Doyle, Watson, and the Purpose of the Corporation

Sherlock Holmes aficionados distinguish between literary criticism that is “Watsonian” in perspective, and criticism that is “Doylist.”  As any fan knows, the stories were written by Arthur Conan Doyle as a first-person narrative; they purport to be the work of John Watson, who is recounting the exploits of his brilliant friend and sometime-roommate, Sherlock Holmes.  Fans who analyze the stories, then, have a choice: They can take an “outsider” perspective and discuss them as works of fiction authored by the real-life person Arthur Conan Doyle, or they can take an “in-universe” perspective and discuss them as the actual literary product of John Watson, unreliable narrator.  Depending on which viewpoint you adopt, you may end up in strikingly different conversations.  For example, a Doylist might look at inconsistencies in how Watson’s wife is described throughout the series, and attribute them to the multi-year period over which the stories were published; a Watsonian might argue that Watson was covering for a gay relationship with Holmes and couldn’t keep his lies straight.  Neither viewpoint is incorrect, but the two fans are talking past each other; in order to communicate, they have to define the relevant playing field.

That’s how I feel about a lot of the conversations currently surrounding corporate purpose, especially the ones you see in popular media. 

We’ve had a lot of soul-searching recently about whether corporations should be run to benefit society overall, or whether they should be run to benefit their shareholders alone.  But that conversation is incoherent unless you first clarify your perspective.  Are you looking at things from outside the corporation, in terms of structuring our overall legal and societal institutions?  Or are you looking at things from inside the corporation, in terms of how corporate managers should understand their jobs and their own roles?

From a societal, or Doylist, perspective, I don’t think there’s any dispute that corporations exist to serve the community as a whole.  We charter corporations, we create rules for their operation, we develop infrastructure to facilitate investing, all because we believe that on balance, corporations are (or can be) a net good.  They are an efficient way of doing business, which means they contribute to innovation and economic development, provide necessary (or even just enjoyable) goods and services, generate wealth not only for investors, but also for workers and governments (through tax payments).  They can provide outlets for creativity and generally contribute to human flourishing.  That is the social purpose of a corporation.

But corporations harness and coordinate labor and capital on a potentially global scale, and thus are very powerful tools.  Any form of power can be misused.  Corporations might exploit and injure workers, or consumers, or the environment, perhaps to the point where the benefits are not worth the costs.  Thus, we need to arrange our societal institutions to minimize these harms, and maximize the benefits.

Corporate purpose debates are not about those principles – on which, I suspect, everyone agrees.  The debate about corporate purpose is a debate about method.  If we agree that corporations exist to benefit society, and if we agree that we need some kind of legal and/or market structure to ensure this occurs, what does that structure look like?

And this is when we switch to the Watsonian perspective, from within the corporation itself.  And here, the question is, is it better that corporate managers understand themselves to be servants of society, and manage the corporation to effectuate that purpose?  Or is it better if managers understand themselves to be serving investors, while other societal institutions – regulation, contract law, and the like – protect the rest of society?

This is a point I’ve made repeatedly, most recently in Beyond Internal and External, but also in Not Everything is About Investors, and I’m hardly the first to do so.  For example, though Henry Hansmann’s & Reinier Kraakman’s essay, The End of History for Corporate Law, has received its share of ribbing for being a bit premature, it lays out this framework very neatly, while arguing that society overall benefits if managers focus on investors, while we reserve other types of regulation to protect non-investor constituencies. 

Unless these premises are understood by everyone in the conversation, it devolves into the same incoherence one would expect from a Doylist discussing Watson’s marriages with a Watsonian.  So, for example, the New York Times recently published a retrospective on Milton Friedman, with soundbites from assorted businesspeople and academics.  You will, I’m sure, be shocked to learn that every business person included argues that corporations should be run to benefit all stakeholders.

Now, there’s a certain banality to this exercise – what CEO is going to say “screw my customers, I’m all about the stock price”? – but more importantly, there is no dispute that corporations should operate to benefit all stakeholders; the issue is how do we make that happen. One method – and only one method – is to rely on managerial largesse to distribute surplus to shareholders and stakeholders alike. (We’ll call this the “Martin Lipton” method*)  But there are other mechanisms to constrain corporate behavior besides managerial largesse, and it’s impossible to talk about the merits of such largesse without acknowledging those mechanisms and discussing how they function.  The question, properly framed, isn’t whether CEOs should consciously operate their companies to serve society, but what are the options we have for making sure they do so, and which mechanisms are more effective than others, and why?  If CEO altruism is one of our options, is it better or worse than other possibilities, and if we are going to rely on altruism, what institutions do we need to generate that altruism and channel it appropriately?

Which is why I find the NYT piece so frustrating, because it’s got the Doylists and the Watsonians all mixed together as though they’re talking about the same thing.  Many of the academics are Doylistically describing the types of societal structures we need to corral corporate power and ensure that capitalism benefits everyone.  Meanwhile, Starbucks’s Howard Schultz, Home Depot’s Ken Langone, J&J’s Alex Gorsky, and BlackRock’s Larry Fink, among others, take the Watsonian view, which is to say, they argue what all businesspeople argue: CEOs should run the company with a view toward serving shareholders because you cannot serve shareholders without serving the rest of society.  From Watson’s perspective, if the CEO is focused on maximizing profits, s/he will make good products that consumers want to buy, and create good jobs that attract high-quality employees, which satisfies Doyle’s desire for a better society overall. 

But by focusing on the Watsionian viewpoint and eliding the Doylist challenge of the academics, these businesspeople avoid any test of the very specific factual claim that undergirds their argument: that the interests of shareholders and the interests of other stakeholders are aligned.  And in order for that to be true – that shareholders cannot profit unless the rest of society benefits – nonshareholder constituencies must be sufficiently powerful Doylistically to extract a price for corporate malfeasance.  They must have, in Galbraith’s words, countervailing power, from labor unions, a strong regulatory system, consumer advocacy groups, and so forth. 

That line of thinking yields two possibilities: We can maximize the benefits provided by corporations, and minimize their harms, by strengthening these countervailing institutions, or we can do it by weakening corporations. The latter, for example, was long the goal of antitrust law, and it’s why there’s so much advocacy around limiting corporate political donations.

Which brings me to Jens Dammann and Horst Eidenmueller, who have written a pair of papers that arguing that co-determination (whereby employees, as well as shareholders, get to vote for corporate directors) may not strengthen corporate functioning but instead weaken it, by creating a type of separation of powers within the corporate form, and that itself may be net beneficial to society.  I made a similar point in Beyond Internal and External, where I argued that the regulatory system shapes shareholders to have divergent preferences in a manner akin to the separation of powers.  The separation of powers has two Watsonian functions.  The first is that it encourages a variety of incentives and goals among corporate managers, which encourages a broader perspective in corporate decisionmaking.  The second is that it impedes any kind of corporate action in the first place, by making it more difficult for corporations to reach a consensus.  In that vein, certain kinds of corporate governance reforms – elimination of dual-class stock, separating chair and CEO roles, and so forth – seem less about ensuring good (profit-maximizing) governance than creating friction in governance, because by impeding the corporation’s ability to act, we necessarily strengthen other constituencies. 

Okay, yeah, so just imagine I have a pithy conclusion here.  Whatever, it’s a blog post.

*no relation