Experts versus representatives? Financialised valuation and institutional change in financial governance

ABSTRACT Valuation devices and institutional change are key aspects of financialisation, but have been largely separated in the literature. This article explores the causal link between both concepts. We draw on the linked ecologies framework to argue that financialised valuation, such as market valuation of assets and liabilities, creates opportunities for financial experts to influence policy processes. Making use of their institutional position, experts frame policy problems to advance further financialisation as a solution. So doing, they perform essential boundary work to tie authority over financial knowledge to their social position as financial market actors. The result is a process of endogenous institutional change, whereby financial experts increasingly displace other actors in governance arrangements. We illustrate this argument providing a case study of pension fund governance in the Netherlands. There, financialised valuation requirements led to a crisis of pension funding, to which policymakers responded by strengthening experts’ positions on pension fund boards, thereby eroding century-old corporatist institutions. We show how the infrastructural entanglements between financial markets and the state in the Dutch pension system not only benefited financial actors’ institutional positions, but also point to limits of financial agency as the state increased its own capacity over a formerly corporatist domain.


Introduction
This article explores the causal link between financial valuation practices and institutional change. When actors introduce financialised valuation practices that value assets or liabilities based on current financial market prices, exposure to financial market volatility increases. This process is known as financialisation (Mader et al. 2020). Financialisation serves as an opportunity for financial experts, who claim to translate risks into profits, to strengthen their position within the marketplace (Pernell et al. 2017). Crucially, these developments are not isolated from financial regulation and governance: we argue that the growing importance of financial market volatilities increases dependency on financial expertise in policy processes. This allows financial experts to build alliances across policy and governance arenas, using financial devices to make sense of financial dynamics, define problems, ascribe responsibilities, and develop solutions (Hirschman and Berman 2014). We draw on the linked ecologies (Abbott 2005) perspective in order to show how financial valuation tools bring together financial experts and governments by 'defining issues and developing strategies across professional groups' (Seabrooke and Tsingou 2015, p. 4) Financial experts are not neutral arbiters of technical knowledge, but seek to further their own position within institutional arenas ('ecologies') for financial regulation and governance (Seabrooke and Tsingou 2014). Importantly, we argue, the elevation of financial expertise within these arenas fortifies a false dichotomy between the expertise ascribed to financial actors and those representing particular political constituencies. Sociologists of professions have used the term professional closure (Abbott 1988, Boussard 2018 to describe processes in which emerging professional groups claim occupational niches, based on the definition of new problems to which they promise effective solutions. The new problems created by financialised valuationensuring profits and liquidity in the face of volatile financial marketsthus allow financial experts to claim jurisdiction over financial governance and are key to their boundary work. From a political economy perspective, the closure of financial governance has profound implications, as it forges endogenous institutional change (Streeck and Thelen 2005). As financial experts gain authority, other actors such as elected politicians or organised interests may be pushed to the side lines. This is particularly important in coordinated market economies (Hall and Soskice 2001), where organised interests play a central role in economic governance. While institutional scholars have stressed the resilience of national-level institutions against macro-level financial pressures (Carruthers 2015), we argue that financialisation can erode institutions by altering governance from within.
We illustrate this argument through a case study of occupational pension governance in the Netherlands. With total pension assets of around €1.7 trillion, the Dutch occupational pension system is one of the largest in the world. Although legally considered a Defined Benefit (DB) pension system, pension levels are de facto tied to pension funds' financial market performance, as indexation has been made contingent upon funding ratios (Van der Zwan 2020). However, the Dutch pension system is also highly coordinated. Not only do social partners (employers and labour unions) bargain collectively over the contents of the pension plan. They are also jointly in charge of administering these plans through occupational pension funds (Anderson 2019, Sorsa andVan der Zwan 2021). Dutch pension funds are characterised by representational governance, whereby pension fund boards are composed of representatives of employers, active members and retirees.
At first sight, the Dutch pension system presents a case of institutional resilience in the face of growing financial dependencies. Our study of recent changes to Dutch pension governance, however, reveals considerable weakening of representational governance and the rise of independent financial experts. In 2013, the government launched four new governance models that reduced the scope of representational governance by employers and labour unions and instead strengthened the position of independent financial experts. Our research reveals profound changes within pension fund boards: as independent financial experts increasingly occupy seats that formerly belonged to employer or union representatives, the pension system has seen significant endogenous institutional change (Streeck and Thelen 2005).
To explain this outcome, we show how the introduction of mark-to-market valuation requirements for pension assets and liabilities in the early 2000s created unforeseen problems. When both share prices and interest rates fell following the 2008 Great Financial Crisis (GFC), Dutch pension funds' shrinking financial assets were met by growing pension liabilities as future pension entitlements had to be discounted using rapidly falling market interest rates. To make sense of the crisis, policymakers invited financial experts to offer problem diagnoses and policy solutions. Using mainstream financial theory to make sense of the crisis, experts and policymakers did not see the system's performance problems as the outcome of pension financialisation, but argued for a stronger role for financial experts to weather the storm of market volatilities.
This article contributes to scholarship on valuation and institutional change by investigating how changes in valuation requirements affect the social positions of professional groups. Whilst pension funds' increasing financial market engagement per se did not harm representational governance institutions, it was the state's introduction of mark-to-market funding ratios that made pension outcomes dependent on financial market volatilities. The growing dependenceor infrastructural entanglement (Braun 2020)of pension policy on financial markets created novel opportunities for financial experts to increase their influence in policymaking as the crisis hit. Financial experts used this opportunity to advocate for a closure of their professional group, gradually pushing representational governance to the margins.
The finding that financial valuation requirements, through their effects on professional groups, can foster endogenous institutional change has important implications for our understanding of the interdependencies between states and financial markets. On the one hand, the institutional transformation of the Dutch pension system represents a threat to democratic input and output legitimacy (Dorn 2014, Nölke 2020. As financial experts grew in power and proximity to the government, social partners increasingly lost their voice and alternative forms of pension valuation became difficult to envision. On the other hand, however, these developments also point to limits of financial agency. Rather than a 'colonisation' (Chiapello 2015) of the state by finance, the Dutch case points to the empowerment of the state vis-à-vis a previously self-regulated domain. Social partners agreed to the stronger role for experts in the hopes of maintaining their independence from the state in pension governance. Paradoxically, their agreement strengthened the role of the state, as it used the pension sector's infrastructural entanglements with financial markets to strengthen its regulatory capacity in the pension sector.
In the following section, we review the literature to carve out the link between financialised valuation, expertise, and institutional change. Section 3 describes the explanandum by mapping the new institutional features of Dutch pension governance. Section 4 then describes the policy trajectory leading to these changes, while section 5 explores how financialised valuation has been translated into governance. The article ends with a discussion and conclusion.

Financialised valuation, expertise, and institutional change
Scholars of financialisation have identified expertise as an important power resource held by financial actors and organisations. In addition to the financial sector's lobbying (Pagliari and Young 2016) or structural power (Culpepper and Reinke 2014), the supposed technical and complex nature of financial issues gives those who possess financial knowledge an advantage over other organised interests. As described for economic experts more broadly (Fourcade 2009, Christensen 2018, the possession of expertise has provided financial actors with privileged access to policy-making arenas and influence over policy outcomes. Financial actors' institutional position (Hirschman and Berman 2014, p. 781) in policymaking or elite networks offers an important conduit for financial power. While these dynamics have been studied widely in relation to banking regulation (Baker 2015) and fiscal and monetary policy (Wansleben 2013, Riles 2018, they have received less attention in social policy (for exceptions, see Dowling 2017, Golka 2019. Investigating the connection between finance and policy, scholars of political economy are increasingly resorting to Abbott's (2005) linked ecologies framework (e.g. Seabrooke 2014, Seabrooke and Tsingou 2014, Helgadóttir 2021. The linked ecologies perspective is an expansion of Abbott's (1988) earlier sociology of professions. Ecologies denote social arenas in which professional actors vie for jurisdiction over tasks. Within ecologies, professional groups gain influence by acting on tasks through the knowledge they claim. For example, financial experts may claim a seat on pension fund boards by mobilising their expertise in financial risk management. Seabrooke (2014, p. 56) uses the term epistemic arbitrage to describe how professionals gain advantage by 'positioning particular forms of knowledge as the most appropriate to deal with particular problems.' One important aspect of ecologies is that dynamics in the realm of knowledge are tied to changes in social positions. As Abbot (2005, p. 246) points out, jurisdictional success of one professional group creates new openings or defeats for adjacent professional groups. By succeeding in the struggle over task jurisdiction, professionals thus perform essential boundary work that helps differentiate themselves and exclude others from their institutional positions (Boussard 2018). Abbott (2005) describes two strategies for establishing links between ecologies that actors mobilise to help their professional struggle: hinges and avatars. Hinges are shared projects between ecologies that reward actors in their respective ecologies. For example, Farrell and Quiggin (2017) describe how macroeconomic policy became a hinge issue between economics and policy that defined both prestige within the economics profession and policymakers' success within the policy domain. Hinges are important sources of collaboration and competition across ecologies, as they affect actors' positional gains within their respective ecology. Avatars, by contrast, are reproductions of one professional group within a different ecology to 'benefit their claim to jurisdictional control' (Seabrooke 2014, p. 53). For example, financial actors may build an avatar in the economics profession to provide academic justifications for mark-to-market valuations rather than historical prices, which, in turn, strengthens their jurisdictional claims in economic governance (Seabrooke 2014). As the financialisation literature has shown, alliances between financial actors and governments, on the one hand, and academic justifications by financial economics, on the other hand, have been important enabling factors for rise of finance in economic governance (Lazonick and O'Sullivan 2000).
However, the linked ecologies perspective has not yet acknowledged that financial knowledge or expertise is not just possessed by the actors themselves, but also embedded in the instruments and practices they employ. Recent years have seen a growth in studies on calculative devices and valuation practices (MacKenzie 2006, Besedovsky 2018. Valuation studies have shown how the adoption of financialised practices and instruments have led to the colonisation of previously non-financialised realms (Chiapello 2015). Financialised valuation is 'equipped by models, instruments and representations belonging to the explicit knowledge (…) of finance professionals' (Chiapello 2020, p. 83). Financialised valuation may serve as a performative engine that shapes former non-financialised realms according to financial models (Lockwood 2015).
While valuation scholarship has produced key insights into financial sector transformations, it has been less concerned with the impact of those transformations on actors' institutional positions. We argue that this is an important gap in the literature, because the interplay between professional positions and valuation devices is key to understanding both professional opportunities arising from new forms of valuation and the processes behind the performativity of devices. In what follows, we combine insights from political economy and financial valuation scholarship. In particular, we show how the introduction of market-based devices to value pension assets and liabilities created a 'manufactured' crisis (Wood and Ausserladscheider 2020), whereby the performance problems experienced by Dutch pension funds after the GFC were largely induced by the valuation device itself. The GFC was not the type of existential crisis that policymakers made it out to be: temporary reductions in investment returns were quickly recovered and the Dutch pension system continued its earlier trajectory of achieving records levels of growth. However, the opening the crisis provided to financial expertise had enduring effects.
Using the linked ecologies perspective, we argue that hinges and avatars linking financial experts and policy were key to reinforcing pension financialisation as a crisis response (Wiß 2019). Financialised valuation devices served as a hinge between financial experts and policymakers as pension plans' funding ratio quickly became the shared benchmark of professional success for both pension fund managers and policymakers. Instead of questioning financialised valuation during and after the GFC (Best 2016), policymakers rendered it part of the 'fiction' of the crisis (Carstensen 2013), further increasing their demand for financial knowledge to make sense of new uncertainties. Increasingly dependent on financial expertise, policymakers invited economists to make sense of the crisis and propose a way out. This created an avatar of financial expertise in pension policy, which translated financial knowledge into a powerful analytic frame that impacted problem diagnosis and the development of new policy instruments (Palier 2007). As a result, policymakers set in motion a broad overhaul of pension governance, whereby the only solution to the perceived financial instability was a further entrenchment of financial expertise in the Dutch pension system. We argue that the rise of financialised valuation and expertise in pension governance should be seen as an instance of institutional displacement (Streeck and Thelen 2005, p. 5), whereby institutions change by replacing existing structures with new organisational forms and practices. The introduction of financialised valuation and independent financial experts marked a profound case of endogenous institutional change: the changes introduced to Dutch pension governance were not only state-led (on state-led financialisation, see Krippner 2011), they also gave the state new policy instruments by which to adapt prevailing governance practices to its own agenda. Here, the Dutch case speaks to emergent scholarship in political economy on infrastructural power (Braun 2020, Trampusch andFastenrath 2021), whereby the infrastructural entanglements created by market-based governance serve as power resources for state and financial actors alike. However, institutional change did not stop here, as financial experts used their new institutional position to advocate for a closure of financial expertise in pension governance, that gradually pushed representatives of the social partners to the side lines.
Representational pension fund governance: institutional resilience or financialisation?
The Netherlands has a long-standing tradition of member representation on pension funds. The first regulation on occupational pensionsa Royal Decree from March 31, 1908mandated employee representation on pension fund boards for contributory pension plans. Employees were entitled to have a voice in pension governance, the legislator reasoned, because they had no choice in contributing their own deferred wages to these plans. The legislator applied the same institutional logic in the 1952 Pension and Savings Fund Act (PSW), which followed 1947 legislation that made participation in occupational pension plans mandatory for most employers and employees. Article 6 PSW stated that industry-wide pension fund boards required an equal number of representatives from employer associations and labour unions. On company fund boards, equal representation of employer and employee members was required. By the late 1990s, moreover, the Labour Foundation (Stichting van de Arbeid) and several elderly organisations agreed to include retiree representatives to advisory bodies within the pension funds, the so-called participation councils (deelnemersraden). In the 2007 Pension Act, the position of retirees was strengthened even further by including them on pension fund boards, thereby sharing half of the board seats with employee representatives.
The representational logic of Dutch pension governance was partly abandoned with the introduction of the 2013 Law to Strengthen Pension Fund Governance (Wvbp) that made four new governance models available to pension funds. 1 These models differ from the traditional paritarian governance model in three respects (see Table 1). First, pension funds were allowed to implement boards staffed with independent (i.e. non-representative) experts instead of representatives. Importantly, funds were given the opportunity to add independent experts even to representational boards. Second, a pension fund may have one of two kinds of advisory bodies. The accountability body (verantwoordingsorgaan, VO) represents active members, retirees and sometimes also employers and former members. The stakeholder body (belanghebbendenorgaan, BO) represents unions, employer associations and retirees. Both bodies are similar, but only the BO holds the additional right to give formal approval to key strategic decisions regarding, for instance, contributions levels. Third, internal supervision of the pension fund board is organised either by a mandatory supervisory board, by external evaluators, or by non-executive board members that are either independent or representatives of the social partners. Table 2 presents an overview of the prevalence of each of the five governance models in the Dutch pension system. For this overview, we have collected information on all 212 pension funds listed in the official register of the pension regulator, the Netherlands Central Bank (DNB), in 2020. We collected information on all funds' board composition by consulting their websites and/or the most recent annual report available (2019 or 2020). Of the 212 funds, 30 were in liquidation or had merged with other pension funds. These were omitted from further analysis. We were unable to access information for three pension funds.
At first sight, the introduction of the new governance models has not threatened the resilience of representational governance in the Dutch pension system. Eight years after the introduction of the Wvbp, more than two thirds of Dutch pension funds still use paritarian models of governance, whereby decision-making power in the board is shared among employer and beneficiary (active members or retirees) representatives. The paritarian model is the most popular in large, industrywide funds that usually feature strong sectoral bargaining from organised labour: of the 20 largest pension funds, 17 have a paritarian model of governance. These funds represented a total of 9 million active and retired members ultimo 2020 (DNB 2021a), compared against a total adult population of 12,5 million people (it is possible to be a beneficiary of multiple pension funds).
On closer look, however, the introduction of the Wvbp has substantially strengthened financialised governance in the Dutch pension system. First, nonrepresentational governance models are increasingly gaining momentum. In 2014, top-5 pension fund PME was one of the first to switch to a one-tier board, while the Netherlands' largest pension fund ABP (Algemeen Burgerlijk Pensioenfonds) will similarly switch to a nonrepresentative model in 2022. Moreover, a growing number of company funds are merging into multi-company funds, which usually feature nonrepresentational models.
Second, the number of financial experts has increased for pension funds with paritarian governance models, as funds increasingly make use of the option granted by the Wvbp to introduce additional independent financial experts to pension boards. Modelled after executive corporate boards, these independent members are thought to strengthen the position of pension funds visà-vis financial intermediaries, such as asset managers. As independent members are not listed in Dutch national pension statistics, their specific number has to date been unknown. However, our data show that 33 out of 120 (27 per cent) paritarian boards have independent members. Independent board members are particularly prevalent among the larger pension funds: 30 out of 50 largest pension funds have them, compared to only 28 in the bottom 100. This means that the rise of financialised governance is not confined to small funds, but takes place across the whole spectrum. Third, the role of financial expertise has been substantially strengthened through national regulation. In 2012, DNB issued a new Policy Rule on Suitability that requires all board members, internal supervisors and BO members to certify their financial expertise with the financial regulators. VO members, who have less formal authority, have no formalised expertise requirements. As the Policy Rule holds that expertise requirements must be met before candidates assume their positions, the Dutch Pension Federation advises funds to establish 'breeding pools' of suitable candidates (Pensioenfederatie 2014). To do so, candidates from multiple organisations undergo joint training in cohorts, often with specialised providers such as SPO. This means that financial expertise is defined by functional rather than organisational boundaries, while the carriers of said expertisefinancial professionalsincreasingly share professionalisation experiences.
In short, the implementation of the Wvbp has led to a weakening of representational (i.e. by the social partners) governance of Dutch pension funds. In contrast to theoretical expectations, the 'governed interdependence' (Rethel 2020) between financial experts and the regulator has rendered them central in Dutch pension funds in three different ways: through the emergence of new independent governance models, through the introduction of financial experts on paritarian boards, and by the tightening of suitability requirements for pension fund board members. As a result, representational governance has been gradually dislocated to more peripheral positions within pension funds.

Financialising pension governance
Why did the Dutch legislator abandon a century-old institutional logic of representational governance solely through employer, employee and pensioner representatives? We argue that this displacement has been an unintended consequence of the introduction of financialised valuation requirements. Until the Global Financial Crisis hit the Dutch pension sector between 2008 and 2010, the development of occupational pensions can best be described as coordinated financialisation. Pension financialisation refers to 'the growing role of pension markets and financial actors in the provision of old-age pensions' (Van der Zwan 2020, p. 2). In the Dutch context, financialisation has proceeded in a coordinated fashion in the sense that increasing financial market dependency coincided with a deepening of representational governance (Anderson 2019), as described in the section above. Embedded in strong institutions of representational governance, the past century has seen three discernable waves of pension financialisation.
A first wave of pension financialisation in the Netherlands saw the rise of pension funds that were still largely shielded off from financial market volatilities. Growing participation in occupational pensions during the postwar years provided an important impetus to the increased centrality of pension funds to the Dutch political economy. After the introduction of quasi-mandatory industry-wide occupational pensions in 1947, the (nominal) value of Dutch pension assets grew from an estimated NLG 193 million (€ 87,58 million) in 1952 to NLG 69.3 billion (€ 31.4 billion) in 1979. Despite their increased financial importance, however, pension funds continued to use financial conventions (Chiapello and Walter 2016) that detached them from financial market volatility. Until the late 1980s, pension funds sought to match long-term pension liabilities with equally long-term investments with minimal risk. As a result, the majority of pension assets was invested in fixed income products such as loans or mortgages, whereas equities rarely made up more than 10 per cent of pension fund portfolios (Van der Zwan 2017).
Fuelled by the 1990s stock market boom and increasing pressure to internationalise, pension funds underwent a second wave of financialisation, as they increasingly engaged in financial markets. The share of equities in pension fund portfolios grew from less than 10 per cent in 1986 to over 40 per cent in 1997. Rather than matching long-term liabilities with predominantly fixed income assets, pension funds became increasingly invested in volatile financial markets. This development is epitomised in the transformation of public sector pension fund ABP from a public lender to the Dutch government to a private financial actor in the 1990s: within only seven years, its share of corporate equities rose from 8 per cent in 1993 to 39 per cent in 2000 (Van der Zwan 2017).
A third wave of pension financialisation resulted from the introduction of new financial regulation between 2005 and 2007. After negotiations with the social partners, the Dutch regulator introduced new solvency requirements for pension funds in 2003 (Van der Zwan 2019). These requirements were formalised in the 2007 Pension Act. Key to the new rules was the introduction of a required funding ratio of 105 per cent of liabilities that included pension entitlements, buffers, projected indexation, and fund management expenses. As pension funds were required to halt indexation or even cut benefits if funding requirements were not met, pension levels were directly tied to perceived funding levels. Thus, a hinge between pension policy and financial markets had been created.
Pension funds' susceptibility to financial market volatility further increased with the introduction of a new regulatory financial framework ( financieel toetsingskader, FTK) in 2007. The FTK required funds to expand financial market valuations (so-called 'fair value') to both assets and liabilities. As Knafo (2021) argues, transformations of liability management are a crucial yet often overlooked dimension of financialisation that tends to focus on the asset side and on dynamics in equities markets. This is particularly true for DB pensions where pension funds face significant liabilities that often span far into the future. To assess the extent to which future liabilities are covered in the present, pension funds discount future liabilities. As the FTK required pension funds to estimate discount rates based on current market interest levelsrather than a fixed discount ratefalling rates directly translated in a potentially significant increase in pension liabilities and hence a reduction in funding ratios. Introducing a dual dependency of both asset and liability values on financial markets, the FTK essentially equated pension funds' funding ratios with financial market performance and thus opened up a key pension policy variable to instantaneous financial market swings. The consequences materialised imminently as the stock price crash was complemented by a significant drop of market interest rates in 2008/9. 2 As average funding ratios dropped from 152 per cent in mid-2007 to 92 per cent in early 2009 (DNB 2021b), numerous pension funds had to take the unprecedented measure to cut pension benefits.
The sharp decline in funding ratios and reduced pension entitlements following the GFC led to two main legislative developments. First, market dependency of pension outcomes introduced uncertainty and volatility into the realm of pension policy. This represented an opportunity for the rise of experts in the policy process, as epitomised by the Uncertain Security report presented by an expert commission headed by former ABP board member Jean Frijns. This development will be discussed in detail in the following section. Second, the sudden drop in funding ratios led to a recalibration of valuation requirements that was introduced with the revised FTK in 2015. In order to make pension funds less susceptible to short-term market shocks, market valuation of liabilities was smoothed to a twelve month floating average. As the regulator imposed its view somewhat against the will of the pension sector that would have favoured a further relaxation (Willis Towers Watson 2018, p. 6), the persistently low interest rates continue to pressure funding ratios. This has led to the paradoxical situation where pension funds are seen as underfunded, although their assets have more than tripled from €550 billion in 2009 to €1.68 trillion in 2020 (DNB 2021b).

Transposing financialisation: from valuation to governance
The idea that independent experts should support pension governance rose to prominence with a 2005 report by the Dutch Labour Foundation (Stichting van de Arbeid), that sought to strengthen pension governance by introducing distinct functions. One of these proposed new functions was so-called internal supervision (see section 3) meant to ensure the quality of pension fund decision-making (Stichting van de Arbeid 2005, pp. 6-11). To strengthen this function, the Foundation suggested a pool of independent experts, which smaller funds in particular could access for technical expertise to ensure effective long-term risk management. By doing so, they would contribute to self-regulation of the pension sector and thus minimise regulatory interference (p. 11). In its 'inventory on good governance principles', the government's main socio-economic advisory body, the Socio-Economic Council (Sociaal-Economische Raad), largely agreed with the suggestions of the Labour Foundation (Sociaal-Economische Raad 2009, p. 28).
The idea to strengthen financial expertise in pension governance gained significant momentum during the GFC. As pension funds had to take the unprecedented step to pause indexation or even cut pension benefits, the crisis of financial valuation quickly translated into a political crisis. Dependent on financial expertise, the Balkenende IV cabinet (2007-2010) commissioned two expert committees to investigate the causes of and possible solutions for the perceived crisis of pension funding following the GFC (Tweede Kamer der Staten-Generaal 2008-2009). The first committee was led by professor of financial economics and former ABP executive Jean Frijns. The Committee Frijns focused on investment policy and risk management in Dutch pension funds. The second report focused on the financial sustainability of occupational pension contracts. It was commissioned to a group of university professors, led by pension economist Kees Goudswaard. Both committees published their reports in January 2010 (Tweede Kamer der Staten-Generaal 2011-2012).
The committees developed their own narrative on the causes of the crisis (Palier 2007), thereby creating an avatar of financial expertise in pension policy. Authored by pension fund managers and economists, the causal analyses presented in the two reports corresponded closely with mainstream thinking in financial economics. The Frijns report mentioned the 'positive and disciplining' (2010, p. 7) effects of the financial valuation techniques introduced with the 2007 FTK reform. To cope with the increasing vulnerability to financial market volatility that these forms of valuation entailed, the report argued that board members should solve the agency problems within their investment chains by gaining a 'grip' on their outsourced fund management (p. 3). The Goudswaard report added to this diagnosis by proposing a transfer of pension funds' new financial risks from employers to participants, as it considered increasing pension contributions as harmful to the competitiveness of the Dutch economy (2010, pp. 3-5). In other words, rather than questioning the destabilising effects of the FTK reform, both committees advised a reinforcement of the existing policy trajectory as a way out of the crisis.
The causal analysis presented by both expert committees had important ramifications. The Frijns report in particular played a major role in setting the cabinet's legislative agenda, as it deemed it 'necessary' that 'at least two experts' sit on the board to form a 'countervailing power' against financial consultants and asset managers (2010, p. 49). Stating that financial expertise had to be demonstrable, the commission added that such expertise could not be obtained by simply following 'a number of training courses' (Frijns 2010, p. 49). For this reason, funds should have independent experts on pension boards and not recruit financial experts from among the stakeholder groups. To achieve this goal, legislative action would be needed to allow pension funds to add independent experts to existing representational boards, and to create additional governance models that allowed for the possibility for boards staffed entirely by independent experts. These new governance models would effectively end the dominance of paritarian boards and remove representational governance from some pension fund boards entirely, transferring it to a yet-to-be-created representational body within the fund (pp. 49-50).
It is noteworthy how the Frijns report neither problematised the power of asset managers nor proposed the upskilling of representational board members. Even where the report did problematise pension funds' engagement with other financial actors, such as 'disappointing' (p. 39) returns and the 'unnecessarily' high implementation costs of pension funds' active investment strategies (p. 43), these concerns did not address the (infra-)structural dependencies of pension funds on financial chains (Arjaliès et al. 2017). Instead, the proposal gained traction to add independent financial experts as the new captains for pension funds, wholike diligent seafarershad been embattled by a financial storm beyond their control. Here, experts mobilised 'strategic ignorance' in order to 'assert expert control' (McGoey 2012, p. 555) as they deliberately ignored the fact that the financial storm had actually been made much more severe for pension funds due to the financial valuation techniques introduced by the regulator.
The experts' financialised interpretation of the funding crisis directly informed the government's legislative agenda both in substance and in justifications. The introduction of independent experts and the new governance models were two cornerstones of the Wvbp. The cabinet legitimised these legislative changes with a dedicated section on the implications of the commissioned reports in an explanatory memorandum presented to Parliament (Kamerstuk 33182 nr. 3,. Quoting the Frijns commission, the government argued that independent experts in pension fund boards would present a 'countervailing power' vis-à-vis asset managers (p. 4). Moreover, the government explicitly cited agreement from the DNB, which had argued for strengthening financial expertise in the face of growing financial market complexities and uncertainty (DNB 2012, p. 1) In the parliamentary and policy debate following the first draft of the Wvbp, the notion of experts and expertise as safeguards of stable pensions was largely accepted and barely problematisedeven where it displaced representational governance. In its advisory statement on revisions to the Wvbp draft, the Socio-Economic Council (2012) held that moving representational actors to nonexecutive board roles was an 'efficient and effective' solution (p. 11). The depoliticised discourse on expertise also translated to the parliamentary debate of the Wvbp: epitomised in a parliamentarian's assertion that 'nobody can be against an increase in expertise', the shared perspective in the debate was that the locus of financial expertise should move from supervisionwhere it was located in the 2005 Labour Foundation reportto executive boards, as defined in the Wvbp (Tweede Kamer der Staten-Generaal 2012-2013). Despite some minor cautioning, there was an implicit consensus that allowing independent experts on the board would be a fruitful way to strengthen expertise in fund management and thus improve the pension system overall.
Within a small number of years, financial experts have witnessed a remarkable professional closure. From the initial suggestion to introduce them in the rather peripheral role of internal supervisors, the hinge created between financial expertise and government through the financialised valuation devices rendered the crisis into an opportunity for the rise of experts. While their positional gains were defined with the Wvbp, the certification of financial expertise through the DNB promoted a closure and professionalisation of financial experts. Thus, the false dichotomy between financial and representative actorspropelled by the experts' avatar in pension policy and epitomised by the Frijns committee's words that expertise cannot be obtained 'by following a number of training courses' (p. 49)was formally institutionalised.

Discussion and conclusion
In this article, we have pointed to the link between a financialisation of valuation and broader institutional change. Drawing on the linked ecologies perspective (Abbott 2005), we have shown how the regulatory requirement of financialised valuation techniques for Dutch pension funds served as a hinge between financial experts and government, and presented a professional opportunity for financial experts once the crisis hit. Pointing to the role of valuation devices as professional opportunities is an important addition to recent scholarship in political economy that is increasingly concerned with professional dynamics (Seabrooke 2014, Helgadóttir 2021. The outcome of these dynamics is an important case of institutional change: at first sight, the relatively slow adoption of the new governance models seemingly points to the resilience of corporatist institutions of representational governance. However, the growing importance of independent experts within representational boards instead points to a rather robust case of endogenous institutional change as representational governance is gradually displaced (Streeck and Thelen 2005).
Yet, the threat of financialisation to representational governance is not solely that it elevates financial expertise over other forms of knowledge. This is an important finding from combining financialised valuation with the linked ecologies approach. What is at stake here, is that financial experts use financial valuation requirements as an opportunity for professional closure (Boussard 2018) by gaining support for their argument that financial expertise belongs to some (financial) actors, but not to (non-financial) others. Such an assumption presupposes that expertise is not the same as knowledge, which can be gained through education or experience. It assumes that that expertise is a quality that belongs to actors, thanks to their structural position within the political economy. In other words: expertise belongs to financial actorsand not representational actorssimply because they are situated within the financial system. For the case of Dutch pension governance, this represents an important infringement of the institutional logic that characterised the Dutch pension system since 1908. Being pushed from boards to accountability bodies and supervisory functions, employers and beneficiary representatives effectively lose decision-making power (Ebbinghaus and Wiß 2011). As a result, the rise of expert governance undermines the input legitimacy of the Dutch pension system.
Our case simultaneously draws attention to the pivotal role played by the state. From the introduction of the new financialised valuation techniques by the regulator between 2003 and 2007 to the Wvbp itself, the state expanded its regulatory authority over a sector with strong corporatist institutions. Whereas pension fund regulation used to be marked by a hands-off approach that prevented interference with self-governance by the social partners, current regulation addresses nearly all financial and non-financial aspects of pension governance. Crucially, the definition of financial expertise is now beyond the remit of social partners and instead with the regulator. This finding corresponds with recent scholarship on the infrastructural power of finance (Braun 2020): while financial experts have been able to strengthen their position within pension governance as the system's dependencies on financial markets increased, they also personified an increase in state authority.
The transformation of Dutch pension policy and governance following the manufactured crisis elicited by the financialisation of valuation is also important with regards to the conflicted realm of finance and democracy (Block 2014). The technicalisation of pension funds' financial management has increased the 'cognitive affinity' (Nölke 2020, p. 430) between regulators and the regulated. It has thereby reduced democratic output legitimacy of the pension system, as some policy options (e.g. altering pension funds' value models) are no longer available. Here, our findings differ from other scholarship that points to institutional inertia (Baker 2015), resilience of financial systems (Carstensen 2013) or even a questioning of financial expertise and metrics (Best 2016) in the wake of the GFC. To the contrary, we expect a lasting entrenchment of financial experts as the underlying infrastructural dependencies persist. This can be seen during the recent COVID-19 crisis where pension funds' average funding ratios dropped from 102.1 per cent in Q4/2019 to 95 per cent a year later, even though asset levels increased by about 7 per cent (DNB 2021b). As the perceived funding crisis is exacerbated by historically low interest rates, the importance of financialisation on the liability side for financial power becomes visible (Knafo 2021).
In sum, we have explored the causal link between financialised valuation and institutional change, thereby bringing meso-level sociological analyses into conversation with the macro-level political economy of financialisation (Karwowski et al. 2020). Additional research is needed to explore this relation further. This includes investigating whether similar processes can be found in other countries with stronger or weaker forms of representation in pension governance. Further research should also explore the link between financial valuation, experts, and governance dynamics in other institutional settings. These cases could include national settings such as commercial or development banking, as well as transnational ones such as asset management and investment funds. If indeed financial and economic experts or models perform financial markets and the wider economy (MacKenzie 2006), then we need to gain a better understanding of the political conditions that allow them to do so.

Notes
1. There are some particularities for company and professional funds, which will be omitted here due to their lower importance in the pension sector.
2. For example, 10-year market interest rates in the Netherlands nearly halved from 4.73% in June 2008 to 2.52% in September 2010 (DNB 2022).