WIXLIB

Models of the Actuarial Balance of the Pay-As-You-GoPension System. A Review and some PolicyRecommendations*María del Carmen Boado-Penas †, Carlos Vidal-Meliá and Junichi Sakamoto‡.P24 May 2010ABSTRACTThis paper reviews the two main methods used by government Social Security departmentsto draw up the so-called actuarial balance of the pay-as-you-go pension system, focusingespecially on results, methodology and actuarial issues. The specific models studied arethose in the United States, Japan and Sweden, and their main differences and similaritiesare highlighted. The authors make some specific policy recommendations that would be ofinterest to public finance economists, social security actuaries and policy makers, and theysuggest that international organizations (ISSA, the World Bank, the OECD and othersinterested in public pension reform) could be supportive in developing and enforcinginternational accounting and actuarial valuation standards for pay-as-you-go pensionsystems(JEL: H55, H83, J26, M49).Key words: Japan, Notional Accounts, Projections, Public Pensions, Retirement,Sustainability, Sweden, USA.*Carlos Vidal-Meliá and María del Carmen Boado-Penas are grateful for the financial assistance receivedfrom the Spanish Ministry of Science and Innovation (Ministerio de Ciencia e Innovación) project ECO200913616. We would also like to thank Ole Settergren for his invaluable help and comments, Richard Jackson forhis information and explanations about the US model, Gauthier Lanot for his comments and Peter Hall forhis English support. Preliminary versions of this paper were presented at the XI Spanish-Italian Congress ofFinancial and Actuarial Mathematics in Badajoz (Spain), the 4th International Colloquium of the IAAPensions, Benefits and Social Security Section (PBSS) in Tokyo (Japan), at two internal seminars organized byKeele University and the Basque Country University respectively and as a table document at the Nonfinancial Defined Contribution (NDC) Pension Conference in Stockholm (Sweden). The opinions expressedin this paper are the authors' and do not necessarily represent the views of the University of Valencia, KeeleUniversity or the Nomura Research Institute. Any errors are entirely due to the authors.† (Corresponding author) Institute for Public Policy and Management, Keele University, Keele, StaffordshireST5 5GB, UK. (e-mail: m.d.boado-penas@econ.keele.ac.uk ). Department of Financial Economics and Actuarial Science, University of Valencia, Avenida de losNaranjos, s.n. 46022 Valencia. (Spain). (e-mail: carlos.vidal@uv.es ).‡ Nomura Research Institute, 1-6-5 Marunouchi, Chiyoda-ku, Tokyo 100-0005 (Japan) (e-mail: jsakamoto@nri.co.jp ).P1

1. Introduction.According to Settergren (2008), one useful way of reaching the objective of betterdesigned and managed public pension systems is to provide better information on thefinancial development and status of those systems. In many countries, reporting on theseaspects in public pay-as-you-go (PAYG) pension systems is not fully established. Thisreduces the accumulation of knowledge as to how the pension system interacts with thedemography and the economy of these countries, reduces informed public debate onpublic pensions, increases the longevity of old misunderstandings and is fertile ground fornew ones. For Bovenberg (2003), transparency about objectives and agreement on risksharing is a key requirement for all pension systems. Explicit agreements as to how to sharerisks become even more important as ageing makes systems more vulnerable to shockssuch as unexpected longer longevity or inflation. Being explicit about risk-sharing bysetting clear rules and having more transparent accounting increases the costs associatedwith breaking the intergenerational contract and thereby reduces the political riskssurrounding public PAYG pension systems.The so-called actuarial balance of PAYG pension systems, Boado-Penas (2008), provides asuitable answer to these issues. It also supplies a positive incentive to improve financialmanagement by eliminating or at least minimizing the traditional mismatch between theplanning horizons of electors and politicians, often only four years, and the system itself.The minimum time horizon should be the system's turnover duration, the normal value ofwhich varies between 30 and 35 years, although actuarial reports often contemplate aminimum horizon of 75 years.This paper looks mainly at what will be described as the actuarial balance of the PAYGsystem. The specific models studied are those in the United States, Japan and Sweden.According to the BOT (2009), a very detailed annual actuarial balance - an actuarial report,in fact - has been compiled in the United States since 1941. From 2002 it has includedstochastic methodology. The so-called “US” actuarial balance drawn up by US SocialSecurity - similar to that published by the authorities in Japan (Sakamoto (2005)) and UK(GAD (2010)) every five years, in Canada every three years since 1966 ((OSFIC) (2007 and2008)) - is not a balance sheet in the traditional accounting sense of the term, with a list ofassets and liabilities.Compiling an official actuarial balance sheet has been normal practice in Sweden since2001. This actuarial balance sheet - in the form it takes in Sweden - has until now attractedlittle attention from academics. This is surprising given that the literature suggests manymethodologies to analyze the viability or sustainability of pension systems or to forecastaggregate spending, and this field is of special interest to a number of researchers1. As faras we are aware, only in the cases of Japan and Spain has the actuarial balance sheet with itstypical structure of assets and liabilities been used by researchers. For the case of Japan,Takayama (2005) uses the actuarial balance sheet as an element to analyze proposals forreforming the pension system, although the list he presents of the items it comprises is notvery developed, whereas Boado et al. (2008) compile an actuarial balance sheet to assess thesolvency of the Spanish system and compare it to the Swedish system. They also develop1The most commonly used methodologies for making projections of spending on pensions or analysing thesustainability of pension systems are: aggregate or growth accounting models, micro-simulation models,general equilibrium models and indirect models. On this aspect see the papers by Lefebvre (2007) and TEPC(2007).2

the concepts of the contribution asset and the average turnover duration for definedbenefit PAYG systems. Vidal-Meliá et al (2009) apply the balance sheet of the Spanishsystem to support the introduction of an automatic balance mechanism (ABM). On anofficial level, the actuarial balance sheet has not been used outside Sweden . At themoment, no other country uses the actuarial balance sheet in the form it takes in Sweden,but some countries do make financial or actuarial projections, and the analysis is verysimilar to that done in the USA, Canada, the UK and Japan .PFPThe aim of this paper is to show the advisability of drawing up an actuarial balance inPAYG pension systems so as to improve their transparency, credibility and solvency. Thisis in line with the trend seen in some developed countries which are trying to introduceactuarial analysis methodology into the field of public PAYG pension system management.The paper also aims to shed some light on the two main methods used by governmentSocial Security departments to draw up the actuarial balance, focusing especially on theirresults, methodology and actuarial issues, and makes some specific policyrecommendations that would be of interest to public finance economists, social securityactuaries and policy makers. As far as we are aware, there is a large gap in the literaturewhich we are attempting fill with this paper because, until now, this area of study has notbeen looked at from the perspective of the models used by public S.S. administrations.2Following this brief introduction, Section 2 describes, defines and presents the basicaspects of the US actuarial balance and the Japanese model. In Section 3 we define andexplain the Swedish model, especially the balance sheet and the actuarial income statement,basic assumptions, actuarial aspects and the most recent results. The paper ends with themain conclusions, focusing on the basic differences and similarities between the Swedish,US and Japanese models and specific policy recommendations.2. The Actuarial Balance: the US and Japanese Models.The methodology used to compile the actuarial balance would best be described as aprojection model of aggregate accounting spending on pensions. Basically it involves usingthe forecast demographic scenario to determine the future evolution of the number ofcontributors and pensioners according to the rules of the pension system. Themacroeconomic scenario that determines the amounts of future contributions and pensionsis exogenous.The US ModelThe Old-Age, Survivors, and Disability Insurance (OASDI) program in the United Statesmakes available a basic level of monthly income when insured workers become eligible forretirement and in cases of death or disability. The OASDI program consists of twoseparate parts that pay benefits to workers and their families - Old-Age and SurvivorsInsurance (OASI) and Disability Insurance (DI). Under OASI, monthly benefits are paid toretired workers and their families and to survivors of deceased workers. Under DI,monthly benefits are paid to disabled workers and their families.The actuarial balance of the OASDI program is aimed at measuring the system's financial2The paper by Plamondon et al. (2002) is a first attempt to conceptualize the actuarial balance of the PAYGsystem, though it obviously does not include the “Swedish” model. The paper by Boado-Penas et al (2010)does include a conceptualization of this model, and highlights the advantages that both models (Swedish andUS) could have for the case of Spain.3