clarify and improve on pension protection, that is, the assurance of continuedprovision of pension payments at their agreed levels under viable alternative modes of pensionmanagement.

Keywords:

Pension provision, pension benefit, pension liability, IPSAS, EPSAS, pension fund management, actuarialevaluation, public sector accounting regulation, public finances

JEL Codes: H55; G23; G28; M41; M48; K23

Acknowledgments: We wish to thank Mme Danile Lajoumard and Mr. John Stanford for their valuablehelp through discussion and documentation provision. We further thank Andreas Bergmann, EugenioCaperchione, Sandra Cohen, Giovanna Dabbicco and the other participants to the EGPA PSG XII 2015Spring Workshop in Zurich-Winterthur (7-8 May 2015) for their useful comments. Usual disclaimerapplies.

7/26/2019 SSRN-id2606547

2/37

2

Accounting for Pension Flows and Funds:

A case study for accounting, economics and public finances

Introduction

Pension funds and obligations have been a critical matter all along the international accountingconvergence process for both private and public sector entities. Their recommended accountingrepresentation has coevolved with political and financial economic strategies aimed to prompt activefinancial markets and institutional investors. These strategies lead a growing number of accountingrepresentations to consider pension obligations as deferred remuneration, to be attached today toindividual pay and expensed through actuarial measurement of future cash outflows. Outstandingliability (related to accrued future obligations) is then expected to be included in the liability-side ofbalance sheet. At the same time, these strategies foster the immediate funding of pension obligations,creating cash funds to be invested, monitored and recognised in the asset-side of balance sheet.

Notwithstanding continued pressure for adopting this actuarial accounting representation, practice and

regulation are still diversified between pension funds and across jurisdictions. Accounting standards-setting implies then a delicate balance between constructing one best practice and acknowledgingexisting alternative options.

In this context, our paper aims at elaborating a theoretical perspective that helps disentangling somekey features of existing practice while elaborating a frame of analysis, i.e., an accounting model ofreference for pension funds and flows over time. This elaboration bases upon illustrative review ofregulation and practice, together with numerical visualisation. This modelling step may enable to betterunderstand current practice, to disentangle concurrent positions surrounding accounting standard-setting proposals, and to provide recommendations for improved accounting representation of pensionfunds and obligations in public and private sector entities. Our analysis focuses on governmental

financial accounts and reports, while scoping out national accounts systems1. Specific attention isdevoted to pension funds and flows in the context of ongoing harmonisation of European Public SectorAccounting Standards (EPSAS) led by European Commission (Eurostat).

The rest of the paper is organised as follows. The first section situates the recent debate on accountingfor pension obligations in the context of political and financial economic strategies aimed to promptactive financial markets and institutional investors, as well as transnational harmonisation andconvergence of accounting standards between private and public sectors. The second sectionelaborates some conceptual tools to better understand pension obligations between accounting,finance and management. The third section applies this frame of analysis to the case of Internationaland European Accounting Standards-making for pension obligations. A summary of main argument and

results concludes.

First Section. Pension obligations framed between the financial placements one best way and

existing variety of inconsistent practices

1.1 Pension reforms and the financialisation movement

According to Josiah et al. (2014, p. 18), which provide further references, over the past three decades,particularly from the mid-1980s, there have been many significant changes in the concept and detail of

1 Reimund Mink (2008), senior advisor at European Central Bank, summarises the main standards of the existingnational accounts systems in a paper titled General government pension obligations in Europe.

7/26/2019 SSRN-id2606547

3/37

3

pension provision in both public and private sectors. These changes are occasioned by governmentpolicy and influenced by capital markets.

Among others, in 1994, the World Bank (1994) developed a framework for pension obligations basedupon three pillars: (i) A publicly managed pension system that requires mandatory participation from allmembers of society but is only aimed at alleviating poverty, not providing a comfortable retirement; (ii)

a privately managed pension system that ideally would cover all members of society; and (iii) voluntarysavings by individuals. According to ApRoberts (2007), this three-pillars approach was furtherendorsed by the European Commission in 19972. This approach does

[] constitute a normative project aimed to prompt capitalisation while limiting pay-as-you-go

mechanisms as mode of funding pension obligations. This is an ideological endeavour whose

objective is to persuade that pay-as-you-go scheme, the main mechanism of funding pension

obligations in most countries, has a too large place []. This choice draws upon a free-market

conception that assumes individuals to act alone, although aiming to make individual saving

compulsory.

relve dun projet normatif visant promouvoir la capitalisation et contenir la rpartition

comme mode de financement de la retraite. Il sagit bien dune entreprise idologique dont

lobjectif est de convaincre que la rpartition, mcanisme principal de financement des pensions

dans la plupart des pays, occupe une place trop importante. La Banque Mondiale avec le

deuxime pilier affiche une nette prfrence pour les comptes individuels dpargne retraite

obligatoires la chilienne quelle contribue instaurer avec un certain succs en Amrique

latine et en Europe centrale et orientale. Ce choix senracine dans une conception librale o

lindividu est cens agir seul, mme sil sagit de rendre lpargne individuelle obligatoire.

According to the World Bank (1994), population aging, which is predictable, makes it very costly toreduce poverty and replace wages though a single pillar that is pay-as-you-go financed. Therefore, theWorld Bank recommended separating the saving function from the redistributive function and placingthem under different financing and managerial arrangements in two different mandatory pillars - onepublicly managed and tax-financed, the other privately managed and fully funded - supplemented by avoluntary pillar for those who want more.

This World Bank policy recommendation clearly illustrates the connection between pension reformsand emergent political and financial economic strategies which aim to foster active financial marketsand institutional investors. In line with this connection, a World Bank study by Holzmann (2005) arguesthat:

Pension reforms should seek to create positive developmental outcomes through increasednational saving and financial market development. (p. 6)

Funded pillars ideally should be introduced gradually to enable [Latin America and Central andEastern Europe countries] to facilitate financial market development. (p. 15)

These strategies intend to facilitate and organise the collection of (supposedly individual) savings to begathered through institutional investors towards active financial marketplaces where those institutionalinvestors manage investment portfolios on behalf of saving individuals. In turn, corporate entities areexpected to access those financial marketplaces to get funding that is required for their ongoingfinancial needs. Financial marketplaces come then to play a key coordination role between institutional

2 In fact, according to Holzmann and Hinz (2005), the World Bank has amended this approach by adding twoadditional pillars which better acknowledge the social assurance dimension of pension obligations. This showsthat the debate is not yet settled in favour of the individual savings account view.

7/26/2019 SSRN-id2606547

4/37

4

investors and corporate entities, while savings remuneration becomes dependent on financial marketdynamics.

These strategies are an integral part of a more general socio-economic movement labelledfinancialisation by recent literature surveyed by van der Zwan (2014) and Carruthers (2015). Accordingto Epstein (2001), financialisation refers to the increasing importance of financial markets, financial

motives, financial institutions, and financial elites in the operation of the economy and its governinginstitutions, both at the national and international levels. Consequently, financial practices and mind-sets have acquired an increasingly pervasive role in economy and society, across countries and

jurisdictions, in recent decades.

In this context, pension funds have been seen as a high-potential policy tool to collect material financialfunds to be employed for financial investment and trading (Berle, 1959; Dietz, 1968; Brown, 2014). Thecase studies of US and UK are especially illustrative of this development.

In US, since the seventies, major pension reforms were designed to prompt funded pension funds tobecome active financial investors by managing portfolios of assets held on behalf of pensionbeneficiaries. Before these reforms, pension funds were not allowed to freely manage their investmentportfolio, since prudential rules and constraints prevented them to invest in risky assets such ascorporate shares (Montagne, 2006). These reforms fostered pension funds to channel increasinglymaterial flows of financial funds through financial marketplaces. Consequently, many independentfunds started delegating financial management to leading financial institutions, which also entered thisbusiness by establishing or developing their own funds. This latter movement confirms the role offinancial institutions in the pension reform, consistent with the financialisation movement.

Concerning UK, pension reforms over the past two decades (1990s and 2000s) have fostered pensionfunds to become active on markets for corporate shares and for long-term bonds, including long-termgovernmental bonds. According to Greenwood and Vayanos (2010), the two major steps of this top-down transformation - the Pension Act of 1995 and the Pension Act of 2004 - expressed concerns withbankruptcy of sponsoring and providing entities. The Pension Act of 1995 response was to establishminimum funding requirements that coincided with accounting reforms linking the valuation of pensionliabilities to market-based discount rates. The Pension Act of 2004 introduced a government-backedguarantee fund expected to act as the rescuer of last resort in case of bankruptcy, while establishingfines for underfunded pension plans. Concerning private pensions schemes accounting, UK FinancialReporting Standard 17, which became mandatory in 2005 and was the UK equivalent of the IAS 19,forced UK companies to recognise pension liabilities in their financial statements at current value(Chitty, 2002; Slater & Copeland, 2005).

1.2 Existing variety of pension practices: evidence from the public sector

Notwithstanding this general movement of pension reform, public pension schemes still provide strongevidence of diversity in existing practices. For instance, let explore the case studies of US, UK andFrance.

In US, state governments pension schemes are only partly pre-funded (Ponds, Severinson, & Yermo,2011; Lav & McNichol, 2011; Rauh, 2010). However, the World Bank report by Holzmann and Hinz(2005, p. 46)3 states an accounting practice that results in substantially unfunded schemes (see alsoBNAC (2009)):

3 R. Holzmann, professor of economics at Universiti Malaya (Malaysia) and UNSW (Australia), was the SectorDirector and Head of the Social Protection & Labor Unit at the World Bank Group between 1997 and 2009.

7/26/2019 SSRN-id2606547

5/37

5

[US] States that accumulate large reserves within their pension funds do not act as though thefunds were available to finance non-pension government operations. This behaviour contrastswith the experience in national-level governments. [They] have attempted to prefund a portionof their public pension liability [but] a large proportion (60100 percent) of the pension fundaccumulation in national social insurance systems is found to be offset by larger deficits inother budgetary accounts. Smetters (2004) claims that the offset for the United States exceeds

100 percent. ()

In UK, most major public pension funds remain unfunded. According to BNAC (2009), the five largestunfunded schemes, accounting for 96% of outstanding unfunded liabilities in 2009, are the NationalHealth System (NHS), the Teachers, the Civil Service, the Police and the Armed Forces. The sixthscheme, which is funded, is the centrally guaranteed, but locally administered, Local GovernmentPension Scheme (LGPS). In 2009, these schemes covered together about 6.4m employees, or about 25%of the UK labour force.

In France, the main schemes for public sector employees are financed on a pay-as-you-go basis, andno liabilities are recognised in public sector financial statements (Ponds, Severinson, & Yermo, 2011).The French pension scheme for central government employees, magistrates and military personnel, andlocal government employees (CNRACL) apply a pay-as-you-go financing method. Only, the RAFP(Retraite Additionelle de la Fonction Publique), which is a complementary pension scheme for centralgovernment employees, local government employees and healthcare employees, applies a fundedfinancing method.

1.3 The one best way overarching ongoing reforms

Fostered by the financialisation movement, an emergent normative pattern has been driving pensionreforms. It is a fundamentally normative process since existing practices are still inconsistent with itsrequirements and underlying views. While the previous section provided evidence from public sectorpension funds, this section will show evidence from a variety of management modes such asgovernmental pension funds, corporate pension funds, independent mutual and defined benefitpension funds, and notional defined contribution schemes.

According to this new norm, pensions are understood as individual saving accounts held on behalf ofpension beneficiaries. According to Le Lann (2010) :

While pension public money used to be mainly understood as an intergenerational system ofsolidarity, it is progressively metamorphosed into a system of individual savings [since thenineties]. [] Progressive introduction of a liability for pay-as-you-go schemes should beunderstood as a triggering consequence of the movement toward financial capitalisation inpension statistics.

Alors que largent public de la retraite tait prioritairement pens comme un systme desolidarit intergnrationnelle, il se mtamorphose progressivement en un systme dpargneindividuelle. [] Linstitution progressive dune dette des rgimes en rpartition doit trecomprise comme lacte gnrateur du mouvement de patrimonialisation des statistiques de lapension.

This emergent view mingles pensions together with other financial investment funds held by individualsfor precautionary or speculative reasons. An ideal pension fund is then supposed to be attached toeach individual and transferable, when his beneficiary leaves the corporate entity that has promised his

future pension payment and has been managing his pension account. Pension is then understood as anindividual saving account that is supposed to stock ongoing inflows paid against accrued pension rights.

7/26/2019 SSRN-id2606547

6/37

6

These inflows are expected to be continuously invested over time to accumulate the final pensionamount that shall be liquidated, in form of lump-sum or life rent, at a certain maturity date scheduledby the individual pension scheme. According to this pension as saving account view, it is somehow theindividual accumulation of cash that is expected to generate the future pension, period after period.Indeed pension becomes another kind of financial placement at the individual level.

From this perspective, according to Bodie et al. (1988, p. 139):

[Pension] Benefit levels depend on the total contribution and investment earnings of theaccumulation in the account. Often the employee has some choice regarding the type of assetsin which the accumulation is invested and can easily find out what its value is at any time.

According to Broadbent et al. (2006), which refer to Barr (2009):

Defined Contribution plans4 () provide employees with much more control, choice andflexibility in terms of how they manage their retirement savings and investment, and indeedhow they manage their financial assets over their lifecycle.

In this context, the accounting representation of pension obligations is expected to enact thisunderstanding by providing actuarial information about the current net value of the joint pension fund,as if each individual were able to claim its current value pension share at will. A pension fund becomesthen similar to a close monetary fund whose shares split ongoing fund current value amongbeneficiaries, after deduction of management fees and other corporate running costs.

Notwithstanding continued pressure to adopt this actuarial accounting representation, financialpractices and regulations are still diversified between pension funds and across jurisdictions. Types ofpension funds exist in virtually all countries and jurisdictions that do not conform to this idealrepresentation. Let explore the cases of governmental pension funds; corporate pension funds;independent mutual and defined benefit pension funds; and notional defined contribution schemes.

Governmental pension funds are still largely unfunded and based upon pay-as-they-go5 schemes.Some countries such as China, Colombia, Brazil, Belgium, France, Germany, Ireland, and Luxemburghave separated unfunded schemes for civil servants pensions (Pinheiro, 2004).

Moreover, notwithstanding changes in their accounting representation through enforcement of the IAS19, some corporate pension funds maintain close ties with corporate sponsors, including in theirinvestment strategies and pension operations. They may then remain substantially unfunded.Moreover, there are risks implied by self-investing: in case of bankruptcy, self-investing strategies canbe harmful as it have been the cases in the United Kingdom for the Robert Maxwells collapse, and inthe United States for the Enrons collapse (Sullivan, 2004, pp. 83-84). As stated by Sullivan, self-investing remains a current practice (2004, pp. 84-85):

Many US corporations make contributions to their employees 401 (k) accounts in the form oftheir own company shares. These companies typically have around one-third of their 401 (k)plans self-invested. This compares with about 19 per cent for all 401(k) plans. While a high levelof self-investment is contrary to the principles of prudent diversification, it is perfectly legal.Although pensions regulation impose a 10 per cent limit on self-investment, the restriction onlyapplies to defined benefit schemes. The 401 (k) plans of some of Americas largest and most

4According to a World Bank report (Holzmann, Palacios, & Zviniene, 2004, p. 10), the notional defined

contribution pension combines the individual accounts of a privately managed defined contribution scheme withpay-as-you-go financing. Please refer to our Box 1 for working definitions of main pension schemes.5 Hereafter, we replace the usual expression pay-as-you-go with pay-as-they-go, to highlight the collectivedimension of this system, where current contributors pay for other peoples pension through time.

7/26/2019 SSRN-id2606547

7/37

7

venerable companies members of the Fortune 500, with thousands of employees have self-investment levels above 60 per cent. For example, in 2002, Coca-Cola and General Electric hadabout 75 per cent of their 401 (k) plans invested in their own shares. The equivalent figure forProcter & Gamble was just under 95 per cent.

Independent mutual and definite benefit pension funds are not managed as close monetary funds. They

may be partly unfunded and provide collective guarantees and transfers that are inconsistent with thepension-saving account representation. Some countries have only partially funded schemes such as US,Singapore, Ecuador, Canada (Augusztinovics, 2002) and Finland (Oulasvirta, 2008). In fact, othercountries such as the United States, the United Kingdom and Sweden have collective guaranteemechanisms, which can be voluntary or mandatory and concern either the private sector or the publicsector. According to ApRoberts (2007), a common fund is essential to achieve a certain equality oftreatment in the group concerned. If a voluntary membership scheme allows a capital output, it makesno pooling of risks.

Despite increased individualisation, notional defined contribution schemes maintain a collectivedimension and are partly unfunded. In countries such as Sweden, Italy, Latvia and Poland, publicnotional defined contribution accounts are maintained along with unfunded public pension liabilities(Holzmann, Palacios, & Zviniene, 2004, p. 10). According to Nistic & Bevilacqua (2012), Italy andSweden moved toward notional defined contribution pension schemes by "[adopting] defined-contribution rules while retaining a payasyougo financial architecture. According to Cichon (1999):

() Notional defined-contribution (NDC) system [consists in] individual social insurance pensioncontribution records [that] are converted into a fictitious savings amount at retirement, whereupon the defined-contribution approach is followed. It is a novel pension policy instrumentrather than a new type of pension formula, and most of its potential financial and distributiveeffects could also be achieved by a classical, linear defined-benefit formula. It is the packagingthat differs and, in politics, that often is what matters.

In conclusion, a clear move has been prompted to consider pension obligations as deferredremuneration, to be managed on behalf of its ultimate beneficiary through a financial investmentscheme, accounted for through an actuarial basis of accounting. However, the current state of pensionobligations affairs does not deliver a clear-cut, consensual view on their concept (what they are), theirmanagement (how they are fulfilled) and their accounting representation (how they are accounted for).Current practice and regulation still vary across funds, countries and jurisdictions.

In this context, accounting standards-setting implies a delicate balance between constructing the onebest practice, and acknowledging alternative options that factually exist. Should accounting regulationfor pension obligations follow that emerging pension as individual saving account view? The followingsection aims at elaborating a theoretical perspective that may help disentangling some key features ofexisting practice while elaborating a frame of analysis, i.e., an accounting model of reference forpension funds and flows over time.

Second Section. Toward a comprehensive frame of analysis for pension obligations

2.1 Accounting implications of the one best way

Political and financial economic strategies designed to develop active financial marketplaces lead agrowing number of accounting representations across jurisdictions to consider pension obligations asdeferred remuneration, to be attached today to individual pay and expensed through actuarial

measurement of future cash outflows (Napier, 2009). Outstanding liability (related to accrued future

7/26/2019 SSRN-id2606547

8/37

8

obligations) is then expected to be included in the liability-side of balance sheet. As summarised byOulasvirta (2008):

[Under US GAAP, IAS/IFRS and IPSAS standards], the whole accrued and unpaid pension debtmust be shown in the balance sheet and not only in the notes of the balance sheet or in theannual report narrative. If the central government has a funding arrangement for state

employee pensions and there is a special separated entity (pension fund) that is responsible ofthe funding, the pension liability and debt recording belongs to this pension fund. When thepension fund belongs to the national government, the national government must also show thedebt in the consolidated whole of the central government balance sheet.

At the same time, these strategies foster the immediate funding of pensions, generating cash funds tobe invested, monitored and recognised in the asset-side of balance sheet. Those funds are evidentlyexpected to be invested and generate financial returns in active financial marketplaces. Referring tofurther literature, a Word Bank report (Holzmann R. a., 2005, p. 47) argues that:

() Funded schemes clearly seem to promote the development of securities markets (Impavido,Musalem, & Tressel, 2003), making them more liquid and deeper as well as more sophisticatedand innovative (Walker & Lefort, 2002).

In a first approximation, we can confidently summarise this alleged one best way as follows: (i)pensions are understood as deferred remuneration, (ii) to be accounted through a balance sheetdetermination of outstanding liability (iii) at its discounting-based actuarial measurement (currentvalue).

According to Le Lann (2010), Donaghue (2003), an IMF economist, provides a clear-cut illustration ofthis view while recommending the treatment of pension obligations as financial liabilities (see alsoFeldstein (1997)). In particular, Donaghue (2003) focuses on the passage of time as the determinantof accruing liabilities and argues that:

A present obligation has been created simply by persons being in observance of the conditionsrequired by the governments pension policy up to the present periodthat is, the passage oftime has created reasonable expectations which leaves the government with no realisticalternative to meeting (at least in the main) those expectations. () The operation ofgovernment pension schemes gives rise to present (constructive) obligations of government,where the obligating event is simply the passage of time while the pension scheme is inoperation. () These criteria fit the definition of a liability [according to IPSAS 1]. [(p.6) boldadded].

Therefore, pension liabilities should be recognized on the balance sheets of government, andthe corresponding transactions (transfers to households) should be included in themeasurement of the governments operating result in the periods during which the pensionobligations accrue. (p.7)

In accounting terms, all government non-exchange pension schemes give rise to liabilities (ineconomic terms insurance technical reserves), and expense and financing transactions. Theliabilities are the net present value of obligations accrued to the current date. (p.11)

Donaghue (2003) argues that by relieving an individual of the present need to make provision forfuture risk, the government is actually providing a current benefit (similar to insurance cover) andshould therefore consider pensions obligations as financial liabilities. The author underpins its

arguments by quoting the US governmental accounting standard:

7/26/2019 SSRN-id2606547

9/37

9

This same point is made in the USA Federal Accounting Standards and Advisory Board SFFASNo. 17 Accounting for Social Insurance which notes that some argue that it is inherentlymisleading to fail to quantify the size of the promise that is being made and on which peopleare told they can rely.

Donaghues arguments are then in line with the concept of pension as entitled saving account:

The corresponding entitlements of individual citizens begin to accumulate when they reach theage of economic independence, and increase until they reach pensionable age (providing theyare in observance of the conditions applying to the scheme), after which they begin to declineas benefits are paid out.

In fact, Donaghue (2003) himself concedes that that the recognition of pension obligations as a liabilityis not self-evident and goes beyond current practice and regulation. This further implies that he hasbeen asserting a normative view that is not based upon a heuristic or comparative analysis of pensionmanagement modes.

The recognition of obligations (and corresponding economic flows) arising from governmentpension schemes recommended in this paper goes beyond that set out in GFSM 2001. (p. 2)

It is important to note that the treatment proposed in this paper will present a very differentview of the economic stocks and flows relating to government pension schemes from pay-as-you-go or similar treatments. (p. 7)

It has been argued by some fiscal economists including at the International Monetary Fund(IMF) that the social pension arrangements do not adequately satisfy the constructiveobligation criterion, due to the relatively soft nature of the commitment. Some fiscal analystswould therefore prefer that there be disclosure of the possible scope of government liabilities,rather than full recognition of the liability in the financial statements. (p. 12)

Oulasvirta (2008) summarises this one best way to treat accounting for pension obligations as follows:

US GAAP, IAS/IFRS and IPSAS standards all have as a starting point that post-employmentbenefits should be recognized and recorded following the principle of pension benefits earnedduring the work. Based on this principle, pensions should be recorded on the following grounds.

The paid salary and the unpaid part of the salary which is the pension benefit earnedduring the accounting period should be recorded on accrual basis

as an expense incurred in the income statement (the profit and loss statement)

and the earned but still unpaid part of the benefits at the book closure date as a liability(debt) in the balance sheet.

The paying of the pension benefit to the recipients are only instalments of the debt(pension liability) recorded on a cash basis (no impact on income statement).

The rest of the section will assess this way against an informed theoretical analysis based upon a reviewof current practice.

2.2 A theoretical analysis based upon the review of existing practices

As a matter of fact, the one best way for accounting for pension obligations is not consistent with allthe existing practices that still characterise pension management and reporting across countries and

jurisdictions. Moreover, important debates still exist even within this accounting view, especiallyconcerning the preferred actuarial method to estimate remote and uncertain flows subject to

7/26/2019 SSRN-id2606547

10/37

10

unforeseeable change and hazard, as well as the appropriate discount rates of reference and relatedupdating over time (Bader & Gold, 2003; Dietz, 1968; Exley, Mehta, & Smith, 1997; Bader & Gold, 2003;American Academy of Acturaries, 2007; Mortimer & Henderson, 2014; Waring, 2009; Keating,Settergren, & Slater, 2013).

Pension management generally occurs through organised entities that perform it on behalf of sponsors

and beneficiaries. This entity dimension is not consistently included in the view of pension obligationsas individual saving / deferred remuneration accounts. A saving account can exist independently frommanagerial delegation to a specialised financial entity, while an entity-held pension account does notnecessarily feature all the characteristics - regarding appropriability, transferability, remuneration andso forth - that functionally define an individual saving account. For instance, an entity-held pensionaccount may not be appropriable or transferable at will, while it may not be fully funded at every time6.

Whatever separately incorporated or not, a financial entity specialised in pension management(generally labelled pension fund) is purposively devoted to pension obligation fulfilment over time.This fulfilment constitutes its constitutive mission. Perform this mission involves two complementaryworking processes to be accounted for:

i. One process does concern the series of cash (cash-receivable, and cash-promised) flows thatpass through the entity from ongoing contributing members (including future expectedbeneficiaries and committed sponsors) to incipient and future beneficiaries. This processinvolves cash and non-cash financial funds and flows to be accounted for. It points to thefinancial dimension. This cash process of entity economy is consistent with a cash basis ofaccounting.

ii. The other process does concern the economic recovery of the outward flow of payments thatare due over time to incumbent beneficiaries. This process involves recognition andmeasurement of ongoing payments and matching contributions, as well as dedicated assetsand outstanding liabilities that are generated by ongoing management of pension entity overtime and hazard. It points to the economic dimension of entity economy and is consistent withan accruals basis of accounting.

From this perspective, an individual saving account approach to pension involves a quite narrow viewon both processes. Accordingly, each individual is expected to collect his own series of cashsettlements, which, through financial placement, are the only way to get future pension payments byongoing financial accumulation that is dependent on cumulated cash funds and proceeds. To expand onthis view, it may be useful to disentangle its implicit understanding of pension management through adualistic approach that identifies couples of contrasting terms. This dualistic approach draws upon areview of existing practice, as follows:

Individualistic vs collective: this discriminating concept distinguishes between individualisticand collective approaches to pension obligations. This concept especially refers to theeconomic process. According to individualistic approaches, each individual is expected to payfor himself. Social solidarity through mutualistic transfers is then excluded, in principle.According to collective approaches, the whole of constituting members assure the coverage ofpension payments over time. Individualistic appropriation is then excluded, in principle.

Funded vs. unfunded: this discriminating concept distinguishes between funded and unfundedpension obligations. This concept especially refers to the financial process. Under fundedschemes, the pension account(s) is expected to contain cash and cash-receivables to beinvested to recover future pension payments. Under unfunded schemes, the pension

6 Bohn (2011) reflects on the ambiguous meaning of full funding, stating that there are several conceptuallydifferent ways to its interpretation.

7/26/2019 SSRN-id2606547

11/37

11

account(s) is not expected to contain cash. It identifies outstanding pension rights andobligations that do not necessarily match some underlying financial investment process.

Stock vs. flow basis of accounting: this discriminating concept distinguishes between the twomost general families of accounting models (Biondi, 2012). A stock basis of accounting adopts abalance sheet accounting approach which gives priority to recognition and measurement of

assets and liabilities as they stand at one point of time, to represent and account foroverarching managerial processes. A current (fair) value accounting measurement is generallyconsistent with this stock basis. A flow basis of accounting adopts an income statementaccounting approach which gives priority to revenues, costs, contributions and expenditures.An historical cost (historical nominal amount) accounting determination is generally consistentwith this flow basis of accounting.

Deferred remuneration vs. social protection: this discriminating concept distinguishes betweentwo alternative understandings of pension rights and obligations. On the one hand, pension isunderstood as deferred remuneration that is due to the individual along with the currentremuneration. In its pure form, this implies that both accrued pension payment and its cash

liquidation are performed in the accruing period when the current remuneration is paid. On theother hand, pension is understood as social protection that is granted by a whole ofconstituencies (pension fund members, citizenship) and delegated to a pension-purpose entity(mutual, governmental). In its pure form, this implies that ongoing pension payments areassured by that entity (including on behalf of sovereign powers) and do not belong tobeneficiaries before they are due and liquidated.

Table I summarises these couples of contrasting terms.

Table 1. Couples of contrasting terms

Dimension of Reference Discriminating concept

Economic process Individualistic vs. collective

Financial process Funded vs. unfunded

Accounting representation Stock basis vs. flow basis of accounting

Overarching definition Deferred remuneration vs. social protection

According to this frame of analysis (Table I), the pension as individual saving account view that has beenrecently affirmed corresponds to an individualistic approach which involves funded financialmanagement and a stock basis of accounting, as for pensions are understood as deferredremuneration. At the opposite side, we can situate the pure unfunded pay-as-they-go scheme that hasbeen generally adopted by public sector pension funds. This latter collective approach fosters anunfunded financial management, it prefers a flows basis of accounting and it understands pensions as

social protection assured by the pension-purpose entity on behalf of the whole union (the citizenship;the mutual community of members).

Our frame of analysis overcomes the alleged one best way that defines pension as an individual savingaccount. Overarching managerial processes exist and are sustainable (depending on conditions andcircumstances) under various models which correspond to, and can be classified through each couple ofdiscriminating concepts. For instance, Table 2 shows a classification of existing practices according totwo discriminating concepts: individual vs. collective, and stock vs. flow basis.

Table 2. A theoretically-informed classification of existing modes of pension management

Stock basis Flow basis

Individualistic Defined Contribution Schemes Individual saving accountCollective Defined Benefit Schemes Pay-as-they-go Schemes

7/26/2019 SSRN-id2606547

12/37

12

Accordingly (Table 2), under individualistic regimes, Defined Contribution (DC) schemes apply a stockbasis of accounting and management. Each individual is then expected to hold a share of the pension

joint stock. Individual saving accounts are by definition personal, but they are generated by progressiveaccumulation of savings flows and related reinvestment proceeds. Under collective regimes, DefinedBenefit (DB) schemes promise continued pension payments on behalf of the whole of constituencies(including beneficiaries and sponsors). Pay-as-they-go schemes make the same promise, but fulfil it

through ongoing matching between contributions and pension payments over time.

In this context, a stock basis of accounting points to the notion and function of money as reserve andmeasure of value (Le Lann, 2010), while a flow basis of accounting points to the notion and function ofmoney as symbol, means of payment and unit of account (Biondi, 2010). In the first case, dedicatedasset portfolio refers to the very existence and accumulation of identifiable assets that are expected topay for future pensions. In the latter case, pension commitments stand as acknowledgement of statedpromises of future payments by the entity responsible for the fulfilment of these promises.

From this perspective, the one best way that understands pension as an individual saving accountappears to be inconsistent with received social meaning of pension. This meaning has been related toprotection granted to the old or the sick. In its pure form, the pension as saving account view holdseach individual - independently from all the others - responsible for its own financial sustainability overold age, making it dependent on the hazardous results of the ongoing financial investment process.However, according to the Merriam-Webster dictionary, pension means an amount of money that acompany or the government pays to a person who is old or sick and no longer works, while its meaningas wage is considered as archaic. According to the Oxford, pension means a regular payment made bythe state to people of or above the official retirement age and to some widows and disabled people.7

Late Middle English (in the sense 'payment, tax, regular sum paid to retain allegiance') derives from OldFrench, as well as from Latinpensio(n-) that means 'payment', frompendere 'to pay'. This current verbsense dates from the mid-19th century.

In this context, notwithstanding discredit that has been claimed against them, unfunded pay-as-they-go schemes can be sustainable as long as current and future contributions from constituencies(including sponsors and future beneficiaries) go on matching current payments that becomes due toincumbent beneficiaries over time. Appendix A shows a numerical illustration denoting thissustainability.

Symmetrically, an actuarial representation of pension obligations can hide significant issues and hazardrelated with pension provision over time. Appendix B shows a numerical illustration denoting theselimitations.

Our numerical analysis further implies that funded schemes do not guarantee better provision andsecurity of pensions. Shortcomings of funded schemes have been occurring especially in the aftermathof financial crises. For instance: in the UK, in 1992, the Maxwell scandal (Augusztinovics, 2002, p. 26); in2000, the insurance company Equitable Life, and in 2007, the pension fund of Allied Steel and Wire. InUS, an illustrative example is offered in 2002 by the Enron bankruptcy and related scandal. In France,the additional pension fund for civil servants named CREF (caisse complmentaire de retraite de lafonction publique), which was partly funded, incurred financial distress and was transferred in 2002 tothe COREM under the supervision of the State8 (Pouzin, 2014). According to Augusztinovics (2002, p.

7 To be sure, Oxford Dictionary also adds the following definition: A regular payment made during a personsretirement from an investment fund to which that person or their employer has contributed during their workinglife.8 Thousands of contributors claimed in court for compensation for their damages and were partially satisfied(Pouzin, 2014). Some lawsuits remain unsettled (Prache, 2008), while the new fund COREM has been reducingpast pension promises as a consequence of the financial distress of 2002.

7/26/2019 SSRN-id2606547

13/37

13

26), the funding requirement can also lead to high administrative costs and low compliance rates as ofthe Chilean pension fund administrators (AFPs).

Since unfunded pay-as-they-go pension schemes can be sustainable (Appendix A), while partiallyfunded, financial-return-based pension plans can be unsustainable (Appendix B), we can conclude thatfunding and sustainability are not necessary linked. According to Augusztinovics (2002, p. 26):

Contrary to the new pension orthodoxys major arguments, there is ample conceptual evidencein the literature to demonstrate that the method of finance and the type of management areno panacea ().

From our perspective, the overarching accounting purpose concerns theprotection of pension promisesthrough enhanced reporting and disclosure. Accountability for pension management involves beingaccountable for the main purpose of that management, i.e., timely and continued provision of pensionpayments as they become due at their previously committed levels.

The pension as saving account model assures this protection through the financial accumulationprocess, which exposes funded pension liability to financing cost and risk, as well as investment costand risk, including misappropriation and misallocation by controlling parties. From this perspective, itsactuarial mode of accounting representation affords the danger to undermine control andaccountability, as for the discounting/unwinding measurement method cannot track actual cash flowsand funds that are involved in overarching managerial processes. Moreover, this actuarialrepresentation introduces subjectivity and volatility of valuation, making ongoing valuation dependenton assumptions over critical variables concerned with the financial accumulation process, includingdiscount rates of reference, and forecasts over very long periods of time (Biondi, 2014; Biondi, et al.,2011).

Unfunded pay-as-they-go model assures this protection through collective responsibility forincumbent beneficiaries, discharged by the managed entity on behalf of entity constituencies. This hashistorically led to a lack of accounting reporting and disclosure by both public sector and private sectorsponsors. Few information (if any) was provided through their accounting reporting, while noquantitative determination was included in their balance sheet concerning outstanding positions.However, Appendix A shows how the ongoing structure of flows and funds can be represented withouthaving recourse to current (discounted) values that are inconsistent with this model.

To conclude, our theoretical approach develops a more comprehensive and neutral perspective onaccounting for pension obligations. Accordingly, accounting standards-making is not so much requestedto endorse one particular mode of pension management, as to rule accounting options that make themconsistently represented and accountable for pension obligations over time. In particular, Appendix Ashows how an actuarial mode of accounting would provide information that is inconsistent withgovernance and managerial needs for pension obligations under pay-as-they-go schemes, whileAppendix B shows how this very method undermines disclosure on cash management by managingentities.

The next section will apply this theoretical frame of analysis to the current debate on harmonisation ofpublic sector accounting standards across Europes Member States.

Third Section. Pension funds under International and European public sector accounting standards-

making

7/26/2019 SSRN-id2606547

14/37

14

This section applies the frame of analysis developed in the previous section to the case study ofaccounting for pension obligations under International and European public sector accountingstandards-making, respectively labelled the IPSAS and the EPSAS.

3.1 Pension obligations under the IPSAS: Issues and Perspectives

The Public Sector Accounting Standards (IPSAS) are developed by the IPSAS Board under the auspices ofthe International Federation of Accountants (IFAC). This privately-run body has been transplanting theIFRS developed by the IASB into the public sector. Since 2014, the IFAC, the Big 6 accountancy firms andinternational institutions such as the Organization for Economic Co-operation and Development(OECD), the International Monetary Fund (IMF) and the World Bank Group established a coalition calledAccountability. Now, in order to to help drive awareness of the critical need for high-quality,transparent, comparable public sector financial reporting [based on International Public SectorAccounting Standards], and of the importance of engaging citizens in the process of holdinggovernments to account (IFAC, 2015).

The IPSAS-Board has been fostering a convergence between public and private sector accountingstandards, although a conceptual framework that is specific to the public sector is under developmentby the IPSAS Board since November 2006 (IPSASB, 2015).

Since 2002 (first exposure draft issuance) through 2008 (issued standard n. 25), the IPSAS-Boardapproach to pension obligations has constituted a paradigmatic example of fostering the one best waydenoted in our previous sections. According to Le Lann (2010, p. 17), accounting standards-makingcontributes then to reshape pension reforms by establishing a single mode of assurance throughassets financial accumulation.

Issued in February 2008, the IPSAS No. 25 deals with employees benefits in general, including pensionbenefits. Accordingly, accounting for pension obligations is based on a dual accounting treatment thatdisentangles defined contribution (DC) and defined benefit (DB) schemes. Accounting treatment forother schemes (such as multi-employers, including employers under common control, state plans andcomposite social security programs) must make reference to one of these alternative treatments (seeIPSAS 25, 32, 33, 25.43 and 25.47).

According to IPSAS No. 25, defined contribution plans are defined in a way that excludes any currentobligation to pay for future pensions (IN6; par. 28). For a pension scheme to be qualified as definedcontribution plan, the only obligation that exists and is then recognised is the current annualcontribution paid by the reporting entity to existing employees as part of their current remuneration inexchange for service rendered by them.

In fact, the series of future contributions may represent an implicit liability under constructiveobligation, but the IPSAS standard excludes its recognition (par. 55). This accounting treatment fordefined contribution plans actually undermines accountability and responsibility of the reporting entityfor ongoing and future pension obligation fulfilment, since no pension obligation is assumed to existunder this accounting treatment. Therefore, potential beneficiaries do not receive informationconcerning the ongoing accumulation of their contributions, and the foreseeable level of pensionpayment that has been reached and may be sustained under the ongoing defined contribution schemeprocess. They do not receive information on ongoing investment policies and their past, current andforeseeable returns over time.

As long as some commitment to pay future pensions exists under legal or constructive obligation,

including through informal practice and social expectation (par. 63), the defined benefit accountingtreatment should be adopted. Termination rights by the sponsoring entity do not exclude this

7/26/2019 SSRN-id2606547

15/37

15

constructive obligation (par. 64), while pay-as-you-go pension schemes are explicitly considered as akind of defined benefit plan (par. 34 (a)).

According to the IPSAS 25, pension obligations should be considered as deferred remuneration whiletheir outstanding liability should be included into the sponsors balance sheet through an actuarialmethod of evaluation labelled Projected Unit Credit Method (see paragraphs 77-78 of IPSAS 25,

(IPSASB, 2008)). This method belongs to the current value accounting family, which relates to a stockbasis of accounting. The recourse to qualified actuarys expertise is then encouraged (par. 68), while thereference to market expectations for choosing actuarial hypotheses is required (par. 90).

Both the IAS/IFRS and the IPSAS have adopted a balance sheet accounting approach (asset-liabilityaccounting model), rejecting the revenue-expense accounting approach (Oulasvirta, 2008). Napier(2009) stresses the historical shift toward this balance sheet approach: early attempts to developaccounting standards were based on a cost orientation and reflected funding considerations. Morerecently, a balance sheet focus led to issues over identification and measurement of pension liabilitiesand assets. The IPSAS-Board has consistently maintained this balance sheet preference and focus overvarious amendments since 2002 to the final publication in February 2008.9

This IPSASB choice mimics the choice done for the standard IAS 19 issued for the private sector by theIASC. However, the specificity of public sector economy and finances casts doubts over thisconvergence regarding technical feasibility and the overall consistency with governmental economyand finances.

From the one hand, technical problems occur with complexity, cost, subjectivity and volatility impliedby the actuarial representation. These problems are common to both private and public sectors, as foraccounting is supposed to facilitate accountability and responsibility by the reporting entities. Theybecome especially sensitive for the public sector, which discloses and discusses budgets along withfinancial reports. Among others, Ouslasvirta (2008, p. 231) argues that:

Even small changes in certain parameters (for instance, the discount rate and life expectancy)may cause tremendous changes in the amounts of liabilities in the balance sheet.

Concerning the discount rate and the hazard of volatility attached to it, Lequiller (2014, p. 24),

counsellor to the OECD Statistics Directorate at Eurostat, expresses concerns on the recognition ofpension obligations in the balance sheet of EU countries:

However, if it was decided one day that pension obligations were to be recorded on balancesheet in the EU, there would be two necessary technical conditions which would need to beimplemented: (1) a common discount rate. Indeed, considering the massive impact of thechoice of the discount rate on the amount of the estimated obligation, and its arbitrariness, itwould be irresponsible to allow using different discount rates among Member States; (2) theheadline surplus/deficit should be protected from the volatility of the changes in the estimateof pension obligations. If not, it would be made useless for fiscal target making.

From the other hand, overall concerns relate to the specific economic nature of the public sector.Private sector entities are expected to go on allocating residual earnings to their shareholding

9 In November 2002, the Public Sector Committee (PSC) issued a draft on Employee Benefits for PSC Review. Then,a revision of this draft was prepared by John Stanford for the IPSASB Paris meeting and discussed on the 25 May2006. In October 2006, the IPSAS Board published the Exposure Draft (ED) 31 for Employee benefits. Commentson ED 31 were requested by 28 February 2007. As at 28 May 2007, thirty comments had been received. Then, ananalysis of these submissions on ED 31 Employee Benefits by John Stanford is available online, date 29 May2007, and was presented in July 2007 in Montreal meeting of the IPSASB. The final standard ISPAS 25 waspublished in February 2008.

7/26/2019 SSRN-id2606547

16/37

16

recipients, period after period. It may then be consistent with their accountability and responsibility toinclude a fair estimation of pension obligations, in view to better protect them against mismanagementand unsustainability over time. This estimation protects pension funds from being appropriated byincumbent shareholders. Biondi (2014) raises similar concerns for covering environmental liabilities.However, public sector entities are not expected to generate or distribute residual earnings. Inclusionof pension obligations as operating expenses is then less significant at the income statement level.

Moreover, public finances systematically employ debt issuance and refinancing to cover for assetpositions (Biondi, 2014). If public sector accounting standards put pressure over entities to includefuture obligations into their balance sheet, these entities may be encouraged or constrained to increasetheir debt position and incur additional cost to fund them, transforming their pension management intoa speculative hedge fund strategy leveraged on debt. On this matter, Benot Coeur (2007, p. 6), Frencheconomist appointed to the Executive Board of the ECB since 2011, casts doubts about converting non-financial pension obligations into straight financial liabilities:

Inclusion of pension commitments implies adding to the portfolio of financial debt anotherportfolio of pension obligations, of a gilt kind but with very long duration. In principle, thisshould lead to significantly reduce the duration objective for the financial debt. In other terms,does it valuable to increase very long-term governmental commitments through either giltslasting thirty or fifty years, or interest rate swaps over similar duration, while the states arealready committed to their employees over those time horizons?

Inclure les engagements de retraite revient adjoindre au portefeuille de dette financire unportefeuille de droits retraite, de type obligataire mais de duration trs longue. En toutelogique, ceci devrait conduire raccourcir significativement lobjectif de duration assign ladette financire. En dautres termes, cela vaut-il la peine daccroitre les engagements trslong terme de lEtat en mettant des obligations 30 ou 50 ans, ou en payant des contratsdchange de taux (swaps) trs longs, quand lEtat est dj engag vis--vis de ses employs de tels horizons ?

By adopting its standard (IPSAS 25), the IPSASB has overridden the issue whether pension obligationsshould be funded, arguing against reasons for underfunding pension commitments and for excludingthem from public sector balance sheets. In 2004, Eurostat argued that the IPSASB invitation tocomment on accounting for old age pensions (Eurostat, 2004, p. 3) does not articulate the reasonswhether and why civil servants unfunded employer pension schemes should be treated differently (i.e.,treated as pension funds) from social security pensions. Eurostat further commented on theimportance of the reference to the IAS in the IPSASB proposal (Eurostat, 2004, p. 3):

It is noted that the International Public Sector Accounting Standards (IPSAS) of the PSC does notinclude guidance yet on the recording of civil servants pensions. It is sometimes assumed thatthe International Accounting Standards Board (IASB)s standard on pensions (IAS 19) applies(see the EDG contribution by IFAC PSC staff).

Moreover, international organisations professionals, researchers (economists and accountants) andnational accounting boards stressed limits and concerns on pension funding and pension recognition inpublic sector balance sheets.

In particular, Rizzo (1990) argues that the analogy done by economists between unfunded pensionpromises and the issuance of governments bonds has its limitations, which Holzmann (2004)summarised as follows:

The creditors in a pay-as-you-go (PAYG) pension scheme do not enter into the agreement

voluntarily, but rather are forced by law to participate. Furthermore, the return on thegovernment bond is known (at least the nominal yield), while the ultimate value of a PAYG

7/26/2019 SSRN-id2606547

17/37

17

pension promise depends on a wide array of variables entering the defined benefit formula aswell as the possibility that the government may change the formula itself in response to otherfiscal demands.

In 2003, in response to Donaghue (2003)s article, pension experts argued against pension obligationsto be recognized as liabilities in financial statements (Aaron, et al., 2003). For instance, Barry Bosworth

stated that, at the national level, there is still a substantial interest in cash flow accounting to measurethe short term economic effects of fiscal actions and recalls that private firms game the system bychoosing among about 15 different formulas for measuring the accrued liability (Aaron, et al., 2003).

Moreover, Murray Petrie, another pension expert, argued (Aaron, et al., 2003) that:

The [Donaghue 2003] paper fails to justify why pensions should be recognized when contingentliabilities should not. [] The paper does not consider the practical effect of recognition on thevalue of the financial statements to users. [...] [Petrie] would expect that well-informed userswould quickly strip out the "pension effect" and calculate some kind of "underlying deficit." Inwhich case, what has been gained? Less well-informed users will find the reported deficitmeasure less informative. [] There is arguably far less information disclosed from simplerecognition on its own compared to the kind of detailed information that can be providedthrough supplementary reporting. []

Petrie further encouraged the IFAC (more exactly, the IPSAS-Board) to comparatively assess existingdifferent approaches on pension obligations:

Perhaps IFAC should be encouraged to put out a discussion document that compares andcontrasts different approaches to pension obligations, rather than an exposure draft of astandard they are heading towards.

Some international organizations servants stressed the limits of funding. Although advocating forfunding, the World Banks report prepared by Holzmann (2005) highlighted its costs and limitations:

While we claim that funding provides some (gross) benefits in many circumstances, we are alsovery much aware that it introduces new or additional costs, most importantly throughadditional risks (such as investment risks), higher transaction costs (such as fees), and fiscaltransition costs (when replacing an unfunded scheme) (p.44).

Palmer (2002) asks whether countries should continue with pay-as-you-go systems, or do individualfinancial-account systems provide a better alternative?. He argues that countries in the OECD havebeen reluctant to make this transition, however, due not only to the high initial cost for the transitiongeneration, but also to the financial risk involved.

A joint work published by an academic, Ponds, together with two OECD servants, Severinson and Yermo(2011), deals with the funding issue in public sector pension plans. This study raises several issuesincluding that, if circularity in government funding occurs, pensions funding systems has little value

added relative to a pays-as you-go system10:

First, to the extent that funding risks can be smoothed over time as they can be shared withfuture generations of tax payers, underfunding in market value terms may be an optimalstrategy (Cui, Jong, & Ponds, 2011; Munnell, Kopcke, Aubry, & Quinby, 2010). Secondly, afunding surplus might also mobilize pressure to increase benefits which in turn leads in the

10 Holzmann (2005, p. 46) mentions differently the circularity of the American Federal pension and social securityplan.

7/26/2019 SSRN-id2606547

18/37

18

longer term to higher funding costs and so underfunding. So for taxpayers it is rational to aim atunderfunding rather than full funding or overfunding. Moreover, a funding surplus will enforcecontribution cuts and once contributions are reduced, it is difficult to get them increased. Theaccountability horizon of pension fund management and politicians is much shorter than thehorizon over which pension promises have to be met by adequate funding. This horizon gapmay lead to pressure to underestimate costs and risks and to overestimate the earning capacity

of assets. Thirdly, to the extent that prefunding leads to investment in domestic governmentbonds, a circularity in government funding may be created, with little added value relative to aPAYG system.

Academic scholars such as Oulasvirta (2008) and Bohn (2011) addressed the pension funding issue.Concerning pension benefit liabilities and social policy cash transfers, Oulasvirta (2008) concluded thatthe way IPSAS standards handles liabilities is not as such optimal for government financial statementreporting. Bohn (2011) further developed an economic model where most taxpayers hold debt andface intermediation costs, returns on pension assets are less than taxpayers cost of borrowing. Pensionfunding is costly and hence zero funding is optimal.

These and other concerns have been raised throughout the public consultation process regarding theIPSAS 25. This consultation gathered comment letters to the Exposure Draft 31 on Employee Benefits,including those by the State of Geneva (2007), the Swedish National Financial Management Authority(2007), the American Academy of Actuaries (2007) and the Quebec Ministry of Finance (2007).11

The State of Geneva (2007) does not support, in general, the Exposure Draft, outlining that "theaccounting treatments proposed do not correctly reflect the economic reality of public pension plans inSwitzerland and that, under its actuarial approach through profit and loss, financial statementswould give no warning signs to users when a public pension plan experiences real financial difficulties.Accordingly, this treatment does not reflect the reality of the obligations of states regarding publicpension plans, as mixed funding by capitalization and repartition has proven sustainable in the longrun. To conclude:

[The State of Geneva] believes that if there is one subject on which IPSAS should depart fromIAS, it would be on pension plans, as the perenniality of states does change the economicreality of obligations on this matter compared to the private sector.

[The State of Geneva] believes that the IPSASB should depart from the accounting treatmentsset by the IASB on pension funds and develop an approach that takes into account thespecificities of public pension plans and their relationship with State Communities.

The Swedish National Financial Management Authority (2007) argues that the proposed Standard ED31, according to postemployment benefit pensions, are not applicable for Swedish Central Government[as for its pension funds] dont have any plan assets according to the obligation and there is also verymuch uncertainty in e.g. demographic assumptions, rates for employee turnover, future salary andbenefit level. Furthermore, according to the Swedish authority:

[T]he actuarial assumptions decided by The Swedish Financial Supervisory Authority aresufficient to measure the amount of the post-employment defined benefit pension obligation inSwedish Central Government.

11 Oulasvirta (2008) stated that the Government Accounting Board (Valtion kirjanpitolautakunta) gave a criticalassessment of the suitability of IPSAS standards for the Finnish central government in 2006. Position expressedby France, UK and Sweden are reported in our Appendix.

7/26/2019 SSRN-id2606547

19/37

19

On a general level [they] believe that the same reasoning is applicable to most countries. Thereare for example no economic implications for setting aside assets to meet specifically long-termemployee benefits. The implications for the financial targets for a public sector organization arenot the same as for a business organization. The targets are not linked solely to the singleorganization itself but to the society.

The American Academy of Actuaries (2007) argues that the proposed standards in ED 31 represent[s] asignificant departure from current United States accounting standards related to public-sector pensionand other post-employment plans. Accordingly:

Last year the Governmental Accounting Standards Board (GASB) in the US published their 2006white paper Why Governmental Accounting and Financial Reporting is and should be Different (2006) explaining their reasons why public pension plans should be treateddifferently than private sector pension plans.

According to this white paper (GASB, 2006, executive summary):

Governments are fundamentally different from for-profit businesses in several important ways.They have fundamentally different purposes, processes of generating revenues, stakeholders,budgetary obligations, and propensity for longevity. These differences require separateaccounting and financial reporting standards in order to provide information to meet the needsof stakeholders to assess government accountability and make political, social, and economicdecisions.

The Quebec Ministry of Finance (2007) argues that the new standards proposed for the public sectorwould be essentially those currently used in the private sector. It explains why such a situation wouldnot be appropriate, in the Ministry view by referring to the perennial character of governments(limiting the risk associated with termination of retirement plans) and the different purpose of financialstatements depending on the different economic nature of private enterprises and governments. Inparticular, the Quebec Ministry of Finance stresses that:

[I]n the case of governments, the market value is an indicator that is not significant.

For retirement plans offering equivalent benefits in the future, the actuarial liability of a publicentity (estimated on the basis of the rate of central government bonds) would be greater thanthat of a private enterprise (estimated on the basis of the rate of high quality corporate bonds,which is higher than the rate of central government bonds). It is absurd that two identicalretirement plans show different actuarial liabilities depending on whether one plan is in thepublic sector and the other in the private sector.

Notwithstanding this unsettled debate, the IPSAS Board decided to adopt an approach consistent withthe one best way by considering that the time value of money should be computed and that pensionobligations are liabilities. This choice seems to imply that other approaches are no longer admissible.Our Box 2 summarises the requirements for pension obligations as stated by the IPSAS 25.

The IPSAS Boards preference for full actuarial representation of pension obligations is clearly expressed

by paragraph 61 of the IPSAS 25 (IPSASB, 2008):

Accounting by an entity for defined benefit plans involves the following steps:(a) Using actuarial techniques to make a reliable estimate of the amount of benefit that employees haveearned in return for their service in the current and prior periods. This requires an entity to determine

how much benefit is attributable to the current and prior periods (see paragraphs 8084), and to makeestimates (actuarial assumptions) about demographic variables (such as employee turnover and

7/26/2019 SSRN-id2606547

20/37

20

mortality) and financial variables (such as future increases in salaries and medical costs) that willinfluence the cost of the benefit (see paragraphs 85 104);(b) Discounting that benefit using the Projected Unit Credit Method in order to determine the presentvalue of the defined benefit obligation and the current service cost (see paragraphs 7779);(c) Determining the fair value of any plan assets (see paragraphs 118120);(d) Determining the total amount of actuarial gains and losses and the amount of those actuarial gainsand losses to be recognized (see paragraphs 105111); ()

In its basis for conclusions, the IPSASB (2008) candidly acknowledges that these requirements areinconsistent with widespread public sector practices:

BC17. The IPSASB acknowledged that applying the requirements of this Standard in relation toliabilities relating to obligations arising from defined benefit plans may prove challenging formany public sector entities. Currently, many public sector entities may not be recognizingliabilities related to such obligations, and may therefore not have the systems in place toprovide the information required for reporting under the requirements of this Standard. Whereentities are recognizing liabilities relating to obligations arising from defined benefit plans, thismay be on a different basis to that required by this Standard. In some cases, adoption of this

Standard might give rise to tensions with budgetary projections and other prospectiveinformation.

This one best way approach to account for defined benefit plans (the only ones that assure futurepension obligations) is consistent with the approach endorsed by the IASB for the private sector at leastsince the issuance of the IAS 19 in 199812.

Street and Shaughnessy (1998) give a historical perspective of private pension accounting in countriesrepresented in the G4+1 working group13. Accordingly, historically, pension accounting standardsprovided for flexibility in the choice of actuarial methods and assumptions (Skinner, 1987). However, in1985, the FASB introduced an approach that has since been adopted by other G4+1 members.

Therefore, the US led the adoption of a balance sheet focus for pensions through the FAS 87 issued in1985, followed by the International Accounting Standards Committee (IASC) that proposed a singleactuarial method in its ED 54 published in 1996 (eventually adopted in 1998 as IAS 19).

In this context, Glaum (2009) explains that pension accounting has caused controversies ever sincestandard-setters started to regulate the recognition and valuation of pension-related liabilities, assets,and costs. The author further argues that the American Accounting Principles Board had to concedethat, with FAS 87, improvements in pension accounting were necessary beyond what was consideredpractical at those times [ (FASB, 1985) bold added].

Glaum (2009) provides then a comprehensive analysis of this harsh debate and resistance raised by this

standard in the US private sector, echoing the overall lack of consensus on the legal-economic nature ofcorporate pension obligations (Klumpes, 2001; Napier, 2007; Blake & Tyrrall, 2008). Consistent withIPSAS 25, the IAS 19 defines accounting for defined benefit plans as follows:

The measurement of a net defined benefit liability or assets requires the application of anactuarial valuation method, the attribution of benefits to periods of service, and the use ofactuarial assumptions. [IAS 19 (2011).66]

12 IAS 19 concerns pension accounting and more precisely the determination of the cost of retirement benefits inthe financial statements of employers having plans. This standard should be distinguished from the IAS 26 dealingwith accounting and reporting by retirement benefit plans themselves. The IASC published its first Exposure DraftE16 on Accounting for Retirement Benefits in Financial Statements of Employers in April 1980, further issued asstandard in January 1983. Actuarial methods were suggested for pension obligations since then.13

The G4+1 was a working coalition of accounting regulators from the Anglo-Saxon world active in the 90s.

7/26/2019 SSRN-id2606547

21/37

21

The fair value of any plan assets is deducted from the present value of the defined benefitobligation in determining the net deficit or surplus. [IAS 19 (2011).113]

The present value of the defined benefit obligation should be determined using the ProjectedUnit Credit Method. [IAS 19 (2011).67-68]

In sum, international accounting standards for both the private and the public sector are consistentwith, and endorse the emergent trend toward pension as an individual saving account and a financialplacement. To be sure, the IPSAS are not currently adopted by any major jurisdictions, although theIPSAS Board has been developing a professionally endorsed best practice that has been having asignificant influence on the ongoing public sector accounting debate and regulation (Brusca,Caperchione, Cohen, & Manes-Rossi, 2015). Contrary to the IAS/IFRS (which have been substantiallyadopted by the European Union for consolidated statements of corporate groups listed in EuropeanRegulated Exchanges), the IPSAS do not have compulsory authority to rule accounting practice on thismatter.

The question whether the standard IAS 19 should be applied to the public sector (directly or throughthe IPSAS No. 25) has already been discussed by students of national statistics (De Rougemont, 2003)14.Through the EPSAS project, this very discussion has shifted from national statistics to governmentalaccounting and reporting. The following paragraph will address accounting for pension obligations inthe context of ongoing harmonisation of European public sector accounting standards (EPSAS) led byEurostat.

3.2 Pension obligations under the EPSAS: Issues and Perspectives

The IPSAS were declared to be an indisputable reference (European Commission, 2013a, p. 8) for theongoing process of harmonisation of public sector accounting standards by the European Commissionwhich leads this project since 2013 (announced in 2011).

Eurostat is in charge of the EPSAS project in name of the European Commission. It is important to bearin mind that, in 1997, the European Commission endorsed the 1994s World Bank three pillar grid analysis while recognizing that, at that time, 88% of EU pension payments are covered by the first pillarwhich is a pay-as-you-go (PAYG) system (European Commission, 1997, p. 5).

In 2010, the European Commissions Green Paper (European Commission, 2010, p. 5) acknowledges theongoing shift from PAYG pensions towards more prefunded private schemes, which are often of aDefined Contribution (DC) nature, in most but not all Member States, in order to lower the share ofpublic PAYG pensions in total pension provision15. In fact, the Commission acknowledges that:

Member States are responsible for pension provision: this Green Paper does not questionMember States' prerogatives in pensions or the role of social partners and it does notsuggestthat there is one 'ideal' one-size-fits-all pension system design.

This same Green Paper (European Commission, 2010) aims strengthening the internal market forpensions, claiming that:

14 The treatment of pension schemes in macroeconomic statistics is analysed by Pitzer (2002).15 A figure in this report presents the share of occupational and statutory funded pensions in total grosstheoretical replacement rates in 2006 and 2046 and shows that funded pensions will provide for a larger shareof retirement income in 2046 that in 2006 (European Commission, 2010, p. 36).

7/26/2019 SSRN-id2606547

22/37

22

Completing the internal market for pension products has a direct impact on the EU's growthpotential and therefore directly contributes towards meeting the Europe 2020 objectives.

The adoption of the Directive on Institutions for Occupational Retirement Provision (IORP) in2003 was a major achievement.

The 2003 Directive on the activities and supervision of institutions for occupational retirementprovision (European Parliament & Council, 2003) deals, in article 15, with technical provision and, inarticle 16, with the funding of technical provision. The Directive requires that:

(26) A prudent calculation of technical provisions is an essential condition to ensure that

obligations to pay retirement benefits can be met. Technical provisions should be calculated on

the basis of recognised actuarial methods and certified by qualified persons.

(28) Sufficient and appropriate assets to cover the technical provisions protect the interests of

members and beneficiaries of the pension scheme if the sponsoring undertaking becomes

insolvent.

Full funding is required only for cross-border activity (Art. 28), while Member States can permitunderfunding for national institutions, while establishing a proper plan to restore full funding andwithout prejudice for protection of employees in the event of the insolvency of their employer (CouncilDirective 80/987/EEC of 20 October 1980).

Concerning public sector accounting, the European Commission (2013b, p. 6) believes that an accrualsbasis of accounting helps assessing fiscal sustainability:

[T]he important advantage of accruals over cash accounting is that both assets and liabilitiesare consistently recorded, making it possible to have a complete and consistent picture of thereal financial position and of whether it is sustainable.

In this context, pension obligations are especially relevant. According to the report prepared byPricewaterhouseCoopers (2014, p. 108) on behalf of Eurostat in the EPSAS project:

Pension liabilities in respect of defined benefit schemes rank among the most significant areasin terms of impact on the opening balance sheet, when making the transition from a cash-based accounting system to IPSAS. () IPSAS 25 on employee benefits would requirerecognition of large pension liabilities on the balance sheet. This could lead to negativereactions from stakeholders if negative net assets are disclosed.

In the previous preparatory report (European Commission, 2013b, pp. 125-126), the IPSAS 25 wasincluded among standards that need adaptation, or for which a selective approach would be needed.Further technical discussion was then encouraged with a panel of accounting experts. This reportfurther stated that the difficult areas are pensions, and to a lesser extent, other long-term benefitssuch as long service leave, which represent a large problematic part of the standard.

It is then clear that accounting for pension obligations constitute one of the main issues underlying thisaccounting process of harmonisation, although it was excluded by the main topics of the first survey ledby Ernst&Young on behalf of Eurostat to provide supporting documentation for the EPSAS project.

The EPSAS project aims at bridging the gap that exists among the different public financial reportingsystems and at minimizing incoherences between the micro level public sector accounting and

reporting framework and the European System of Accounts (ESA) macro level financial system.

7/26/2019 SSRN-id2606547

23/37

23

In this context, it is relevant to pay attention to the previous work done on this matter by Eurostat,especially: (i) Eurostats point of view on the current international trend of integrating pensions as aliability in the balance sheets of national accounts, and (ii) Eurostats prevailing requirement onpensions.

Eurostats opinion on the treatment of pension schemes in macroeconomic statistics (point (i) above)

was included in the Eurostat communication to the Advisory Expert Group (AEG) on national accountsissued in 2004. The AEG was the result of an Electronic Discussion Group (EDG) on the treatment ofpension schemes, including social security and unfunded employer pension schemes established by theIMF Statistics Department. The discussions in this EDG resulted in a report that was presented in theFebruary 2004 meeting of the Advisory Expert Group (AEG) on National Accounts. This ElectronicDiscussion Group (EDG) was lead because differences existed in treatment between differentmacroeconomic systems16.

In its communication to the AEG, Eurostat (2004) summarises the EDG main proposal and was veryreserved on it (bold in the original):

The EDG main proposals for changes of the SNA are (1) to treat unfunded employer pensionschemes similarly to funded schemes, (2) to use actuarial valuation for flows and stocks and (3)to allocate the net assets of pension funds to the sponsor.At the same time, it can be questioned whether unfunded or pay-as-you-go schemes areeconomically the same as funded schemes. This concerns the less solid nature of the claimasits value can be unilaterally altered by the debtor. Moreover, it should be taken into account towhat extent this value depends on all kinds of assumptions on uncertain future events andwhether or not it can be estimated within narrow margins. In order to reconcile the importanceof providing information on pensions liabilities in the SNAand perhaps in the accountsthemselves with the hesitation to give to such liabilities the same status to others, someinnovative alternative accounting option.() It is strongly suggested the EDG needs to examine more in detail the other identifiedoptions, with the aim to indicate the pros and cons of each of those options also in view of

the concerns and borderline issues mentioned above.

Concerning current European requirements for the Member States (point (ii) above), they includepublication ofa special compulsory table in which government pension liabilities. In 2008, the finalreport of the Eurostat/ECB Task force on the statistical treatment of pension schemes summarised thecompromise that was reached in the new SNA (Eurostat, 2008, p. 17) by arguing for six basicprinciples. For sake of simplicity, only two principles are reported here:

(iv) Concerning government sponsored systems: Pension entitlements of unfunded, pay-as-you-go government sponsored systems which provide the basic social safety net type of provision,sometimes referred to as pillar one type provision, will be only recorded in the supplementarytable (but not in the core account);(v) The recommendation of the new SNA regarding the recording of unfunded pension schemessponsored by government for all employees (whether private sector employees orgovernments own employees) will be flexible.

These principles were largely supported by senior statistical staff in summer 2006 (Eurostat, 2008, p.17). In 2013, this supplementary table was included in the Chapter 17 dealing on social insurance

16 Indeed, the System of National Accounts 1993 (SNA 1993) the United Nations system and the

European System of Accounts (ESA95) as well as they were based on the SNA 1993 did not recognizepension obligations as liabilities of the schemes while the IMF's Government Finance Statistics Manual2001 (GFSM 2001) recommended that that stocks of government liabilities for all employer schemes.

7/26/2019 SSRN-id2606547

24/37

24

including pensions in the ESA 2010 (Eurostat, 2013). This table provides detailed and reliable estimatesof stock and flow data on both pensions which are included in core accounts and pensions which arenot included in them. The [supplementary] table also covers stock and flow data not fully recorded inthe core national accounts for specific pension schemes such as government-unfunded defined benefitschemes with government as the pension manager, and social security pension schemes (Eurostat,2013, p. 379). In this way, detailed and comparable information is provided at the level of each Member

States, while impact on governmental accounts, affecting measurement of public sector deficit anddebt, is excluded (Dabbicco, 2015).

In this context, it should be remembered that pension provision modes differ across Europes MemberStates. Since the issuance of the IPSAS No. 25, very few European countries have adopted its actuarialaccounting approach. As stated in PricewaterhouseCoopers (2014) report, which examines existingaccountin