Written evidence from David Kershaw (CGV0120)

The comments below respond to several, although not all, of the terms of reference of the Committee’s inquiry. This does not reflect a view that the other issues are unimportant rather this reflects my time constraints for this submission. The submission has not been drafted as an academic paper, and, for brevity’s sake, I have kept academic references to a minimum. I would be more than happy to provide more detailed references in relation to this submission, if it would be of assistance.

Directors Duties

Is company law sufficiently clear on the roles of directors and non-executive directors, and are those duties the right ones? If not, how should it be amended?

The question both invokes consideration of the role specification of the executive and non-executive directors in the board but also the question of the nature and the extent of the duties to which all directors are equally subject, but which adjust to take account of the role of the respective director.  I will contain my answer here to role specification and address duties below.

First, company law itself is not at all clear on these roles; indeed it says nothing about the composition of a board of a UK private or public company, requiring only one director for a private company and two directors for a public company. Particular roles and divisions of responsibility are not set forth in the Companies Act 2006 or related legislation. As the Committee is aware, role distribution for listed companies is set forth in the UK Corporate Governance Code, with which such companies must comply or explain. Role distribution in the Code is clear. There is no need to revisit the Code in this regard unless, connected to the consideration of employee board participation (see below), the Government elected to explore the option of a two-tier, instead of the current one-tier (“unitary”), board.

For publicly traded companies who are not subject to the Code, which would include companies traded on AIM, again if the question relates to role specification then I don’t think there is any evidence that there is any need for regulatory intervention in this context. I presume that the Committee does not have its sights in this regard on the smaller private companies that often do not have outside directors and have concentrated ownership.

Is the duty to promote the long-term success of the company clear and enforceable?

Yes, the duty set forth in section 172 of the 2006 Act is clear. It combines a requirement to promote the interests of shareholder where corporate power is exercised, and a care requirement (subject to section 174 of the Act) to consider the interests of all groups affected by corporate action when making that decision. There is no doubt that this is a pro-shareholder provision and no doubt that it is wholly consistent with the approach English company law has taken in relation to the exercise of corporate power since the inception of incorporation by registration in 1844. It is also consistent with English law’s long established approach to the exercise of delegated power in multiple areas of the law, namely that it should be exercised honestly and in good faith for the purpose that power was delegated.

The duty is clearly enforceable, however, the nature of the duty itself which focuses on the subjective intent of the director makes it difficult to enforce. As long ago as 1478 courts noted that: “the devil himself knows not the intent of a man.”[1] What courts have typically done when faced with the difficulty of assessing subjective standards is to ask whether the action taken is plausibly or rationally connected to the interest which the agent is asked to further. In some contexts, for example where a company approaches insolvency, the nature of the insolvency context increases the likelihood that the duty will be breached. We can identify several modern examples of such breach. However, outside of those contexts, section 172’s subjective standard is inevitably a very undemanding standard and easy to comply with, or to create the appearance of compliance.  However, there are good reasons, which courts have recognised for centuries, which underpin the subjective nature of the standard. In particular, that if courts (who are less skilled at reaching and understanding business judgements and who judge with hindsight) can impose a more demanding standard on directors – such as the requirement to make objectively reasonable decisions – then skilled individuals will refuse to serve as directors and those who are willing to serve may become far too frightened of risk taking.

Does this mean that sometimes egregious decisions are made that deserve public opprobrium and yet are not sanctioned by corporate law? Yes it does. But to make the standard more demanding to sanction such decisions runs a significant risk of damaging UK business and making the UK corporate form an unattractive vehicle. 

Section 172 and the balance of shareholder and employee interests

As the Committee will be aware, a long standing and contentious debate surrounds the issue of whether the duty in section 172 should be amended to allow directors when acting to act in the interest of other corporate constituencies (such as employees) even when such action may not be deemed to be in the shareholder’s interests. With such a duty the interests of shareholders and stakeholders are deemed to have equal priority. This type of duty is often referred to as a “pluralistic” or “multiple purpose” duty. Many jurisdictions, including Germany and several US States adopt this approach. This approach contrasts with the “shareholder first” focus of section 172.

In my view there are several potential advantages of adopting such a pluralistic duty, including the awareness amongst employees that their interests matter as much as shareholders. Such an awareness can, in theory, contribute to a culture of firm “membership” or “ownership;” a culture which can generate employee commitments which in turn generates positive productivity benefits. Such a duty would also contribute more broadly to a more inclusive idea of modern capitalism. However, there is a counterbalancing and legitimate concern that such a duty would enable transfers of shareholder value to other groups, increasing the cost of companies’ equity capital and expropriating value from shareholders. However, one needs to be careful not to overstate the extent to which a move to a pluralistic duty could have any such effects.  It is central when thinking about any change in corporate law that one understands that corporate law rules interact.  The extent to which a rule change has effects, intended or otherwise, is a function of how a rule change interacts with existing rules. In this case a change in the nature of section 172 to provide a pluralistic approach to the corporate objective would have limited impact as such a duty runs counter to, and is undermined by, the core structure of a UK company. 

The structure that I refer to here is the strong pro-shareholder orientation of UK company law. No other advanced economy provides for such strong mandatory shareholder rights, including, amongst others: rights to remove directors; rights to call shareholder meetings; rights to vote on substantial transactions; and pre-emption rights. This structure of rights sets the control parameters within which board power is exercised and makes it clear that the board is answerable to shareholders. A pluralistic duty would not alter these control parameters. Paradoxically, this means that the risks of changing this standard are relatively low if no other changes are made to this structure. That is, such a change would provide political capital for the governments “new capitalism” agenda, without having much of an “on the ground” effect. Less cynically, one could say that it would make some contribution to employees feeling more committed to their workplace without putting shareholder value at risk.

The case for providing a pluralistic duty is particularly strong in relation to banks and financial institutions, which - because of the interaction of the incentives of diversified shareholders, limited liability and the “too-big to fail” problem - have strong incentives to take excessive risks which are not disciplined by credit markets. Put differently, bank shareholders have strong incentives to exploit the “too-big-to-fail” subsidy. The idea that boards of banks have a primary responsibility to shareholders, who have incentives to take risks that are not aligned with society’s interest in having stable financial institutions, is perverse.[2]

Shareholder Rights

Although beyond the scope of this consultation, there are in my view strong reasons to alter the nature of some of the shareholder rights referred to above.  Such reforms would both help to generate real effects of a (reformed) pluralistic section 172, but would also address the effects of corporate law on corporate short termism.

As the review does not ask for commentary on this I will be very brief. The Kay Review did not effectively consider the ways in which corporate law rules facilitate pressure on boards by active shareholders. Such pressure may be benign – holding the board and managers to account – but it may also be malign – forcing managers to adopt shorter term shareholder preferences at the expense of long term investments and research and development. For too long, UK lawyers, politicians and commentators have been complacent about the nature of UK company law, viewing it as a model form of governance regulation. But as the nature, form and incentives of shareholders have changed, there are real concerns that these rules may drive negative outcomes. Any review of corporate governance needs to place an exploration of these issues on the agenda.

Employee and Consumer Representation on Boards (“Co-determination)

My comments here focus primarily on employee representation. I deal briefly below with consumer board representatives.

There are clear economic benefits that may be obtained from employee co-determination. To connect to the debate on a pluralistic corporate objective discussed above, board representation may contribute to a culture of “membership” or “ownership” of the firm (in a social not a legal sense); a culture that can generate positive productivity benefits arising from employee morale and human capital investments in the firm’s operations and processes.  Such a culture may be a superior and cheaper way of generating employee effort and commitment than the promise of financial incentives. Board representation can also contribute to increased trust in the firm amongst employees and managers by enhancing the willingness of employees to believe what managers tell them. This can facilitate, and reduce the costs associated with, firm restructuring and wage negotiations.  However, there are also costs that may be associated with such employee representation. The design of the form of co-determination is, therefore, centrally important to ensure that these benefits exceed the costs.

With regard to employee co-determination, the economic literature identifies two primary problems.[3] First, that board representation elevates the interests of employees in corporate decision making and, thereby, may detrimentally affect shareholder value where the interests of employees and shareholders interests are not aligned. This could result in investors paying less for equity capital to the extent that the above economic benefits are not deemed to outweigh such losses. The second concern relates to the effective functioning of the board, which also translates into value but not as a result of direct employee/shareholder interest trade-offs.

There are two reasons why co-determination may undermine the effective functioning of the board. First, the presence of board employee representatives may result in more limited and controlled information flow to the board in relation to any issue that involves confidential or proprietary information. The concern here is that such information is more likely to leak, particularly where employee interests are implicated. A less informed board is a less effective board, both as a decision maker and a supervisor. The second concern about co-determination and board effectiveness is that the presence of employee representatives increases the heterogeneity of interests on the board, thereby making the board a less efficient governance mechanism.  For example, the more heterogeneous the interests on the board the slower decision making is likely to be, and important decisions may be compromised as employee representatives hold-up decisions to extract other benefits for employees. This latter concern is a particular problem during periods when restructurings and wage and benefit negotiations are taking place.

It seems clear that the extent to which codetermination generates costs is a function of the nature of board structure. In Germany, where we see the most extensive form of co-determination, large companies have a two-tier board with a supervisory board and a management board. Co-determination only applies to the supervisory board, where in the largest of companies half of the supervisory board consists of employee and union representatives, and the other half of shareholder representatives. In the event of deadlock the chairman (appointed by the shareholders) has the casting vote. German corporate law makes it very clear that the supervisory board has no responsibility for the operational management of the company. The power to manage and direct the company is transferred by German corporate law directly to the management board, which consists only of a company’s senior managers.  The supervisory board can, and typically does, retain certain key veto powers which generate some operational influence, but, that said, German company law is very clear that the management board manages and exercises corporate power, whilst the supervisory board supervises.

The concerns noted above that co-determination undermines the function of the board are more apposite for one–tier or unitary boards. With a two tier board the concern that employee interests are over-weighted in the managerial decisions that unitary boards make evaporates because the board on which the employee representatives sit does not make operational decisions. Similarly, in the absence of having to make such operational decisions the risks associated with confidential information leakage are reduced (although as commentators on German corporate law will observe, this still remains a concern) and the risk of hold-up and delay in relation to key decisions is reduced. One might respond in this regard that large UK company unitary boards do not make many managerial decisions and so these risks are overstated. It is true of course that most managerial decisions are delegated to senior management, however, important managerial decisions are made by unitary boards beyond the decision to appoint and remove management. There is as a result a different managerial bias in a large company unitary board as compared to a supervisory board.

That said, although supervisory boards do not make operational decisions, as the management board is appointed and removed by the supervisory board the presence of employee representative enhances the interests of employees and the responsiveness of managers to those interests. Accordingly, although the supervisory board is operationally disempowered, the co-determination of these boards contributes to the benefits identified above.

The Committee should also take evidence on the perception of co-determination in countries that deploy it. Here is not the space to provide a detailed account of that evidence. Anecdotally, co-determination is viewed unfavourably by German managers and shareholders for the reasons outlined above. Indeed the Committee should note that there are large German companies that do not have co-determination. I understand that this is possible where companies do not have works councils (Betriebsraete). In contrast, when one crosses the border to Austria, where co-determination also exists, it is not viewed as unfavourably. There are two primary reasons for this. First, co-determination in Austria consists of a maximum of one third of the supervisory board and second, in contrast to German co-determination, there are no union representatives just company employee representatives. These differences reduce both the influence of employee representatives and the lack of incentive alignment amongst members of the supervisory board. It also reduces the confidentiality problems outlined above.   Clearly, the Committee would be well advised to seek testimony from German and Austrian lawyers and businessmen on the operation of co-determination in these jurisdictions.

In conclusion, therefore, there are both benefits and risks associated with introducing co-determination. The risks can be reduced in several ways. First, and most obviously, by low numbers/percentage representation, for example, one employee board member. However, it seems likely that such low number solutions are likely to reduce the positive effects of co-determination more significantly than they alleviate the concerns – which would continue to exist particularly in relation to confidential and employee sensitive information. With regard to the benefits, requiring only one representative risks the reform begin viewed as ineffective tokenism by employees, more a regulatory gimmick than a contribution to a culture of membership. 

The second, preferable, way of reducing the risks associated with co-determination would involve providing for a two tier board. A two tier board is, in theory, perfectly possible under existing UK corporate law. It could be designed through the corporate constitution - the articles of association - but it is rarely deployed. The Companies Act 2006 assumes a one tier board and creates some design obstacles that would inhibit the functioning of a two tier board, with its clear divisions of authority and lines of accountability.

Given this, the Committee should explore the facilitation of employee representation rather than mandating it. And doing so in accordance with tried and tested UK regulatory techniques. Facilitation could take place through:

(i)                 Legal reforms of the Companies Act 2006 that rendered it easier for companies to provide for and transition to a two tier board with co-determination if they elected to do so.

(ii)               An informal 20-30% employee representative target on a voluntarily adopted supervisory board – coupled with a strong steer from government to companies and shareholders that companies need to consider adoption. Where companies do not adopt, they would be expected / required to periodically report on their consideration of the matter (a softer version of comply or explain)  and outline the reasons for non-adoption. 

(iii)            The precise way in which co-determination would be implemented – including procedures for selecting / electing the employee directors would be left up to the company to design.

With regard to the proposal to include consumer representatives on the board, I think it is difficult to view this positively. Consumer interests are not fully aligned with those of the company. Would consumers prefer a company to cut prices on its products or maintain higher prices to fund uncertain research and development; would they prefer the company to cut prices or to boost pay for employees in an uncertain attempt to improve employee morale and productivity? Furthermore, a set of problems can be envisaged similar to those discussed above in relation to employees.  First, consumer representatives would likely result in restrictions on information flows to the board when problems affecting consumer issues arise. Precisely the time when the board needs to be well informed, in order to intervene and address the problems.  Second, the presence of more heterogeneous interests on the board could undermine the quality and efficiency of board decision making. However, in contrast to the case of employee representatives, there are no clear counterbalancing company value benefits arising from the presence of consumer representatives on boards. Of course, taking greater account of such interests could have positive regulatory benefits (more competitive markets; greater regulatory compliance etc), but this would be an unusual and highly untargeted way of addressing such regulatory concerns.

Executive Pay

It may well be the case that increases in executive pay in some companies reflect an excess or an abuse of power by senior managers and failure of board accountability. But this is only one of several drivers of pay inflation which the Committee needs to take account of. In my view it is important that the Committee gives consideration to the range of possible drivers before reaching any regulatory conclusions or recommendations. I will not address drivers related to market pay levels or the internationalisation of the UK management here, as I am sure the Committee will receive ample evidence on this. Rather my submission makes one point about the fact that pay regulation, together with the context of outrage about pay levels, may be a hidden driver of pay inflation.

There is a correlation between pay regulation, pay processes and increases in pay. Put simply: pay regulation itself is an important driver of pay inflation. Place yourself in the shoes of a non-executive director sitting on an all independent non-executive remuneration committee of a UK listed company considering the pay deal of an incumbent executive director (for simplicities sake I will assume an incumbent director, but a similar analysis could be applied to a new executive director). The NED has two key goals: first, where the executive director is perceived to be performing well, to retain that executive and to incentivise her to maximise her efforts; and second, to stay below the radar of politicians, the media and institutional shareholders.[4]  To achieve the first goal she needs to pay the executive what is perceived to be the market rate. Thanks to pay regulation and pay consultants who analyse that regulation, remuneration committee NEDs have a wealth of data on the pay rates of industry and extra-industry competitors. Having identified that market rate, the remuneration committee is likely to agree (indeed to want to provide) a pay deal that is a market-plus deal. The “plus” in an average company case is likely to be slightly above the current market rate and, where the director is acting in the best interests of the company, will be justified by the perceived importance of retaining the executive and as a reward for the efforts of the manager; or as an incentive to do better. Of course the “plus” could also reflect self-serving influence of executive directors over weak or corrupt NEDs. But we need to acknowledge that much of the time such benign considerations will be in play and that where there is no such abuse of power there will be a “plus” for the other reasons – reasons that are legitimate and in the corporate interest.

Importantly, the “plus” is unlikely to be significantly above the market rate precisely because the non-executive director wants to avoid any attention from politicians, the media and institutional shareholders in the contemporary highly-charged pay environment. To avoid such attention new pay arrangements must stay within the close orbit of existing market norms, both structurally and in terms of absolute amounts. Any such negative attention will damage the reputation of the non-executive directors of the company and particularly those who sit on the remuneration committee. Such reputational damage will undermine their ability to be reappointed and to obtain other NED appointments.

Although the “plus” component is, for the reasons outlined above, modest it goes into the next market-rate calculation carried out by the next company and pay consultant. Slowly (and not always that slowly) but surely pay levels rise. If such “ratcheting” is an important driver of pay inflation then legislators need to take this into account when considering their regulatory responses. The primary regulatory responses on pay for the past 25 years have been based on the assumption that pay inflation is a result of an abuse of power. If it is not, then the regulatory logic that underpins pay regulation is missing its target. More of the same may satisfy short term needs of politicians to be “seen to be doing something” but it will not stop pay inflation. 

It is very difficult to know what if anything can be done. Pay ratios – either ratio disclosures or mandatory ratio limits - are an intuitively attractive response and also have the added benefit of recognising a different basis for regulation – not abuse of power but pay equity. However, they are a clumsy tool to achieve pay equity and many would have significant reservations about the ability of government to identify appropriate pay ratios for different industries. Furthermore, direct or indirect absolute pay restrictions will damage retention and recruitment in UK companies. Of course legislators may decide that this is a price worth paying in furtherance of equality and social equity goals. There are also other well versed reservations about pay ratio limits and disclosures. First it may drive the outsourcing of low pay jobs in order to improve pay-ratios, driving more employees outside of the firm. Second, the nature of the company and the industry will drive pay ratios resulting in unfair public treatment of some companies and the obscurity of others.

It seems probable that corporate pay regulation has little more to offer. It has contributed to the problem but clearly the clock will not be turned back on these reforms. If government decides that it requires a response in relation to perceived excessive pay the better targeted one would be a taxation response. Of course such tax-based responses generate a raft of other difficulties beyond the scope of this submission. They also in this field have an infamous history of unintended consequences – the non-deductibility of pay over $1,000,000 in the United States being a good example.

Professor of Law

London School of Economics

26 October 2016

4