Ignore causation, correlation is enough

The submissions in response to the BEIS corporate governance inquiry that have been posted so far make for interesting reading, especially on the board gender diversity question. 

Mr Michael Romberg, "a former civil servant who has held posts in HM Treaury and the Home office", observes that: "The biggest noise in this arena comes from bourgeois women." 

And one of the noisy bourgeois women, Helena Morrissey, somewhat astonishingly states that:

"There is plenty of empirical evidence (McKinsey, CSFB, Citibank, SocGen have all published extensive analyses, based on Global, Australian and European companies respectively) pointing to a positive correlation between gender diversity on boards and company performance. It is irrelevant that causality can’t be proved – arguably the smart companies ‘get it’ around diversity and are therefore likely to be more likely to be forward looking in other areas."

So, in this post-referendum topsy-turvy world, just as opinion is enough, never mind the evidence of experts, it seems that correlation is enough, never mind causality.

Here are some enlightening correlations: http://tylervigen.com/spurious-correlations

The Parker Review

So now we have another knight of the realm reviewing the state of the British boardroom. Sir John Parker says:

"This is not an exercise of tokenism; the recommendations are underpinned by strong industrial logic and the need for UK companies to be competitive in the increasingly challenging global marketplace."

What is this "strong industrial logic"? Let's apply some logical thinking to the consequences for board composition. If a board is required to increase diversity based on gender and ethnicity, how is its nomination committee going to balance this with the need for directors with the particular skills and experience relevant to the strategy of the company? 

And why should diversity be limited only to gender and ethnicity? What about age? Boards are now characterised as "male, stale and frail" so presumably the frail bit is next on the agenda. (And let's not ignore the fact that neither Davies nor Parker is female or in the first flush of youth...)

And perhaps someone can point me in the direction of evidence that links board composition to competitiveness...

BEIS consultation on corporate governance

The government launched this inquiry in September with very wide terms of reference and a short time frame for comment. I picked out the question that interested me most:

• What evidence is there that more diverse company boards perform better?

- and, doubting that anyone else would submit anything critical, sent this off. (The paper by Boivie et al, with its splendid title "Are Boards Designed to Fail? The Implausibility of Effective Board Monitoring", is especially useful as it contains a comprehensive list of all the relevant literature, noting for each paper the theoretical framework used and summarising the findings.)

---------

What evidence is there that more diverse company boards perform better?

1. To date, academic research into board diversity has focused principally on gender diversity, and on its impact on corporate performance, rather than  specifically on board performance, although some may consider the two to be effectively synonymous.  Boards are generally assumed to perform two roles - advisory and monitoring - and the latter has received the most attention.  

A recent paper by Boivie et al (2016) provides a comprehensive review of the literature which addresses the oversight role of the board. They focus primarily on the board’s role as information processor and identify ten barriers to effectiveness relative to this. They address the issue of board diversity (p338ff) and observe that the literature suggests that, while there may be benefits from board diversity:

“….diversity increases certain challenges for group interactions within boards…. diversity may make it more difficult for the group to work together thereby creating barriers that affect boards’ ability to be effective information processors.”

(p339)

2. The literature on the impact of board diversity on corporate performance is scattered across a range of different academic disciplinary areas: it uses varying theoretical approaches and research methodologies and has been conducted in a variety of national contexts.  The results are inconclusive: correlations between board gender diversity and financial performance have been identified but the direction of causality cannot easily be inferred. Authors of these studies are careful to include this caveat although those who cite their work often ignore it.

A comprehensive statistical meta analysis of the literature is provided by Post and Byron (2015) who note that:

“..our results suggest that board diversity is neither wholly detrimental nor wholly beneficial to firm financial performance.” (p1563)

They warn readers:

“However, we cannot—nor do we—claim causality; we can merely claim that we have taken a few extra steps to rule out some plausible alternative explanations for our results. “(p1562)

3. Rhode and Packel (2015) provide a narrative review of the literature. They cite a number of empirical studies (p393ff) which indicate correlations between gender diversity and improvements in board process, corporate governance oversight and corporate reputation but they also point out that:

“…correlation does not demonstrate causation, and it could be that well-governed corporate boards are more committed to diversity and seek greater gender parity.” (2015:401)

4. These three recent and substantial studies of the literature on board gender diversity therefore conclude that the evidence is mixed and that there is no clear causal relationship between board gender diversity and corporate performance.

Are the right questions being asked?

5. Observing that there is little evidence to support the business case for diversity based narrowly on board/corporate performance, Fanto et al (2011) argue for broadening the scope of the debate to include other justifications for diversity. While the evidence to support policy with regard to “the business case” for board gender diversity is inconclusive, there may be sound social/political reasons for pursuing this. These have not been well articulated in the UK debate which has moved to focus on how to achieve gender balance rather than why this is necessary (Spira, 2013)  

This lack of challenge to fundamental assumptions may lead to unforeseen problems in the future as economic and political environments shift. The Committee may find it helpful to consider the more fundamental question: does mandating board composition improve corporate governance?

The work of Boivie et al cited above also suggests that the evidence to support such policies is not clear cut.  They cite empirical studies conducted in the US dating from the 1990s that question this approach. The UK experience of mandating the appointment of independent non-executive directors (NEDs) suggests that outcomes are difficult to assess and unintended consequences may follow.

6. Although it has taken several decades since the publication of the Cadbury Code to embed fully the idea that board oversight will be improved by the appointment of a specified number of independent NEDs,  the assumption underlying this requirement has not been challenged.

Identifiable and measurable criteria signifying independence stand as proxies for the independence of mind that is being sought in board members, and the strength of this relationship has not been tested. Similarly, the identifiable and measurable characteristics of diversity (gender, age, race, etc) may not provide the diversity of thinking that is posited to improve board and corporate performance.

7.  Evidence has been available for some time that independent NED appointments do not necessarily have the expected effect. This requirement for board composition in UK quoted companies has led to significant structural change. Boards have become smaller and are now predominantly non-executive. When executive directors and NEDs no longer meet around the boardroom table, contact between the two groups is in effect mediated by the CEO. The impact of this change on corporate governance outcomes has yet to be assessed but intuitively one might expect communication processes at board level to have changed. Requiring further alterations in board composition to accommodate gender and other forms of diversity may have more complex and unanticipated effects.

How does research influence policy?

7. In a reflective essay on the potential usefulness of research to policy makers, Ferreira, a leading scholar in this area, observes:

“Finally, it has never been clear to me why the board is the place to start with policies that aim to promote better female representation in business. Most proponents of board quotas believe that, if we smash the glass ceiling at the board level, we will also reduce discrimination at lower levels. I am not

sure that this conclusion follows. If a group is more likely to be promoted to the top, perhaps employers will become more demanding when first recruiting from this group. I would like to see more empirical and theoretical research on this issue.” (Ferreira, 2014)

8. But even where a significant body of rigorous and robust research is available, it may not be put to the most effective use. Eagly (2016) explores the relationship between research findings, advocacy and policy making around gender diversity on boards. She probes critically the use of available research by advocates and policy makers, demonstrating that simple arguments are chosen by these groups which do not accurately reflect the complexity of the underlying research but then become received wisdom and remain unchallenged.

“…simplistic renditions of scientific findings on diversity continue to find favor among diversity’s advocates and the legions of practitioners and consultants engaged in helping organizations meet their diversity goals. Presented as if they were evidence-based findings, broad claims about the advantages of diversity for group and organizational performance appear regularly in promotional materials of consultants and advocates. Also, their scientific allies may engage in selective citations of those studies or portions of studies that have shown the hoped-for performance gains, without

hinting at the general pattern of findings across studies.” (p207)

While Eagly urges researchers to adopt a more proactive stance in relation to the ways in which their findings are interpreted, improved policy development could also result from more detailed consideration of the robustness of the arguments on which policy decisions are based.

References

Boivie, S., Bednar, M.K., Aguilera, R.V. & Andrus, J.L. (2016) Are Boards Designed to Fail? The Implausibility of Effective Board Monitoring, The Academy of Management Annals, 10:1, 319-407 http://dx.doi.org/10.1080/19416520.2016.1120957

Eagly, A H (2016) When Passionate Advocates Meet Research on Diversity, Does the Honest Broker Stand a Chance? Journal of Social Issues, Vol. 72, No. 1, 2016, pp. 199--222

http://dx.doi.org/10.1111/josi.12163

Fanto, J.A., Solan, L.M. and Darley, J.M. (2011) Justifying Board Diversity

North Carolina Law Review  Vol. 89, 902-936

Ferreira, D (2014) Board Diversity: Should we trust research to inform policy?

Corporate Governance: an International Review 23(2): 108–111

http://personal.lse.ac.uk/FERREIRD/CGAIR.pdf

Post, C and  Byron, K (2015)

Women On Boards And Firm Financial Performance: A Meta-Analysis

Academy of Management Journal Vol. 58, No. 5, 1546–1571.

http://dx.doi.org/10.5465/amj.2013.0319

Rhode, D.L and Packel, A.K  (2015)

Diversity on Corporate Boards: How Much Difference Does Difference Make

Delaware  Journal Of  Corporate  Law 39, 377-426

Spira, L.F (2015) If women are the solution, what is the problem? economia, March, 66-67

http://viewer.zmags.com/publication/e8c7bb38#/e8c7bb38/66)

The End of Accounting

As a change from ploughing through the Man Booker prize shortlist, I recently read this. Once you get used to the style – a bit matey, very slightly patronising - and structure - rather like a text book, short chapters with more complex material in appendices and chapter endnotes into which some of the more interesting information is tucked away – it’s readable and, indeed, a quite convincing polemic.

The authors set out their case clearly, first demonstrating the declining usefulness of the accounting information in corporate reports and then assessing the information investors really want: on the basis of this they propose a new form of statement – the Strategic Resources and Consequences Report – which they illustrate using case studies. This report is more closely aligned to the business model perspective and focuses on the strategic resources of the business, identifying the way they are used to create value (their creation, preservation and deployment) using a cash basis and thus stripping out much of the estimation that underpins current financial reporting. You can see an example of the proposed report here.

A couple of things irritated me about this book,  As usual with US authors, little attention is paid to what goes on in the rest of the world. The reader is told that accounting is pretty much standard everywhere and the only past suggestion for change was Ijiri’s triple entry bookkeeping proposal. This ignores the very thoughtful debate in the UK prompted by “The Corporate Report” (1975) and “Making Corporate Reports Valuable” (1988)  The debate at that time addressed the usefulness of financial reports to a wider range of stakeholders than investors: this book takes a narrower approach. And that debate did generate a new form of report: the Value Added statement. Its rise and fall are discussed here.  Secondly, the authors seem to treat accounting and financial reporting as synonymous. I don’t think they are.

The first few chapters of the book set out the data and analysis which support the authors’ contention that the usefulness of published financial information has declined. They attribute this decline to three factors: the increased importance of intangibles and the problems in accounting for them; the increased level of estimation underpinning accounting numbers; and delays in reporting important business events. It seems to me that these are all linked to the inexorable rise of the accounting standardisation project but they don’t seem to discuss that much.

They extend their argument by attempting to assess what information investors want, basing this on a content analysis of conference calls, identifying a major focus on questions on the use of strategic resources.  They then develop their proposed report (asserting that it is based on economic theory but this basis was not at all clear to this reader). And then they provide illustrative case studies of how the report might be constructed in practice.

The proposed report is certainly interesting. It includes both quantitative and qualitative information. It is industry specific. I think it would be difficult to make comparisons between companies (or across time periods) but then relying on the reduction of complex information to standard KPIs which appear to allow comparison can also be misleading.  But with the best will in the world, I can’t see boards of directors wanting to adopt this very different way of reporting, because disclosing information about their strategic resources could damage their competitive advantage. The authors argue that much of this information is already public but not readily available in a coherent form, and they propose a further incentive in the reduction of other regulatory reporting requirements. Convincing investors to demand change might be a more effective approach but it’s taken a very long time for investors to take an active role in demanding corporate governance change.

It will be interesting to see if this book prompts any significant debate.