1
THE PERFORMANCE OF
PENSION FUNDS
THE ROAD TOWARDS MORE ADEQUACY,
SUSTAINABILITY AND LEGITIMACY
Jolinne van Baar – s1402994
Thesis Public Administration: Economics & Governance
January 9, 2019
Advisor: N. van der Zwan
Second reader:
2 TABLE OF CONTENTS
ABSTRACT ... 3
1. INTRODUCTION ... 4
2. LITERATURE REVIEW ... 8
2.1 The institutional paradigm of welfare state development ... 8
2.2 The policy framework in pension systems ... 10
2.2.1 The stakeholders in supplementary pensions ... 12
2.3 Socially Responsible Investing ... 13
2.3.1 Definitions and categories of SRI ... 14
2.3.2 The drivers of SRI ... 15
2.3.3 Fiduciary duty ... 16
2.3.4 Paying a price for doing good? ... 17
3. METHODOLOGY ... 20
3.1 The variables ... 20
3.1.1 The dependent variable ... 20
3.1.2 The independent variable ... 23
3.1.3 The control variables ... 24
3.2 The cases ... 24
3.3 The methods used... 25
3.4 Validity, reliability and generalization ... 26
4. CASE DESCRIPTION ... 28
4.1 The Netherlands ... 28
4.2 The United Kingdom ... 31
4.3 Switzerland... 35
4.4 Summary ... 38
5. THE PERFORMANCE OF PENSION FUNDS ... 41
5.1 Adequacy ... 41 5.2 Sustainability ... 45 5.3 External legitimacy ... 48 5.4 Summary ... 53 6. ANALYSIS ... 55 7. CONCLUSION ... 62 Bibliography ... 65
3 ABSTRACT
Pension funds are in the spotlight: their vast (and growing) amount of capital within pension funds have the capacity to stimulate positive change, especially in light of demographic pressures, financial crises and climate change. Furthermore, the question arises of how present and future benefits of beneficiaries can be guaranteed? This thesis investigates the possible correlation between regulation and the performance of pension funds. The bigger issue at stake is the effectiveness of regulation more broadly. Then one can answer: does regulation have the possibility to be a (positive) influence on the market of pension funds? National legislation, regulatory activity, civil regulation and EU-level legislation are incorporated as features of regulation. Moreover, stakeholder engagement (e.g. of labour unions and business actors) could have influenced the level of (civil) regulation. This thesis creates a tangible and measurable concept of the normally abstract ‘performance of pension funds’ in connecting factors of adequacy, sustainability and external legitimacy of pension funds. I investigate the differences between three mature multi-pillar pension systems with a highly capitalized occupational pillar: the Netherlands, the United Kingdom and Switzerland. The results of the covariational analysis show, however, similar levels of regulation and similar performance levels. In this way, a strong correlation cannot be established. Yet some remarkable aspects beyond national legislation – such as stakeholder engagement, civil regulation and the role of the EU – might have explanatory power in the performance of pension funds.
4 1. INTRODUCTION
Investment policies of pension funds are changing since the 21st century. Socially responsible investments (SRI) have grown rapidly and have become more mainstream in the last few years (Sievänen, Rita & Scholtens, 2012; Eurosif, 2016). SRI falls under the broader practice of incorporating environmental, social, and governance (ESG) elements in investment decisions. In this way, SRI seeks “opportunities for the creation of long-term value and society” (Eurosif, 2016a, p.2). At the start, this practice was aimed at risk management, mostly in excluding specific companies and sectors that do not comply with general norms in society. SRI moved to the more positive perspective in actively screening and incorporating companies that act in a conscious manner, instead of only excluding the norm-violating companies (Eurosif, 2016a, p. 2). A recent example is the commercial of a Dutch insurance company with the message of a young girl that an investor has the responsibility look beyond financial returns to incorporate the long-term interests of their clients along with future concerns (ASR, 2018). This positive and active attitude is also remarkable at the EU level, for example in the Capital Markets Union (CMU) Action Plan of the European Commission (of 2015) to focus on integrating environmental, social and governance (ESG) issues in the investment decision-making process, as well as a focus on the long term. The European Parliament (EP) agreed upon this Action Plan by focusing on the incorporation of ESG issues in investments. Moreover, the EP’s position is in line with the focus on active and long-term investment perspectives (Eurosif, 2016, pp. 32-33).
The growth of the SRI market is however not directly compatible with the primary goal of pension funds; to achieve the highest possible financial return, while keeping risk at an acceptable level. Mainstream economic rational theory state that institutional investors like pension funds choose assets that are the most attractive in terms of time and risk. The substance of the asset (which company, what does it do, and how?) is inferior to the asset’s financial return, since the focus is on the beneficiary receiving the highest possible pension when he or she retires. Integrating ESG factors into the decision-making would hamper this strategy, as the returns could potentially be lower. On paper, the ‘best match’ would be exchanged for another, lower-ranked, choice. The hotly debated topic of ‘what does it cost to be socially responsible’ is therefore not surprising. Multiple scholars have written about it, and found diverse inconclusive results1.
This thesis seeks to address the puzzle of the performance of pension funds by analysing the role of legislation and regulation in the growth of the socially responsible investment market of pension funds in three political economies: the Netherlands, the United Kingdom, and Switzerland. Both national and EU-level regulations could be important drivers of SRI strategies, as regulation can encourage and discourage behaviour of pension institutions (Majone, 1997). The European Sustainable Investment
1 See for example El Ghoul & Karoui, 2017, pp. 53-54; Capelle-Blancard & Monjon, 2012, p. 239; Dam &
Scholtens, 2015, p. 114; Sethi, 2005, p. 106; Dam & Scholtens, 2015; Wen, 2009; Salomon, 2006, Renneboog, Ter Horst & Zhang, 2008
5
Forum (Eurosif) found in their biannual survey, for instance, that regulation is one of the most important influencers of SRI practices among European countries (Eurosif, 2018). Consequently, exploring the national and European influences for the SRI market is relevant to optimise this growing responsible market.
SRI strategies and ESG issues have become more and more significant components for institutional investors’ decision making. From the investor’s perspective, ESG factors influence the risk and return of the investment portfolio. From the policy-maker’s perspective, it could be beneficial to harness the financial capital of institutional investors to support global arrangements such as the Sustainable Development Goals (SDGs) and the Paris Agreement. The large sum of money captured yearly by pension funds (to illustrate, in the Netherlands the amount is as high as 800 billion Euros, opposed to the GDP of 600 billion) has an enormous potential to change for the better. Surprisingly, these progressive sounds do not only originate from scholars as the business-sphere is advocating change as well (Sethi, 2005; Kellermann, 2012). From the citizen’s perspective, civil society is also exercising pressure on the investors to use their weight and capital to change ESG practices at companies where they have assets in (OECD, 2017, p. 9).
The ‘Universal Owner’ approach is also related to this perspective of the potential positive influence of pension fund capital. Pension funds own a relevant share of global markets and are therefore ‘universal owners’. Consequentially, they are vulnerable to the overall state of the market. Pension funds are concerned about economic crises and economic decline, since returns on investment will decrease. In order to reduce the risk of lower returns on investments, pension funds exert pressure on companies to improve their attitude and actions. Besides, pension funds have the duty to influence it for the better by taking long-term factors into account so that ESG-related risks are reduced. This explains the engagement of pension funds in the market. In this way, the ‘Universal Owner’ approach gives pension fund investors an important role in the SRI development: engaging in companies to stimulate positive change. This approach relies more on inter-generational issues and the sustainability of the economy, which changes will have an influence on future returns on investments (OECD, 2017, p. 24; Kellermann, 2012). Traditional financial investors externalize costs related to ESG-factors, but ‘universal owners’ see this as passing costs to other companies that are in the investment portfolio as well. In the end, the portfolio does not perform better, but worse (OECD, 2017, p. 24).
The starting point of this thesis is the assumption that the responsibilities of the state and the responsibilities of private actors (in this case: pension funds and related actors such as asset managers) determine the potential for pension funds to act effectively and responsibly (regarding e.g. SRI). Moreover, I assume that the policies of pension funds are influenced by institutional characteristics, such as the pension fund system and the level of capitalization, the strength of social partners, national regulation and EU-level legislation. The thesis uses a broad concept of regulation, incorporating three
6
elements: national legislation, regulatory activities by supervisory agencies, and civil regulation by the pension sector itself and non-state actors (such as labour unions, interest groups and other private actors) influencing it (Van der Zwan et al., 2018, p. 2; Vogel, 2010, p. 69). Then, I assess the performance of pension funds regarding their policies in terms of (1) adequacy for the current generation to guarantee poverty protection and income provision, (2) financial sustainability on the long-term in light of the demographic challenges and (3) external legitimacy regarding SRI practices that minimize negative externalities.
The research compares the performance of pension funds for pension funds in the three cases of the Netherlands, the United Kingdom and Switzerland. The data and evidence is collected by means of document analysis. The covariational analysis is used as the method of data analysis, since it controls for variation on the independent variables across a small number of cases. In this way, the covariational method allows for the identification of a correlation between an independent variable and a dependent variable. Supported by existing theory, the relationship can be recognised.
The three cases resemble each other, but also show variation on a number of potential explanatory factors. All three share the history of a multi-pillar pension fund system and the high capitalization of the second pillar (of occupational pensions). All three are categorized as a ‘mature’ multi-pillar pension system, as developments are relatively not recent (Ebbinghaus, 2011). This means that the basic public pension is supported by funded occupational pension plans and even a third option of personal pension plans, which are often voluntary. Both the Netherlands and the UK have a history of investing according the ‘prudent person rule’, whereas Switzerland has not. Moreover, Switzerland is not a member of the European Union, while the other two political economies are. Another factor of comparison is stakeholder engagement. Employer groups and labour unions – together the ‘social partners’ – could have significant bargaining powers and exert influence on regulation related to occupational pensions. Unlike Switzerland and the Netherlands (where the social partners are strong and embedded in the decision-making structures), the UK is not familiar with the social partners having a say in the policy-making area. The resemblances on several factors of the pension system and capitalization of the second pillar combined with differences in investment traditions, stakeholder engagement and membership of the EU establish an interesting comparison to study the three countries in depth.
The outline of the thesis is as follows. The next chapter discusses the scholarly literature on the welfare state development in general, continuing with specifying theories on pension systems development and socially responsible investing (SRI). Several hotly debated topics are reviewed; issues regarding definitions of the term and related ones – for example sustainability, CSR, and ESG factors –, principal-agency dilemmas in pension funds relations, the meaning of fiduciary duty, the prudent person rule, and the potential costs or benefits of ‘doing good’. The third chapter outlines the methodology regarding the variables used, the case selection and the covariational analysis that is employed. The chapter also
7
discusses the validity, reliability and generalization of the research. The following chapter elaborates on the multi-pillar pension system of each country in detail, as well as the regulation that shapes the SRI possibilities for the pension funds. It also shines light on the level of stakeholder engagement in the three cases. The fifth chapter details the performance of pension funds within the three countries, regarding adequacy, sustainability and external legitimacy in terms of SRI. Then, the analysis links the findings from the two earlier chapters to the hypotheses formulated earlier. Hence, existing theories of the theoretical section are tested. The thesis will conclude by formulating an answer to the research question as well as suggestions for further research, ending with a policy recommendation.
8 2. LITERATURE REVIEW
2.1 The institutional paradigm of welfare state development
Pension schemes serve two main functions: to protect older people against poverty and to sustain income after people retire (Immergut & Anderson, 2007, p. 18). Pension systems are therefore an important part of the welfare state as a whole. Multiple institutions (to regulate, monitor and administrate pension plans) developed over the years, with growing welfare state expenditures and more specifically, pension expenditures, as a consequence. However, in order to measure the scope of the welfare state in society, it is too narrow to focus only on welfare state expenditures or social expenditures (Huber, Ragin & Stephens, 1993, p. 712; Esping-Andersen, 1990). The most basic definition of the welfare state is that it is revolved around state intervention to secure a standard of living (welfare) for the citizens within that state (Esping-Andersen, 1990, pp. 18-19). The essence is that the welfare state makes sure that people are not fully dependent on the market. Huber, Ragin and Stephens describe this as they speak about ‘constitutional structures’ such as direct democracy and different electoral districts and the impact these structures have on the welfare state (1993, p. 719). The welfare state thus encompasses more than expenditures: it is about the extent of social protection, effective governance and different social risks.
This is the beginning of the institutionalist paradigm as many scholars point out (Weaver, 1998; Esping-Andersen, 1990; Korpi, 2006; Bonoli, 2000; Scharpf, 2000). The paradigm highlights that (political) institutions matter for the scope and success of governmental action in the sphere of the welfare state. Esping-Andersen (1990) argues that the origin and expansion of the welfare state is due to the concept of ‘power resources’ that institutions have in terms of electoral numbers and collective bargaining. The scholar was ground-breaking in his categorization of welfare states in the liberal, conservative and social democratic welfare state and the institutions that go with each category (Esping-Andersen, 1990, pp. 25-26). Scholars have argued that change in these regimes occur in line with institutional arrangements that have already been in place for a longer period of time. Therefore, change is regime-dependent and path dependent (Pierson, 1996).
Quickly after Esping-Andersen’s research, other scholars of the employer-centred approaches elaborated and criticized the theory. Swenson (1991) was one of the most important contributors to this debate, making the point that the focus on the welfare state was too narrow. The political economy as a whole (literally “bringing capital back in”) would change the theory. Employers played a significant role in centralizing the interests of labour unions instead of the notion of the passive and/or weak capital (Swenson, 1991, pp. 543-544). Business, according to Swenson, could be supportive of social policies as well – an argument that was previously ignored. Korpi (2006) however criticized Swenson and elaborates on the theory of Esping-Andersen in distinguishing between first-order and second-order preferences. If certain first-order interests are at stake, an actor (the protagonist) fights for the issue and
9
the actor does not accept a compromise. Consenting to a deviant option is also not possible in the light of first-order preferences. But in case of second-order preferences, the actor (consenter) would not actively push for the desired outcome and slightly different outcomes are accepted as well. Second-order preferences therefore reflect lower levels of policy preferences. This was the case for business actors in Swenson’s theory, according to Korpi (Korpi, 2006, p. 182). These preferences can also change over time. Limited power resources would more likely result in consenters, while protagonists require more power resources. In this way, business actors with less power resources would be more likely to consent to social policies, which contradicts Swenson’s notion of business actors advocating these policies (Korpi, 2006, pp. 181-182).
In contrast to the expansion of the welfare state which institutionalism can ‘easily’ explain, welfare state retrenchment captures a different development. Retrenchment is not the ‘mirror image’ of expansion, because of changing goals and contexts as Pierson (1996) describes in his ‘New Politics’. Retrenchment occurred from the 1970s onwards, prior to the more recent privatization period. Welfare state expansion is mostly in direct advantage of citizens, with more benefits in terms of retirement income. However, retrenchment involves cutting the existing benefits and is considered unpopular. Retrenchment necessitates politicians to behave differently due to this unpopularity, with strategies of blame avoidance and lowering the visibility of the cutbacks (Pierson, 1996, pp. 145-147).
The success of reform depends on key stakeholders. Their acceptation of reform and the conditions under which they accept reform decide upon the impact of the reform itself. Labour unions are such an example of key stakeholders. In the line of the thesis of Pierson’s blame avoidance, reforms may have been more successful if those were easily accepted by labour unions than when they opposed the reform. Corporatism and ‘social pacts’ could therefore have had a significant impact on the politics of reform (Visser & Hemerijck, 1997). Bringing labour back in the analysis of social policy was already advocated by Rainwater as early as 1986, emphasizing the role of firms and social partners.
Moreover, the context has changed in the sense that interest groups beside politicians and labour unions have become relevant in society (Pierson, 1996, pp. 150-151). This creates another layer of difficulty for politicians to reform the existing pension system. All these interest groups try to exert influence on the existing pension system once it is institutionalized. Due to this institutionalization and the long-term aspect of pensions (since arrangements involve planning on the long-term), ‘lock-in’ effects exist which defend the status quo (Immergut & Anderson, 2007, p. 11). The focus then tends to be on political stability. However, this does not explain the reform period in the 1990s. Nonetheless, no intervention of the state (or the tendency to avoid blame) can also have major consequences due to ‘policy drift’: “changes in the operation or effect of policies that occur without significant changes in those policies’ structure” (Hacker, 2004, p. 246).
10
Above perspectives describe different trajectories of the expansion, retrenchment and the increasing role of private actors in the welfare state and especially pension systems. Generally accepted is that pension reform would not occur unless objective ‘problem pressure’ exists. The pressures are widely acknowledged: the population is ageing, birth-rates are falling, and employment rates and economic growth is fluctuating (Immergut & Anderson, 2007, p. 16). These developments, as well as the two financial crises in the previous decade, pose a challenge to the financial sustainability of pension systems. In PAYG systems, for instance, less workers have to make sure more retirees have their pension income. These problem pressures do not exclusively explain the pattern of reform among all pension systems. Politics, decision-making of relevant societal actors and governance play a pivotal role in addressing these pressures. In this way, the increasing role of occupational pensions (with more responsibility for the private pension funds) raises fundamental issues of the consequences of the changing responsibility for the private sphere and participatory rights of beneficiaries when no official government body is tasked with the organization of private pensions (Ebbinghaus, 2011, p. 3).
2.2 The policy framework in pension systems
Studying pension systems with a focus on institutional change, Ebbinghaus and Gronwald (2011) found that timing and sequence of different developments were relevant and contributing to unique consequences for the public-private mix in different countries. The public-private mix entails an analysis of the country’s welfare regime, labour relations, and the market economy (Ebbinghaus, 2011, p. 11). The outcome is a combination of public and private schemes (Rein & Rainwater, 1986). Each country stands out in their public-private mix, leading to different and responsibilities of the state as well as private actors (Ebbinghaus & Gronwald, 2011, p. 23). Countries with the same welfare state regime differed in the reforms implemented to change the pension system – differences in contents, timing and sequence –, which led to ‘path departure’ instead of ‘path dependency’ (Palier, 2007, p. 87; Esping-Andersen, 1990; Pierson, 1996; Skocpol, 1985). Due to these trends and shifts, pension systems have become more and more privatized and the interpretation of the aspects of coverage, benefits, financing and administration has changed. Moreover, these four aspects are intertwined. For example, the financing mode of the pension programme has implications for the administration and benefits of beneficiaries. The privatization has led to a shift in responsibility for old-age pensions from state to private actors, specifically employers and employees. The state proposed and enforced reforms of privatization, but the impact depends on the adaptation of the relevant private actors to take responsibility. These reforms and the adaptation thereof modify the public-private mix, mostly in the recent trend of privatization.
If private actors fail in taking responsibility for their new tasks, a gap arises (for an example, see Bridgen & Meyer, 2005). Subsequently, the scope for public regulation and societal control of private pensions
11
increased while the state pulled its hands of their responsibility to finance public pensions. This is what Leisering et al. (2002) describe as the ‘paradox of privatization’ (Ebbinghaus & Wiss, 2011, p. 351). Even if the ‘market’ is more prominently visible in private pensions, the role of the state in regulation and supervision is accepted to achieve social justice (Hyde & Dixon, 2009, p. 126; Bridgen & Meyer, 2009, pp. 135-136). The debate on reforms of pension systems focuses – following this line of argument – more on regulation, since the action of the state and non-state actors determine the situation and social rights of the (future) retirees (Ebbinghaus, 2011, p. 7; Bridgen & Meyer, 2009, p. 130). The regulation and governance of the (more or less) privatized pension systems will thus be crucial for the impact of the public-private mix.
The shifting balance in the public-private mix calls for a new typology of the different pension systems that exist in the empirical world (Ebbinghaus, 2011, p. 9). The most common typology is described by Immergut and Anderson (2007, p. 21) and is set up by the World Bank (1994). The outcome is a three pillar model, each pillar with several tiers. ‘Pillars’ indicate the sector in which the pension scheme is organized, ‘tiers’ are used to describe the different layers of income protection due to different functions that retirement can serve in society (Immergut & Anderson, 2007, p. 23). The difference with the broad theory of Esping-Andersen (1990), for example, is that this pillar system explicitly makes clear which institution is responsible for which part of welfare. In this way, the first pillar indicates public pensions, the second pillar is organized by employers and/or social partners and constitutes of occupational pension schemes, and the third pillar is the individual one. Then, the first tier assures a minimum of income, the second tier is earnings-related and the third tier is the ‘extra’ topping-up of individual savings and investments (Immergut & Anderson, 2007, p. 22). Supplementary pensions are all non-state pensions in addition to the public pensions that are mostly mandatory (Ebbinghaus, 2011, p. 9).
The categorization of pillars and tiers in practice result in different pension systems. The two main functions described in the previous part – poverty relief and maintenance of income – led to ‘two worlds’ of retirement income systems: Beveridge-type basic pensions and Bismarckian social insurance schemes (Bonoli, 2003, p. 400). Beveridgean pension systems are mostly found in the Nordic countries, with basic income protection as well as supplementary pensions that are related to earnings (that could be public or private). The Bismarckian tradition then is characterized by the earnings-related public pensions that cover most occupations, while the private influence is rather limited. Third, the ‘mature multi-pillar’ pension system is characterized by on the one hand public pensions for all citizens, and on the other reasonably advanced private pensions that are often funded (Ebbinghaus, 2011, p. 10; Ebbinghaus & Wiss, 2011, pp. 355-356).
Another categorisation of pension systems is in terms of the financing mechanism. Two institutional differences are predominant in the literature: on the one hand the ‘pay-as-you-go’ (PAYG) financing
12
and on the other hand the discussion between ‘defined-benefit’ (DB) and ‘defined-contribution’ (DC) benefit systems (Immergut & Anderson, 2007, p. 18). In the PAYG design of most public pensions, the current working population pay contributions that are transferred directly to the retirees, ensuring their pension. Demographic pressures put this traditional PAYG financing system under pressure (Ebbinghaus, 2015, p. 56). Simply put, in PAYG systems with an ageing population, more retired people have to be supported by less working people, so the ‘burden’ on the ‘strong and young’ increases. Indeed, scholars plead for a shift towards a partly capital-funded (a DB or DC design) pension architecture to ensure stability and affordability (Ebbinghaus & Wiss, 2011, p. 373; Ebbinghaus & Whiteside, 2012, p. 267; Bridgen & Meyer, 2005). Workers within the funded schemes pay contributions that are invested on the long-term, so that their money now will be their pension in the future. In a DB scheme, pensions depend on the annual income and the number of years of contributions; in this calculation, this benefit payments are set. In a DC scheme, the benefit is not determined; only the level of contributions is indicated beforehand. As the contributions are invested, the benefits are determined by the returns on invested contributions (Immergut & Anderson, 2007, pp. 18-19; Ebbinghaus & Wiss, 2011, p. 351). Changing to another financing system poses the problem of the ‘double-payment’ (Pierson, 1994), since the cohort in the transition period have to pay contributions to the current retirees (as the PAYG design prescribes) as well as contributions for their own pensions in the future (in DC schemes).
2.2.1 The stakeholders in supplementary pensions
Capital-funded pension systems generate principal-agent dilemmas. Following Moe, the principal-agent model is “an analytical expression of the agency relationship, in with one party, the principal, considers entering into contractual agreement with another, the agent, in the expectation that the agent will subsequently choose actions that produce outcomes desired by the principal” (1984, p. 756). This contractual agreement can originate from a lack of expertise or the extensive size of the task. However, the agent has his own interests at heart, beside the interests of the principal (thus: diverging interests). Difficulties emerge because of information asymmetries (or moral hazard (Dixit, 2002)): the actions of the agents might not be fully transparent and could be controlled by the agent. In the case that the information asymmetries are so great, principals are not on the driver’s seat anymore and cannot control their agents (Shapiro, 2005, pp. 263-267, p. 277). Moreover, agents want to minimize their costs in working for the principal, and are therefore risk-averse. As principals are more risk neutral, this poses some differences as well (Shapiro, 2005, p. 265).
In the case of pensions, principal-agency problems result from the fact that the pension beneficiary always relies on an agent that performs tasks for him or her, being the principal (Ross, 1973, p. 134; Dixit, 2002, p. 697). The advantages are superior investment skills, pooling of risk and economies of scale (Tonks, 2006, pp. 460-461). Some disadvantages, however, emerge as well: misplaced incentives,
13
multilevel principal-agency relations, lack of competence and information (Besley & Prat, 2005, p. 121; Hollanders et al., 2013, pp. 94-95).
The principal-agent theory applied to supplementary pensions are more visible because of the financial responsibility that is externalized to the independent pension funds. Internal organization of pension schemes are better protected from principal-agent problems since state regulation is direct and extensive (Ebbinghaus & Wiss, 2011, p. 367). Information asymmetries emerge as a consequence of the externalization: the beneficiary may not completely understand the investment issues involved and probably does not have full access to information on investments and related decisions (Laboul & Yermo, 2005, p. 502). On top of these problems, the boundaries of being an ‘agent’ or ‘principal’ are not that strict. An actor can be an agent and a principal at the same time, and these roles can vary over time (Shapiro, 2005, p. 267). An example is the vertical relationship between beneficiary, investor and pension fund manager which is known as the ‘double agency’ or multilevel agency problem. Here, on the one hand, the pension fund is the agent of the beneficiary, while on the other hand the pension fund plays the part of the principal of fund managers (Tonks, 2006, pp. 460-461, Lakonishok et al., 1992, p. 341).
In order to overcome the problems regarding the principal-agent relations and foremost the asymmetries in information, trust is extremely important. As Shapiro also states: “These fiduciaries face a problem as well: why would anyone ever trust them when their conduct is so unrestrained?” (2005, p. 278). Good governance, relevant regulation and information provision can also attribute to increase trust and lower the risk of mismanagement due to agency problems. Implemented institutional arrangements by the international community (the European Union, the OECD and others) are the installation of supervisory agencies, codes of good practice and information rights (Ebbinghaus & Wiss, 2011, pp. 353-354; Besley & Prat, 2005, pp. 132-133; Kakabadse et al., 2003, p. 376; Laboul & Yermo, 2005, p. 502). Another perspective is to extend the ‘voice’ mechanism of beneficiaries, in the way that they have participatory rights and have some degree of influence on the pension fund involved instead only having the ‘exit’ option of switching to another fund (Ebbinghaus & Wiss, 2011, p. 354). In addition to governance and regulation arrangement, supervision could be beneficial for the security of beneficiaries against agents that act in their self-interest (Ebbinghaus & Wiss, 2011, p. 371). These solutions – and mostly the governance arrangements – would guarantee that the promises made in the ‘contractual agreement’ are kept on the long term (Besley & Prat, 2005, p. 120). Thus, regulation could lower the information asymmetries and make sure that the participatory rights of the beneficiaries are incorporated.
14 2.3.1 Definitions and categories of SRI
Due to the challenges of demographic change, economic downturn and financial crises of 2001-2002 and 2007-2008 as described in the earlier parts of this chapter, participants and investors in pension funds have begun to pay more attention to (social) responsibility and democracy in the markets. This goes hand in hand with more focus on transparency and accountability of participants in those markets, since the confidence of the public in the financial markets dropped (Scholtens & Sievänen, 2013, p. 605; Clark & Hebb, 2004, p. 158). The underlying idea is that in today’s globalized world, the role of the state has changed drastically (and may be even weakened) as market actors such as private pension funds play an increasing role in decision upon citizens’ pension and pension governance. As Banerjee rightfully states: “If corporations are to carry out activities once the purview of governments then there is a need to examine the processes and outcomes of corporate involvement in political and social domains” (Banerjee, 2010, p. 265).
This increased awareness and changed public opinion of the market could be seen in the same realm of the growth of socially responsible investing (SRI) in the last decade. SRI is the inclusion of social, corporate governance, ecological and ethical criteria in making investment decisions for a variety of reasons. However, SRI is not a new phenomenon (Soederberg, 2009, pp. 212-213). SRI originated as a religious tradition in medieval times, but in the 20th century became more of a practical outcome of ethical and social principles of individual investors – such as opponents of the Vietnam War and protestants of the South African apartheid. Today, SRI has become more relevant due to ethical consumerism, where participants in the (pension) market pay a price for products that fit with their personal beliefs (Renneboog et al., 2008, p. 1725).
Proponents and opponents of SRI have such different perspectives of what exactly entails SRI. This makes it difficult to establish legitimacy and efficiency of the term (Sethi, 2005, p. 101). Often, scholars accept the definition as written down by the Social Investments Forum, which is the association of the SRI industry in the United States (or the European equivalent Eurosif):
“Sustainable and Responsible Investment (“SRI”) is a long-term oriented investment approach, which integrates ESG factors in the research, analysis and selection process of securities within an investment portfolio. It combines fundamental analysis and engagement with an evaluation of ESG factors in order to better capture long term returns for investors, and to benefit society by influencing the behaviour of companies” (Eurosif, 2018, p. 12).
Many definitions exist in the scholarly literature (as well as versions of the definition in each Eurosif report), differing in context (time and place), relevant elements and meanings of SRI. This becomes accurately clear in some key wordings of the term: ‘socially responsible investing’, ‘ethical investment’,
15
‘responsible investment’ and ‘sustainable investment’, to name a few (Capelle-Blancard & Monjon, 2012, p. 240; Sievänen et al., 2013, p. 138)2. Boatright (1999) has coined a definition that has been cited
often: investing which takes account of ‘people’ and the ‘planet’ (Scholtens & Sievänen, 2013, p. 605). Other scholars use multiple measures for SRI, as they distinguish between a ‘broad’ and a ‘narrow’ SRI (Hollanders et al., 2013, p. 83). Some critical perspectives neglect the ‘investing’ element in SRI, as they see it as a form of philanthropy, which is not the case: the focus is inaccurately on non-financial criteria (Sethi, 2005, pp. 100-101). The true essence of the concept is therefore contested: it involves multiple dimensions and serves the needs of multiple investors that are not a homogeneous group (Bengtsson, 2008, p. 969; Derwall et al., 2011 p. 2145; Woods & Urwin, 2010, pp. 2-3; El Ghoul & Karoui, 2017, p. 62). At this point of definitional ambiguity, Woods and Urwin (2010) rightfully ask whether SRI is an ‘essentially contested concept’ or a ‘radically confused’ one (as formulated by Gallie, 1955). The scholars conclude that SRI is radically contested, since the core of the term is contradictory: separate goals and processes exist. If this view is accepted, then this ambiguity in the literature can be acknowledged as well and the different aims may be more transparent and clear (Woods & Urwin, 2010, pp. 3-4).
Different aims imply different types of SRI. The first is negative screening: clearing the investment portfolio of poor ESG performance (mostly tobacco and alcohol sectors). Second, positive screening exists: seeking those proactive companies that perform well in creating positive external effects (Hollanders et al., 2013, pp. 84-85; Nofsinger & Varma, 2014, p. 182). This tactic is also known as ‘best-in-class’ (BIC) investment, assuming that greater responsibility will have the effect of higher value (Hockerts & Moir, 2004, p. 86; Bengtsson, 2008a, p. 164). Related is integration, meaning including ESG-issues in the investment decision-making process and analysis, to influence the social responsibility of firms and to optimize the risk-return balance (Sievänen et al., 2013, p. 140). Corporate engagement and stakeholder activism is a third category of SRI, consisting of shareholders acting, speaking and voting more actively on specific resolutions within the company to change it for the better. The ownership aspect of being a shareholder makes this possible (Scholtens, 2006, p. 25; Bengtsson, 2008a, p. 161; Clark & Hebb, 2004, p. 144).
2.3.2 The drivers of SRI
Internal (firm-specific) and external (environmental) factors can be distinguished that would be stimuli for SRI (Sievänen et al., 2013, p. 141). Often mentioned in the latter category is country-specific regulation (Cox & Schneider, 2010, p. 257; Bengtsson, 2008, p. 166; Renneboog et al., 2011, p. 564;
2A related concept to SRI is corporate social responsibility (CSR). In the literature, the two terms sometimes are used interchangeably, yet a difference is visible. CSR is directly related to the firm itself and to which level that firm incorporates environmental, social, ethical, and governance issues to its agenda, whereas SRI is indirectly related since SRI is at the level of the investor.
16
Sandberg et al., 2009; Scholtens & Sievänen, 2013). Haverland (2007) even argues that it is the most important element that determines the scope for SRI since path dependency complicates institutional change and harmonization (p. 887). Other external factors are economic openness, the size of the pension industry and cultural values (Scholtens & Sievänen, 2013; Sievänen et al., 2013). Examples of internal factors are improved reputation management (Clark & Hebb, 2005, p. 2021). According to the perspective of the firm (Jones, 1995), incorporating social initiatives may reduce transaction and agency costs, which is more a financial driver for SRI in the internal category.
Individual values and moral perspectives can also be a stimulator for more SRI. Research shows that investors care – to some extent – about certain non-financial elements in investing (Borgers et al., 2015, p. 113; Renneboog et al., 2008, pp. 1723-1724). In this way, investors derive utility from investing in a socially responsible manner, and react less heavily to gains or losses in comparison with investors who do not incorporate any SRI in their policies. So funds investing in a socially responsible manner are less sensitive to past returns than conventional funds who do not incorporate any SRI (Bollen, 2007; Nofsinger & Varma, 2014, p. 182; Renneboog et al., 2011, p. 563). This is in line with the Prospect Theory of Kahneman and Tversky (1979) which entails the idea that investors are more negatively affected by losses than they are positively affected by an improvement of similar size. Investors, then, are risk averse. Hence, they would opt for downside protection and willing to give up some return in periods of economic growth, since the utility curve in the loss section is steeper than the curve in the gain section (Nofsinger & Varma, 2014, p. 182, p. 185). This shows that investors who care about certain non-financial elements in investing derive some utility from this behaviour and therefore can be a stimulator for more SRI in the field.
2.3.3 Fiduciary duty
When applying SRI to pension funds, a relevant topic is the ‘fiduciary duty’ of pension funds. This duty functions a safeguard for participants in a pension fund so that they will be provided an old age income when they retire. Pension fund investors are obliged through this duty to ensure that they make the investment decisions that are in the best interests of the beneficiaries (Clark & Urwin, 2008, p. 38). Two important elements of fiduciary duty are loyalty and prudence: loyalty requires to keep the interests of beneficiaries in mind and prudence requires the investors to decide with care, expertise and diligence so that the ‘best’ investment decisions are made (Woods & Urwin, 2010, p. 13). Conventionally, the best interests are interpreted as narrow financial interests (Sethi, 2005, p. 99). In the case of firms performing well in the narrow financial sense, but badly in the face of negative externalities, what is then the best decision? Conventional fiduciaries would invest in the firm, while SRI proponents would not. Hence, according to conventional thought, SRI would legally not be possible since incorporating SRI practices would harm the fiduciary duty. Sethi (2005) criticizes this view and proposes more SRI (in fact; he states that it is the ‘only alternative’), since the scholar points out that the long term aspect of the fiduciary
17
duty is not often accounted for, as well as other factors of potential impact on the returns (Sethi, 2005, pp. 105-106).
In line with the fiduciary duty, pension funds in some countries (mostly Anglo-Saxon) have the legal obligation to act according to the ‘prudent person rule’ (PPR). The PPR is a flexible regulatory norm aimed at optimizing the quality of the investor of the pension fund; investors are expected to act with prudence to achieve the best investment outcomes. These prescriptions are mostly abstractly formulated. The PPR is opposed to few quantitative prescriptions that set on absolute boundaries and amounts on categories of investment (Hollanders et al., 2013, pp. 89-90). Under the PPR, investors have substantial freedom in strategizing their investments (Haverland, 2007, p. 891; Guardiancich & Natali, 2012, p. 308). This ‘prudent person rule’ is also reflected in the EU Directive on the Institutions for Occupational Retirement Provision (IORP). One could state that the prudent person rule is the common sense version of fiduciary duty (Clark & Urwin, 2008, p. 40).
2.3.4 Paying a price for doing good?
A second debated topic is the financial performance of SRI-based investment decisions. In other words: do you pay a price for doing good? Multiple works have been published that say that funds incorporating SRI do have lower returns than conventional funds. However, at least the same number poses evidence that shows the other way around or state that there is no significant difference in performance levels and returns (El Ghoul & Karoui, 2017, pp. 53-54; Capelle-Blancard & Monjon, 2012, p. 239; Dam & Scholtens, 2015, p. 114; Sethi, 2005, p. 106; Wen, 2009; Salomon, 2006, Renneboog et al., 2008). Roughly two opposing strands can be categorized, as Derwall et al. (2011) group them into two hypotheses. The first one is the ‘shunned-stock hypothesis’ and argues that socially controversial stocks have superior returns because SRI-based investors (also values-driven investors) literally ‘shun’ these stocks, resulting in lower demand for the stock and henceforth a lower price. The relative value therefore would go up (Derwall et al., 2011, p. 2138; Renneboog et al., 2008, p. 1734). Hong & Kacperczyk (2009) find precisely this: sin stocks have higher returns. However, this hypothesis is solely based on the effects of negative screening on returns, which is only a part of the whole SRI concept (Sethi, 2005, pp. 102-103). The second hypothesis proposes another mechanism, called the ‘errors-in-expectations’ (Derwall et al., 2011, p. 2138) or the ‘outperformance’ hypothesis (Renneboog et al., 2008, p. 1734). The reasoning here is that the market does not recognize the positive effects of strong CSR elements on future expected returns just yet, so that the actual value of socially responsible stocks is higher than the value that the market predicts. Value-relevant information is therefore crucial (Derwall et al., 2011, p. 2139).
It seems that the two hypotheses are diametrically opposed. However, these studies often only see one aspect or type of SRI and do not recognize the fact that SRI has become a multidimensional (and even
18
‘radically confused’ as described earlier) concept. Therefore, the two hypotheses may both be true as the effects relate to different forms of SRI. An example is that SRI funds incorporating negative screening perform not as well as funds that incorporate good governance elements and positive screening (Renneboog et al., 2008; Bollen, 2007; Nofsinger & Varma, 2014). Moreover, Nofsinger and Varma (2014) and Renneboog et al. (2011) argue that SRI funds perform better during crisis periods due to being less sensitive to risks (as those firms generally are stronger in governance and reputation). In this way, the ‘no result’ finding of studies can be the effect of both hypotheses cancelling each other out (Derwall et al., 2011, pp. 2145-2146). Dam and Scholtens (2015, pp. 114-115) solve the contradicting hypotheses (or ‘paradox’ as they call it) by stating that while returns may be lower, the market value is higher, if SRI is incorporated in the investment strategy.
This increased market value of SRI is exactly what Soederberg (2009) criticizes as Dam and Scholtens (2015) elaborate on how investors value the internalization of external effects. Soederberg argues that the neoliberal paradigm of the current debate neglects the whole idea of SRI to go beyond the narrow financial aspects in investments (2009, p. 216; see also Capelle-Blancard & Monjon, 2012, p. 246). Hence, the entire debate of financial performance and ‘paying a price for doing good’ forgets that the ultimate goal is to achieve social justice in the first place. The scholar states bluntly: “… the marketisation of social justice transforms human suffering into a cost-efficiency calculus and an exercise in risk management” (Soederberg, 2009, p. 212). Whereas the general opinion is that SRI can promote economic progress and social development (Sievänen et al., 2013, p. 140), Soederberg states that this is solely based on the neoliberal assumption that the market is able to resolve social issues (2009, p. 220). The market is not complete and information is not transparent. The general transition to more market forces and financialization in social matters has to be closely guided and structured by government to prevent citizens from harm (Soederberg, 2009, p. 226; Haverland, 2007, p. 887; Woods & Urwin, 2010, pp. 14-15). Here, we are back at the point where state regulation is favourable to ensure (a higher level of) social justice.
Taking into account both the theoretical discussion of the growing importance of state regulation in multi-pillar pension systems, as well as the growing importance of SRI in pension investment, I formulate the following hypotheses. First, I expect national legislation to be correlated with pension fund performance, following the statements of multiple scholars3. Extensive national legislation would
lead to a better performance of pension funds on the three indicators, especially the external legitimacy. Second, I expect that EU-legislation will lead countries within the EU to incorporate more SRI strategies to enhance their external legitimacy compared to non-EU members. Compliance with EU Directives
3 For example; Haverland, 2000; Haverland, 2007; Renneboog, Ter Horst & Zhang, 2011; Scholtens &
19
lead to an improved image of the country and facilitates integration with the rest of the EU. Third, I expect that the more stakeholders engaged in pension funds and investor practices (so more principals, agents and consequentially more dilemmas between the two), the more effort is put in civil regulation and ultimately leading to a higher level of external legitimacy. If stakeholders such as social partners (both labour unions and business) have extensive influence, the more pressure is put on investors to ‘do the right thing’ for society as a whole. I theorize that labour unions would want to create more equality in the whole society and therefore rewarding good behaviour of businesses (through more civil regulation and codes of conduct), whereas strong business stakeholders would want to strengthen the internal policymaking and corporate strategy of a firm by engaging in it. In this way, strong labour unions would advocate the ‘positive’ categories of SRI more (the ‘best-in-class’, norms-based screening and integration); whereas in the case of only strong business stakeholders, I expect the SRI category of ‘engagement’ to be large. In what follows, I will investigate whether these theory-based expectations applicable to mature multi-pillar pension systems will be in line with the empirical findings.
20 3. METHODOLOGY
A qualitative comparative study is performed to answer the research question of whether national and EU-level legislation leads to a better performance of national pension funds in terms of returns and spending on SRI. Such a comparative design combines in-depth within-case evidence with cross-case interpretation (Toshkov, 2016, p. 258). Besides, the goal is to uncover the effects of causes, namely the effects of national and EU-level legislation (Blatter & Haverland, 2012, p. 35). In this chapter, the methods used in this study are outlined. First, the relevant variables are defined and operationalized. Second, case selection and method of data collection are explained. Then, the method of covariational analysis is presented to elaborate on the method of analysis. Finally, the validity, reliability and scope for generalization is discussed.
3.1 The variables
3.1.1 The dependent variable
The dependent variable is the ‘performance of pension funds’. This variable captures three relevant elements. The first is the adequacy of the pension outcome, namely being able to provide the benefits promised and having a sound system in place. Then, the sustainability of the pension system is discussed, meaning the viability for future generations in light of (for example) demographic changes. The third element of the dependent variable is the external legitimacy of the behaviour of pension funds to society, captured by the extent and type of socially responsible investments (SRI). Each of the three elements are given a value of low, moderate or high. In this way, the dependent variable captures the ‘traditional’ view of delivering financial returns and therefore assuring old-age pensions in addition to the more ‘modern’ view of including externalities and investing in a more socially responsible and legitimate manner. In the following paragraphs, the operationalization of the three elements of the performance of pension funds is discussed.
Measuring adequacy
The most popular indicator to assess the adequacy of pension funds is the replacement rate. This indicator copes with measurement error, for example the assumption of uninterrupted working lives and only focussing on the first pillar of the pension system (EC, 2012, p. 13). This makes international comparison quite difficult (Fisher, 2010). The OECD, EU, Allianz, Mercer and other organizations have tried to capture at least a part of the overall adequacy of pension systems across the world, but caution should be placed on using these rankings due to the different definitions, measurement assumptions, and data availability (De Deken, 2013, p. 276; Peeters et al., 2014, p. 20). In particular, the theoretical indicator of the replacement rate is useful to compare pension schemes across countries, but not national
21
pension systems as a whole. This is a limitation of the research by the OECD (Peeters et al., 2014, p. 24). These quantitative rankings provide simple understanding of the complex world, which is in theory very beneficial. This process is called ‘commensuration’:
“Commensuration transforms qualities into quantities, difference into magnitude. It is a way to reduce and simplify disparate information into numbers that can easily be compared. This transformation allows people to quickly grasp, represent, and compare differences. One virtue of commensuration is that it offers standardized ways of constructing proxies for uncertain and elusive qualities. Another virtue is that it condenses and reduces the amount of information people have to process, which is useful for representing value and simplifying decision-making” (Espeland & Stevens, 1998, p. 316)
The risk of bias through ‘commensuration’ is decreased in this study through using multiple measures, indicators and critical scholars to provide the right nuance. I start with the indexes that are available, providing a starting point. Mercer is an organisation that publishes its index every other year. According to Mercer, the net investment return over the long-term is a crucial factor to determine whether or not such an adequate retirement income could be provided (Mercer, 2018, p. 6). Allianz developed another one, calling the compound indicator ‘RIA’: the Retirement Income Adequacy indicator (Allianz, 2015). This index captures the replacement rate, the coverage rate, elaborates on the enforcement (whether it is mandatory, voluntary or has the element of auto-enrolment), the capitalization level, the contributions by employers, non-pension wealth, spending needs and the transition from work (Allianz, 2015, p. 7).
Then I elaborate on the indicators that are used and if they are relatively strong or weak in explaining adequacy. Lastly I take all information together and try to give a nuanced image of the adequacy of the pension systems in the three countries. An example is the work of De Deken (2013) in building a new compound index with the focus on the private pillars. Four quantifiable indicators are taken into account: the theoretical replacement rates of public plans, the capitalization of the funded pension schemes, the expenditure on private pensions, and the coverage of private pension plans (De Deken, 2013, p. 270). Different career paths are taken into account to measure the replacement rate: he drew scenarios with replacement rates based on 50%, 100% and 150% of the average wage (De Deken, 2013, p. 278). Taking this and other critical researches into account, the adequacy score is determined. The low, medium or high score in terms of adequacy is a reflection of the rankings available (including the checks on these rankings as discussed above).
Measuring sustainability
Macroeconomic changes – such as increasing life expectancy and lower fertility rates – put pressure on the pension system to remain financially sustainable (Alessie et al., 2013, p. 308). Together with the two financial crises of the last decade, pension reform is a salient issue. Several elements are decisive for the long-term sustainability of national pension systems, such as the old-age dependency ratio (population of 65 and over as a percentage of the population until 65), the labour market participation,
22
the employment of older people (and the incentives or the lack thereof to increase this percentage), the effective exit age of workers and all other measures taken by the government to increase the sustainability of the pension system (for example, increasing the retirement age).
The sustainability of the pension system is often at odds with the adequacy of the system. What is more, increasing the sustainability often lead to a lower level of adequacy, for example in increasing the contributions for members of the scheme. In this way, the relative benefits are cut, whereas the future generation will profit from the increased contributions (EC, 2012, p. 10). Another example is the case of the perception of declining adequacy in pensions. If the public opinion thinks the pension system lacks in adequacy, social tensions will rise. Consequentially, reforms will not be popular, and the government will lose its credibility too. This all endangers the financial sustainability of the system, since the reforms would not be accepted. This makes clear that adequacy and sustainability need to be balanced, yet it is hard to do so.
To assess the sustainability, I will use the Pension Sustainability Index (PSI) as formulated by Allianz (2016). The RIA, as described in the previous section, captured the adequacy of a pension system to a large extent; the PSI tries to do the same in the perspective of sustainability for future generations. This PSI captures different elements of pension systems with different influential factors, policy changes and macroeconomic development. The focus of the PSI is on the first pillar, which directly points to the limitation of this index. Therefore, additional information on demographics, employment numbers and policy changes (of the OECD and the EC) provides a more complete image of the sustainability of the three cases of the Netherlands, the United Kingdom and Switzerland. The low, medium or high score in terms of sustainability is a reflection of the rankings available (including the checks on these rankings as discussed).
Measuring external legitimacy
The performance of pension funds is not only dependent on its adequacy and its sustainability, but also on its investment decisions. According to theory, investing in a socially responsible manner would lead to less negative external effects and potential costs for other companies or society as a whole (see for example Sethi, 2005; Derwall et al., 2011; Renneboog et al., 2008). Therefore, a pension fund performs better when it incorporates SRI in its investment strategy. The definition of SRI of the European Social Investment Forum (Eurosif) is used:
“Sustainable and Responsible Investment (“SRI”) is a long-term oriented investment approach, which integrates ESG factors in the research, analysis and selection process of securities within an investment portfolio. It combines fundamental analysis and engagement with an evaluation of ESG factors in order to better capture long term returns for investors, and to benefit society by influencing the behaviour of companies” (Eurosif, 2016, p. 9).
23
The indicator for this element is the sum of money spent on SRI. The data available (of Eurosif) to measure the SRI market, however, also captures a broader set of actors than only pension funds. The data of Eurosif compromises of both institutional investors (69.23%) as well as retail investors (30.77%) (Eurosif, 2018, p. 76). Moreover, pension funds only cover 39.4 percent of the institutional investment branch (EFAMA, 2018, p. 5). This two-way correction is applied to the data. Moreover, the size of the SRI market is compared to the size of the pension system and its assets under management. The low, medium or high score in terms of external legitimacy therefore depends on i) the size of the SRI market in general relative to other OECD countries as discussed in international rankings (of the OECD and the EC) and ii) the relative size of the SRI market compared to the total managed assets (to correct for the size of the system). According to OECD standards (2018), a ‘low’ score is where a country does not spend more than €50.000 million on SRI; ‘medium’ indicates a size between €50.000 million and €500.000 million; and more than €500.000 million spent on SRI is indicated as ‘high’. This score is also compared to the total assets managed of the pension system. A high score on total assets is when the number passes the USD 1 trillion (and the funded assets relative to the GDP higher than 100%); a medium score lies between USD 0.2 and 1 trillion (and the funded assets relative to the GDP between 50% and 100%); a low score is below USD 0.2 trillion (and the funded assets relative to the GDP below 50%) (OECD, 2018, p. 7). These scores are combined into the indicator of external legitimacy.
3.1.2 The independent variable
The independent variable is legislation (as well as regulation), both measured at national level and European Union (EU) level, which has an (direct or indirect) effect on national pension funds. Simply put, all relevant regulations, implementations in national legislation and EU-directives of the last eight years are taken into account (measuring from 2010 onwards). Formal legislation is the first indicator, regulatory activity by agencies is the second one, and the third indicator is civil regulation. Civil regulations are “private, nonstate, or market-based regulatory frameworks” to govern private institutions and networks (Vogel, 2010, p. 69). Public authority has no part in this; the legitimacy, governance and implementation of civil regulation results from actors outside the state. Global codes of conduct and voluntary directives are concrete examples of civil regulation (Vogel, 2010, p. 69). Each of the three indicators of national regulation receive a score of low (with no more than one legislative piece, no regulatory activity and/or no civil regulation regarding codes of conduct), moderate (with two legislative pieces, one relevant regulatory body and/or at least one relevant form of civil regulation) or high (with more than two legislative pieces, more than one relevant regulatory body and/or more than one relevant form of civil regulation).
24 3.1.3 The control variables
Beyond the dependent and independent variables as described above, the study includes some control variables that have (to a significant extent) similar values among the cases selected – the cases will be outlined below. The first one is the pension fund system as defined by Ebbinghaus (2011): all three cases are mature multi-pillar pension systems. Such a system has two key elements: public pensions that are available to all citizens and privately funded pensions that have developed significantly (Ebbinghaus & Wiss, 2011, pp. 355-356). The second control variable is the capitalization of the system. With this, the degree of funding relative to the GDP of the country is meant. In all three the countries, the capitalization of the second pillar is high compared to other OECD countries. Besides, the high capitalization of the second pillar is an indicator of a multi-pillar system. Third, stakeholder engagement is incorporated in the thesis. Stakeholders such as labour unions and business actors could possibly put pressure on politicians as well as pension funds to improve the behaviour of pension funds. Therefore, the context of a high level of stakeholder engagement could have an influence on the extent of (civil) regulation and ultimately on the behaviour of pension funds. In the Netherlands and Switzerland, the social partners (labour unions and employer organisations) have significant impact and are embedded in social policy institutions. Here, the level of stakeholder engagement is high. In the UK, however, only business interests are strong as business actors are large shareholders in pension funds. Here, the level of stakeholder engagement is lower and substantially different. In this way, the context variable of stakeholder engagement is an indicator of the influence of civil regulation.
3.2 The cases
In this research, a small-N approach is mostly suitable, since complex theories, several mechanisms and historical developments play a part in shaping pension fund policies. Hence, in-depth analysis is necessary to find an answer to the research question. However, case selection significantly shapes results, and therefore needs to be justified. The score of the control variable of the multi-pillar pension system determines the selection of cases in this study (Blatter & Haverland, 2012, pp. 42-43). This research comprises three cases, to be exact three welfare states in advanced political economies in Europe: the United Kingdom (UK), the Netherlands and Switzerland.
All three countries fall into the category of mature multi-pillar pension system, and have similar private pension plans with a high level of capitalization in addition to the basic public pension. Moreover, these three countries have the highest level of funding relative to other European welfare states. The universe of cases is therefore countries with capital-funded multi-pillar systems. The UK and the Netherlands are most similar cases, which will function as a paired comparison (Tarrow, 2010, pp. 243-244). Switzerland obviously is not a member of the EU, but will serve as a counterfactual: if a country had not been a member of the EU, what will the outcome then be? (Toshkov, 2016, p. 259) If it turns out that the outcome is the same along the three countries, the conclusion is that EU-level directives and regulations
25
do not have full explanatory power in explaining the performance of pension funds. Another contextual variable is stakeholder engagement, thus presenting another paired comparison of the Netherlands with Switzerland. Here, the stakeholder engagement is high (of both labour unions and employers institutions), in contrast to the UK were only business actors are strong.
Table 1: Variation across the three cases
Netherlands United Kingdom Switzerland Pension system type Mature multi-pillar
pension system Mature multi-pillar pension system Mature multi-pillar pension system Capitalization of second pillar
High High High
Stakeholder engagement High – of social partners Medium – of business interests High – of social partners
National legislation Medium Low Low
EU-level legislation Yes Yes No
Interestingly enough, the three countries have not been subject to a study regarding the three indicators of the performance of pension funds. Larger studies of big research institutes have been performed, of course, but always with all European (or EU) countries included (SIF, Eurosif, WorldBank). Other studies have placed a few countries in the spotlight, for example the Scandinavian countries (Bengtsson, 2008, 2008a) or the United States and the United Kingdom (Clark & Hebb, 2005; Cox & Schneider, 2010). The mature multi-pillar system is the most developed pension system with a diverse range of pension schemes: public, occupational and personal pension schemes. The regulation thereof is also diverse, since multiple actors (state actors and private actors) are involved. Especially, a lot of private actors are involved in the funding of pension assets (asset managers, pension funds managers, trustees, etc.). Current and future challenges will have an effect on these actors, possibly affecting pension fund performance as well. In this way, the study of mature multi-pillar pension systems is theoretically relevant: in incorporating diverse forms regulation (legislation, regulatory activity and civil regulation) as well as the incorporation of different actors. This study will present a new perspective regarding the countries chosen as well as the conceptualization of the dependent variable. The temporal boundaries will lie between 2010 and the present day, to study the effect of policies and the strengthening of the EU. The bigger issue at stake is therefore the effectiveness of regulation more broadly. Then one can answer: does regulation have the possibility to be a (positive) influence on the market of pension funds?
3.3 The methods used
The method of data collection is document analysis, based on both primary sources and secondary sources. Primary sources such as historical documents (on for instance (inter)national conferences) and legal documents (in which the national and EU-legislation has been laid out) are used. Secondary sources such as research reports will support the findings in the primary sources. Examples of institutes
