Managing Metropolitan Futures: The Role of Information Intermediaries

This paper reviews the literature and the experience of the development of Community Statistical Systems deriving from the original Community Information Network movement in the 1980s. The paper traces the evolution of these systems into the second generation of Neighborhood Information Systems1 , or Community Statistical Systems, which with the development of technology provide powerful integration of multiple databases for improved community outcomes. This paper provides a conceptual review of the development of Community Statistical Systems and their precedents in Neighborhood Information Systems and Community Information Networks. The paper presents findings for the development of a new approach to the effective deployment of Community Statistical Systems.2 Geographic Information Systems technology enables mapping of any number of neighborhood trends and patterns. By combining layers of information about a place, GIS enables comprehensive evaluations of the area and the development of Neighborhood Information Systems (NISs) that build integrated data sets. There are two levels of NIS use that fulfill different purposes: "transactional" use based on individual points of data and "analytical" use based on the transformation of data into information. Typically, neighborhood information systems are used by community groups to find specific information about individual property parcels. That, in and of itself, can help promote better community development. However, information about individual property values, for example, has limited use in revealing patterns for the neighborhood as a whole. On the other hand, in the aggregate, analysis of house values can be used to develop neighborhood price indexes, which in turn can be powerful indicators of the relative economic health of different neighborhoods across time and space. The two uses have different functions. One supplies raw data about land or housing parcels, while the other provides information on the community. However, in general, community statistical systems and neighborhood information systems cannot easily supply the totality of these data sets to support this second function, nor can they update these data in a timely fashion. We propose here the development of a new approach and a new model for neighborhood information systems. To date, most such systems are hosted at a single location by a single server provide by a data intermediary. The power of an NIS derives from the integration of multiple data layers for interpretation, analysis and discovery of patterns; the relevant data sets change, depending on the user. Space and cost limitations make database storage of all relevant data sets unwieldy and probably infeasible. The alternative and better approach is to develop a distributed web strategy with a different role for the data intermediary. Instead of hosting all requisite databases (and having to continually update them), the databases would instead remain with the original data providers and the intermediary would facilitate the connections between different databases and the end-user. The intermediary would also develop compatibility protocols to ensure that all databases can be integrated into the same GIS template. In addition, protocols for automatic updating of data sets and for determining access would be functions of the intermediary. The potential for this new model of NIS is great since ultimately access could be provided to relevant administrative data sets, currently used for the day to day business of governments, which could then be deployed for informing and improving community development policies. 1 The Wharton Real Estate Department Geographic Information Systems (GIS) Lab of the University of Pennsylvania developed a Neighborhood Information System (NIS) prototype with the support of a grant from the Fannie Mae Foundation. 2 An annotated bibliography of relevant literature is provided in Appendix 1.