Introduction
Geographic Information Systems (GIS) is represented in this discussion not just as software, but as incorporating visualization and communication tools, information systems, analysis procedures and the mechanisms for decision support.
This discussion will be drawn from observations of the role of GIS as a tool supporting urban neighborhood revitalization. Progress in this arena has been restricted by - limited resources, small local organizations with nonprofessional staffs and boards, the degree of "distance" between grass-roots organizations and government or business sectors and fundamental political differences among many players. Opportunities for GIS tools to overcome these limitations are often overstated.
A number of models for community information and community research activities – often including some degree of GIS – have emerged over the last decade. Universities have sponsored many of these programs. A number of others have been demonstrations funded by foundations. In Milwaukee, the Nonprofit Center of Milwaukee Data Center program began as a more independent entrepreneurial model. Most of the initial funding came from community clients.
Much community information experience has been episodic – affecting communities for only short periods of time and in fairly limited ways. Few efforts have been comprehensive. Some programs have organized detailed information, but played a limited role in assuring that the data is used. Other programs have stressed service, but have struggled with access and resources to improve information systems. Many programs have been sector specific – limited to housing, health, education or environmental concerns.
As models emerge to address this challenge, it is necessary to frame critical evaluation questions that may be used to critique and refine these programs. It is important to begin critiquing the experiences. How do we assess the efforts? What expressions of broader objectives for community change might also guide the way we use this tool in the service of community?
There are three basic contexts for evaluation –
What is the value of the results?
Is the process managed well?
How has the process supported broader
community development principles?
I will review topics within each of these contexts. I will occasionally identify projects of the Nonprofit Center of Milwaukee Data Center program to demonstrate success, limitations and failures along these dimensions.
Is the material produced appropriate to the tasks community organizations are addressing? Can the data be organized to match the issues that are to be addressed?
Available data is often limited. The data may not be sharply focused. For example, a coalition of organizations seeking to change the process for renewing alcohol licenses could protest the concentration of licenses in specific areas of the city, but could not clearly isolate the problem license sites within this data. They were also not able to link crime data to license site data because the level of resolution of crime data was restricted at the time to census tract aggregates. Data may miss important elements. Many vacant lots in Milwaukee’s central city are sold to adjacent property owners obscuring the record of the growing number of vacant lots. Data may not make important distinctions. A bank loan reported in the HMDA (Home Mortgage Disclosure Act) may be to a speculator rather than to a home owner. Data may be biased toward a particular perspective. Evidence that housing stock has been increased within a neighborhood may avoid the question of whether it is affordable to local residents. Data may cover too much ground. Examination of crime rates without recognition of the different categories of crime may limit insight into prevention strategies. Data may represent an incomplete picture. A Children’s Hospital may have a complete database on children treated for asthma, but a substantial number of children may have been served elsewhere.
Actionable
Is the data actionable? Can the organization receiving the information use it to support decisions, enhance communication or persuade others?
Data should be detailed enough to inform actions. "Indicators" projects tend to focus upon a small set of numbers which may be used to persuade someone that action is necessary, but will offer little information to program planning and implementation.
Data needs to be scaled to the level of action. Even census block data is not enough for an organizer seeking to organize a block. The organizer requires "face-block" information.
Data often requires substantial digestion before it can be used effectively. The HMDA data on the RTK Internet site is an example. Although basic statistics are accessible, research into community patterns would require extensive structuring of the material.
Data should help to refine quantities of data and provide a focus. A map of the location of incidents of lead poisoning will be enhanced by using statistical analysis to isolate "hot spots" within the data to permit prioritizing and targeting of programs.
Data that only adds to existing evidence of a problem may not be at all useful when an organizational response is not feasible. For example, neighborhood organizations only tend to be interested in crime information when they are in a position to use that information in a local program.
In the mantra of information science – data must become information, information must become knowledge. Knowledge must be translated into decision support – informing choices.
Fits schedule of activities/ priorities
For organizations to adapt to better use of information, it should be available within the schedule of the organization. Strategic planning may be easily managed as an occasional activity with substantial lead-time. But for GIS to fit the management frame of organizations, material needs to be available quickly. For GIS to be relevant to the work of staff, information may need to be available immediately – at the staff member’s desktop. For the power of GIS to be fully realized in group decision-making, probing requests for additional information, scenario development and comparative analysis tools should be available in real time – while participants are exploring options. This final step is embedded in the vision of PPGIS. But it has yet to be realized in a pragmatic setting.
The timeliness of information is also an issue. In 1998, U.S. Census information is outdated for many purposes. Efforts to monitor changes in crime rates, in program participation rates, etc. are compromised when the change in rate may only be a reflection of the change in population in a neighborhood. When organizations wish to identify how they may respond to population shifts within communities, the information needs to be much more current.
The schedule of a local organization may be different than that of a technical assistance resource. For example, university calendars and perspectives toward time may seriously limit the capacity of universities to contribute to short term needs of community organizations. Timeliness is also compromised by the delays often encountered accessing information, negotiating for access, waiting for preprocessing work to be done by the organization maintaining the data and by the frequent need to rework the data for new uses.
Community organizations set priorities each year among the many possible subjects that they may focus upon. These priorities should not be distorted by which project has immediate access to data. But frequently, the research programs of data resources are shaped by the pragmatic recognition that some information is available for work and other material is not.
Accurate
How accurate are results? In community settings, information is frequently limited. But a high degree of precision may not be required. However, at a neighborhood level, accuracy is often more important than at larger settings. If ten percent of addresses can not be geocoded, that may be in part a result of a street name change that could represent a large error from a neighborhood perspective. Extractions of lists that seem complete enough for a community wide view may be deficient at a local level where individual items on the list are important.
Many barriers to accuracy are built into the data. Often administrative data retains little or no historical information. Data may be collected on an as-needed basis, with substantial variations in the age of the information. Those who collect data may introduce errors. Strategies to address these issues over time are also important.
Many times, local knowledge can be used to clear up errors and limitations in large data sets. Little has been done to structure larger data systems to incorporate changes that may be contributed by local level actors.
The level of detail can matter at all levels. Often zip code data is too crude to generate accurate results because of the high degree of heterogeneity within the areas. Even block level data may be inappropriate when organizations work with face-block information.
Insightful
Because many who work with neighborhood data do not know the neighborhood well, data and maps leave the impression that the neighborhood has been revealed. Data is often a weak reflection of reality. It is often used by those who understand an issue or place to communicate to others some of what the local organizations already know.
How often does data inform those with an intimate perspective? To what extent are myths about a community challenged by the facts? Local perceptions may be different than reality. For example, neighborhoods are often convinced that crime problems are substantial when only small increases have occurred.
The search for patterns in data can suggest unanticipated relationships. But these can be the result of errors in data or models or preparation of materials. Generally face validity is a useful check against the meaningfulness of results
How important is it that work reveal something new? Much of the time it will not happen. But the results should be valued when they do.
Offers Perspective
The substantial improvement in data access and mapping tools may mask the fact that historical information may be very limited. Such information may be more difficult to organize. But often trends are important. Is a problem increasing or stable?
It may also be important to place the experience in one neighborhood in perspective. How does it compare to other neighborhoods in the region? Less frequently addressed, how does it compare to neighborhoods in other cities? In the absence of mechanisms for sharing much more than census information, U.S. neighborhoods are woefully ignorance of how their community compares to others.
Synergistic
When data and research is focused upon one sector, an advantage may be lost. From the perspective of neighborhoods, problems are often inter-locking. Data available from only one sector, for example, housing files, may offer an incomplete perspective. Those organizations or departments that collect data tend to focus only upon the data that they collect.
An important synergy can result when information from several sources is brought together. Information clearinghouse objectives address this question by creating a central location to bring data together. Additionally, a clearinghouse may be able to explore relationships in data across data sets that by linking records at a level that is not available to the public.
Combines Qualitative and Quantitative
There is often a mismatch between those who work with the quantitative interpretation of data and those who approach an issue from a direct knowledge of persons affected by the issue. Organizations who know the story best often reject the use of data. They have often learned that political arguments can be won with a personalized approach to the issue. But the quantitative and analytical approach to data can also be an important political tool. Ultimately, "stories" can be disregarded is the listener does not accept how typical the cases may be.
Creative ways should be found to link the two perspectives. This is often done by leaving the final responsibility for reports in the hands of those who know the stories. This requires a substantial investment by all parties to learn to work together.
How does a program support itself? Is the support model sustainable?
Foundation supported programs are often trapped by the built in assumption that the funding is short term. This limits the capacity of the system to design for a long term role. Grant based budgets may also lead to substantial swings in resources from excess to substantial cutbacks.
Programs expected to raise revenue through fees can also be limited within this arena. Funding options may drive the priorities of a program toward serving the needs of well endowed organizations who can afford the service and who already appreciate the value of the work.
One element of a funding model should be to find ways to reduce costs and to deliver basic services for less and less costs. Development efforts should use grant funding to create the procedures that routinize long-term delivery of service.
A sustainable program is enhanced by a capacity to expand and contract resources as required by demand.
The costs of a program can be addressed through a number of innovations – use of variable and low costs resources – students / volunteers, involvement of the staff of organizations being served, and procedures to routinize repetitive work. Additionally, the system of resources need to be changed so that the costs of information are built into grants, programs, technical assistant pools, and other local budgets.
Replicable
Many GIS and analysis tasks are complex, even tedious, the first time they are done. When the tasks are likely to be repeated, an extra effort may be valuable as a way to reduce long term efforts. More critical is whether the work can be designed to make the process simpler the next time.
Often universities may do the opposite. Certain basic tasks are viewed as valuable experiences for new students to learn. From this perspective, it is better to start at the beginning each time. Template construction may also be a more difficult task for students to understand when they have just learned basic procedures.
Replication efforts are constrained by the wide variety of needs. Predetermined reports, map series and templates may limit an organization with a perspective and priorities different from previous client organizations.
Efficient
Given the limited resources and potentially large demand for services, the efficiency of the process can be important. More efficient procedures make repetition more likely.
An important objective should be to invest additional energy creating programming procedures that substantially reduce the cost and complexity of common tasks. When 80% of the work can be accomplished this way, resources are available for more difficult work.
Some tasks may be very difficult to make efficient. For example, the creation of a complex neighborhood assets map may require the skills and patience of a cartographer to ensure that it is comprehendible to the intended audience.
Timely
Data uses range from long term research which may take several years, to strategic planning often requiring several months, to program planning and to program management that are much shorter term. As local organizations build GIS and information into their daily work, data should be rapidly available.
Convenient and responsive "intermediary" structures also increases the likelihood of use. Data will be most used when those unaccustomed to the resource do not have to wait long periods of time for results.
Immediate
Ideally, access to information should be immediate. Material should be available to inform the decision making process as it occurs. In some settings, small groups have met with a computer expert around the computer while reviewing various options. The idealized vision of PPGIS suggests hardware, software and data resources brought to the center of an active decision process with citizens directing the use of the tools to support the discussion.
At a more basic level, organizations that can use GIS information on a routine basis in their work would benefit from its integration into the local organization. In these cases, GIS would be only one element in case/ client management or information and referral or routine monitoring or investigation activity. In Milwaukee, six small community based organizations have collaborated in the development of software for their individual use. The software expands upon traditional MIS designs to incorporate linked information about the entire neighborhood being served – a CIS (Community Information System.) And the concept is being further extended by introducing a GIS module. The challenge is this effort is the need to change the style of work within these organizations to fit the information routine.
Sophisticated
GIS and the related techniques can be complex. Often what is called GIS is merely map making. When a series of addresses are presented as a layer of points, the analysis may have only begun. What is the pattern? What does the pattern correlate with? Are certain locations above or below the expected value given correlated effects?
On the other hand, complex techniques can be difficult for laymen to understand. Additional efforts are required to insure that the results are clear.
Not every task requires sophisticated techniques. Sometimes the appropriate product is simple. Reducing complexity may also lead to more approachable data. It has been argued that newsprint and crayons may be the best tools to engage citizens in the use of maps for neighborhood visioning. The technology is then not a barrier to citizen understanding.
When GIS systems are designed to be easy to use, the design may compromise complexity. Or the more complex procedures may be so much more difficult than the more automated ones, that laymen will avoid them even when they are important. Popularizing GIS software in simplistic modules may lead to a generation of misuse or under-utilization of the real potential of GIS. It may be argued, for example, that the introduction of "Community 2020" as a GIS solution allows a fast start that ultimately limits the flexibility of community users.
Does the process support a community building agenda?
No episodic efforts to work with community data can substitute for a concerted effort to create within each local community the elements of a working information system. Because this objective is so difficult to achieve, it should be a part of early initiatives. Because this objective requires concerted action, it should be a public part of all communication with others. A consensus should be encouraged toward a community mission to support the information infrastructure. No national or state initiatives can replace the local effort, because critical local data is the responsibility of local organizations.
A working local information system includes –
Data sources that have developed efficient, accurate data base systems, a recognition of a public responsibility to allow access to the data, a local cost procedure to share the data, and a set of protocols to ensure the confidentiality of data that should not be public.
A clearinghouse operation – likely to be independent of data providers – that can guide the creation of a community information system, acquire data from all sectors, serve as a reliable custodian of confidential data, work through the technical problems of data organization and linkage, archive and consolidate data, generate summaries, trend lines and indicators, cosponsor cross sector research and support access by others to public data.
A service provider – a technical resource to less technical organizations – providing consultation, education and product to others on demand.
Community research analysts examining the data to identify patterns and suggest policy conclusions – driven by questions raised by community activists.
Community organizations and activists accustomed to integrating community data and research into their strategic planning, program development, administration, evaluation and impact assessment.
Sources of support to meet the costs of these activities. Few organizations have allocated resources to information functions. Many local organizations have few resources to spare.
Increase access to information and ensure the right of access
Access to data can be a significant problem. When one organization has resolved that problem by negotiating for access, that may not improve the general right of access by others. The barriers to data access may be most often the costs of data. It is important to cover the reasonable costs of data in a way that reduces the costs for others.
Internet is viewed as a critical vehicle to ensure broad access to data. But most of the models for access are limited to either access to individual records or to a crude aggregation tool with little flexibility. Much more powerful on-line analysis tools are required to allow these sites to be of value. And, ideally, selective extractions of the data for local use would be available as well.
When local organizations do not have access to resources that would assist in organizing and interpreting data for their own needs, access to data is not sufficient. Only the strongest organizations tend to benefit when data is only available in undigested formats.
Keep priority setting in the hands of community
As the needs for expertise increases, the capacity of local organizations and coalitions to guide agendas should not be compromised. Many other perspectives may compete to control priorities: Universities are often influenced by research agendas that may be independent of immediate local issues. Foundations tend to set "themes" that drive the work of community organizations eager to be funded. Local government agencies set priorities that may conflict with local neighborhoods. But these organizations often control the resources.
Mechanisms are required that allow local organizations to participate in the process from the beginning. The "Policy Research Action Group" (PRAG) at the University of Loyola, Chicago is one excellent example of the process. Research round-tables involve both academic and community leadership. The Milwaukee Data Center program also benefits from its independence. The program is a part of an association of nonprofit organizations. The 280 members of the Nonprofit Center of Milwaukee represent most all of the community based nonprofit organizations in the city.
Address process objectives
Data - even when packaged in sophisticated ways -may have little effect if organizations are not able to understand how to use it. If organizations are to be transformed by embracing these new tools, then attention need be paid to that transformation.
What is the role of education – usually informal – in the relationship between local organizations and those receiving service? Such education may begin with efforts to help organizations become better consumers of information. They may extend to empowering staff of local organizations to conduct much of the analysis work themselves. Education is often complicated by the rapid turnover in small community-based organizations. Ways need to be found to educate organizations as well as specific persons within them. Today, much of the educational process seems to require endless cycles as personnel changes.
Recognize the value of co-production
Joint activities between local organizations and larger research focused organizations may lead to much richer data. "Co-production" uses the resources of a community to expand the data gathering process and to involve users in the creation of data. Surveys of housing condition, current retail uses and resident priorities are often best performed by the local organizations.
Recently, a number of cities have developed "youth mapping" exercises, involving youth in a neighborhood in the development of an inventory of assets in their community. In these cases, the outcomes are not merely the gathering of data, but the broader benefits to the involvement of youth.
Increase the capacity of local community system to use the technology
Community development proponents frequently follow the adage - "Teach a man to fish and he’ll eat for a lifetime." These days options range from "Teach him how to select the best fish in the marketplace" to "Find out who owns the pond and buy it."
Given the technical challenges involved, an "intermediary" role is often important. As the intermediary educates others, how important is the transfer of knowledge and skills?
Generally, technology transfer is important only to the point that consumers are better at utilizing the results of technology. That suggests that a better consumer of services is most important.
Too much emphasis upon empowering local organizations to do their own work may divert the energy of organizations from organization and advocacy work best performed at that level.
When substantial differences over method and perception are likely, the control over the technical process is more important.
Certain functions may need to be organized on a centralized basis. Data clearinghouse and data development activities would be more complex if fragmented. They require the highest professional skills and the time frame to build capacity over many years.
Increase the power of less powerful neighborhoods and organizations
Are priorities set to identify organizations that can most benefit from significant support? It is often much more convenient to work with organizations who have the strength from the beginning to assimilate the new techniques of this technology. This may serve to widen the gap between sophisticated and more grass roots organizations.
Integrate into a broader community development process
The use of data is incidental, of course, to broader community development objectives. It is important that data initiatives be linked to these other processes. Use of GIS to serve single programs may be less valuable than service to coalitions and neighborhood wide strategic planning. The tools can be more meaningful when they assist with decisions that allocation resources rather than set speculative priorities or demonstrate needs that will not likely be met.
Neighborhood based organizations should be empowered to work with information on a casual basis – not just program management data, but neighborhood wide data. Organizations and local leadership need to develop the capacity to respond to the challenges facing their community on an independent, locally controlled basis.
Concluding Comments
The questions raised in this discussion are not meant to suggest mandates for the ideal community information system model. A number of contrasting models appear to be viable ones. Often, instead of suggesting a good and better way of operating, the questions point out a dilemma. How far should models go to transfer capacity to local organizations, how complex or how centralized an activity is appropriate?
The issues raised should be used less as a "scoring" system and more
as an assessment tool. It may not be appropriate to compare "apples to
oranges", but as advocates for more effective use of information and GIS
tools we should be seriously critiquing the whole "fruit basket" of alternatives
as they proliferate.