Decentralized Judgment vs. Leadership

In a recent post on the Harvard Business School 'Working Knowledge' site, professor Jim Heskett - referencing James Surowiecki's excellent book 'The Wisdom of Crowds' (see right) - raises pointed questions about how decentralized wisdom and market-based forecasting methods challenge traditional notions of 'imperial' corporate leadership. Money quote:

"...these studies could help us understand why some organizations are more successful than others, why organization size can matter, why decentralization and diversity works, and why lines of communication that enable members at all levels of an organization to make their views heard are so important. If this is the case, why is so little attention paid to this kind of research by corporate leaders? And what does it say about the true nature of effective leadership?"

My own theory, as noted here, and here, is that this phenomenon is still in its embryonic stages. As such, only the boldest and most personally secure leaders will be willing to acknowledge the cumulative prediction powers of their organizations (and their customers!) as potentially superior to their own. Prediction markets could finally make the business case for 'walking the walk' of previously vague theories and intentions around knowledge management, diversity, and empowerment - to name just a few recent fads.

Though he is hardly alone, one thing I take issue with in Heskett's otherwise insightful treatment of the subject is his too-easy conflation of prediction with decision-making. As with AI and computer modeling (for example), it's not necessary to tightly couple a source of input and an analytic method with the executive decision-making process itself in order to acknowledge the promise of a new class of decision-support tools.

Comments

[John N. Kelly]

Individual event prediction by markets can, even if initially inaccurate, gradually contribute to organizational resilience. Organizations can nuture resilience if they repeatedly use a disciplined, sensitive form of a predictive market AND ask for, record and publish (perhaps with elective anonymity) the reasoning of participants for making their predictions.
Over time, a relatively stable group of market participants working in the same general area of business, technology or policy acquire 'market derived' expertise. The challenging discipline is to remain open to novel reframing in light of market data while continually refining conceptual models of the field under investigation.

[Art]

My colleague John makes an excellent point - that prediction markets ought to be but a part of a larger framework for information-sharing, strategic dialogue, and decision-making within the enterprise. -Art