Organizations are occasionally faced with technology-based and accident-triggered crises that may cause costly disasters if not handled properly. Questions arise: How should organizations, with their complex processes and human involvement, be designed if they are to perform well in such crises? Would organizations benefit from structural changes during crises? From a neo-information processing perspective that views organizations as composed of cognitively restricted, socially situated, and task-oriented actors, we argue that the causes and consequences of crises may be better understood through the systematic examination of both environmental and organizational factors. We address our research questions using a rather unique approach: a matched analysis of 80 real organizational cases and 80 computer-simulated organizations. The findings show that a crisis can present critical challenges to organizational performance both externally and internally, and that there is no design guarantee that a high-performing organization will continue to perform well during a crisis situation. In addition, when organizations restructure to adapt to crisis situations, they often face the serious challenges of having to understand not only the external environment, but also organizational design traps. Key words: organizational performance; organizational design; computational modeling; real-crisis cases Whether theories of organization can be applied to nonconventional events or crisis situations has largely been assumed but certainly not fully explored (Carley

Organizations sometimes face critical situations or crises that can induce severe consequences or even disasters if wrong decisions are made. The bulk of crisis management research has relied heavily on case study methods yet often with rhetorical or even inconsistent suggestions. With an exclusive focus on crisis prevention, the issue of how organizations can maintain good performance when faced with critical situations has largely remained unexplored. There is also an apparent lack of consideration regarding how aspects of organizational design and task environment interact and affect organizational performance under critical situations. In this paper, we attempt to address this issue from an open system’s perspective and integrate techniques of computational modeling with the analyses of two crisis cases, the Vincennes incident and the Hinsdale incident. The use of a computational model with strong organization theory foundation has provided a systematic mechanism for abstracting empirical information and generating theoretical results, thus complementing conventional case analyses, which thrive on in-depth information but are often limited by the lack of analytical ability to provide theoretical insight that goes beyond empirical descriptions. For the two crisis cases, the study shows, through detailed quantitative illustrations, that the computer model can be very effective in predicting organizational performance and suggesting designs that organizations can employ to mitigate the impact of crises. This study has demonstrated that our approach of computational case analysis can be very successful in providing systematic and explicit guidance for effective crisis mitigation both theoretically and empirically.

Organizations are occasionally faced with technologically based and accident triggered crises that can be extremely costly. An example is Bhopal, a major chemical spill. In the aftermath of such a disaster, organizations, both the one that suffered and others in the same or similar industries, often reexamine how they are structured. The questions arise: how should organizations be designed if they are to perform well in such crises, and would organizations benefit from structural changes during crises? We address these questions using a combination of computational analysis and archival data on 69 real organizations faced with crises. For each crisis incident, we contrast the organization’s predicted and actual performance in a crisis situation. We find a high level of validation for the computational model. Using the validated computational model we then go on to address the hypothetical question: what is likely to have happened if the real organization had responded differently to the crisis. The findings show that there is no guarantee that a well performing organization in a general situation will continue to do so during a crisis situation. In addition, how to design or restructure an organization to mitigate the impact of crises will depend on the various design factors the organization employs. This work further demonstrates that often, the lessons learned by organizations when responding to crisis situations may be exactly the wrong lessons.