Towards an index of private pension provision

Abstract

Privatization entails a number of quite different aspects that overlap only to some extent: (1) responsibility in provision; (2) possibilities of exit from statutory plans; (3) the transfer of the administration and management of schemes from public bodies acting in a unitary non-competitive administrative environment to private sector actors acting in a more fragmented competitive market based environment; (4) the transfer of the risks of retirement from the collective to the individual; and (5) the ‘recommodification’ and ‘financialization’ of retirement risks. The paper develops a compound index of private pension provision based on a number of quantifiable variables (theoretical replacement rates of public plans, assets of funded pension schemes, private pension expenditure and coverage by private pension plans) that tries to tap into these different aspects. It can be used to empirically categorize the extent to which Organisation for Economic Co-operation and Development (OECD) countries have privatized their pension provision.

Keywords

comparative political economy financialization mandating pension privatization recommodification welfare state regimes

Privatization is often at the centre of debates on welfare state reforms in general, and of studies of changes in systems of retirement provision in particular. It therefore comes as a surprise that only few attempts have been undertaken to investigate exactly what is involved in this form of institutional change. Most research on the boundaries between ‘public’ and ‘private’ social policy has centred on normative perspectives (see Gilbert, 2004; Pearson and Martin, 2005) and only a few attempts have been made to assess to what extent one can make hard claims that the privatization of social risks has led to the transformation of welfare states (Hyde and Dixon, 2010; Leisering, 2012). The purpose of this article is to develop an index that can measure how countries differ in the extent to which ‘empirically the pendulum has swung from “public” to “private”’ (Seeleib-Kaiser, 2008: 1).

We will start by defining the public–private divide in general, and situate it in the discussion on capitalist diversity. Next we will discuss the different aspects of this form of institutional change in the context of pension provision. Finally, we will propose an index based on quantifiable indicators that can measure the extent to which pension schemes have been privatized in the Organisation for Economic Co-operation and Development (OECD) countries. Our approach differs from earlier attempts that have sought to address the issue by relying on expert interviews (see Hyde et al., 2006) or only on expenditure data (see Adema et al., 2011), in that we will use a combination of OECD statistics and social rights indicators (Scruggs, 2007).

One of the problems with empirically assessing privatization seems to be that ‘private’ can mean a number of things, or at least that it entails a range of quite different aspects, which only to some extent overlap with one another. This article explores these issues by examining the characteristics of national systems of retirement provision. The main purpose is to devise an index of private pension provision that seeks to tap into these different aspects and that, in further research, can be used to categorize OECD countries. The purpose of this index is not to measure the distributive consequences of welfare regimes as such, but rather to indicate to what extent statutory rights and obligations have been replaced by a more voluntaristic coordination of welfare preferences via the price mechanism.

Distinguishing public and private welfare

In the debate on welfare state reform, privatization has in general been defined as ‘a movement away from a purely public state of social welfare under which benefits are funded by tax revenues and directly provided by government agencies’ (Gilbert, 2004: 100). This might be an appropriate definition in an Anglo-Saxon context, but in most continental welfare regimes many non-state organizations, such as trade unions, employers’ associations, mutual benefit societies or the church, have traditionally been involved in the delivery and administration of statutory benefits and tax-financed social services (Alber, 2003; Crouch, 1999; Trampusch, 2006). This kind of ‘private’ provision in subsidiary welfare states was never subordinated to profit motives and market exchange. Instead, these non-state organizations have always been governed by a complex system of corporatist arrangements that produce a style of ‘public’ decision unknown to most English-speaking countries (see Whiteside, 2006).

What may appear similar, but on closer examination turns out be quite different, is the involvement of private charitable organizations in the financing and delivery of social services. Whereas the involvement of non-state actors in subsidiary welfare states is embedded in statutory regulations about eligibility and entitlement, in the case of charities, government has very little control over these matters. This does not mean that the state does not play a role in supporting such voluntary initiatives by granting generous tax deductions to private donors, but the way these funds are allocated is largely dependent upon private decisions of charitable organizations.

Addressing the question by looking at who is covered under what conditions, Van Gunsteren and Rein have proposed to reserve ‘public’ for welfare arrangements that are in principle for all citizens, whereas ‘private’ schemes are limited to the members of particular collectivities (such as the employees in a particular industry) or are individual arrangements. Whereas ‘private–collective’ arrangements often still entail an element of coercion, ‘private–individual’ arrangements solely result from individual decisions to take part in a scheme (Van Gunsteren and Rein, 1985: 130). To the extent that such private arrangements arise from standard employment contracts embedded in labour law, they can be almost as binding and unavoidable as statutory social security legislation. The problem is that according to this approach, most status-based schemes in ‘Bismarckian’ welfare states, which cover specific segments of the population under differentiated conditions, would have to be qualified as ‘private–collective’ arrangements. An exclusive focus on the scope and content of coverage, the involvement of semi-autonomous non-state organizations and the extent to which arrangements are binding does not really seem to allow us to pinpoint the distinction between ‘public’ and ‘private’ or to tap into the format of changes in the responsibilities for social risks.

This requires a closer examination of the terms and conditions under which non-state providers operate, and the kind of relationships they have with the population at risk. Privatization can be defined as a reassignment of the property rights of the organizations that are involved in the provision of welfare (Starr, 1988). Non-state organizations become part of the private sector when they are allowed to produce a residual reward to their owners, often in the form of profits, and when their survival depends upon generating this surplus. Rather than meeting approval of the political process, private welfare providers have to be able to compete with one another in the marketplace, and stand the test of profitability. Privatization thus can be related to a strengthening of the role of markets and of the profit motive in the generation of welfare. In a way it can be understood as the flip side of ‘decommodification’.

In what follows, ‘privateness’ will be conceptualized in terms of the extent to which welfare arrangements have embraced market principles including the coordination via the price mechanism, competition, profit making, voluntarism and formal freedom of choice, and the individualization of risks. In an ideal-typical market based system the relationship between the organizations that are responsible for providing welfare benefits and individuals in search of social protection is no longer mediated by statutory rights and obligations, but is coordinated by the price mechanism. Welfare provisions are treated as commodities that can be bought and sold on the market. Whereas in a ‘public’ welfare system non-state welfare providers operate in a hierarchically structured context under the auspices of the state, in a ‘privatized’ welfare system they operate in a competitive market and are allowed to generate profits.

In the real world, even the most liberal systems will end up regulating the marketplace to prevent market failures, such as a tendency of providers to establish monopolies and exploit information asymmetries, or the problem of myopia among the population. These regulations, however, are intended to make the market work, not to reverse its theoretically expected outcomes.

The spread of market principles weakens the rights and obligations that are attached to a standard employment relationship. Firms are being freed from their responsibility to finance welfare provisions and they no longer have to back-service the costs of unexpected contingencies.

Markets can also enter provision of welfare when the production and delivery of welfare is transferred to the private sector. This is, for example, the case when hospitals are privately owned and operate on a for-profit basis; when government agencies pay voluntary and for-profit organizations to provide goods (such as low-cost housing or meals for the elderly) and services; when the government issues vouchers so that recipients can purchase the goods and services they need in the marketplace; when owner occupation of housing is promoted by the development of exotic mortgage products; or when pension plans outsource their administration or the investment of the pension fund’s assets to the financial services industry.

The different aspects of private pension provision

Privatization of pensions entails a number of quite different aspects, that only to some extent overlap with one another: (1) a shift in the responsibility for providing, from public authorities to voluntary occupational or individual forms of provision; (2) the transfer of administration and management of schemes, from public bodies acting in a unitary non-competitive administrative environment to private sector actors acting in a more fragmented, competitive and market-based environment; (3) the creation of possibilities of exit from statutory arrangements; (4) the transfer of the contingencies associated with retirement from the collective to the individual; and (5) the ‘recommodification’ and ‘financialization’ of retirement risks. The following sections will elaborate on these different aspects by examining what these changes entail in terms of old age pension provision.

Responsibility in providing: who initiates and organizes the scheme?

In a public pension system this is done by the state, who sometimes may decide to delegate the actual implementation to some neo-corporatist non-profit actors, but the state still retains control over the overall orchestration of the scheme, including the decision of who is required to join under what conditions, who has to pay for the scheme, and who is entitled to what form of benefits. In private schemes the decision-making power on these issues lies primarily with private sector actors: usually the sponsoring employer and/or a financial firm. This shift in responsibility to provide does not necessarily need to imply a complete retreat of the state from the world of pensions. It rather involves a kind of change of the role played by public authorities: instead of directly providing pensions, the state engages more in the role of regulator and subsidizer of various forms of private provision.

The shift in responsibility for providing is often brought about by a retrenchment of state provision, or by a failure to upgrade existing public pensions in line with rising expectations or perceptions of need. This is said to create a ‘social protection gap’ (Bonoli et al., 2000: 46), which in turn is expected to lead to demands for compensation through expanded private arrangements. In essence, the underlying idea behind this ‘passive privatization’ (Bridgen and Meyer, 2007: 223) is a reversal of the earlier thesis of ‘crowding out’ private arrangements by generous public schemes (Künemund and Rein, 1999). In common with this last perspective, it supports the assumption that public and private pensions are in essence substitutes in quantitative terms, and that therefore they form functional alternative means to achieve the same ‘adequate’ income after retirement (for a critique of the validity of the ‘crowding out’ thesis see Pedersen, 2004). State subsidies for this kind of private provision can take on different forms. The most prevalent form of such subsidies consists of tax exemptions. Yet some countries, most notably Germany, have gone one step further by also paying out grants specifically to encourage individuals to take up private pensions contracts, to compensate for planned reductions in statutory pensions.

A shift in responsibility can also be brought about by various practices of mandating. These range from direct unilateral mandating by the state (for example the Obligatorium in Switzerland) (Bonoli, 2005), a concerted effort by trade unions and the state (for example the Australian Superannuation) (Bateman and Piggott, 1998) and the administrative extension of collective agreements negotiated between employers’ associations and trade unions (for example the industry-wide pension plans in the Netherlands) to the numerical strength of trade unions (for example occupational collective pension insurance in Denmark). In all these cases, non-institutional arrangements that are in principle alien to a liberal market economy (Schröder, 2008) boost a form of welfare provision that is normally considered to be part of the liberal world of welfare capitalism (see Jackson and Vitols, 2001).

Mandating social partners to establish their own institutions with their own entitlement rules and autonomous financing mechanisms ought to be distinguished, though, from obliging employers and employees to jointly finance and jointly administer statutory schemes in which all the rules remain formulated and enforced by the state (such as is the case in most parity-based ‘Bismarckian’ social insurance bodies) (Rein and Turner, 2001).

Transfer of the administration of the pension plan

A third aspect of privatization of pensions concerns the transfer of the administration and management of schemes from public bodies, acting in a unitary non-competitive administrative environment towards private sector actors acting in a more fragmented competitive market-based environment. Even though mandated schemes in countries such as the Netherlands, Switzerland and Denmark might be governed by (one of) the social partners, they often end up contracting out the management of the scheme (Denmark) or of the accumulated contributions (the Netherlands) or of the assets collected by a statutory agency (such as the Swedish Premium Pension Authority) to commercial financial services companies.

Exit from statutory plans

A fourth aspect of privatization consists of rendering the participation in a pension plan a matter of voluntary choice rather than of statutory obligation. Such defection from statutory programmes can be encouraged by introducing ‘opting out’ clauses, that allow the affluent winners of the increased inequality, experienced in most industrial countries in the course of the past decades, to organize their own provision and pay less for the deprived sections of the population. As such it relieves them from their obligation to participate in the system of collective risk redistribution (De Deken et al., 2006). The most radical form of ‘exit’ consists of the full replacement of collective social security retirement provisions by a system of individual private pensions savings accounts that individualize the risks associated with retirement completely.

Offering more individual choice and increasing personal freedom is often presented as one of the main attractions by the proponents of privatization. Yet such a choice is often an attractive option only for a minority, and making it available might end up undermining the sustainability of the preferred alternative of the majority of the population (see Erskine, 1997).

A variation on the ‘exit’ strategy consists of the deliberate under-provision by statutory programmes: keeping replacement rates for middle and high income groups low implicitly encourages them to complement their public pensions by private supplementary forms of provision on an individual or on a collective basis.

The transfer of the risks of retirement from the collective to the individual

Privatization can be seen as part of a broader strategy that seeks to diminish collective responsibility for pension provision (see Hyde et al., 2006). This includes such aspects as the extent to which individuals are made responsible for individually coping with misfortunes that might arise from the contingencies associated with retirement. These contingencies can be modelled on the basis of two dimensions: individual versus the generational and economic versus the demographic dimension (De Deken et al., 2006). Individual demographic risks include, for example, the risk of living too long after one has lost the capacity to perform paid work (a risk that has a distinct gender dimension to it). Generational demographic risks include ageing of the population and the possible risk that future generations might be hard pressed to sustain the retirement income of future pensioners. Individual economic risks include risks such as unemployment, work incapacity or performing non-standard work that is not adequately covered by social security legislation, resulting in an incomplete contributory history in statutory schemes, and thus eroding the total entitlement accumulated over the life course. Collective economic risks include the risk of low national growth rates or stagnant productivity increases that may again undermine the sustainability of pension promises in the future. What is at stake here is the likelihood that other more fortunate individuals might be called upon to collectively shoulder the burden of these risks, or even that third parties such as shareholders of firms (part of the rationale behind employers’ contributions) might be asked to partly foot the bill.

The ‘recommodification’ and ‘financialization’ of retirement risks

The reassignment of the risks of retirement we described in the previous section can be (and often has been) implemented through a series of parametric reforms of public retirement systems. Individual responsibility for labour market risks has always been part of actuarially orthodox insurance systems that made earnings-related benefits conditional upon a lengthy employment or contribution history. The link with the labour market career of beneficiaries weakened only to the extent that those schemes granted extraneous insurance benefits. In this regard, there has always been a contradiction in Esping-Andersen’s operationalization of the concept of ‘decommodification’ (Esping-Andersen, 1990): in order to score high on indices that are supposed to capture the ‘decommodifying’ nature of pension schemes, national systems have to provide both generous unconditional benefits and offer high earnings-related replacement rates – the latter only being possible in cases where the level of benefits is linked to past labour market records. This increases the role played by (labour) markets in the building up of retirement income. If during the post-war period the actuarial orthodoxy of many social insurance schemes has been dramatically watered down, the recent move towards a so-called ‘notional defined contribution’ system in countries such as Italy and Sweden has again expanded the role of labour market records (Cichon, 1999). Even without such radical reforms, however, most public schemes have limited the recognition of phases equivalent to contributory periods, tampered with indexation mechanisms, and reduced the relative share of employers’ contributions.

Such measures can be politically painful, and one option through which politicians can successfully engage in ‘blame avoidance’ is to increase the mediating role of markets in managing these risks. In essence this involves ‘recommodification’, not only in terms of an increased importance of a record of labour market participation as a precondition for pension entitlements (Esping-Andersen, 1990), but also of an increased reliance on private financial institutions to secure the resources necessary to pay for the liabilities that arise out of entitlements – in particular by counting upon the allegedly superior returns of stock market investments to foot the bill of the contingencies of pension provision (see Minns, 2001).

‘Financialization’ introduces a more pertinent market contingency into the retirement system than labour market ‘recommodification’ and the parametric reforms discussed above. It expands the role of financial motives, financial markets and financial institutions in the operation of the pension system. It is a change in the funding of retirement that also has also far reaching repercussions in the way that pensions are governed (Clark and Whiteside, 2003). If the accumulated assets of pension schemes are invested through the intermediation of the financial services industry, the monitoring of the performance of those assets on financial markets becomes a ‘life strategy’ for people from all walks of life (Martin, 2002). They are then led to adopt the identity of a ‘self-disciplined investor subject’ (Langley, 2006). ‘Financialization’ of retirement provision thus leads employees and their trade union representatives to stake their long-term welfare on the ability of the finance industry to reap high returns on investment (see Engelen et al., 2008).

Other aspects of pension systems conflated with the public–private divide

In debates about privatization, pensions systems tend to be presented as packages that consist of bundles of arrangements that inevitably have to go together or are at least institutionally complementary to one another (see, for example, World Bank, 1994). The distinction between ‘public’ and ‘private’, however, is relatively independent from these other institutional or functional dimensions of pension regimes. These other dimensions include whether pension benefits are flat-rate or earnings-related; whether the schemes are organized on a defined benefits (DB) or a defined contributions (DC) basis (this distinction is crucial in the individualization of financial and labour market risks); whether plans are externally funded, run as an insurance, based on book reserves or are run on a pay-as-you-go (PAYG) basis (with or without a sizeable reserve buffer); and whether the scheme is provided as part of an employment relation, or whether it originates in a voluntary individual decision from the plan’s participant and/or the plan’s sponsor. Some of these dimensions, though, are intimately intertwined with the ‘public’ versus ‘private’ divide in which we are interested. For example, most private pension plans tend to involve some kind of pre-funding (via external pension funds, through setting apart internal book reserves or by delegating the funding requirements to insurance companies). The only exceptions to this rule seem to be the French occupational schemes for wage earners, ARRCO (Association de Regimes de Retraite Complémentaires) and AGIRC (Association Générale des Institutions de Retraite des Cadres) (Blanchet and Pele, 1999). Yet these French second pillar plans are only capable of guaranteeing their PAYG benefits because they are embedded in a long-term government mandate (and hence their ‘privateness’ can be contested, which is one of the reasons why they are often presented as an integral part of the French public system). On the other hand, it must be recognized that ARRCO and AGIRC use private asset managers to invest the limited reserves these schemes use as a buffer for their PAYG commitments.

This necessary condition does not seem to operate in the other direction though. Funded schemes can be administered by general government, as the case of central provident funds (CPF) in some Asian countries demonstrate (the most renowned example being Singapore’s CPF), and can be part of some transitional arrangements such as the reserve funds in the second pillar ATP (Allmän tilläggspension) system in Sweden (until the most recent pension reform in that country), or the old age funds in Belgium (the Zilverfonds) and the Netherlands (the AOW-fonds) that solely invest in state bonds to anticipate costs for the statutory pension scheme when the peak of ageing will be reached around 2030.

However, in Europe, North America and Australia, publicly-controlled funded schemes that invest their savings on the capital market appear to be the rare exception so far, as governments in these parts of the world seem to be wary of giving the state such a strong leverage upon investment capital. In these countries funded schemes tend to take the form of autonomous pension funds or are administered by private sector insurance companies.

Another feature that is hard to separate from the public–private divide is the granting of an unconditional indexation of benefits to prices or wages. This can only be assured in the shadow of the compulsive power of public authorities that can mobilize the necessary resources to back-service the financing of these benefits over different generations. In certain cases full indexation (and for that matter the unconditional DB schemes) introduces a PAYG element into even the most privately funded schemes. Without the looming shadow of the state and its capacity to impose a tax or a mandate, private plans can only guarantee DC plans.¹

Towards an index of private pension provision

Given the multiple aspects of privatization, operationalizing the distinction between ‘public’ and ‘private’ forms of provision may be less clear-cut than it might appear at first sight. In addition, it is sometimes impossible to obtain comprehensive cross-national data on some of the institutional characteristics that are conceptually of central importance.

We decided to devise a compound index that consists of four variables, for which comparable cross-national data are available. Those variables tap into different aspects of the private nature of pension provision. Our index does not capture all the aspects in which we are interested, however important they may be, because so far these aspects have not been documented in a systematic way for many countries. For example, the privatization of market induced risks could be best measured by the extent to which pension plans are organized as genuinely DB systems, and most clearly deviate from the orthodox DC model that individualizes both labour market and financial market risks. The problem is that data on the distinction between DB and DC plans are only available for a few countries. In addition, the application of this distinction is far from straightforward. It is not really a dichotomous distinction, but rather ought to be seen as a continuum between two polar extremes, as there are many different ways of giving additional guarantees on benefits going beyond the orthodoxy of a DC model (of central importance here is the conditionality of indexing pension benefits and pensions accruals) rather than a binary classification.

Even the variables we will be using for the construction of our index, which have been reasonably well documented by organizations such as the OECD, are, however, still marred with manifold measurement problems and ambiguities in terms of classification. This is particularly true for countries, such as Finland, that developed a hybrid pension system that is hard to classify; but it is our expectation that by combining four different variables into a single index, we might succeed in ironing out some of these ambiguities. In part this can be expected because most of the data involved are provided by national governments, some of which seek to present their national systems in the most favourable light in view of the different kind of benchmarks supranational organizations such as the OECD and the EU use, depending upon the policy goals that are compared. Thus, if one looks at the data reported on public spending many governments seek to play down the cost problem of their pension provision by labelling as ‘private’ as much spending as possible. Through such practices of ‘financial profiling’ they seek to comply with criteria agreed upon in the context of intergovernmental agreements such as the stability pacts of the European Union; or to reassure financial market actors such as rating agencies. On the other hand, when governments report on the social quality of their systems (which includes generous replacement rates) in order to demonstrate best practices in the context of procedures such as the EU’s Open Method of Coordination), some governments are tempted to bend the rules in such a way that, as much as possible, benefits are attributed to the ‘public’ category. In what follows we will give some examples of such practices.

A second reason to use a compound index is related to the intertemporal nature of pensions systems. Some indicators tell us more about the structure of retirement position in the past, rather than what the present division between public and private is, or to what extent the employed population is currently accumulating pension rights. Two of our variables reflect primarily pension arrangements in the past: current expenditure on private benefits and the funding rate of private pensions. Two indicators measure the importance of current and future private pensions: the coverage of private pension plans and the theoretical replacement rates of public pensions.

Replacement rates of public pensions

As the first variable we will use the standard theoretical replacement rates offered by ‘public’ pension schemes. As we argued, the idea behind this indicator is that generous statutory schemes will ‘crowd out’ private forms of provisions, and form an indicator of the presence of the implicit option of exiting from statutory programmes. Calculating replacement rates, however, turns out to be a complicated matter, which is reflected in the very different estimates that can be found throughout the literature (European Commission, 2006; Organisation for Economic Co-operation and Development, 2005, 2007, 2011). Some estimates include only the first pillar of strictly public pensions, whereas others seem to treat the sum of ‘public’ and ‘mandated’ schemes as statutory benefits.

The problem is that such additions are not done in a consistent way for different countries in one and the same cross-national study. Thus, for the Netherlands, the OECD (2005) adds the public basic pension of the AOW (Algemene Ouderdomswet) and the benefits of the mandated occupational funds (even if in the SOCX expenditure database, as we will argue further on, the OECD considers the latter not to be ‘mandated’, but to be part of ‘voluntary’ private programmes). For Switzerland, on the other hand, the OECD only takes into account the basic pension of the AHV (Alters- und Hinterlassenenversicherung), even if in its SOCX expenditure database, the benefits paid by Swiss occupational schemes (Pensionskasse) are considered to be part of the ‘mandated’ category. Only in the most recent issue of Pensions at a Glance has the OECD started to calculate separate replacement rates for public programmes, mandated programmes and voluntary programmes (Organisation for Economic Co-operation and Development, 2011).

In Figure 1 we report replacement rates for someone who earned 100 percent of the average wage and someone who earned 50 percent or 150 percent of this wage. The countries are ranked on the basis of the average of these three rates, and it is this average that we will retain for our compound index. The bars for each country stack the replacement rates for the three wage groups.

Figure 1.

Cumulative replacement rates in 2011.

Towards the top of the figure we also report the replacement rates for countries that, on top of public pensions, also have a mandated ‘private’ tier (‘Pub-Man’ refers to cumulative replacement rate of public and mandated private schemes). We have not, however, retained those rates for our index, because even though mandating can have a considerable impact on the distributive consequences of private schemes, it also extends the part of the pension system that is administered directly by private actors such as the sponsoring firm, or commercial intermediates like for-profit insurance companies. Mandating privately funded pensions also extends the reach of the ‘financialization’ of retirement risks. Yet, it is obvious that including the mandated arrangements would reduce the expected room for privatization significantly and that countries in which those practices prevail would score lower on our index of privatization if the cumulative replacement rates had been used in our calculations.

Pensions scheme assets

A second indicator builds upon the observation that private pensions tend to be pre-funded. Funding pension plans render pension entitlements more contingent upon financial market performance and hence reinforces the ‘recommodification’ and the ‘financialization’ of retirement risks. The magnitude of private pensions can thus be measured in terms of their financial assets as a percentage of a country’s gross domestic product (GDP). Here again our indicator is marred by a number of methodological and interpretation problems. A first problem is which assets are to be included. Most studies only take into account funding by autonomous pension funds. That would leave out most of the assets accumulated in countries that use pension insurance contracts as the main financing vehicle for funded pensions (for example Denmark, Sweden, Norway, Ireland and Belgium) or that also rely upon book reserves (for example Germany, Austria, Luxembourg and Canada). Recently, the OECD has finally started to report in its financial markets statistics the magnitude of these other financing vehicles, but reporting remains incomplete (Organisation for Economic Co-operation and Development, 2006).

A second problem is that if one takes this indicator at face value, some schemes that were classified as ‘public’ in our first indicator now end up in our ‘private’ category. This is for example the case for the funds accumulated by the statutory wage earner scheme in Finland (TEL or Tyontekijain elakelaki), which though initiated and governed by statutory legislation (and hence in that regard can be considered as statutory), relies on a form of pre-funding where the accumulated assets are administered by private investment companies (and from this perspective can be considered to amplify the role of private markets in the country’s pension provision) (Lassila and Valkonen, 2000). Finally, a third problem is that our second indicator is sensitive to two factors that are not directly related to the differences in the current role of markets. The size of assets reflects (1) the maturity of the funded schemes (and not only the current setup of the pension arrangements); as much as (2) the strictness of the rules the legislator or the regulator imposes upon those schemes (with more strictly regulated – and hence less privatized – schemes having more assets than less regulated regimes). It therefore does not come as a surprise that countries where both factors play an important role (such as the Netherlands and Switzerland) come out on top of our ranking, whereas countries where pre-funding is of recent origin (such as Finland, or for that matter Germany) have sizeable but less impressive assets. It also explains that the Netherlands, Switzerland and Denmark score higher in terms of this index of ‘privatization’ than more archetypical liberal market economies such as the USA and the UK. Figure 2 rank-orders countries on the basis of the volume of their financial assets broken down according to the type of investment vehicle in the year 2007.²

Figure 2.

Assets of private pensions according to investment vehicle in 2007.

The type of investment vehicle is also, to some extent, indicative of the degree of ‘financialization’: autonomous pension funds tend to be complementary to a system of financial market intermediation, whereas book reserve finance allows companies to access finance without issuing stocks. In other words, book reserves, though a funded form of pension finance, are less likely to reinforce the role of capital markets and are more complementary to a coordinated market economy. Pension insurance can be seen as a form of provisions situated somewhere in between these two poles.

Private pension expenditure

As a third indicator of the importance of private pensions we use the ratio between public and private expenditure as reported by the OECD (SOCX). This ratio can be considered to indicate to what extent the administration of pension benefits has been transferred to private sector actors. In this context we follow the distinction between ‘public’ and ‘private’ as made by the OECD in its nomenclature on pensions (and which guides the classification in its expenditure database). A pension plan is considered ‘public’ when ‘statutory programmes [are] administered by general government (that is central, state and local governments, as well as other public bodies such as social security institutions)’; ‘private’ pension plans, on the other hand, are those that are ‘administered directly by a private sector employer acting as the plan sponsor, a private pension fund or a private sector provider’ (Organisation for Economic Co-operation and Development, 2005: 12).

There are some problems with the OECD’s application of these principles. Thus, even if the earnings-related part of the Finnish pension system is administered by private insurance companies, the OECD considers the expenditure that is involved as ‘public’ because it is the state that directly mandates inclusion into the scheme as well as the conditions of entitlement. The private administration of the Finnish scheme only affects the sponsors of the scheme, that is, primarily the employers:³ in so far as that the larger the income that the provider (who is selected by the employer) achieves, the greater the amount it is capable of granting employers in bonuses, which reduces the final pension contribution (Risku, 2003: 88). Paradoxically, what in Figure 3 is represented as ‘private’ expenditures for Finland are actually expenditures paid to former civil servants by autonomous state sponsored funds (that is, VEL, KVTEL and KIEL). This peculiar classification is a consequence of the application of the guidelines for System of National Accounts: ‘social insurance schemes organized by government units for their own employees, as opposed to the working population at large, are classified as private funded schemes … and are not classified as social security schemes’ (System of National Accounts, 1993: §8.63). In other words, if pension payments for former civil servants go via autonomous funds (such as separate pension and/or insurance companies) and the government ‘does not make up the deficit on a regular basis’ (Queisser et al., 2007: 553), they are considered to be a form of private expenditure (even if the state is the sponsor of those funds and the beneficiaries are former civil servants). On the same grounds, the OECD categorizes civil servant pensions in Australia, Canada, Denmark, Sweden, the UK and the Netherlands as a form of ‘private’ social expenditure.

Figure 3.

Private expenditure as percentage of public expenditure in 2007.

Coverage by private pension plans

In order to capture the accumulation of future entitlements we combined two variables: the percentage of the active population covered by a private pension plan, and the contribution rates levied by ‘the largest private pension schemes’ (Organisation for Economic Co-operation and Development, 2007: 77; 2011). This is somehow indicative of the extent to which private sector actors are responsible for providing future retirement income. The OECD data only seem to take into account funded ‘private’ schemes: for France, for instance, the figure does not take into account the ARRCO and AGIRC schemes. For a country such as Finland, on the other hand, the privately managed pension fund assets are part of the statutory earnings related ‘public’ benefits, which explains the low score of this country on our fourth indicator (in other words, schemes such as TEL are not included here). We use the average of the minimum and maximum rate. In Figure 4 the countries are ranked on the basis of a factor, that was calculated by multiplying contribution rates to second pillar pensions (‘contribution’) by the percentage of the dependent labour force covered by those schemes (‘coverage’) (that is, the two factors depicted in the figure). It is this factor that will be incorporated into our compound index.

Figure 4.

The coverage and contribution rates of private pension schemes.

In a way this fourth component mirrors our first component: it tells us something about the way in which the current generation is provided for, by measuring to what extent private schemes have not been ‘crowded out’. Of course in neither case can one assume for sure that those entitlements will be granted in the end: pension rights in public schemes could be subject to retrenchment by the time people reach their retirement, and there is no guarantee that what is paid into privately funded schemes actually leads to benefits being paid out, as the assets may be eaten away by a greedy financial service sector or by a financial crisis.

The compound indicator

Because each of our four indicators is marred with measurement reliability and classification errors for some of the countries, we decided to develop a compound index, which should at least reduce this problem. As we argued, two variables (REPLACEMENT and CONTRICOV) measure the crowding out ofthe building up of future private entitlements, whereas the two other variables (FUNDING and EXPENDITURE) measure aspects related to how in the recent past private entitlements have been accumulated for the current generation of pensioners. For estimating the compound index we standardized each of our four indicators on a scale from 0 to 10 in order to maintain the relative distance between our cases. Subsequently we calculated the mean of these four scores into one single score, thus attributing the same weight to each indicator. For the replacement rate indicator we reversed the values, so that the lowest value reflected the least room for privatization. Figure 5 stacks the four scores, but ranks countries on the basis of the average of the four components.

Figure 5.

The compound indicator for the importance of private pensions.

Table 1 gives an overview of the original non-standardized values of the four components of the compound indicators, the sum of the standardized values and the compound score that averages the scores of the four aspects.

Table 1.

The composition of the compound index of pension privatization.

	Replacement rate of public pensions: mean of 0.5, 1.0 and 1.5 × average production worker wage	Assets of funded pensions as % of gross domestic product	Private expenditure as % of public expenditure	Contribution rates multiplied by coverage private pensions	Compound index of pension privatization (mean score)
	Replacement	Funding	Expenditure	Contricov	Compound index of pension privatization (mean score)
Luxembourg	96.0	1.2	0.07	80	0.3
Spain	84.2	12.2	0.05	76	0.8
Italy	71.8	3.6	0.11	23.5	1.1
Portugal	71.0	26.0	0.02	12	1.3
Austria	88.6	18.8	0.05	61.25	1.7
France	60.9	6.9	0.02	27	1.9
Germany	55.5	17.9	0.09	210	2.2
Finland	67.2	78.1	0.02	55	2.4
Belgium	56.5	14.4	0.40	135	2.6
New Zealand	49.7	11.1	0.31	80	2.6
Norway	53.0	54.5	0.09	180	2.9
Japan	42.4	20.0	0.37	162	3.2
Sweden	33.3	57.4	0.23	225	4.0
Ireland	38.2	93.6	0.40	520	5.3
Canada	52.1	103.4	1.07	331.5	6.3
UK	42.1	96.4	0.78	531	6.3
Denmark	41.4	140.6	0.30	1120	6.5
USA	50.8	124.0	0.82	486	6.6
Australia	20.4	119.5	0.72	810	8.4
Netherlands	40.2	149.1	0.67	1305	8.5
Switzerland	42.7	151.9	0.84	1125	8.7
Mean	55.1	61.9	0.4	359.8	4.0
Standard deviation	19.1	53.7	0.3	403.6	2.7

Conclusion

The task of constructing an index measuring the extent to which countries have privatized their pension provision is laden with conceptual and empirical difficulties. The distinction between ‘public’ and ‘private’ remains essentially contested and it entails many aspects. In the index that we have developed, we decided to privilege the aspect of market orientation because we considered this to be a defining feature of the changes in the post-war settlement. Private actors and labour market records have at all times played a role in the operation of even the most decommodified welfare regimes of the ‘golden age’. What has changed in most countries, and what continues to form a source of cross-national variation, is the environment in which private actors are implicated in the governance of pensions: in Bismarckian welfare states, and to a lesser extent in Social Democratic policy regimes, private actors have always played an important role, but only in a non-competitive administrative environment that functions in the shadow of the coordinating power of the state. In more privatized regimes those actors operate in a much more fragmented competitive market-based environment. Various forms of mandating have an ambivalent effect on pension provision. On the one hand they can create the possibility of more inclusiveness and enhance the degree of solidarity, but on the other hand they induce ordinary people, who otherwise probably would refrain from taking part in an equity culture, to render a sizeable part of their old age security contingent upon developments in the financial markets. In many of the countries that score high on our index of private pension provision these non-liberal practices have been in place for some time, and hence have contributed to maturation of their private pension pillars.

The compulsive and persuasive power of the state, however, also seems increasingly to be used as a policy tool to transform Social Democratic or Conservative regimes. Thus the 1999 pension reform in Sweden ‘was intended to universalize equity ownership’ (Belfrage, 2008: 283), and the Riester reforms in Germany seemed to be more geared towards boosting the position of Germany as a financial centre than contributing to solving the politically and demographically induced imbalances in the German PAYG pension system (De Deken, 2002). Within a decade, such a policy contributed to a doubling of the total amount of pension assets relative to GDP in Sweden (from 26.38 percent of GDP in 1998 to 56.36 percent in 2009). The fact that the German reforms were less successful in this regard can in part be attributed to what in essence remained the voluntary nature of the state induced drive towards capital funding: in spite of the massive federal direct and indirect subsidies, the German government never opted for a true form of mandating and over a period of 10 years the size of pension assets only marginally changed (from 3.35 percent of GDP in 1999 to 5.20 percent in 2009) (OECD Pension Indicators, various years). Sweden now scores almost as high on pension privatization as classical liberal welfare states such as the US and Canada, while Germany remains closer to the group of countries where private plans play only a very subordinate role.

The index we propose privileges certain aspects of privatization over others: it is more effective in assessing the extent to which pension arrangements embrace market principles than in measuring the degree to which retirement risks are individualized. The level of individual responsibility for retirement provision is hard to quantify with the indicators that are available and comparable on a cross-national basis. The individualization of retirement risks is to a significant extent determined by the scope of the regulation of the private sector. Hence, in order to obtain a more complete picture of the extent to which the pension system of a country is privatized, the results reported in this article need to be complemented by more qualitative oriented case studies that will allow us also to qualify some of our findings.

Practices of mandating that limit individual choice and individual responsibility are likely to make a country score higher on our index than if participation in private pension plans is contingent upon voluntarism. A more encompassing regulation that obliges the providers of pension products to maintain larger reserves to back up their pension promises can, in some cases, also lead to a higher score on our index than if the private pension system is less regulated. On the other hand, the paradoxical consequence of limiting voluntarism and tightening regulation is that more people are drawn into market-based pension arrangements and that financial markets become more pervasive within the daily life of the population. In this regard countries with a mandated and more tightly regulated market-based pension system can indeed be said to be more private.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Notes

References

Adema

Frohn

Ladaique

(2011) ‘Is the European Welfare State Really More Expensive? Indicators on Social Spending, 1980–2012’, OECD Social, Employment and Migration Working Paper 124. Paris: OECD.

Alber

(2003) ‘Recent Developments in the German Welfare State: Basic Continuity or a Paradigm Shift?’, in Gilbert

Van Voorhis

R.A.

(eds) Changing Patterns of Social Protection. New Brunswick, London: Transaction Publishers.

Antón

(2007) ‘Distributional Implications of Tax Relief on Voluntary Private Pensions in Spain’, Fiscal Studies 28(2): 171–203.

Bateman

Piggott

(1998) ‘Mandatory Retirement Saving in Australia’, Annals of Public and Cooperative Economics 69(4): 547–69.

Belfrage

(2008) ‘Towards “Universal Financialisation” in Sweden?’, Contemporary Politics 14(3): 277–96.

Blanchet

Pele

(1999) ‘Social Security and Retirement in France’, in Gruber

Wise

(eds) Social Security and Retirement Around the World. Chicago, IL: University of Chicago Press.

Bonoli

(2005) ‘Switzerland: Adapting Pensions within Tight Institutional Constraints’, in Bonoli

Shinkawa

(eds) Ageing and Pension Reform Around the World. Chelthenam: Edward Elgar.

Bonoli

George

Taylor-Gooby

(2000) European Welfare States: Towards a Theory of Retrenchment. Cambridge: Polity Press.

Bridgen

Meyer

(2007) ‘Private Pensions versus Social Inclusion? Three Patterns of Provision and their Impact’, in Meyer

Bridgen

Riedmüller

(eds) Private Pensions Versus Social Inclusion? Non-State Provision for Citizens at Risk in Europe. Cheltenham: Edward Elgar.

10.

Cichon

(1999) ‘Notional Defined-Contribution Schemes: Old Wine in New Bottles?’ International Social Security Review 52(4): 87–105.

11.

Clark

Whiteside

(2003) Pension Security in the 21st Century. Oxford: Oxford University Press.

12.

Clark

R.L.

Mitchell

(2002) ‘Strengthening Employment-based Pensions in Japan’, Pension Research Council Working Paper WP 2002-1. Philadelphia, PA: Wharton School, University of Pennsylvania.

13.

Crouch

(1999) ‘Employment, Industrial Relations and Social Policy: New Life in an Old Connection’, Social Policy and Administration 33(4): 437–57.

14.

Davanne

Lorenzi

J.H.

Morin

(1998) Retraites et Épargne. Paris: Conseil d’Analyse Économique.

15.

De Deken

J.J

. (2002) ‘Pensions and the Reduction of Non-wage Labour Costs. Modelling a Decade of Reforms in Germany’, Journal of European Social Policy 12(4): 277–92.

16.

De Deken

J.J.

Ponds

E.H.M.

Van Riel

. (2006) ‘Social Solidarity’, in Clark

G.L.

Munnell

A.H.

Orszag

J.M.

(eds) The Oxford Handbook of Pensions and Retirement Income. Oxford: Oxford University Press.

17.

Engelen

Konings

Fernandez

(2008) ‘The Rise of Activist Investors and Patterns of Political Responses: Lessons on Agency’, Socio-Economic Review 6(4): 611–36.

18.

Erskine

(1997) ‘The Withering of Social Insurance in Britain’, in Clasen

(ed.) Social Insurance in Europe. Bristol: Policy Press.

19.

Esping-Andersen

(1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

20.

European Commission (2006) Adequate and Sustainable Pensions: Synthesis Report. DG Employment Social Affairs and Equal Opportunity, Luxembourg: Office for Official Publications of the European Communities.

21.

Gilbert

(2004) Transformation of the Welfare State: The Silent Surrender of Public Responsibility. Oxford: Oxford University Press.

22.

Goudswaard

Beetsma

R.M.W.J.

Nijman

Th.E.

Schnable

(2010) Een sterke tweede pijler: Naar een toekomstig stelsel van aanvullende pensioenen. Den Haag: Commissie Toekomstbestendigheid Aanvullende Pensioenregelingen (study commissioned by the Dutch Ministry of Labour Affairs).

23.

Hyde

Dixon

(eds) (2010) Comparing How Various Nations Administer Retirement Income: Essays on Social Security, Privatisation, and Inter-Generational Covenants. Lewiston: Edwin Mellen Press.

24.

Hyde

Dixon

Drover

(2006) The Privatisation of Mandatory Retirement Income Protection: International Perspectives. Lewiston: Edwin Mellen Press.

25.

Jackson

Vitols

(2001) ‘Between Financial Commitment, Market Liquidity and Corporate Governance: Occupational Pensions in Britain, Germany, Japan and the USA’, in Ebbinghaus

Manow

(eds) Comparing Welfare Capitalism Social Policy and Political Economy in Europe, Japan and the USA. London: Routledge.

26.

Künemund

Rein

(1999) ‘There is More to Receiving than Needing: Theoretical Arguments and Empirical Explorations of Crowding In and Crowding Out’, Ageing and Society 19(1): 93–121.

27.

Langley

(2006) ‘The Making of Investor Subjects in Anglo-American Pensions’, Environment and Planning D: Society and Space 24(6): 919–34.

28.

Lassila

Valkonen

(2000) ‘Prefunding in a Defined Benefit Pension System: The Finnish Case’, in Feldstein

Siebert

(eds) Social Security Pension Reform in Europe. Chicago, IL: University of Chicago Press.

29.

Leisering

(2012) ‘Pension Privatisation in a Welfare State Environment: Socialising Private Pensions in Germany and the United Kingdom’, Journal of Comparative Social Welfare 28(2): 139–51.

30.

Lindell

(2003) ‘Financing pensions’, in Hietaniemi

Vidlund

(eds) The Finnish Pension System. Eläketurvkeskus: Finnish Centre for Pensions.

31.

Martin

(2002) The Financialisation of Daily Life. Philadelphia, PA: Temple University Press.

32.

Minns

(2001) The Cold War in Welfare: Stock Markets versus Pensions. London: Verso.

33.

Organisation for Economic Co-operation and Development (2005) Private Pensions: OECD Classification and Glossary. Paris: OECD.

34.

Organisation for Economic Co-operation and Development (2006) Pension Markets in Focus, October 2006, No. 3. Paris: OECD.

35.

Organisation for Economic Co-operation and Development (2007) Pensions at a Glance. Paris: OECD.

36.

Organisation for Economic Co-operation and Development (2011) Pensions at a Glance. Paris: OECD.

37.

Pearson

Martin

J.P.

(2005) ‘Should We Extend the Role of Private Social Expenditure?’ OECD Social, Employment and Migration Working Paper 23. Paris: OECD.

38.

Pedersen

A.W.

(2004) ‘The Privatisation of Retirement Income? Variation and Trends in the Income Packages of Old Age Pensioners’, Journal of European Social Policy 14(1): 5–23.

39.

Queisser

Whitehouse

Whiteford

(2007) ‘The Public–Private Pension Mix in OECD Countries’, Industrial Relations Journal 38(6): 542–68.

40.

Rein

Turner

(2001) ‘Public–Private Interactions: Mandatory Pensions in Australia, The Netherlands and Switzerland’, Review of Population and Social Policy 10: 107–53.

41.

Risku

(2003) ‘Investment of the Statutory Earnings-related Pension Scheme’, in Hietaniemi

Vidlund

(eds) The Finnish Pension System. Eläketurvkeskus: Finnish Centre for Pensions.

42.

Schröder

(2008) ‘Integrating Welfare and Production Typologies: How Refinements of the Varieties of Capitalism Approach Call for a Combination of Welfare Typologies’, Journal of Social Policy 38(1): 19–43.

43.

Scruggs

(2007) ‘Welfare State Generosity Across Space and Time’, in Clasen

Siegel

(eds) Investigating Welfare State Change: The ‘Dependent Variable Problem’ in Comparative Analysis. Cheltenham: Edward Elgar.

44.

Seeleib–Kaiser

(2008) ‘Welfare State Transformations in Comparative Perspective: Shifting Boundaries of “Public” and “Private” Social Policy’, in Seeleib-Kaiser

(ed.) Welfare State Transformations in Comparative Perspectives. Basingstoke: Palgrave Macmillan.

45.

Starr

(1988) ‘The Meaning of Privatisation’, Yale Law and Policy Review 6(1), 6–41.

46.

System of National Accounts (1993) System of National Accounts 1993. Brussels/Luxembourg, New York, Paris, Washington, DC: CEC-EUROSTAT, IMF, OECD,UN and the World Bank.

47.

Trampusch

(2006) ‘Industrial Relations and the Welfare State: The Different Dynamics of Retrenchment in Germany and the Netherlands’, Journal of European Social Policy 16(2): 121–33.

48.

Van Gunsteren

Rein

. (1985) ‘The Dialectic of Public and Private Pensions’, Journal of Social Policy 14(2): 129–49.

49.

Whiteside

(2006) ‘Adapting Private Pensions to Public Purposes: Historical Perspectives on the Politics of Reform’, Journal of European Social Policy 16(1): 43–54.

50.

World Bank (1994) Averting the Old Age Crisis. Oxford: Oxford University Press.