In The Netherlands, police ICT has always mirrored the organization of the police system. Until 1993, the Netherlands had 144 local police forces at its disposal, which were supplemented by a national police force. Since 1994, when the 1993 Police Act was enacted, the Netherlands has had 26 police forces. Twenty-five of those are regional forces, and one provides a few specialist police services on a national scale. Although the number of forces has declined steeply since 1994, a heavy stress is still put on regional autonomy, as the 1993 Police Act knows no provisions for cooperation among forces. Until 1993, police informatization was primarily a local matter, but after the 1993 Police Act was enacted, it became a regional affair. The Police Act, therefore, did not put an end to the existing situation characterized by so-called islands of automation. The Dutch police still use many different information systems that often are incompatible, which seriously hampers the information exchange among forces. Because of the sharp focus on regional autonomy and the lack of legal or other incentives to encourage cooperation, it is remarkable that the police have been striving toward the creation of a uniform and concern-wide information management in recent years. In this article, we argue that with this effort, a federative common pool resource (CPR) is called into being that can be seen as a form of administrative innovation in which horizontal intergovernmental cooperation through self-regulation is the central point. Such horizontal cooperation is of huge importance to the Dutch police system, as it is highly decentralized and as central steering, which often has failed in the past, would come with high transaction costs. A CPR can be defined as a shared facility that supplies goods or services to those participating in it. Characteristic for a CPR is either that it diminishes from use and/or that its creation and preservation depend on the participants’ collective actions. The federative characteristics of the developments we describe are best interpreted by using Davenport’s (1992, 1997) typology of types of information management. In this typology, five different models are distinguished: anarchy, feudalism, federalism, monarchy, and technocratic utopianism. Davenport’s (1992, 1997) main message is that in practice, many organizations struggle with a shift from feudalism toward federalism, the model that he deems superior to all others, as it enables us to create a common information system without the use of central steering, which is so difficult to many organizations, including the Dutch police. Feudalism, the current situation with the Dutch police, is highly unwanted, as it comes down to organizations not sharing information, which seriously obstructs many activities. Currently, the Dutch police are undergoing a transformation of the kind Davenport (1992, 1997) describes. Not long ago, its information management displayed strong feudal traits, but under pressure from central government over the past few years, the 26 Dutch police forces collectively have pursued the realization of a uniform concern-wide information management with, in Davenport’s (1992, 1997) terms, federal characteristics. Closely associated with these developments is the rise of a new institutional paradigm that differs strongly from the one currently existing in the Dutch police field and that already is influencing the (legal) base of the Dutch police system.