The Second Time as Farce? The EU Takeover Directive, the Clash of Capitalisms and the Hamstrung Harmonization of European (and French) Corporate Governance


  • The author would like to thank the Garnet Network of Excellence ‘Global Governance, Regionalization and Regulation: the Role of the EU’ for the ‘mobility award’ funding a research sabbatical at CERI, Sciences-Po, Paris in Spring 2007 which enabled the fieldwork to be carried out. He would also like to thank the Department of Politics and International Studies at the University of Warwick for granting study leave and providing financial research support.

Ben Clift
Department of Politics and International Studies
University of Warwick
Gibbet Hill Road


This article focuses on the EU Takeover Directive and its transposition into French law. French outcomes diverge from European Commission aspirations for greater clarity and uniformity. The clash of European capitalisms as well as heightened uncertainty and differentiation in takeover regulation exacerbate problems of asymmetric vulnerability of EU states (and firms) to the European Commission's liberal reform agenda. This explains the failings of EU-level harmonization of varieties of capitalism and corporate governance.


Originally proposed in 1989, the European Commission (EC) has long aspired to a liberal harmonization of European takeover regulation (and a common EU takeover code), seeing it as crucial to the achievement of an integrated European capital market by 2010. In the 2004 version of the Takeover Directive, the EC harmonization agenda took a new turn, seeking to create a ‘level playing field’ across the EU. However, just as there are diverse national varieties of capitalism within the EU, so there are varieties of takeover markets (Jackson and Miyajima, 2007), underpinned by different institutions, corporate governance norms and ownership patterns.

The market for corporate control (and regulation surrounding it) is potentially a highly significant mechanism of change within corporate governance (and capitalism) (Clift, 2007). This is why European Commission attempts at takeover harmonization (along liberal, UK-influenced ‘Anglo-Saxon’ shareholder value lines) have provoked what Callaghan and Höpner (2005) call a ‘clash of capitalisms’, crystallizing into ‘contending conceptions of the internal market’ and ‘contending conceptions of capitalism in Europe’. The resultant ideological struggle pits ‘the neoliberal project’ against ‘regulated capitalism’ and Europe as an ‘organized [economic] space’ (Hooghe and Marks, 1999, pp. 74–9 and 82–91; Wincott, 2003, pp. 292–4). Normative contestation surrounds the relative import of these elements and explains the failings and limitations of the Takeover Directive detailed below.

Differing national ‘supply side’ takeover market conditions are deeply entrenched, institutionally embedded in national legal cultures and frameworks and long-established firm practices. This makes the ‘level playing field’ a forbiddingly difficult public policy goal. The patchwork quilt of different corporate governance and takeover market institutions and norms generates an ‘asymmetric vulnerability’ (Knudsen, 2005, p. 524) of firms and national corporate governance regimes to attempts at shareholder, value-oriented, EU-level regulatory change. This generates political tension, fuelling the ‘clash of capitalisms’.

This article focuses on the European Takeover Directive, its long gestation process and the impact of its transposition into national (specifically French) law, assessing the impact of EU-level initiatives relative to the ambitions of reformers. After the failed 2001 attempt, the EU Takeover Directive was finally agreed in April 2004. However, during its legislative passage, it encountered impediments arising from the asymmetric vulnerability of different European varieties of capitalism, and specifically of different Member States and firms, to its provisions. As a result, its key elements were made optional, meaning the new regime lacked ‘teeth’.

The French model is employed here as a test case in assessing the relative importance of the EU regional level, and more specifically the EC, as an actor inducing harmonization within varieties of capitalism and corporate governance. The French case was selected because its takeover regime was, perhaps surprisingly, already aligned with a UK City Code and its open, liberal regulatory framework. This ‘goodness of fit’ with the EC model (the Takeover Directive being inspired by the City Code approach) augured well for harmonization (see Featherstone, 2003, pp. 15–17; Radaelli, 2003, pp. 44–6; Börzel and Risse, 2003, pp. 59–61). However, in the Takeover Directive case, whilst the EC did have a significant impact on European takeover regimes, it did not induce the desired liberal harmonization. Rather, paradoxically, French transposition involved departures from pre-existing liberal norms.

Recent comparative capitalisms work has emphasized the heterogeneity within national models of capitalism (Deeg and Jackson, 2007, pp. 155–6; Crouch, 2005). Deeg and Jackson note the ‘differential adoption of “old” and “new” business practices according to sectoral and firm-specific characteristics' (2007, p. 155). The hard law optionality of the Takeover Directive, now enshrined in national law, allows Member States (or firms) to ‘opt in’ to new practices, creating the potential for increased differentiation between firms within ‘national capitalisms’. In relation to takeover regulation and corporate governance, the optionality renders more significant the mediation of capital market integration through national and firm-level corporate governance structures, institutions and norms. This generates increased diversity of outcomes, which can only exacerbate the asymmetric vulnerability of Member States (and their firms), which have hitherto hamstrung the harmonization of EU corporate governance.

I. EU Takeover Regulation, Corporate Governance and the Clash of European Capitalisms

Varieties of Capitalism identifies two prevailing ideal-types of capitalism, the liberal (LME) and the co-ordinated market economy (CME) (Hall and Soskice, 2001, pp. 1–70). Corporate governance is a crucial element within any variety of capitalism, a ‘nexus of institutions’ (Cioffi, 2000, p. 574) through which governments, state and firm actors organize and regulate corporate and economic activity. Its impact on actors' behaviour, norms and incentive structures shapes how capitalism works in any given setting. The role of financial markets and their relation to corporate governance is significantly different within Hall and Soskice's two ideal types. These map onto the comparative political economy distinction between ‘shareholder’ and ‘stakeholder’ capitalism (Kelly et al., 1997). CME and LME varieties of capitalism thus ‘provide a broader institutional context within which stakeholder and shareholder models of governance can be analysed’ (Vitols, 2001, p. 338).

Within the LME ideal-type, Hall and Soskice emphasize ‘competitive market arrangements’ to co-ordinate firm activities, and more specifically, ‘highly competitive markets’ to ‘organize relations’ with firms’‘suppliers of finance’ (Hall and Soskice, 2001, pp. 8–9). This is market-based firm financing, reliant on capital markets rather than bank- or institutionally-based processes. In CMEs, by contrast, ‘non-market’ interactions, institutions, relationships and ‘modes of co-ordination’ are crucial. Firms rely on these non-market oriented networks to ‘co-ordinate their endeavours’. Long-termist relations with banks and other private, public or para-public financing institutions are characteristic CME modes of firm financing. These lead to ‘network monitoring based on the exchange of private information inside networks’ (Hall and Soskice, 2001, p. 8). The cross-shareholdings which formerly sustained the ‘protected capitalism’ logic (Schmidt, 1996) of France's ‘financial network economy’ (Morin, 2000) are typical CME mechanisms.

These two different approaches to the firm financing/corporate governance nexus have different implications for the likelihood, prevalence and impact of corporate takeovers. This is why the apparently technical issue of takeover regulation is of great political economic significance because, at root, different approaches to takeovers reflect different models of political economy. Takeover regulation is a battleground in the clash of capitalisms within European corporate governance. Broadly speaking, to the extent that market mechanisms can operate unimpeded, the LME approach prevails. To the extent that states and firms are able to introduce or rely on impediments to a free market for corporate control, a more co-ordinated, regulated or ‘institutional economy’ (Crouch and Streeck, 1997) approach prevails.

The potential for takeover, in particular hostile takeover, brings market discipline to bear on the behaviour and operations of management (Manne, 1965). Below par performance can be punished in the capital markets by takeover bids, rewarding shareholders with higher share prices and sanctioning poor management (often the first to go in subsequent restructuring). Thus a freer functioning market for corporate control in theory results in increased emphasis on ‘shareholder value’, the core organizing principle of ‘Anglo-Saxon’ LME capitalism (Sternberg, 1998, pp. 34–5). In the UK case, the liberal City Code means the threat of hostile takeover has an important impact on firm and executive behaviour.

Thus much hinges on the presence (or not) of impediments to the free play of takeover market forces. There are two broad (not necessarily mutually exclusive) categories of impediments to takeovers: ‘pre-bid’ and ‘post-bid’. One set of pre-bid takeover defences are clauses in the internal constitutions of companies which insulate management. These include differential shareholder voting rights such as ‘golden shares’, multiple votes, voting ceilings and limitations on certain investors' voting rights. Perhaps the most notorious takeover defence is the ‘poison pill’, which involves ‘rights or warrants issued to shareholders that are worthless unless triggered by a hostile acquisition attempt. If triggered, pills give shareholders the ability to purchase shares from, or sell shares back to, the target company’ (Monks and Minow, 2004, p. 236). This raises (perhaps exorbitantly) the cost of a bid and can derail hostile takeovers. These mechanisms empower ‘insiders’ at the expense of ‘outsiders’ and insulate incumbent management. It is to counter such insider protection that the liberal principle of ‘one share, one vote, one dividend’ is espoused within Anglo-Saxon influenced calls for ‘good’ corporate governance (OECD, 1999 and 2004) as a battering ram to allow takeover market forces to penetrate these protective defences. More generally, patterns of concentrated (as opposed to dispersed) ownership hinder takeovers, so share buy-backs can operate as anti-takeover devices as firms seek to put more of their shares in their own hands.

Post-bid takeover defences involve measures that target boards can take to derail a bid once launched. The range of ‘frustrating actions’ include divestitures (selling off or ‘locking up’ prized assets to make the target less attractive), share buy-backs, the search for friendly ‘white knights’ to buy up significant stakes and the issuing of new shares (with unequal voting rights) to friendly shareholders. The liberal principle of ‘board passivity’ seeks to outlaw such attempts to undermine the free market for bids. In similar vein, ‘mandatory bid’ rules clarify takeover processes by requiring a bid for the whole company to be launched once a controlling right is acquired. This prevents ‘creeping control’ and protects minority shareholders. These different approaches to takeover regulation rest upon shareholder (LME) and stakeholder (CME) models of capitalism.

Thus the ‘clash of capitalisms’ explains the political economic causes of divergences between the EC and some Member States over specific takeover regulation proposals. Facilitating hostile takeovers at the EU level, if successful, would augur transformation of the nexus of European corporate governance institutions, towards LME shareholder value norms. Höpner and Schäfer have recently argued that European integration has entered a new phase in which it ‘systematically clashes with national varieties of capitalism’ but also ‘asymmetrically targets’ the institutions of CME stakeholder capitalism, with the result that ‘political resistance in the organized economies leads to a crisis of political integration’ (2007, p. 6). The findings here offer some support to Höpner and Schäfer's thesis. The politics of EU takeover reform involved a ‘clash of capitalisms’ logic (Callaghan and Höpner, 2005) of defence of particular national approaches to corporate governance and takeovers from the perceived threat of LME-oriented re-regulation on the part of state actors in more CME-oriented economies (for example, France, Germany and Sweden). This illustrates how European initiatives present opportunities to policy actors in their import and mediation by institutions, governments and national politics.

However, whilst we agree with Höpner and Schäfer that the ‘clash of capitalisms’ is refracted through the institutions of EU economic governance, we would question the degree of efficacy, coherence and internal consistency of ‘European integration’ of corporate governance posited in their account. It is erroneous to talk of a‘European’ corporate governance reform agenda, either in relation to takeovers, or more broadly. A wide range of views, rooted in different models of political economy, generate ongoing normative conflict about appropriate corporate governance reform. As Wincott points out, much Europeanization literature is ‘too ready to identify the EU – ontological – level as ordered, coherent and consistent, providing a clear basis from which to develop claims about “Europeanization” ’ (2003, p. 300). Any assumption of coherence and shared understandings underpinning ‘the EU level’ in relation to takeover reform is misplaced.

The EC is one key player, but policy elites from many Member States also play important roles. Amongst this multiplicity of actors, there is no accepted, agreed ‘European’ approach to corporate governance reform. Rather, diverse elements endure within the process of European integration, generating ongoing political struggles between, at root, contending models of political economy. Thus the corporate governance regime within the European Union is ‘always still in formation, built through political contests and struggles’ (Wincott, 2003, p. 300). The differential outcome of this ‘clash of capitalisms’ (Callaghan and Höpner, 2005) in different areas of corporate governance is all part of the variable geometry of Europeanization.

II. The Sisyphean Task of European Corporate Governance Harmonization

The scale of EC corporate law interventionism is impressive, with nearly 50 directives or regulations brought in since 1968. Its impact is much more debatable, with the desired harmonization of European corporate governance proving elusive. Much EC legislative effort has struggled to make its relevance felt within national corporate governance regimes. As Lannoo and Khachaturyan argue, ‘the more [the EC] tried to harmonize “corporate governance”[. . .] the less successful they were’ (2003, p. 5). The ‘triviality thesis’ highlights ‘under-enforcement’, ‘sporadic enforcement’, ‘sporadic’ judicial interpretation by the European Court of Justice (ECJ) and ‘parochial interpretation’, such that ‘one can doubt whether anything really worth calling EC corporate law exists “off the books” ’, not least because ‘the Commission has traditionally lacked the resources to monitor Member States’ compliance with corporate law directives' (Enriques, 2006, p. 12; Enriques and Gatti, 2006, p. 5).

Enriques argues that ‘most EC rules can be categorized as optional, market-mimicking, unimportant, or avoidable’ (2006, pp. 1–2 and 6–11). European company law ‘tends to be implemented and construed differently in each Member State, according to local legal culture and consistently with prior corporate law provisions’ (Enriques, 2006, pp. 1–2). These ‘nationalistic tendencies in the interpretation of EC corporate law’ mean that ‘the prevailing interpretation of any given directive in each jurisdiction is, wherever possible, an interpretation compatible with the existing legal culture’ (Enriques, 2006, pp. 17–19). Halbhuber concurs, identifying how national ideational and institutional structures ‘filter European legal materials’, rendering it unlikely that EC corporate law ‘means the same for lawyers in different Member States’ (2001, p. 1385).

There are some qualifications to the thesis (Enriques, 2006, pp. 44–5), notably the ‘sporadic’ ECJ activism has given some rules meaningful impact, such as ECJ rulings on Centros and freedom of establishment with implications for a European ‘Delaware Effect’ and increased regulatory arbitrage (see Höpner and Schäfer, 2007). The ECJ ruling against France over its ‘golden shares’ in Elf-Acquitaine is another case in point (European Court of Justice, 2002). Overall, however, there is substantial empirical support for the Triviality thesis. Enriques identifies ‘major instances of implementing rules that are clearly at odds with the text of the directives’ which ‘can be found throughout the EU’ (2006, p. 13).

Discerning the degree to which corporate governance reform is ‘European’ in origin is a useful analytical exercise, but it has to contend with a co-existence of national and European processes of company law and takeover reform. The waters are muddied further by ‘hindsight bias’, where EC corporate law regulation merely restates already existing regulation at national level (Cheffins, 1997; Enriques, 2006, pp. 20–3). In France, for example, much of the ground covered by the Takeover Directive was already subject to comparable national-level regulation. Of course, the ‘European’ reform agenda is informed by national policy elites rooted in the experience of the corporate governance practitioners. The influential ‘high-level corporate governance working group’ (of 2001–02) and the subsequent European Corporate Governance Forum were sites of a two-directional travel of ideas.

Indeed, whilst a potentially useful analytical distinction, the compartmentalization of ‘national’ and ‘EU’ level does not make sense here in practice. Comparative capitalism and corporate governance scholars have noted how national corporate governance regimes are today ‘institutionally incomplete’ because ‘national institutions are becoming overlain by a growing set of European and international institutions’ (Deeg and Jackson, 2007, p. 155). Thus a multi-levelled governance of corporate activity prevails, which spans the regional level (such as EU takeover regulation) and the global level (such as the World Trade Organization), as well as the national. With increased cross-border flows, activities and actors, European corporate governance only makes sense if understood as a patchwork of national regimes. Relations between these national regimes are crucial to the 2004 EU-level attempt at takeover regulation harmonization, since the EC model of a ‘level playing field’ for transnational corporate activity (in this case takeover deals) rests on an elusive concept of ‘reciprocity’ (see below), which only makes sense at the inter-national level.

In this instance, these analytical problems do not present insurmountable obstacles. The EC (notably the internal market directorate general) has a well-established takeover regulation agenda, as laid out in its draft proposals for 2001 and 2004 Takeover Directives. This ‘European’ takeover agenda, modelled on the UK City Code, is rooted in ‘one share, one vote’ and, more broadly, shareholder value corporate governance norms. The analysis below thus assesses the degree to which these LME shareholder value-oriented preferences prevail in the attempted harmonization of European takeover regulation.

III. European Commission Corporate Governance Harmonization and the Market for Corporate Control

European Commission enthusiasm for liberalizing takeover regulatory reform dovetails with commitments to establish a fully integrated European capital market. The Commission's desired harmonization of European takeover regulation seeks to ensure an unimpeded market or corporate control in Europe, based on an assumption (for which there is at best equivocal support, see e.g. Clarke, 2006, p. 361; Jackson and Miyajima, 2007, pp. 19–21) that takeovers enhance efficiency. The Commission thus takes an explicit LME position within the European clash of capitalisms, aligned with the ‘Anglo-Saxon’ OECD good corporate governance agenda (Commission, 2003, pp. 12–14). In 2007, the Commission accentuated the positive aspects of takeovers as ‘efficient drivers of value creation’ which ‘facilitate corporate restructuring’. The recognition that takeovers ‘are not always beneficial for all (or any) of the parties involved’ was relegated to a footnote (2007, p. 3). Assuming that takeovers deliver efficient restructuring, the Commission's liberal aim has been to dismantle the defences described above that insiders use to derail takeovers.

The 2001 draft directive was viewed in some quarters as embodying an excessively ‘Anglo-Saxon’ view of takeovers as a necessary discipline against inefficient management which ignored valued ‘stakeholder’ and CME-oriented European corporate governance traditions. Thus, conflicting models of capitalism (LME versus CME) were the root cause of the Takeover Directives' troubled passage. Pre-bid defences, such as multiple and double voting shares and concentrated ownership, were recurrent sources of disagreement between the Commission and Member States, as well as between Member States themselves. The reason was the different conception of the firm financing/corporate governance nexus which these mechanisms and institutions embodied. For the same reason, the liberal principle of board passivity, and the UK City Code-style limitations the directive sought to place on defensive measures by target boards subject to hostile takeover attempts (Article 9), generated resistance. Outlawing these (or allowing them only with shareholder approval) went against the grain of the ‘protected capitalism’ logic of some European CMEs. There were also concerns about insufficient protection of workers in target firms.

Furthermore, Germany felt particularly vulnerable having engaged in reforms facilitating the development of a market for corporate control (abolishing multiple voting shares) in the 1998 KontraG Law (see Cioffi, 2002). Given that the directive did not force all others to follow suit in abolishing multiple voting shares, or remove other defences such as golden shares, Germany felt that the 2001 Takeover Directive failed to deliver a level playing field (Knudsen, 2005, pp. 510–11). With Ford eyeing up VW at the time, the German Government sought to prevent more hostile takeovers of its national champions, as had happened with Vodafone-Mannesmann in 1999.

The first attempt at the Takeover Directive was rejected (just!) by the European Parliament on 4 July 2001 in a dead-heat vote of 273 to 273. The ‘clash of capitalisms’ was the underlying cause of rejection (Callaghan and Höpner, 2005, pp. 313–15). The stakeholder preferences of many Member States and European firms were directly threatened by the EC's shareholder value approach. The EC's not too subtle (and failed) attempts at ‘divide and conquer’, seeking to isolate Germany did not help. This episode demonstrated the formidable obstacles to coercive harmonization towards a single European Corporate governance framework. It was the Commission's last heroic effort to ‘hard wire’ a harmonized European corporate governance regime. Its failure is of course wholly consistent with the triviality thesis. In the face of the demonstrable inefficacy of the Commission as a coercive harmonization-inducing corporate governance actor through ‘hard’ law, the strategy shifted to a soft law approach. The 2003 Corporate Governance Action Plan begins by recognizing the need for national actors to ‘reconcile’ the reform process to their ‘domestic circumstances’ (Commission, 2003, p. 1). The Commission consistently recognized the need to be ‘flexible in application’ of the programme (Commission, 2003, p. 4).

The Commission convened a ‘high level’ working group of corporate governance and European company law experts, headed by Jaap Winter, to chart progress towards ‘good’ corporate governance within the EU. More specifically, they were charged with developing measures to achieve the new priority of the ‘level playing field’ (one of the issues which torpedoed the 2001 Directive), and drafting new Takeover Directive proposals. The Winter group's report assumed takeovers to be efficiency-enhancing, and emphasized shareholder decision-making, and their new concept of proportionality between risk-bearing capital and control. The working group's recommendations adhered to ‘one share, one vote’ norms as an integral element of good corporate governance, and recommended abolishing multiple voting rights (Knudsen, 2005, p. 519). The group thus took an ideological position on one side of the ‘clash of capitalisms’ divide described above. The LME shareholder value-oriented approach was later characterized by Bolkestein as promoting ‘best corporate governance practice’ (Bolkestein, 2004, p. 3).

The new recommendations were, in fact, more radical than the earlier defeated measures, arguing strongly for their particular conception of a ‘level playing field’. The idea was to ensure equivalent safeguards for shareholders in all European public listed companies. Furthermore, arbitrary differences in governance structures across the EU should not distort the corporate restructuring process (Clarke, 2006, pp. 355–6). Yet the proposed solution did not remove the problem of asymmetric vulnerability of particular firms or Member States. Rather, some argued, the new proposals ‘would hit dominant shareholders and incumbent managers around the EU unevenly, prohibiting some structural defences while leaving others untouched’ (Enriques, 2006, p. 63).

The Group's recommendations fed into another proposal presented by the Commission in October 2002 (Commission, 2002). This new Directive's guiding principles were that shareholders get equal treatment, are given sufficient information and time to reach an informed decision, and that the target board acts in the interests of the company. The dissemination of information to employees of the target firm was also required. Mandatory bid provisions were included, with thresholds defined at the national level.

The new proposals still contained the controversially LME-oriented Article 9 ensuring board passivity (preventing post-bid frustrating actions). Seeking to overcome pre-bid defensive institutional engineering designed to impede hostile takeover, the revised proposal in autumn 2003 included a new ‘Breakthrough rule’ (Article 11). Designed by the Winter group, this established a threshold of 75 per cent of the target's shares; ‘once a bidder successfully reaches a threshold of shares in accordance with national company law, the new majority can annul the structural defensive devices’ (Knudsen, 2005, p. 519; Menucq, 2006, p. 229).

The ‘Breakthrough rule’ would have the liberalizing effect of neutralizing pre-bid defences including multiple voting rights, voting ceilings and restrictions on transfers of securities. The aim was to create a successful takeover market in Europe despite differential capital and control structures across European countries and firms. The rule applies ‘one share, one vote’ even to multiple voting shares in the shareholder annual general meeting (AGM) or extraordinary general meeting (EGM) which decides on defensive measures (as per Article 9) and the first post-successful bid shareholder meeting decides the new structures and constitution of the company (Clarke, 2006, p. 367). ‘Breakthrough’ was the battering ram the shareholder value enthusiasts used to breach the defences of European CMEs with concentrated ownership and ‘stakeholder capitalism’ norms.

Hostility to the ‘Breakthrough rule’ was particularly intense in parts of Scandinavia, especially Sweden. There, the A & B dual stock tradition (with A stock carrying many times more voting rights than B stock) was seen by Swedish MEPs as integral to the Swedish model of capitalism (Knudsen, 2005, p. 519). Furthermore, the Swedes pointed out, the new proposals were not likely to improve progress towards a level playing field because the proposals exempted France's double share tradition (since these were not a separate class of share). The new proposed Directive proved unacceptable to Member States and agreement could not be reached in the European Council meeting on 19 May 2003. The ‘one share, one vote’ principle was a major stumbling block (Ipekel, 2005, pp. 342–3) because it went to the heart of the differing LME and CME conceptions of corporate governance.

The global level playing field issue remained contentious, with French and German negotiators pointing out that US boards retained a range of defensive measures including ‘poison pills’ (Monks and Minow, 2004, p. 236). The proposed Takeover Directive, in depriving European firms of defence mechanisms, would not only undermine the coherence of co-ordinated capitalism, but also skew the playing field against European firms. Again Germany argued that because ‘one share, one vote’ and a lack of takeover defences was not the norm everywhere, the ‘playing field’ would be unfair. As Clarke notes, ‘the ghost of the hard-fought hostile takeover of Mannesmann in Germany may have haunted the negotiating table’ (2006, p. 374; Knudsen, 2005, pp. 510–12).

The European Parliament finally adopted a watered down Takeover Directive 321–219 on 16 December 2003. Crucially, it makes the key Articles 9 (board passivity) and 11 (‘Breakthrough’) optional for Member States. Furthermore it introduced the notion of reciprocity regarding board passivity and ‘Breakthrough’. ‘Reciprocity’ is the clumsy concept (of questionable practicability) designed to address the level playing field issue. This new concept in EU company law, set out in Article 12(3), specifies that ‘Member States may, under the conditions determined by national law, exempt companies [from applying board passivity and/or the Breakthrough rule][. . .] if they become the subject of an offer launched by a company which does not apply the same Articles as they do, or by a company controlled, directly or indirectly, by the latter’ (Commission, 2004). It falls to national securities regulators to adjudicate the equivalence of measures, and very little guidance is offered, generating a lack of clarity about how reciprocity will be interpreted by Member States.1

The Winter report assumed that both Articles 9 and 11 would be required in tandem to create a level playing field; neither on its own would deliver that objective. The optional nature of these elements in the final Directive means that the ‘level playing field’ remains unattained. Reciprocity potentially addresses the lack of a level playing field between European and American firms. However, the Directive is unclear whether reciprocity applies in relation to non-EU bidders (Menucq, 2006, p. 233; Clarke, 2006, p. 373). The Directive generates uncertainty for bidders as to which restrictions apply in particular cases by enabling individual firms to opt back in to board passivity and ‘Breakthrough’, even if ‘their’ Member State opts out. Firms may thus ‘opt in’ to a more liberal regime voluntarily, for example, to facilitate takeovers where they are bidders.

The new Takeover Directive came into force in April 2004, with transposition into national law required within two years. The optionality and à la carte nature of the Directive stripped it of its desired harmonizing effect. It fell far short of the Winter group's ambitions for convergence on the LME model, facilitating as it did the enduring diversity of European capitalism in relation to corporate governance and takeovers. Commissioner Fritz Bolkestein recorded his hostility, regarding the directive as a testament to national protectionism which he deemed not worth the paper it was written on (Callaghan and Höpner, 2005, p. 311).

IV. Harmonizing European Takeover Regulation? The Impact of the Takeover Directive in the French Case

Despite being an example of co-ordinated or statist capitalism (Schmidt, 2003) with long-established dirigiste norms of state intervention (Shonfield, 1965; Hayward, 1973), the pre-existing French takeover regime was remarkably liberal. Crouch has rightly pointed to ‘the diversity of economic institutions’ within capitalism, with empirical cases including forms drawn from different capitalist types (2005, p. 440). This aspect of the French case illustrates the point well. Since the late 1980s, the principles guiding the legal regulation of takeovers have been aligned closely with UK norms – including board passivity (unlike in many European countries).

The reason for this is the neo-liberal turn in French politics in the 1980s (Jobert and Théret, 1994). The 1980s saw extensive liberalization of the French model of capitalism, the state-orchestrated liberalizing re-regulation of the French financial system and a revitalization of French stock markets (Cerny, 1989; Clift, 2004). The creation of a Second Marché of unlisted securities and the development of a commercial paper market allowed companies to raise capital directly from the public through private bond issues. Takeover regulation underwent similar liberalization. From 1987 onwards, bids no longer required ministerial authorization. French banks and large enterprises' desire for clearer takeover regulation culminated in the 1989 law.

The 1989 law includes a mandatory bid rule (a requirement to bid for 100 per cent of the company), with a trigger threshold of 33 per cent of capital (Journal Officiel, 1989). The legislation outlaws ‘secret understandings’ and ‘covert changes of control’ (Economist Intelligence Unit, 2006, pp. 16–18). French law thus already prohibited many measures which ‘locked up’ the capital of firms in the case of a takeover bid, and ensured equitable treatment of all offers. The principles of sufficient information and opportunity for shareholders to decide upon the merits of an offer were written into both the Autorité des Marchés Financiers (AMF) and before that Commission des Opérations de Bourse (COB) regulations as well as French financial law. This suggests a (perhaps unexpectedly) conducive environment to corporate governance harmonization along the UK City Code lines desired by the EC.

However, this new liberal environment was reconciled to French ‘protected capitalism’ norms through a range of enduring impediments to takeovers, which greatly reduced its significance and impact (see below). Paradoxically, the privatization process expanded these. Finance Minister Balladur's hand-picking of the benefactors of privatization between 1986 and 1988 deliberately reinforced the noyaux durs (hard cores of investors) within France's ‘financial network economy’ (Schmidt, 1996, pp. 369–92). Thus, despite the liberalized takeover regime, the French state retained the orchestrating mechanisms to protect French firms through cross-shareholding and interlocking board directorships.

Whilst the overall legal regime is broadly conducive to takeover activity, a recurrent theme of re-regulatory activity in France has been the desire to carve out scope for dirigiste interventionism in ‘strategic’ sectors. In the 1980s and 1990s, ‘golden shares’ in privatized firms were one means to this end, though these have become of marginal significance (Commission, 2005, pp. 14 and 22). Laws were passed in 1996 (Journal Officiel, 1996), 2003 (Journal Officiel, 2003) and December 2005 (Journal Officiel, 2005) specifying that French state approval is required for takeover or investment in ‘strategic sectors’ such as national defence, public health or public order.

The revision to the French monetary code required governmental authority for takeovers in 11 ‘sensitive’ sectors (see Menucq, 2006, p. 230; Commission, 2006). The Breton Law of July 2005 created an obstacle to the hostile takeover of Renault, in requiring a bidder for a target ‘parent company’ to also bid for overseas ‘subsidiaries’ (in which the target holds more than a one-third stake – this, in Renault's case, includes Nissan). This kind of targeted protection remains a feature of French takeover regulation. Furthermore, the very substantial financial assets of the Caisse des dépôts et consignation (CDC) have often been deployed strategically by the state, investing in large French firms deemed in the national interest. Although not as central in the wake of privatization, the French state's shareholder role in a number of large French firms is still a feature of French capitalism.

In addition to legislative interventions, dirigiste French state actors and politicians (on both Left and Right) have recently re-engaged dramatically in public debates surrounding hostile takeovers. With a few specific exceptions, the French state lacks the policy mechanisms to back up its rhetorical hostility to takeover as the case of Arcelor demonstrated).2 Nevertheless, vociferous ‘economic patriotic’ critique can act as a dissuasive measure (as in the case of Pepsico and Danone in 2005).

A variety of obstacles stand in the path of the free play of takeover market forces. Concentrated ownership remains widespread in French capitalism, in contrast to the dispersed ownership patterns in the UK and US (Barca and Becht, 2001; O'Sullivan, 2003). The erosion of the noyaux durs since the mid-1990s is considerable but not complete (Loriaux, 2003, p. 116; Morin, 2000, p. 39). Their legacy remains in the capital structure of many firms. In 2002, networks of influence constructed around three big banks – BNP, Société Générale and Crédit Lyonnais remained in place and 30 directors enjoyed between them 160 seats on the boards of major French firms (Orange, 2002). Of France's 20 largest companies, eight still have a controlling shareholder. Family ownership of large firms remains a common phenomenon within French capitalism (Philippon, 2007, pp. 14 and 51–70) and another variant of concentrated ownership.

Concentrated ownership operates through a range of the ‘pre-bid’ defensive measures described above, contained within company statutes. These bolster the position of the incumbent management or dominant shareholders and make hostile takeover more difficult (Enriques and Volpin, 2007, p. 117). Mechanisms include differential voting rights, clauses limiting the right to designate board members, shareholder agreements, voting ceilings as well as restrictions to enable controlling shareholders exercising control without necessarily owning a large proportion of the cash flow rights. All of these protect ‘insiders’ and align with ‘stakeholder’ and CME approaches to the firm financing/corporate governance nexus.

French double voting shares are another key element in protecting ‘insiders’. The double voting right is a reward for loyalty, awarded to all without discrimination who have held shares for over two years. As well as double votes, and a range of voting ceilings limiting the voting rights of certain investors, there is also a prevalence of shareholder pacts, notably those restricting transfer of securities. All these allow management to ‘create a friendly shareholder group’ with the effect of ‘seriously obstructing a change of control’ (Fanto, 1998, p. 74). Both unequal voting rights and voting ceilings are much more prevalent in France than in any other major economy. Furthermore, these practices are increasingly prevalent, incrementally replacing the noyaux durs as instruments to dissuade takeovers (Magnier, 2002, pp. 73–4; Goyer, 2003, p. 197 and Table 6.5). In theory, the bottom line is that all of the dissuasive measures are public knowledge and therefore get factored into the price offered by the bidder or bidders. However, the collective impact of these mechanisms is to insert a good deal of ‘viscosity’ into the system.3

Some impediments to takeovers in France relate not to strategic design by feather-bedding management or dominant shareholders, but are a by-product of the historical development of French capitalism. France has lacked a culture which invested in financial savings instruments such as pension funds or mutual funds. Companies have not traditionally looked to financial markets for investment. Thus, from the ‘supply’ and ‘demand’ sides, financial markets as a source of capital for enterprises remained underdeveloped. The historic reliance on institutionally allocated credit (orchestrated through the state) rationed industrial investment. This led the French economy to be caricatured as ‘capitalism without capital’ (Stoffaes, 1989, p. 122). It accounts for the comparative lack of medium-sized firms French corporations. This industrial profile had an impact on takeovers, reducing the number of targets for anyone but the richest bidders. This, combined with the reasons above, explains why there are relatively few hostile takeovers in France, despite a permissive legal environment. Jackson and Miyajima note that there have been only 18 hostile takeover attempts in France between 1991 and 2005. This compares with 176 in the UK, and 332 in the USA in the same period (2007, p. 47, figure 21).

V. French Transposition of the Takeover Directive in March 2006

Transposition took place in a charged political context, with talk of ‘economic patriotism’ in spate as it gushed from the mouths of Prime Minister Dominique de Villepin and Finance Minister Thierry Breton. The (ultimately successful) hostile takeover by Mittal of Arcelor, begun two months earlier, had been greeted with thinly veiled hostility by the French government. This case was governed by Luxembourg, not French, law and therefore would not be affected by the bill. Nevertheless, this economic patriotism raised the stakes of what might otherwise have been a technical debate. French policy-makers used transposition and opt-outs to protect valued elements of French capitalism (such as shareholder pacts). Moreover, their dirigiste urges led them to expand their scope for interventionism in corporate takeovers. The poison pill, a key provision of the new law (see below), had not been present in earlier drafts of the bill. The departures from pre-existing liberal norms were a clear response to the perceived vulnerability of large French firms following the Mittal-Arcelor takeover bid.

In not opting out of Article 9, France retained its established board passivity norms, restricting the frustrating action detailed above and encouraging the free market of bids during an offer period (Belze, 2005). The March 2006 Law (Journal Officiel, 2006) transposing the directive involved some significant changes to France's regime, notably the requirement of prior shareholder approval during the bid period for takeover defences. Previously, prior shareholder delegation of powers to target boards (provided it met certain exigent conditions, including that the action would not make the bid fail) could be used in a bid period. Now, measures taken before the bid, which might impact upon it (and do not come under ‘normal’ operations of the company) must also be approved. Although the law's provisions governing shareholder meetings facilitates convening them at short notice, shareholder approval is an exacting requirement and hastily-convenedshareholder meetings are rare in France. This of course does not outlaw frustrating action if the shareholders endorse it.

Furthermore, hitherto outlawed frustrating action becomes possible under certain conditions due to French endorsement of the reciprocity principle. Under reciprocity, board passivity does not apply if the bidder's national laws do not contain similar board passivity provisions. The AMF has latitude to judge what constitutes the level playing field, given the conditions under which the bidder operates. It also determines the equivalence of provisions in foreign company law and thus what is permissible for French firms targeted by overseas investors (Lamy, 2006, p. 6). The EC's clumsy attempt at ‘levelling’ the playing field thus opens up the possibility, in the French case, for a move away (given shareholder approval within the previous 18 months) from the EC's favoured board passivity norms. The new law employed a broad interpretation of reciprocity, enabling companies to dispense with Article 9 if the bidder was not subject to it. Furthermore, if the bid is a case of concerted action, if one of the bidders does not apply the relevant article, boards are not constrained by passivity restrictions. The Directive text is unclear as to whether this was the intention (ANSA, 2006, pp. 6–7).

Article 18 legislates part of the ‘Breakthrough rule’, instituting a ‘one share, one vote’ principle for any meeting called to authorize defensive measures, and the first AGM after a successful bid and provided the bidder holds 75 per cent of the shares (Enriques, 2006, pp. 24–6; Menucq, 2006, p. 231). Statutory voting restrictions are also suspended for the first AGM after a successful bid. In France the COB (precursor to AMF) had long enforced a similar rule once a threshold of 66 per cent of either capital or voting rights was reached (Marini, 2005, pp. 100–1). Thus, unlike in Germany and Sweden, this contentious part of the new EU regime presented no difficulties to French policymakers.

The French legislator did exercise optionality in relation to other parts of the Directive's ‘Breakthrough rule’. The pre-existing AMF code suspended statutory restrictions on the transfer of securities during an offer period. This enters into French law in the transposition of the directive in Article 15 (Marini, 2005, pp. 101–2). However, parts 11 (2) and (3) of the Breakthrough rule challenged the central role of shareholder pacts and contracts between shareholders. These non-market mechanisms are seen as crucial cement in the networks that co-ordinate corporate activity within French capitalism. France permits such ‘flexible financing and control’ structures, but requires transparency surrounding them (Novelli, 2005, pp. 12–13). These kinds of ‘contractual agreements’ are often defended as enabling founders of medium-sized companies to retain control of the company when listing (Menucq, 2006, p. 231). For this reason, to avoid liberal harmonization and preserve the pre-existing role of shareholder pacts within its corporate governance, France has, as the Directive permits, opted out of parts of the ‘Breakthrough rule’. Crucially, the suspension of statutory restrictions on voting rights, the suspension of ‘contractual’ restrictions (shareholder pacts) in relation to both transfers of securities and on voting rights is left optional in Articles 16, 17 and 19 of the new law.

One aim of the ‘Breakthrough rule’ was to tackle multiple voting rights, and different classes of share which are one of the major obstacles to hostile takeover in some contexts, such as the US. This could potentially challenge another valued particularity of French corporate governance, the double voting right. However, because they are not a separate class of share, and as they lose their double status on sale, the Directive does not affect them. Their place within French corporate governance is defended by the government and many actors within the corporate governance milieu.4 This entrenched attachment to this departure from ‘one share, one vote’ explains why double votes were successfully defended in the context of both the drafting and the transposition of the Takeover Directive. This ensured the ongoing distinctiveness of the French model of capitalism in this important respect.

There was, however, one area where the French government did draw inspiration from US takeover regulation. At first glance this might appear grist to the mill of those asserting an Anglo-Saxon convergence within French capitalism. Yet, the American market for corporate control is nothing like the free and open play of bids disciplining management to prioritize shareholder value that the LME ideal-type describes. Rather, US corporate governance law has a very well-developed anti-takeover arsenal (Monks and Minow, 2004, pp. 42, 110–20, 232–9). In this case, the French government engaged in policy transfer of a US-style ‘poison pills’ or anti-takeover devices in order to bolster their ‘economic patriotic’ capacity to protect French firms.

The new French legislation included the possibility for target boards, with shareholder approval, to issue new securities (in both the pre- and post-bid period) at discounted price to existing shareholders. These ‘bons de souscription’ (also known as ‘bons Breton’, named after the initiator of the law) are modelled on American ‘poison pills’. This is the first time anything of the sort has existed in French law. Under reciprocity, if the bidder does not apply Article 9, these and other defensive measures can be issued or pursued by the board at its discretion provided that authorization has been gained from a shareholder AGM within the previous 18 months. Their effect is to dilute the effect of any holdings the acquirer already has in the firm and to raise (perhaps dramatically) the price of the bid. By making control considerably harder to achieve, and raising the price, the aim is not so much to warn off the bidder as to raise the price of the offer and to facilitate negotiation on terms more favourable to the target firm. Its likely effect, recognized by Breton, was to increase the price, rather than to make the bid fail outright (Marini, 2006, pp. 11–12).

Finance Minister Breton's overall assessment was that the Law ‘enabled French firms to have the same weapons’ and prevented French firms from being penalized but introduced a clearer rule of law governing their overseas expansion (Agence France Press, 2006, p. 1). In the shareholder AGM season immediately following the new law several companies, including Bouygues, Eurazeo, Saint-Gobain and Suez obtained shareholder approval to implement BSAs Breton. In total 19 large French firms voted these poison pills onto their statutes in 2006 and a further 15 were introduced in the 2007 AGM season. Any firm subject to a bid could, of course, follow suit given shareholder approval. These poison pills, combined with the new reciprocity provisions, in practice expand options for management defensive measures. The likely effect of reciprocity is to create a very variegated picture in terms of anti-takeover defence measures.


The clash of European capitalisms has been fought on the terrain of takeover regulation. As a result, far from delivering LME-oriented harmonization (dismantling impediments to the market for corporate control), transposing the Takeover Directive in fact introduced new forms of protection (EC 2007, pp. 6, 10–11). The battleground for European takeovers remained every bit as uneven a surface as it had been prior to the Directive. The EC's ‘level playing field’ remains illusory. Furthermore, given the permitted national derogations, the crucial opt-outs (Articles 9 and 11) and the reciprocity rule, the terrain will become more undulating and nationally variegated (with heightened ambiguity and opacity). This finding necessitates a counter-intuitive rejoinder to the ‘triviality thesis’. The European Commission as a corporate governance actor has had a demonstrable impact, its initiative presenting opportunities to policy actors in their import and mediation by national law and politics. The Commission has not, however, induced liberal harmonization of European takeover regimes.

Recent re-regulation within French corporate governance has not uniformly emulated shareholder value norms. French policy elites, in transposing the Takeover Directive, used the opportunity to rebalance LME and CME elements within the diversity of French capitalist institutions to increase their interventionist scope to protect French firms. Paradoxically, the most significant shift away from a freer takeover market was introduced as a direct result of transposing the LME-oriented EU Takeover Directive. The French government institutionalized resistance to the liberal takeover model through the ‘economic patriotic’ policy transfer of US poison pills, ironically expanding the range of anti-takeover defences available to target boards. This demonstrates how acceptance of some liberal ‘shareholder value’-oriented elements co-exists with enduring attachment among French policy elites (of left and right) to ‘stakeholder capitalism’ institutions and norms and to co-ordinated, ‘statist’ French capitalism. Although the French context remains relatively open to takeover bids compared to many other European countries, the increased prevalence of voting ceilings and double votes, along with the introduction of poison pills, indicate the potential for more defensive institutional engineering. New obstacles impede takeovers in France and the French corporate governance and takeover context is more ambiguous and unpredictable in the wake of transposition.

The clash of capitalisms, overlain with the ‘asymmetric vulnerability’ (Knudsen, 2005, p. 524) of firms, and national corporate governance regimes, to EU-level regulatory change, generates resistance to the EU liberalizing takeover reform agenda at the national and firm level. This constitutes a major obstacle facing supra-national reform of corporate governance, given the ability of national level actors to reshape and amend takeover reforms, both during their passage through the European Parliament and in the transposition into domestic law (Commission, 2007, pp. 10–11). Both countries and firms can opt in and out, increasing the heterogeneity within national varieties of capitalism. This increased differentiation creates a feedback effect which will increase problems of asymmetric vulnerability.

Future attempts at the EU-level to reform corporate governance will continue to be hamstrung by this (increasing) variety and complexity of institutional and legal environments of corporate governance in Europe, as well as the multitude of defensive measures at European firms' disposal. As the Takeover Directive demonstrated, Commission exhortation to endorse supra-national corporate governance harmonization is insufficient to overcome entrenched corporate and political opposition. This is in part because the assumed superiority of the Commission model of political economy, rooted in LME institutions such as shareholder value, and ‘one share, one vote’, is not accepted. Underlying this lack of consensus is an ongoing contestation surrounding European models of capitalism and their relative merits. This clash of capitalisms explains why the Takeover Directive was emasculated through optionality, and why the Commission failed in its bid to develop an EU Takeover Directive with teeth.


  • 1

    In interviews in May 2007, French securities regulators indicated their ongoing uncertainty as to the operationalization and practicability of reciprocity.

  • 2

    The successful hostile takeover of Arcelor is unusual in the French context, in that Mittal's team went round winning over shareholders one by one, and the board remained opposed, almost until the bitter end. The success of this hostile takeover could, like the BNP Paribas case, be interpreted as a turning point in French capitalism. However, this is a time-consuming and very costly approach. To succeed, it required someone with Mittal's resources on the one hand, and a series of clumsy responses from the Arcelor CEO (notably approaching a Russian oligarch as a potential ‘white knight’) to effectively drive its shareholders into the Mittal camp.

  • 3

    The term was used in interview by a leading French investment banker, April 2007.

  • 4

    Interviews with corporate governance advisors, lawyers, investment bankers and policy advisors, Paris, April to May 2007.