Discussion: View Thread

CALL-OSWC-XIV

February 7-10, 2008

On the Organization Science of Extreme Events
 

We are very pleased to announce the Fourteenth Anniversary Organization Science Winter Conference (OSWC-XIV). OSWC-XIV is sponsored by Organization Science in order to stimulate theory and understanding about organizations through experimentation and boundary-crossing conversation. It combines the leading-edge ideas of strategy and organization scholars, executives, entrepreneurs, and interested non-business scholars, in a community-enhancing setting. The time and place are February 7-10, 2008 at The Resort at <st1:place w:st="on"><st1:city w:st="on">Squaw Creek</st1:city>, <st1:state w:st="on">California</st1:state></st1:place>.

Most cases used in the MBA classroom are about situations that, in fact, are rare, outlier, or extreme examples of effective or ineffective organizational and/or managerial functioning. The executive-oriented books written by academics or practitioners are inevitably about outlier examples of good or bad organizations. Why is this? First, because many practitioners actually face a world of dynamical changes and challenges embodied in rare events, outlier cases, and extremes having positive or negative outcomes. Many examples come to mind: Toyota, Control Data, Monsanto, Ikea, and eBay at one end, with Enron, Parmalat, Katrina/New Orleans, 9/11, Challenger, and Pioneer disasters at the other. Second, managers don't really need the advice of organization science scholars when faced with "average" situations. It is when they confront extreme events, emergent outcomes, irregularities, or crises that managers should find it useful to learn from organization scientists. However, a study of the empirical research in the social sciences will find that the vast preponderance of empirical papers are based on Gaussian statistics, which assume "normal" distributions, averages, and "normal" confidence intervals. Such empirical approaches are not very effective for studying coevolutionary dynamics or extreme events that result in Pareto distributions with long tails, unstable averages, variance which is (nearly) infinite and consequent wider confidence intervals.

For OSWC-XIV we wish to focus on the idea that extremes are of critical importance to practitioners, and therefore we wish to explore the organization science of extremes and related coevolutionary dynamics. Consequently, the CALL for participation in OSWC-XIV centers on the following question:

How to open up traditional organization science research and its reliance on Gaussian statistics to research studying extreme events, with both an equivalent scientific rigor and improved relevance to the world of managerial practice?

The list of potential topics that follows hopefully will provide a better sense of what we mean by the study of extremes.

What defines an extreme or rare event in organizations? What are the constitutive attributes, differentiations, possible degrees or classifications of extremes?

Studies of organizational phenomena showing extreme positive or negative outcomes. Studies of disasters   are welcome, as are studies of positive or negative extreme events in general.

Studies of how managers/organizations learn from extreme events in organizations (processes of learning    and challenges of learning in the midst of, and from, extreme situations, etc.).

What are the challenges from a managerial perspective for coping with extreme or rare events? What             implications, risks, or organizational changes can occur from extreme or rare events?

What impacts do extreme and rare events have on behavioral aspects of the organization and its members? Studies of structure and process impacts are also welcome.

What are the implications of extreme events for the current educational approaches and methods of Universities and MBA programs and the practices of consulting firms?

The growing interest in complexity leadership suggests some possible themes such as how should leaders, CEOs especially, cope with and take advantage of scalability and fractal dynamics and, what are implications for strategic thinking in a fractal world?

We especially invite papers incorporating "scalable" phenomena. "Scale-free theories" are used to explain the foregoing dynamics and appear applicable for studying and explaining a variety of organization science phenomena. The following are some examples involving scalable phenomena:

Studies showing power law phenomena. We have a particular interest in studies that attempt to investigate underlying conditions that strengthen or weaken the power law signature. Power law studies could signify fractal structures within organizations, fractal structures of industries, and fractal structures of broader predator/competitor, resource, and ecological phenomena.

Studies that address fractal dynamics and the tell-tale tiny initiating events-also called the butterfly effect-and other scale-free phenomena giving rise to extreme outcomes and power law phenomena.

Positive-feedback-based coevolution is the best known scale-free theory in organization studies, but there are many others such as Zipf's "least effort" and Barabási's "preferential attachment" theories. Studies offering evidence of scalability and scale-free dynamics, with accompanying attempts to explain such phenomena with scale-free theory, or that test existing scale-free theories, are of interest to this conference.

Studies about existing methods, problems, and approaches that could or should be used to research coevolutionary and other scale-free dynamics spiraling into extreme and rare events in organizations.

Since extreme events often appear to be samples of one (or very few), we invite studies that pick up on the challenge of advancing the scientific methodology for researching N = 1 samples. We invite papers using multiple observations of the one event; applications of methods such as hermeneutics and coherence theory, and novel advances for studying N = 1 extremes. More generally, studies developing Paretian methods are most welcome. Building from how geologists study #9 quakes wouldn't be a bad idea.

Agent-based computational models are often coded to produce agent interactions having some possibility of generating extreme outcomes. We welcome agent-based models that study scalable dynamics and extreme events.