ABSTRACT
Two sharply contrasting mindsets have shaped scholarly research on computing in organizations. The first of these we might call the "theoretical" approach—in the loose sense by which that term is applied to ideas derived from the classic sociological tradition of Saint Simon, Marx, or Adam Smith. The second hazard is that discovery of statistical "effects" may lead us to imagine that we know more than we really do about how the association in question comes about—that is, about which specific mechanisms are responsible for such results. Few commentators have been able to resist characterizing the technology as "revolutionary"—though closer attention to their words makes it plain that they have many different sorts of revolutions in mind. In the quest for aggregate evidence of cost-effectiveness across the organizations in our panel, our analyses showed little evidence that computerization enhanced cost-effectiveness in any way that we could measure. The sort of rationality the authors identify here is strictly formal or instrumental.
