Corporate governance and social welfare – mapping the explanatory domain

In the third and final part of the book I address limits to my theory’s explanatory domain and explore factors that the theory suggests might reinforce stability, or catalyze change, in these corporate governance systems moving forward.  Each of these subjects has prompted intriguing thoughts and challenging questions throughout the book club.  I’ll address the explanatory domain here, and then turn to stability and change in my next post.

Both Danny Sokol’s post and Brett McDonnell’s initial post emphasize the distinction drawn between dispersed ownership systems (typified by widely dispersed, passive shareholders) and concentrated ownership systems (typified by blockholders), and question my case selection in this light – notably observing that Australia and Canada exhibit somewhat higher levels of share concentration than the UK and the US do.  Why are these countries included, while others that appear to exhibit high levels of dispersal are excluded? 

As I explore in the last substantial chapter of the book, the distinction between dispersed and concentrated systems is critical to the development of a comparative theory of shareholder-centrism in corporate governance because their respective regulatory postures toward shareholders are radically different.  In concentrated systems the default regulatory posture is to constrain the innate power of controlling shareholders to safeguard the interests of other stakeholders – Germany’s co-determination system (giving employees in large companies substantial board representation) representing the paradigm case.  In dispersed ownership systems, on the other hand, the default regulatory posture is to protect minority shareholders to foster the development of deep, liquid equity markets, as each of the common-law jurisdictions that I’ve examined has endeavored to accomplish. 

It is critical to recognize, however, that various countries in fact fall along a wide spectrum in this regard, as Peter Gourevitch and James Shinn emphasized in their important book.  Gourevitch and Shinn measure levels of blockholding by reference to the market value of listed companies that have 20% blockholders, expressed as a percentage of the total market cap of that country’s stock markets.  By this measure, they report 4.1% for Japan, 5% for China, 15% for the US, 20% for the Netherlands, 23.6% for the UK, and 27.5% for each of Australia and Canada. Setting aside the fraught issue of how ownership concentration is most appropriately measured, in relative terms these figures raise intriguing questions for the approach I’ve taken in my book.  Why do Japan, China, and the Netherlands exhibit substantial share dispersal, yet not exhibit the corporate governance traits that I associate with it?  Conversely, why do Australia and Canada, with somewhat higher levels of concentration than the others listed here, nevertheless exhibit such corporate governance traits?

Part of the explanation is that in Japan, China, and the Netherlands, de facto concentration has historically been far greater than these figures suggest, due to substantial government control in China, the separation of control and cash-flow rights in the Netherlands (through placement of stock in “trust offices”), and the keiretsu system of cross-shareholdings in Japan (though the latter have unwound over recent decades, giving rise to the natural experiment that I described in a prior post).  As to Australia and Canada, then, I acknowledge their somewhat higher levels of concentration and address potential explanations for their policy trajectories.  In each case some degree of historical, cultural, and/or geographic influence was undoubtedly exerted by influential US and UK models – perhaps facilitated by lack of good market data (at the relevant point in their regulatory histories), facilitating assumptions that their equity markets were as fully dispersed as the US and UK markets.  However, this at most bounds the range of alternatives that policymakers in these countries might have considered – it would not explain, for example, how they might have selected between attractive, yet fundamentally different, US and UK corporate governance models.  That Australia shifted from a US-style takeover regime in the 1980s toward a UK-style takeover regime in the 2000s – a period over which their social welfare system itself gradually stabilized – contradicts the notion that their corporate governance system can be attributed to historical path dependence upon UK legal origins.  Similarly, that Canada overcame the substantial pull of geographically proximate and influential US models in some of the most consequential areas of corporate governance itself suggests a decidedly internally determined policy trajectory – which I believe was impacted by the social welfare-oriented logic that I describe. 

Another aspect of the project’s scope relates to how I’ve selected and conceptualized the relevant policy domains.  Matt Bodie’s post submits that I’ve effectively reduced “corporate governance” to takeover regulation, and “social welfare” to health care, and suggests that my working definition of each ought to have been clarified and expanded.  His observation is well-taken, and I’ll expand a bit here on why my approach takes the form it does in each context. 

I observed in my last post that takeovers loom large in this analysis because they so starkly pit the interests and incentives of shareholders against those of other stakeholders, including employees.  This does not, however, mean that other forms of corporate governance power discussed in the book are unimportant or inoperative.  Indeed, takeovers are themselves partially a function of strong removal power, which constitutes the ultimate source of power the bidder aims to acquire.  Others, meanwhile are highly illuminating theoretically – a noteworthy example being the UK shareholders’ authority to “direct the directors.”  While there may be no need, as a practical matter, to assemble the 75% super-majority required to do this (given that threat of removal requires only a simple majority), this power nevertheless dramatically illustrates the fundamentally shareholder-centric structure of the UK corporation, which conceptualizes board authority as a literal delegation of power from the shareholders themselves.

I do define “welfare state” protections in quite general terms, as Matt indicates, referring to “a baseline set of protections relating to income, health insurance, employment insurance, and various other ‘social services.’”  To some extent this degree of abstraction is inevitable in a comparative study of this sort, as different countries can approach these broad sets of concerns through quite different regulatory structures.  Health care looms particularly large in my comparative analysis, however, because it represents the most stark and consequential divergence across the countries investigated – and because it tends to be of great practical concern in times of social and economic crisis, as both the takeover battles in the 1980s and the post-crisis reforms in the 2000s have demonstrated here in the US. 

Matt raises a very important point in emphasizing the relevance of labor law as well, to which the book admittedly devotes less attention.  I do, however, acknowledge that corporate law, social welfare policy, and labor law are all intrinsically interrelated, in so far as each provides means of addressing risks faced by employees in large corporate enterprises – as vividly illustrated by Australia’s handling of employee entitlements in the insolvency context.  Australian labor law imposes an obligation to pay employee entitlements in insolvency, raising the issue of how best to ensure payment.  Director liability under the Corporations Act proved ineffectual because it was generally difficult for employees to show subjective intent to deprive them of their entitlements.  Priority over unsecured creditors proved ineffectual because there was often little left once secured creditors were paid, and proposals for priority over secured creditors and mandatory insurance proved to be non-starters.  In the face of these challenges, the solution adopted by the conservative Howard government was the General Employee Entitlements and Redundancy Scheme (GEERS) – a taxpayer-funded employee protection fund.  Essentially the Howard government made instrumental use of social welfare policy to relieve political pressures from other regulatory domains – including both corporate law and labor law – demonstrating keen awareness that each of these policy domains presented alternative means of addressing employee vulnerabilities in this context. 

I’ve effectively taken single-tier board structures without employee representation as a given, largely as an incident of the explanatory domain described above.  As to why employees haven’t pushed for direct governance power themselves in the countries examined, settling rather for the “second-best” of social welfare protections – another important question that Matt raises – I suspect there is no singular answer.  Much presumably turns on idiosyncrasies of labor politics at a given place and time.  In the UK, for example, a push in the late 1970s to extend board representation to employees in large companies is thought to have failed not solely due to opposition by shareholders and employers, but also due to union fears that board representation might undercut their efforts to make gains through arms’-length collective bargaining.  

I’d note, however, that in much the same spirit that I reject any implication that regard for stakeholders through corporate law provides a functionally equivalent means of social welfare protection, I likewise do not intend to suggest that social welfare protection provides a functionally equivalent means of advancing broader employee interests – a topic I’ll take up further in my next post.