The Urban Institute has launched the National Neighborhood Indicators Project (NNIP), a multi-year initiative to develop measures of changing social, physical, and economic conditions of neighborhoods in cities throughout the United States. The overall objective is to help local institutions develop a comprehensive and technically sound set of indicators of neighborhood conditions, so that community residents, public officials, and civic leaders can better plan appropriate strategies to improve their communities. Two central principles underlie the project. First, the indicators must be formulated in a participatory process that includes residents and experts. Second, the indicators must be capable of affecting citizen action and public policymaking. In this paper we will examine previous indicator experiments, review the scholarly literature on this topic, and reflect on the opportunities and challenges faced by the neighborhood indicators projects.

Raymond Bauer defined the term social indicators as "... statistics, statistical series, and all other forms of evidence ... that enable us to assess where we stand and are going with respect to our values and goals, and to evaluate specific programs and determine their impact" (1966, 1). In this paper we use the terms social, urban, and neighborhood indicators. We differentiate between the geographic unit of analysis for which the data are gathered (household, census block, block group, tract, neighborhood, city, county, state, nation), and the scale at which the data are reported (all neighborhoods in one city, all states in the nation, etc.). Historically, the three terms, social, urban, and neighborhood indicators, have implied as their focus the nation (social), its metropolitan areas (urban), and small areas within a city or metropolitan area (neighborhoods).

The idea that indicators are tied to specific levels of geography can play a passive or active part in indicator development; that is, the conscious choice of geographic level may or may not be a key part of the indicator system. For example, the Annie E. Casey Foundation (1994) issues an annual report of indicators measuring the welfare of children. The data get reported for the nation as a whole, for many states, and for some city subareas such as New York City. In this case, geography plays a somewhat passive role. The unit of analysis is a matter of administrative convenience. By choosing to focus on states, the foundation is not implying that the state is the most important governmental unit making policy pertaining to the welfare of children. States are simply a convenient reporting unit.

In contrast, advocates of neighborhood indicators argue that the geographic place plays an active role in altering the levels of the indicators, and thus in changing the lives of the people living in those neighborhoods. The neighborhood is not viewed simply as a convenient unit of analysis. In practical terms, the neighborhood as a unit of analysis is quite inconvenient. The neighborhood is chosen because some believe that the neighborhood-level indicator can be a tool to change people's lives.

Historically there have been few attempts to develop neighborhood-level indicators. The focus has been on larger units of analysis such as cities, counties, states, and nations. However, two recent factors have combined to create a climate for the increasing use of neighborhood indicators. The first is the development of low-cost, high-powered microcomputing, including desktop geographic information systems (GIS) software. GIS software permits address-matching, the ability to take a data record, like a birth, and place the record's address on a latitude-longitude point in space and then on a map (Cooke and Maxfield 1967; Drummond 1995a). In addition to the address for the birth, the record might contain the baby's weight, extent of pre-natal care, mother's condition at birth, and mother's demographic profile. With many records located in space, the GIS can then aggregate them to any level of geography: city blocks, neighborhoods, census block groups, tracts, municipalities, and counties. Table I shows examples of administrative records used by The Atlanta Project Data and Policy Analysis group in support of neighborhood activities.

The second factor causing interest in neighborhood-scale indicators is the shift of responsibilities for social and economic welfare from the federal to the state and local levels, and the simultaneous emphasis on public-private partnerships and neighborhood empowerment (Wallis 1994). These approaches are the latest attempt to forge new alliances for small-area improvement.

To be successful, the new participatory approaches to neighborhood revitalization must be based on information about the social and economic conditions of these small areas and their inhabitants. This information can be supplied with modest amounts of effort by neighborhood-based information technology staffs. Thus, the focus of the Urban Institute project and this paper is on the implementation of a set of neighborhood indicators. Though "neighborhood" can be defined in many ways, our use of the term implies something less than a municipality but more than a few city blocks. Traditionally, such an area had roughly 5,000 to 10,000 inhabitants with largely similar levels of education, income, and ethnicity, and with a neighborhood elementary school at its core. Neighborhood indicators are an outgrowth of the strong interest that local leaders have maintained in urban indicators. We begin, therefore, with a review of the urban indicator movement.