Thursday, December 13, 2012

Design for Regulation

One source of complexity in organizational design is the requirement for various forms of regulation and governance. This requirement results in significant levels of bureaucracy and management overhead, as seen across much of the public sector as well as in many large commercial organizations.

It is tempting to think that much of this regulation and governance is redundant, and that such organizations would be far more efficient and effective if these layers of bureaucracy were eliminated. Of course that will sometimes be true, but it would be architecturally wrong to eliminate any specific regulatory mechanism without understanding the purpose and systemic effects of this mechanism.

In general system terms, some form of regulation may be required to maintain integrity of structure and behaviour. The architect needs to understand the abstract need for regulation, without prejudging any specific mechanism or institutional solution. This understanding calls for the Cybernetic Viewpoint.

The Cybernetic Viewpoint allows us to consider the existence, purpose and style of regulation, while deferring the question of responsibility for regulation and its technical mechanisms.

There is a considerable debate about the proper style of regulation. There are many voices speaking for principles-based or performance-based or risk-based regulation, typically contrasted with rules-based or process-based regulation.  The need for precise specification and enforcement in a given situation may depend on the degree of trust.

The existence of some form of regulation may be inferred from some form of regularity in the structure and behaviour of a system, even if we don't always know the precise mechanism that produces this regularity. This inference may follow Stafford Beer's POSIWID principle - the Purpose Of a System Is What It Does.

Aside from the style of regulation, there is even greater debate as to the responsibility for regulation. There are emergent forms of regulation (such as market forces or consumer choice), collaborative forms (such as industry self-regulation), and institutional forms (such as independent regulatory bodies). We can identify four main categories as follows. (The labels for these categories are copied from the System-of-Systems world.)

  • Directed regulation - central regulation by a single regulatory authority
  • Acknowledged regulation - where there are multiple aspects of regulation handled by specialist regulatory bodies. 
  • Collaborative regulation - where regulation is distributed around the ecosystem, such as self-regulation
  • Virtual regulation - where a regulatory effect is produced by crowd behaviour, such as market forces

People often have strong preferences between these categories, partly based on political affiliation, and partly based on the degree of trust in the players.

To sum up, architects need to think about regulation both in abstract terms (where market forces, self-regulation and statutory procedures may be regarded as alternative and potentially interchangeable solutions to a common regulatory requirement) and in specific terms (where the implementation of a given regulatory instrument may have a significant impact on system and organization design).

John Kay, Regulation by Rules or Regulation by Values (Feb 2000)

Arnold Kling, Why We Need Principles-Based Regulation (American Enterprise Institute, May 2012) via EconLog

James Surowiecki, Parsing Paulson (New Yorker, April 2008)

Robert Teitelman, The paradoxes of rules-based regulation (Deal Economy, Sept 2011)

No comments: