Configuring the IS Function in Complex Organizations

Abstract

Abstract How to configure the Information Systems (IS) function in large, complex organizations appears to be an eternal management issue (Dickson et al. 1984; Niederman et al. 1991). This is because it is inherently difficult and because circumstances change. Commonly phrased as a centralization versus decentralization question, it is also often perceived by managers as the issue which most impedes their organization’s ability to get value for money from IT. The two basic arguments are often expressed thus. Centralization of the IS function is necessary to reap economies of scale, ensure the ability to integrate applications or share data, and optimize the use of scarce resources. Conversely, decentralization of the IS function is necessary to ensure that IS responds to real business needs, to encourage managers to get involved with IS, and often to add control of IS resources to the autonomy that local units possess. These arguments may appear to be no different from the normal tensions of the centralization versus decentralization debate at large. However, often there are at least two special characteristics at work. First, information technology does raise important questions of scale, infrastructure planning, risk, and change. Secondly, information systems can often be an emotional issue. Since information is often at the heart of power, arguments about the control of information processing can seem critical in the politics of organizational design. Indeed, nowadays information systems are something everybody has a concern about and thus can be guaranteed to stimulate a debate—at least in complex organizations. By complex organizations we mean large, multi-business unit corporations. The authors studied fourteen such companies in Europe in 1987.