Neighborhood Indicators: Taking Advantage of the New Potential

Monitoring Trends at the Neighborhood Level. And finally, though in fewer places so far, there is monitoring of changing neighborhood conditions within cities, predominantly by the partners in NNIP. Their first work for this purpose was completed in the early 1990s, well before many other indicator initiatives that still maintain data only at the city level—see in particular Coulton, Chow, and Pandey (1990), and Bailey (1991). As noted earlier, institutions in a number of other cities are now at various stages of the process of developing similar capacities.

"What is your recommended list of neighborhood indicators?" This is a question participants in NNIP are asked frequently. The response is normally that there is no one correct list. Stakeholders in each city ought to get together and devise a list based on their own local goals and the issues that are critical to them at the time. They should also expect that the list of indicators to be monitored will change over time with shifts in observed conditions and policy priorities.

Table 2 offers an illustrative list, prepared by neighborhood groups participating in the Cleveland Community Building Initiative in a process facilitated by NNIP's Cleveland partner (Milligan, Nario-Redmond, and Coulton, 1997). This list highlights an important issue in this field at this point. It identifies 110 individual indicators the group would like to monitor but only about half of them can be derived from existing data sources (census and local administrative files). If they are to be monitored, the rest will require special surveys—always expensive. Clearly, even with the advances noted earlier, many of the possible indicators that are likely to interest local stakeholders can not as yet be incorporated at low cost.

Using Neighborhood Indicators to Change Things. Although monitoring trends in outcomes is the application most discussed in the literature on indicators, it is not the most important use of the systems of NNIP partners or others who have developed them. Instead, they have focused much more on getting their data used directly in policy change and action program initiatives, particularly in poor neighborhoods.

In doing this, two features of their approach have been noteworthy. First, even though many are trained researchers or planners, they have generally avoided using the data in writing research reports or developing plans themselves. Alternatively, they have focused on helping relevant community and citizen groups use the data (and think with it) directly. Rather than functioning as policy analysts, they function as facilitators of policy analysis by local stakeholders. This is what underlies their term, "democratizing information," especially as they work with residents in distressed communities.

Second, they have found a balanced way to deal with the tensions in planning between the desire for rapid results and the desire for comprehensiveness. Most planners have been trained to value comprehensiveness, and today's community building is giving that theme new currency with its increased recognition of the interrelatedness of the many problems of poor neighborhoods. Yet, comprehensive analysis and planning processes often leave the participants exhausted. Instead, many have advocated starting smaller, working on only one or two issues at a time, and achieving real results rapidly so that they can motivate stronger participation in the future. Some of the most successful practical uses of neighborhood indicator systems have worked in the latter mode, but with an important addition: They have worked on each issue explicitly in a manner that links it to others and leads toward more comprehensive accomplishment over time. Examples below illustrate this approach: (9)

The link between vacant structures and crime in Camden, NJ. Many of the more important cases start with the preparation of just one, or a few, simple maps. In one of these, a community-based church organization in Camden felt that the growing number of dilapidated and vacant housing units in its neighborhood was conducive to crime. Members started by preparing a map with dots indicating recent crime locations overlaid on shades indicating the housing vacancy rate in each block, and then presenting it at a public meeting. The map showed that the group's perceptions were correct: high vacancy rates did imply substantially more crime. But it also helped to motivate a constructive response. Schmitt (1997) notes:

The church group went on to develop and analyze additional (more detailed) data on this issue. This was used in their joint work with city agencies and others in designing and implementing a program to deal with it.

A more comprehensive approach to neighborhood physical renovation in Atlanta. A second example is an effort by the Atlanta Project over the past few years. Project staff used their information system to facilitate work by several neighborhood groups. For each area, they prepared maps and tables showing parcel-level data on tax delinquency, code enforcement violations, and other property conditions. Just looking over the maps, community residents saw opportunities for action they had not seen before.

Their analyses became the basis for several initiatives: targeting assistance to elderly homeowners in jeopardy of losing their homes due to outstanding tax liens; selective community reinvestment in key properties found to be ripe for redevelopment; working with city agencies to shift code enforcement strategies to crack down more effectively on absentee property owners with decaying and abandoned properties; and motivating the state legislature to pass new laws expediting foreclosure processes when communities are prepared to redevelop sites with nonprofit housing.

Cleveland's planning for welfare-to-work and related requirements. A third application was directed toward policy at the metropolitan level. NNIP's partner in Cleveland began by working with automated data on welfare cases, examining the characteristics of different cohorts of county AFDC recipients and sorting out those that would be imminently vulnerable to losing benefits under welfare-reform time limits. They estimated and mapped the numbers of these recipients by census tract and then used geographic data on employment to analyze and map spatial patterns of recent entry-level job openings in the area.

They found that the residential locations of vulnerable AFDC recipients were tightly concentrated in space, mostly in a few inner-Cleveland neighborhoods. In contrast, the entry-level employment opportunities likely to be relevant for these prospective job seekers were largely in metropolitan fringe areas. They went farther to estimate tract-level income losses likely to occur under welfare reform and to calculate commute times that would be required for AFDC recipients to access various shares of new entry-level jobs.

These basic findings were not surprising, but the contrasts were striking and the fact that the analysts had been able to quantify and map this "spatial mismatch" made a critical difference. The maps, with associated hard numbers by neighborhood, cast powerfully memorable images. They captured the attention of the local media and then of policy makers. In response, the state subsequently allocated substantial funding for transportation assistance in Cleveland's welfare-to-work efforts and local transportation planners have worked with the analysis team to test alternative strategies for moving vulnerable recipients to jobs more rapidly.

Since then, the team has begun assembling related neighborhood data; for example, on the locations and capacities of day-care centers and job-linkage services, and the pattern of rental housing affordability. Again, preliminary indications are that the production of solid data that can serve as a basis for sensible response strategies may well prove to be a critical step in motivating local actors to develop such strategies.

Ten Lessons

Evidence on the recent emergence of neighborhood indicator systems and their uses remains patchy at best. Certainly, none of it has been subjected to careful evaluation. Nonetheless, it would seem on the surface that the developments discussed here hold considerable potential. The ten early lessons below are offered in the spirit of stimulating dialogue about how cities might best proceed to take advantage of this potential.

1. Design indicator systems for the explicit purpose of changing things—not just to monitor trends. We have noted that monitoring trends in outcomes is the most commonly understood purpose of indicator systems. The NNIP partners, however, do not see that as an end in itself, but only as an instrument to contribute to their true objective: improving social outcomes, that is, changing things. For them, the focus has been on addressing urban poverty, particularly concentrated poverty in inner-city neighborhoods, although they are clearly interested in other outcomes (for example, improving environmental conditions) as well.

More than anything else, this means that you need to plan ahead so that as you finish the initial (monitoring) part of the process (stakeholders reviewing new indicator values to see what is getting better and worse), you will directly link into follow-on steps that will select priorities for action and design new policy and programmatic responses to what you have learned. If they are to be productive, those steps are likely to rely further on the indicator system, probably using more detailed data about selected issues to explore them further and test out ideas about alternative ways to address them.

This may seem obvious to some or like nit-picking to others, but surprisingly little of the growing literature on indicators acknowledges this point. Why is it important? Among other things, indicator systems cost money, and it seems unlikely that local funders will support them over the long term unless they prove themselves useful. Monitoring indicators must move beyond being an interesting exercise and show that it can contribute to better solutions to real problems. There are already a number of examples of importantly successful local initiatives that would never have been discovered or designed without indicators. Future work in other cities ought to follow those examples.

2. Develop a single integrated system that can support one-stop shopping. What happens now in most cities is extremely inefficient. Most community groups and service providers now recognize the need for data, to prepare winning grant applications if not to prepare competent plans. Some city representatives we have interviewed describe the scene as one of a large number of local players constantly "falling all over each other," all spending a great deal of time and effort trying to assemble the woefully inadequate data presently available, but with none of them able to take on the task of building an adequate system on their own.

Assigning that task to one intermediary (individual institution or partnership) and getting an adequate system built will of course entail some cost, but it is almost sure to represent a net savings in relation to the current resources so many local groups are now spending on data with such unsatisfying results. And this is to say nothing of the substantial benefit that should be realized with all users having access to data that are much richer and of higher quality than what are available now.

3. Develop indicators at the neighborhood level—not just for the city as a whole. Closely related to the first recommendation above is the conclusion that citywide indicators, while they can be valuable for some purposes, turn out not to be very useful in designing solutions to many of today's urban problems. Suppose, for example, we find that an indicator's value for the city as a whole, such as the teen pregnancy rate, fell by 0.5 percent over the past year. What would you do about that finding? Sit back and relax? Hardly! It is well known that teen pregnancy rates vary tremendously across neighborhoods in almost all American cities. It is possible that the problem became much worse in some communities and much better in others. That could imply that the average change for the city did not represent what happened in any actual place.

To know whether and how to take action to address the issue, you need to be able to assess how rates for all of the city's neighborhoods increased or decreased, and by how much. The issue of teen pregnancy (like so many others) is not one you deal with by sitting in an office at city hall or by spreading resources equally across neighborhoods. You need to know where to deploy your resources and in what proportions, and for this, neighborhood-level data are essential.

This observation would not be very helpful, of course, if assembling automated neighborhood-level data was still many times more costly than collecting the same indicators for the city as a whole. Not too many years ago, that was the case. But with the dramatic technological advantages noted earlier, it is no longer so costly to collect neighborhood data. Today, if you are serious about using local indicators, it is hard to see any justification for not doing so at the neighborhood level.

4. Build a data "warehouse"—not just a set of files on indicators. The list of possible outcome indicators that might be of interest in any city or neighborhood is quite long. Indicator gurus normally advise that, to operate the periodic monitoring function successfully, you must cut the presentation list down to a manageable number or you will you risk overloading review panels and the public. What is seldom mentioned, however, is the difference between the contents of the system you ought to maintain and the list of indicators you want to present to the stakeholders in their recurrent reviews. The former ought to be much more expansive than the latter. If it is not, given today's technology, the system is not likely to be cost-effective.

An example should help to clarify why this is the case. Automated vital statistics files (records of births and deaths) are at the core of the systems maintained by all NNIP partners. They can be used to calculate many indicators that have proven valuable for a variety of uses; for example, births to teen mothers, infant mortality rates, and teen deaths due to violent causes. They also can provide such data by race as well as location.

At any point in time, your streamlined list of indicators for trend monitoring might contain only one or two derived from the vital statistics files. But it is very likely you will want to use more of the data from those files for more detailed policy analysis later as new issues emerge. It is also likely that policymakers will revise their short list of indicators for recurrent monitoring as their priorities change, and you might need to use other data from those files to meet their new requests. You have to go to some expense to regularly obtain, clean, and integrate the whole file in order to update those one or two indicators you are now monitoring, and the costs of storing the whole file are negligible. In these circumstances, it would be extremely wasteful to throw the rest of the file away. It makes sense to keep the whole file handy so that you can respond quickly as new data needs are expressed.

This is the concept of a data warehouse. All of the NNIP partners actually operate data warehouses from which indicator reports can be derived, rather than just indicator systems. Each has a sizeable collection of large data files, all parts of which it can access quickly and efficiently when it needs to do so. Only a small share of the data in the warehouse is likely to be in use at any time—the rest is just sitting there. But since the costs of warehousing are now so low, and the benefits of rapid responsiveness in bringing good data to bear on new issues are so high, it clearly pays to operate in this manner.

5. Serve multiple users but emphasize using information to build capacity in poor communities. The NNIP experience suggests that, once an integrated system of neighborhood-level data exists in a city, there will not be a problem in finding users. For the users, the efficiency associated with being able to obtain a wide variety of neighborhood indicators all in one place (all carefully checked and in a consistent format) are substantial. All the NNIP partners provide data to a number of types of users; government agencies, nonprofits, and private firms. For users that can afford it, they sometimes charge fees for data assembly and analysis, which help cover their basic operating costs. All of them have recognized, however, that their data systems are importantly a device by which they can promote equity. In the past, poor inner-city neighborhoods have been disadvantaged in their lack of access to information. Accordingly, all NNIP partners give the highest priority to using their data to support community building in these neighborhoods.

6. Democratize information—help stakeholders use information directly themselves. As noted earlier, all NNIP partners see their role primarily as facilitating the direct use of data by the stakeholders in the issue at hand, rather than themselves serving as the primary actors in policy analysis and plan making. This contrasts sharply with the traditional model in which researchers and planners prepare the report for the users to read and review after the fact. In fact, one of the strongest motives for democratizing information arose from the recognition that a large share of such reports is ignored.

The stakeholders concerned with an analysis may not be as strongly motivated to follow up on it unless they themselves have helped to create it, step by step. They need to feel they own the findings and conclusions. And they may well come up with different, and better, answers than professionals who might do a study for them. They understand the nuances of the situation and are able to see options that the professionals might not recognize. They can benefit from professional facilitation as they do the work–for example, advice on how to handle and interpret data correctly, or policy and program ideas that have worked well in similar situations elsewhere. But they need to make the decisions because they–not the professionals–will have to live with them.

7. Help stakeholders use data to tackle individual issues, but do so in a way that leads toward more comprehensive strategies. There is now a broad consensus that, given its multifaceted nature, the problem of concentrated inner-city poverty is not likely to be addressed effectively by single-purpose social programs. It will ultimately require a holistic approach (Kingsley, McNeely, and Gibson, 1997). But, as suggested by the earlier discussion and examples, that does not mean that starting with a comprehensive strategy is always essential or even desirable. In many cases, it may make sense to start by using indicators to help design and implement solutions to one or two pressing issues. The hope is that this will yield early results that will build confidence, encouraging a more ambitious agenda in the next stage. Indicator analysis can be particularly helpful in showing how current issues, and the means of addressing them, can link to others down the line.

Alternatively, the circumstances may suggest that the stakeholders are up for more comprehensive strategy making at the outset. If so, an integrated data system can of course provide strong support for accomplishing that effectively. But, if this approach is selected, it will be risky to let it postpone too long at least some of the actions that will begin to change things for the better.

The most important conclusion here may be that planning and implementation should no longer be viewed as separate stages, but rather as interrelated concurrent processes that influence and alter each other as they move along.

8. Use information as a bridge to promote local collaboration. The NNIP partners also regularly take advantage of the opportunity to use indicator systems to establish and further collaboration between urban groups and actors that have often been at odds in the past. Collaboration is tough. If players from different groups come to the table holding tightly to their old beliefs and policy prescriptions, it may not work. Something is needed at the outset to shake up the old ways of looking at things. On a number of occasions, NNIP partners have been able to use fresh presentations of data on local conditions and trends to accomplish that (see National Neighborhood Indicators Partnership, 1996, Chapter 4). People can, of course, really disagree. Yet it is surprising how often urban policy disagreements exist only because of faulty perceptions of what is really going on (maybe by both parties), and those can be fixed by facts.

9. Use available indicators but recognize their inadequacies—particularly the lack of sufficient data on community assets. Initiatives to improve conditions in poor neighborhoods need to mobilize residents around a positive agenda, making them recognize that they do have assets and can use them to change things (McKnight and Kretzman, 1993). Accordingly, indicator systems that support such initiatives ought to emphasize measures of assets. Yet the bulk of the indicators that can be derived from administrative records deal with negative events (crime, infant mortality, etc.). Most of the asset-oriented indicators of interest to the Cleveland Community Building Initiative (Table 2) can only be obtained via special surveys. The indicators available now (mostly from administrative data) are largely negative.

How should planners and community-building practitioners respond to this issue? First and foremost, they should use all the data presently available to track things, but keep reminding residents of the assets that are out there that have not yet been incorporated into the system. In other words, use the negative measures in the system but do not let them move residents away from an asset-oriented agenda. Because we all must face reality, you do not want to suppress those measures, but it is necessary to keep them in perspective. This is easier to do now that crime rates, teen pregnancy rates, and several other negative indicators are falling in many urban communities, but NNIP partners were able to accomplish it even back around 1990, when trends were not so benign. Second, planners and others engaged in local public policy should be raising funds to support the surveys needed to expand asset measures in their local indicator systems.

10. Assure integrity in the data and in the institution that provides them. In the 1960s, if anyone had asked who should be responsible for building an information system with neighborhood-level data, the most common answer probably would have been the city planning department or some other unit in city government. In this light, it is interesting that none of the NNIP partners are city agencies. One is a center affiliated with a university, and the rest are community foundations or other nonprofit intermediaries whose missions center around a broad view of furthering the public interest in their cities. This type of institution does offer advantages for operating local information systems serving multiple users. Most importantly, they are not seen as beholden to any short-term political interests that might have incentives to either withhold or misrepresent the facts. Also, a good local system needs to obtain data from county and state agencies, and possibly some suburban jurisdictions, as well as city departments. A nonprofit might be able to connect all those sources more effectively than the city.

Ultimately, however, whether the central institution is public, private, or nonprofit is not as important as how it behaves and the reputation it develops. The institution (or partnership) that operates the system must maintain the trust of both data providers and a wide array of users over the long term. The NNIP partners are very careful about cleaning the data, maintaining confidentiality, and guiding the responsible use of their data. While they advocate using data in policy debates that are often controversial, they avoid taking sides in those debates.

End Notes

1. NNIP is a collaborative effort by the Urban Institute and its partners to further the development and use of neighborhood information systems in local policy making and community building. The six original partners (and their managers) are the Atlanta Project (David S. Sawicki), the Boston Foundation's Boston Persistent Poverty Project (Charlotte Kahn), the Center for Urban Poverty and Social Change at Case Western Reserve University in Cleveland (Claudia J. Coulton), the Piton Foundation in Denver (Terri J. Bailey), the Urban Strategies Council in Oakland (Cheryl Taylor), and the Providence Plan (Patrick J. McGuigan). A seventh, which has recently joined NNIP and is developing similar capacities, is the DC Agenda Project in Washington, DC (Rob Richardson).

2. Nonetheless, these are conclusions of the author, who remains responsible for any errors in fact or interpretation. NNIP partners may or may not agree with them.

3. Several works were published to promote the concept and explore its potential: see, for example, Bauer (1966), Cohen (1968), Bell (1969), and U.S. Department of Health, Education, and Welfare (1969). For a more recent assessment see Miringoff and Miringoff, 1997.

4. At least one researcher proclaimed that the era of large-scale urban models had ended (Lee, 1973).

5. This is the Community 2020ô software package developed by the HUD's Office of Community Planning and Development. See Kingsley et al (1997).

6. A listing and description of 42 different types of automated administrative data files typically available in most cities was prepared by Claudia J. Coulton and published in Kingsley et al. (1997).

7. In addition to the NNIP partners cited earlier, neighborhood data systems are already being operated by The Nonprofit Center of Milwaukee (Michael Barndt, manager) and the United Way/Community Service Council, working jointly with the Polis Center at Indiana University-Purdue University, in Indianapolis—the latter is the Social Assets and Vulnerabilities Indicators (SAVI) system (manager of development was Robert Hoek). The Delaware Valley Grantmakers Associations (Philadelphia) and the Association of Baltimore Area Grantmakers are now in the midst of processes to design and fund similar institutional capacities for their cities. A survey of 33 cities during the first year of NNIP found that key institutions in most of them were aware of concept and interested in exploring it further (National Neighborhood Indicators Partnership, 1996).

8 The National Civic League's Healthy Communities Handbook (1993) describes an approach to the development of local indicators and notes cities that have implemented the approach. One that is highlighted is the work of the Jacksonville Community Council (1994). See also, Andrews (1996).

9. These examples are highlighted in Kingsley, et al (1997). Chapter 3 of that report provides brief case studies of a total of 17 experiences in using automated information systems to address neighborhood and city-wide policy issues. Other cases are presented in Mitchell (1997).

List Of References

Andrews, James H. 1996. Going By the Numbers: Using Indicators to Know Where You've Been and Where You're Going. Planning (September): 14-18.

Annie E. Casey Foundation. 1996. Kids Count Data Book: State Profiles of Child Well-Being. Baltimore, Md.: The Annie E. Casey Foundation.

Bailey, Terri J. 1991. Poverty in DenveróFacing the Facts. Denver: The Piton Foundation.

------. 1997. Building Community Capacity to Use Information. National Neighborhood Indicators Project Working Paper. Washington, D.C.: The Urban Institute. October.

Bauer, Raymond A., ed. 1966. Social Indicators. Cambridge, Mass.: The MIT Press.

Behn, Robert D. 1997. Linking Measurement and Motivation: A Challenge for Education. In Advances in Educational Administration, Vol. 5, edited by Paul W. Thurston and James G. Ward. Greenwich, Conn.: Jai Press Inc.

Bell, Daniel. 1969. The Idea of a Social Report. The Public Interest, 15 (Spring):72-84.

Child Trends, 1998. Indicators of Child, Youth, and Family Well-Being: A Preliminary Inventory of Existing Projects. Baltimore, Md.: The Annie E. Casey Foundation.

Cohen, Wilbur J. 1968. Social Indicators: Statistics of Public Policy. The American Statistician, 22 (October): 14-16.

Coulton, Claudia J. 1995. Using Community Level Indicators of Children's Well-Being in Comprehensive Community Initiatives. In New Approaches to Evaluating Community Initiatives: Concepts, Methods, and Contexts, edited by James P. Connel, Anee C. Kubisch, Lisbeth B. Schorr, and Carol H. Weiss. Washington, D.C.: The Aspen Institute.

------. 1997. Potential and Problems of Developing Community Level Indicators of Developing Community Level Indicators of Children's Well-Being. In Indicators of Children's Well-Being, edited by Robert M. Hauser, Brett V. Brown, and William R. Prosser. New York: Russell Sage Foundation.

Coulton, Claudia J., Julian Chow, and Shanta Pandey. 1990. Analysis of Poverty and Related Conditions in Cleveland Area Neighborhoods. Cleveland: Center for Urban Poverty and Social Change, Mandel School of Applied Social Sciences, Case Western Reserve University.

Federal Interagency Forum on Child and Family Statistics. 1997. America's Children: Key National Indicators of Well-Being. Washington, D.C.: Federal Interagency Forum on Child and Family Statistics.

Flax, Michael J. 1978. Survey of Urban Indicator Data. Washington, D.C.: The Urban Institute.

Jacksonville Community Council. 1994. Life in Jacksonville: Quality Indicators for Progress. Jacksonville: Jacksonville Chamber of Commerce. November.

Kingsley, G. Thomas, Joseph McNeely, and James O. Gibson. 1997. Community Building: Coming of Age. Baltimore, Md.: Development Training Institute and The Urban Institute. April.

Kingsley, G. Thomas, Claudia J. Coulton, Michael Barndt, David S. Sawicki, and Peter Tatian. 1997. Mapping Your Community: Using Geographic Information to Strengthen Community Initiatives. Washington, D.C.: U.S. Department of Housing and Urban Development. October.

Klosterman, Richard. 1994. Large-Scale Urban Models: Retrospect and Prospect. Journal of the American Planning Association. 60 (1): 3-6.

Land, Kenneth C., and Seymour Spillerman, eds. 1975. Social Indicator Models. New York: Russell Sage Foundation.

Lee, Douglas. 1973. Requiem for Large-Scale Models. Journal of the American Institute of Planners 39 (3): 163-78.

McKnight, John L., and John P. Kretzmann. 1993. Building Communities from the Inside Out: A Path Toward Finding and Mobilizing Community Assets. Evanston, Ill.: Center for Urban Affairs and Policy Research, Northwestern University.

Milligan, Sharon E., Michelle R. Nario-Redmond, and Claudia J. Coulton. 1997. The 1995-1996 Cleveland Community Building Initiative: Baseline Progress Report. Cleveland: Center on Urban Poverty and Social Change, Case Western Reserve University. Fall.

Miringoff, Marc L. 1993. Index of Social Health: Monitoring the Social Well-Being of the Nation. Working Paper. Tarrytown, N.Y.: Fordham Institute for Innovation in Social Policy, Fordham Graduate Center.

Miringoff, Marc L., and Marque-Luisa Miringoff. 1997. Context and Connection in Social Indicators: Enhancing What We Measure and Monitor. In Indicators of Children's Well-Being, edited by Robert M. Hauser, Brett V. Brown, and William R. Prosser. New York: Russell Sage Foundation.

Mitchell, Andy. 1997. Zeroing In: Geographic Information Systems at Work in the Community. Redlands, Calif.: Environmental Systems Research Institute, Inc.

National Civic League. 1993. Healthy Communities Handbook. Denver: National Civic League.

National Neighborhood Indicators Partnership. 1996. Democratizing Information: First Year Report of the National Neighborhood Indicators Project. Washington, D.C.: The Urban Institute. March.

____________________. 1997a. 1996-97 Progress: National Neighborhood Indicators Project. Washington, D.C.: The Urban Institute. October.

____________________. 1997b. Stories: Using Information in Community Building and Local Policy. Washington, D.C.: The Urban Institute. October.

Oregon Progress Board. 1992. Oregon Benchmarks: Standards for Measuring Statewide Progress and Government Performance-Report to the 1993 Legislature. Portland: Oregon Progress Board. December.

President's Council on Sustainable Development. 1996. Sustainable America: A New Consensus. Washington, D.C.: U.S. Government Printing Office. February.

Prosser, William R., and Matthew Stagner. 1997. Children in Dire Straits: How Do We Know Whether We are Progressing? In Indicators of Children's Well-Being, edited by Robert M. Hauser, Brett V. Brown, and William R. Prosser. New York: Russell Sage Foundation.

Schmitt, Brian. 1997. Community Organizations and GIS Implementation: From Experience to Theory. Piscataway, N.J.: Center for Social and Community Development, Rutgers University.

Tayman, Jeff. 1996. The Accuracy of Small-Area Population Forecasts Based on a Spatial Interaction Land-Use Modeling System. Journal of the American Planning Association. 62 (1): 85-98.

U.S. Bureau of the Census. 1985. Tiger Tales: A Presentation of the Automated Geographic Support System for the 1990 Census. Washington, D.C.: U.S. Department of Commerce.

U.S. Department of Health, Education, and Welfare. 1969. Toward a Social Report. Washington, D.C.: U.S. Government Printing Office.

Walsh, Joan. 1996. Stories of Renewal: Community Building and the Future of Urban America. New York: The Rockefeller Foundation.

The author is a senior research manager in housing, economic development, and urban policy at the Urban Institute where he also serves as director of the National Neighborhood Indicators Project.