1 Introduction

Collective Impact (CI) strategies bring together nonprofit organizations and governments in a structured way to move the needle on social issues using shared agendas, activities, and communication strategies (Kania and Kramer 2011). A major emphasis of these efforts is on measuring outcomes and impacts. Doing so requires gathering data from the sometimes hundreds of organizations involved and triangulating this data with more specific research studies, as well as neighborhood- and community-level economic impact assessments. This data collection is a form of developing the collective intelligence of organizations involved with the hope of serving the common good. CI efforts are rapidly growing in popularity, both in the form of grassroots organizing strategies among organizations (“grass-tops,” e.g., the Dudley Street Neighborhood Initiative’s Promise Neighborhood) and policy approaches (top-down, e.g., the federal Promise Neighborhoods policy initiative).

My study focuses specifically on one arena in which CI is being used, the Promise Neighborhood. Promise Neighborhoods (PN) are a US federal program of the Department of Education’s Office of Innovation and Improvement. The program began in 2010 and was inspired by the apparent success of the Harlem Children’s Zone (HCZ) wrap-around services strategy for making headway on the achievement gap. The basic principle of the HCZ is to provide coordinated services to all families in a particular geographic region deemed necessary to help all children in the region to graduate from high school and continue on to higher education. Their theory was that better classroom instruction alone would never close the achievement gap. Necessary to any real change was to take seriously the many barriers to education that exist in the lives of poor families of color. HCZ used neighborhood organizers and coordinated monitoring of individual children and their families to track and rapidly intervene in the lives of families in the region in ways they believed would lead toward the eventual success of their children. The program eventually had success in graduating many children who otherwise would likely have dropped out of school (Tough 2009). Adopting this strategy, the Obama administration sought to replicate the effects of the HCZ model in other areas around the country. Beginning in 2010 and continuing through 2016, up to 56 million dollars have been given out annually by the federal government to promote the growth and sustenance of PN initiatives around the country, resulting in fifty funded projects thus far (many of which have not continued beyond the planning phase). The federal guidelines for PN closely resemble those of CI (US Department of Education 2016), and organizations that work alongside Promise Neighborhoods directly associate them with CI strategies (Promise Neighborhoods Institute 2014).

Strategies such as CI, and in education specifically, the federal PN initiative, are now investing heavily in data collection and sharing platforms. These platforms collect and share data, often instantaneously, about individual young people and their families across dozens of service providers, with the goal of intervening in any possible way to ensure progress against a series of metrics, some academic and others related to child development more broadly. From a social service provisions point of view, these systems appear significantly more efficient at moving the needle on major social issues, like graduating children from high school and closing the so-called achievement gap. However, these systems, with greater efficiency and efficacy than ever, also apply—even impose—a very particular model of child development to every young person in their reach. My research is concerned with the degree to which these models of child development have been modeled on psychological research done primarily on white, middle class, male kids in the USA and are being applied to families of color around the country. Thus, my project seeks to understand the ways that the now constant and thorough surveillance of the lives of young people living in PN affects their sense of agency, identities, and imagined futures. In doing so, it hopes to understand the ways that this form of collective intelligence aims to build toward common goods, and how, by addressing some present challenges, it might better do so.

2 A brief history

Social welfare in the USA and the UK have collected and utilized data since their inception. The English Poor Laws, originating in England in the sixteenth century, were a way of managing an emerging class of “beggars” (Trattner 1999). The law was also adopted in the British colonies of the Americas and carried into the formation of the USA. Town councils in both countries managed lists of the poor and indigent, making decisions about who was worthy of receiving housing or money from the townspeople and who would be “warned out” of the city and expected to find help (or, more often, further hostility) in other places. Over the years, these systems evolved alongside changing political landscapes and evolving beliefs about the causes of poverty. However, whether these beliefs situated the causes of poverty in individual moral failings, poor genetics, or social and systemic failures, the corresponding systems developed as government and private responses to poverty collected data as a way to track the poor and, very occasionally, to advocate on their behalf.

Toward the end of the nineteenth century, the Gilded Age in the USA shifted social welfare provision toward the concept of “scientific charity,” which for the first time, sought to use data not only to track the poor, but also to develop more effective interventions (Hansan 2013). While even the founders of these efforts eventually admitted that they were poorly conceived and mostly ineffective, they sparked a trend in social services that can be traced to modern forms of evidence-based practice (Abramovitz 1988). Data, from this moment forward, would be used by social service providers to determine needs, track service users, and develop interventions (Abramovitz 1988).

Each of these changes represents a major turn in the collection and use of intelligence in an effort to promote the common good. Each also remains an important arena of social welfare efforts, where similar ideologies (e.g., “if we only collect enough of the right data, we can use it efficiently and effectively end poverty”) are dressed in new clothing.

CI strategies share many of these roots, but are rhetorically relatively recent inventions. Conceptually, they seem quite simple: determine a set of shared goals, coalesce resources to address these goals, and make sure these resources get to those who need them. However, in practice, these efforts require significant energy and organization. Social service agencies are often entrenched in addressing particular issues and, cash-strapped and without incentive, find it difficult to pivot toward new objects or spend time coordinating with other organizations. And the first and perhaps most challenging step—getting all organizations to agree on a set of common objectives—takes significant effort and skillful leadership. Nevertheless, these strategies have become a recent trend in the social services sector, embodied perhaps most prominently in the USA by PN initiatives modeled after the Harlem Children’s Zone (Tough 2009).

These initiatives draw upon this history of data collection and use, but also overcome some significant historical barriers. I therefore argue that they represent the beginning of a sea change in terms of the use of collective intelligence for the common good in social service provision.

3 Conceptual framework

This study applies two lenses for understanding the effects of PNs on the lives of young people. The first referred to as “child development” brings a critical lens to the claims many PNs make about the development of young people. It is applied as an analytic frame in contrast to implicit claims many PNs often make about their moral position. The second frame referred to as “community-based and anti-ageist” adopts an alternative ethical stance, suggesting that the priority of collective intelligence efforts such as Collective Impact and PNs ought to be rooted in community perspectives—in everyday practice, governance, and ethos. These community perspectives must include the presence of young people, who because of youth-discriminatory notions often associated with the framework of “child development” are excluded from making decisions about their own lives (Lesko 2012). These frameworks are detailed next.

3.1 Child development

The influence of developmental psychologists like Jean Piaget on education has and continues to strongly support the belief that children go through sequential developmental stages.Footnote 1 Piaget claimed there are four universal stages that name the ways children’s thinking and making sense of the world evolve. Most importantly, a child’s capabilities are bounded by their particular developmental stage. Following Piaget, educators, and service providers are increasingly expected to identify developmental stages and know whether children are “developing normally” or whether stage-specific interventions are required. While Piaget was an early contributor, theories of child development have expanded dramatically since.

Though the origins of theories like these were developed almost entirely on studying white, western boys, these theories have evolved and been applied almost universally to all children (Henrich et al. 2010; Lesko 2012). A recent meta-analysis of psychological studies of another popular child development model—the construct of resilience—found that 84.8% of studies were conducted in the “west,” 64% had populations that leaned somewhat or were heavily female (Lee et al. 2013). Given these weaknesses in the origins and continued studies of child development, it is concerning that these theories have been so broadly and deeply applied to the ways we raise children (and the ways we judge them as “successful” or “normal”) with few questions about their generalizability.

The influence of such theories results in a self-fulfilling prophecy in which educators find not what children are capable of, but what it is they expect to find that children are capable of. This example reflects a broader trend in which scientific studies that cannot be separated from their cultural context are taken to reflect universal biological realities of human development.

To better understand the historical construction of childhood, it is important to recognize conceptions of childhood as historically and culturally contingent (Mintz 2004), emerging in the USA through the convergence of child labor laws and the advent of Sigmund Freud and G. Stanley Hall’s conceptualization of child and adolescent development.Footnote 2 For a time, these conceptions of childhood (new to white, Western society) helped enable a shift toward what we now see as a more humane treatment of children—they were used in advocacy for child labor laws, for example. However, these constructed age categories—which for the first time in the West changed the image of child from small-bodied, impetuous and hot-tempered adult to a developing person—were based on very narrow, mostly anecdotal, and often wildly inaccurate, observations of children and youth (Schneider 1992). Despite this unstable foundation, a mixture of socioeconomic, cultural, and intellectual forces conspired to create childhood and adolescence—age categories now taken-for-granted by child development researchers and by educators (Smith 2010). Unfortunately, because of their uncertain and unstable empirical basis, these theories are questionable in their generalizability and especially in their broad universality (Wyn and White 1997).

However, these theories also form the root of many PN strategies. An amalgamation of new scientific studies correlate positive measurements of child development with better standardized testing scores, thus encouraging holistically minded approaches like PNs to adopt metrics around child development as an influence on desired academic achievement outcomes. History and questionable scientific basis aside, these theories are perhaps most concerning in the ways they limit our conceptions of young people’s capabilities. If, based on these theories, we believe adolescents are not capable of making decisions of the same importance as adults, we are not likely to value their opinion in the governance and direction of our organization, thus excluding them from making important decisions that impact their everyday lives. Worse, as the data show, such exclusions become the default. This approach is countered by the second framework presented, called here “Community-Based and Anti-Ageist.”

3.2 Community-based, anti-ageist

This second approach stands as a counternarrative to the first. It begins with a democratic stance that believes that individuals and communities ought to have a voice in issues that impact their lives. In the least, they should not be excluded because of their marginalized status (Finn and Checkoway 1998). This approach is rooted in a long history of community-based work, influenced in the USA by the Settlement House Movement (Addams 1893), the Highlander Folk School (Horton 1997), Asset-Based Community Development (Arefi 2004; Mathie and Cunningham 2003), Participatory Action Research (Torre and Fine 2011 ) and various social movements through history.

As an approach to working in communities, this framework values the voices of individuals and communities, rejects deficit- and damage-oriented perspectives (Tuck 2009), favors systemic framings and interventions over individual fixes, and places its ethical commitments at the center (Torre et al. 2012). These commitments include invitation, participation, shared decision making, and transformation of the social order. While these values are consistent with what many understand to be central tenants of social work practice (Healy 2001), my research suggests that they are often sidelined in favor of the child development framework presented above, which suggests that young people are not yet capable of participating in these sorts of activities.

Utilizing this framework as a contrast, this study investigates the ways that Promise Neighborhood efforts could be changed to promote a more deeply inclusive common good.

4 Methodology

In this study, I examine the ways that PNs, as one example of CI strategies, collect, use, and share data in an attempt to develop collective intelligence toward common goods. This study is especially interested in answering two questions: (1) what notion of “collective intelligence” is promoted by these approaches to social service provisions, and (2) whose version of the common good is forwarded? These questions are further addressed by considering the following: What arguments are being made about data collection and analysis? How are these movements using data to measure and justify activities? Who manages these data and how do they do it? How does data collection, analysis, and visualization shape movement efforts and stakeholder opinions and investments?

To conduct this analysis, I examined hundreds of public documents about the 52 programs that have thus far been funded by the federal Promise Neighborhood program. These documents included meeting minutes, official publications, scholarly analysis, and other documents. While I treated this document analysis as the beginnings of an ethnographic investigation, looking for common themes, similarities, differences that caught my attention, I looked in particular at several questions I expected to offer insight into my research questions. This included examining the demographic constitution of their boards of directors, their governance structures, services offered, funders, and their investments.

Utilizing a list of funded projects provided by the Department of Education, I gathered the unique identifying information for each organization that received funding and matched these identifiers with Form 990 tax information provided by the ProPublica Nonprofit Explorer API (Application Programming Interface). This provided information on board membership, salaries of key staff, revenue streams, and expenditures. Though this information was useful to build context for the organizations involved, it was also relatively limited. A primary limitation for this purpose is that Form 990s do not provide information on specific revenue streams or expenditures, but rather provide summative data. This is fairly informative for organizations whose only function is the PN. However, for many involved organizations, the PN was one of many service offerings, and therefore financial data for the PN strategy and other strategies were indistinguishable.

As I examined each organization in depth, I took care to search for information on their boards of directors and governance strategies. I kept track of whether board members were (1) representing community interests only, meaning they were both a community member and were not identified as serving on the board because of employment elsewhere, like city government or a funding corporation, and (2) identified as a young person representing neighborhood youth. While some organizations provided significant information about boards and governance, others provided only board member names in their Form 990s. For the latter, I attempted to track down each board member individually through Google and LinkedIn searches. When I could successfully track down all board members (50% of the total organizations), I included that organization’s data in the final analysis.

Additionally, I searched each organization’s documents for information on whether they used a known data sharing system or created a bespoke one, and some information on how they put these systems to use. Most organizations were not forthright with this information. For those that were, it was often because they were part of a data sharing system’s advertising strategy, or because they referred to the database as part of their communication with partners.

There are four findings emerging from the analysis of the publicly available documents for Promise Neighborhoods thus far. These are presented in the next section.

5 Findings

5.1 Data collection, sharing, and use

CI strategies are based on two premises: that sharing a set of key metrics will bring together all stakeholders and service providers around a common set of achievable goals, and second, that sharing data toward these goals will increase the likelihood of achieving them. Historically, two major barriers have prevented the latter premise from being a possibility. First, social services often utilize legacy technologies, often built as leftovers from more financially rewarding corporate technologies. These legacy technologies often remain in place for much longer than they should, leaving social service workers and data analysts using significantly out of date technology that hinders easy sharing of resources. Many organizations still utilize paper files as a primary method of retaining data over time. Second, data sharing among organizations has been almost nonexistent. This is due both to legacy technologies, but also concerns about privacy, proprietary information (there is competition among nonprofit agencies, even if it differs from corporate competition), and regulation. Most organizations are therefore very closed with their data, and even if they desired to share their data about services provided or clients, it would be difficult to overcome these technical and regulatory barriers.

CI strategies are starting to shift both of these challenges. The demand for more, better, and regular data is part of the ideology of CI strategies. With more aggregate information, many are arguing, we can make better decisions about where to dedicate scarce resources. With more information shared about individuals, we are better able to intervene in their lives at critical moments and are able to do so most effectively. We are also better able to develop research about such interventions, meaning that we will be able to put greater resources into interventions that are more efficacious. Therefore, up-to-date technologies and data management processes are a key factor in collective impact strategies, often supported financially by what are referred to as backbone organizations—groups that support a particular collective impact initiative with financial and other resources (Preskill et al. 2014). Additionally, these efforts are finding ways to overcome regulatory and cultural barriers to data sharing.

The data collected and shared about young people and their families in many of these programs are not only used to shape and measure the direction and success of the organization. Significantly, individual data are aggregated and shared between service organizations, meaning that service providers now have significant context about individual youth—which, of course, is offered through a very specific lens of the data we have chosen to collect. Further, an essential part of the PN strategy is to have community workers who utilize these data to ensure via various incentives and encouragement that the data lead to specific interventions. While taken individually, the offerings in each of these areas are useful, meaningful, and often fundamental to people’s well-being. When all efforts are made to push young people and their families through a series of hoops to achieve outcome metrics, a very thin line between supporting and manipulating comes into play.

One of the major databases used by PN cites itself as “Holistically address[ing…] individuals need to solve their behavioral issues.” The framing is thus that academic achievement is correlatively, perhaps causatively, related to student compliance with adult authorities’ notions of correct behavior. While we might assume that major goal setting and tracking features of these databases can be used to broader purposes, it seems like the focus of the creators of the database is on behavioral control. Related, several of the systems used for data tracking that I encountered are focused on the goal of developing “self sufficiency” in client users by manipulating client behaviors to act in service of the broader metrics. This operates on a classic assumption of social services since its inception that individuals are broken because they need help and that fixing their need requires changing faulty behaviors. Overall, most of the organizations that had database information available used this information to measure primarily academic outcomes, such as grades, afterschool participation, truancy, and graduation rates. However, some used assessments of behavior and mental health to influence interventions. Many shared data between service providers related to mental health, or factors believed to be related to mental health and healthy development, such as family “health.”

5.2 Youth agency within organizational structures

One of the core practices of many collective impact initiatives is the use of street workers—people who draw on available data about clients to provide necessary services to them, preferably on a rapid and continuous timeline. Unlike many social service offerings, which depend on users to seek services, these workers use data systems to locate those they judge in need of services and to offer those services, without the client needing to initiate these proceedings.

More than ever before, this means that data are being collected by many diverse organizations—housing support, food supplements, supplementary income, mental health services, STI prevention, education, and so forth—and shared. Information that was once private among a teacher, child, and family, or remained locked in a file folder in the basement of a clinic, is now being registered in electronic databases and shared with other service providers. Increasingly, these initiatives know a great deal about individuals and their families and are intervening in their lives in significant ways, sometimes even without their knowledge. The use of this “collective intelligence” to create impact on what these organizations believe to be the common good is therefore perhaps greater than at any other point in history.

The information collected is quite significant. Social workers have shared that they collect demographic data as well as deeply personal details about clients’ lives, for example, financial information, medical and mental health diagnoses, social services previously received, and similar data about family members. Other groups, like teachers, collect attendance data, issue grades or assessments of behavior, and track any issues they view as behavioral problems, as well as the methods used to intervene to create change. Though much of these data are already collected by social service agencies, sharing these data between organizations means that individual organizations now know significantly more than they would otherwise be justified in collecting on their own.

However, the “common good” that this information is being used toward is often assumed with little debate. For the Harlem Children’s Zone and other PN strategies, the common good of the communities they serve is primarily the successful graduation of their youth from high school. However, these strategies are sometimes achieved at any cost. The Harlem Children’s Zone, for example, actively seeks to transform the primarily black communities of Harlem into culturally white and middle class (Tough 2009). The values and actions associated with these groups are those associated with the eventual school success of these children and youth (Tough 2009).

It is not clear that young people are offered many choices within these programs about their own goals and desires, especially when these do not conform to the ideas the programs have for them. PNs have specific outcome measures, often related to the categories “Health,” “Family,” “Education,” and “Community.” Taken individually, the offerings in each of these areas are useful, meaningful, and often fundamental to people’s well-being. For example, services related to health often include vaccination, clinic check-ups, and mental health services. Family resources include child care and case management. Education services include tutoring and afterschool programs related to academic outcomes.

Making this diverse pool of resources more available to families is a key goal of PNs. However, two dynamics limit the range of choices available under a PN. As PNs consolidate resources, programs that better fulfill their metrics often receive funding priority. Second, when specific metrics are desired for each young person involved, there are demands placed on workers to ensure these metrics are met, thus limiting the range of choices available. If a student is struggling in math, they are more likely to be pointed to tutoring than the arts program they are excited about. They may even be pushed via a series of incentives and/or punishments to participate in these programs. Decisions about the young person’s life will naturally become directed toward achieving these outcomes, regardless of the interests of the young person.

Further, some of these services may be unintentionally misdirected. The student struggling at math may be seen as struggling because of their inability to pay attention—perhaps related to an ADHD diagnosis they should be treated for, or because of behavioral issues for which they should receive therapy—rather than because they are intentionally resisting their math course because they perceive the teacher as racist. When all efforts are made to push young people and their families through a series of hoops to achieve outcome metrics, a very thin line between supporting and manipulating comes into play.

5.3 Leadership and governance

Given that these organizations make significant decisions for young people and their families beyond the standard role of school in the development of academic capacities, one might expect these organizations to involve families and even youth in making decisions about their values and priorities. However, this appears far from true for many of these organizations. My research has shown that the boards of directors of most of these organizations do not have formal community representation, have little formal involvement from parents, and almost no formal involvement from youth. Of the 52 organizations that were accepted to do Promise Neighborhood planning grants, only three had an identifiable young person on their board of directors. Only one of these organizations had more than one young person on the board. Based on fairly extensive research into each organization, I found none that had an advisory board staffed by youth. There was only a small handful that had advisory boards staffed by parents. These, of course, have little formal authority, and their role within the governance structure was rarely apparent, especially because they did not appear to have representatives to the board of directors.

Many organizations did not even have a member of their board of directors representing only community interests without obligation to another organization. Instead, the boards resembled those of many other nonprofit organizations, having significant involvement of private funders and representatives from for-profit businesses. They also often included child development and psychology experts, lawyers, and local dignitaries. However, very few involved representatives from the community that were not also attached to other nonprofit or corporate interests.

In the formal governance structures, there is therefore little room for youth and families to determine the goals, values, and operating principles of these organizations. While this is fairly ordinary in many nonprofits, the direct, robust, and thorough involvement of these organizations in family and community life should necessitate greater efforts to involve these community members in efforts of governance. In the present situation, there is little accountability to community values, with the boards of these organizations placing a greater emphasis on outside expertise and demands for data collection by funders.

5.4 Financial foundations and impacts

At present, significant amounts of money are coming from corporate America to support these programs, and that significant amounts of money are going back to corporate America in the form of special charter schools, private administrative support services, data management systems, and of course, future labor. While constituent programs within PNs often draw from a variety of local, state, and federal public funding sources, the coordinating entities themselves receive significant private dollars—and thus their priorities, as well as their structure of accountability, seem at least partially determined by their financial ecosystem.

Some of these initiatives were underway prior to receiving funding from the federal government. They seemed to often be nascent coordinating structures informally organized under what CI terminology calls a “backbone organization”—someone with enough power and money to convene and facilitate a number of stakeholders. Either way, it seemed like few had much in the way of financial resources. PN grants initially offered $500,000 for planning purposes and then several million to a few organizations after this process. This seems to have sparked the ability of PNs, especially those receiving the bigger follow-up grants, to raise additional money. Some of these dollars came from local public agencies. However, significant amounts of money came from private foundations, family and corporate.

While their specific budgets are difficult to access, the information I have been able to track down about funders indicates that for the PNs that have continued post-planning, the amount of money from private funders amounts to between 1/3 and 2/3 of their total operating budget. Both public grants and private dollars have specific, desired outcomes, most of which seemed related to academic achievement. This places significant burdens on PNs to move young people and their families through the hoops to achieve these outcomes, at the risk of losing future funding.

Perhaps most significantly, my research is pointing toward the possibility that private and public dollars are now being re-routed through PNs, rather than given directly to service organizations in the neighborhoods. As the PNs gain power, both through the rhetoric of uniting services and through their ability to wield attractive data, money is increasingly given to these projects that might have gone elsewhere, thus narrowing the diversity of program offerings to those that conform to PN activities. As my early ethnographic research has showed, these funding demands create powerful changes in program administration. I watched one staff team attempt to re-organize their entire program offering to ensure that data were collected properly to satisfy the PN administrators, even though they also clearly saw that such changes would dramatically reduce program quality.

Additionally, it seems that sizable portions of the dollars these organizations raise go toward funding their data collection and sharing systems. These systems, most often developed and operated by for-profit corporations, are seen as vital to the success of PNs. There appear to be at least nine data systems that licensable by groups for these purposes. But there are likely many others I have yet to find, and further, many groups have elected to design bespoke solutions on top of existing systems or from scratch. I located several examples of the latter situation. Total dollars spent on these systems is not typically public, but likely ranges from $500,000 to $1,500,000 over a five year period using the calculator provided by the National League of Cities. It is likely actually higher than this, because this calculator is focused mainly on afterschool/school integrations, but does not include other social service providers.

While this research by no means offers conclusive evidence, it opens questions as to the source and use of financial resources in PNs. It points to the potential consolidation of program resources under the governance of one organization, rather than many diverse organizations. As shown earlier, this governance is almost never under the influence of the community. Moreover, given the source of financial resources, governance, and desired outcomes especially are more likely influenced by funders and by the data collected in these databases (suggested by the database creators and government initiatives in addition to funders) than they are the local community. This is yet another way that young people and their families are stripped of agency in the determination of their own desired outcomes and futures.

6 Discussion

The four areas examined in this study can at the best demonstrate several perspectives from which to understand PNs, and by extension, CI efforts. They cannot be argued as a “holistic” assessment. However, they attempt, based on data available, to drive at the question of how collective intelligence can be developed narrowly, and toward narrow goods, and perhaps open up whether an alternative is possible.

It is clear from the information shared here that PNs largely live in a framework of individual, child development. They are focused on developing youth in specific ways that involve individual betterment along trajectories predetermined by adults. They are often given a more limited menu of choices regarding their time outside of school. Young people are almost entirely excluded from decision making regarding the organizations that have increasing power over their lives. As is typical in child development, it is corporations and government that have the greatest say over their lives—even parents are mostly excluded.

These demonstrate a model of collective intelligence for the common good that is likely to become increasingly common. Some may argue that, given all who are excluded, these do not represent a “collective” intelligence, but rather a narrow intelligence based on the opinions, ideas, and theories of a small group of people. This research supports such an argument; however, on the surface these organizations seem to represent a collective. After all, they are constituted by dozens, perhaps even hundreds, of organizations coordinated around a common goal using an ever-increasing plethora of data to generate ever more intelligent approaches to their work. They might also argue that they represent a common good: the graduation of children from high school and college. But we are forced to ask, whose common and whose good? This research shows that there are some groups considered part of the common and some groups who are acted upon.

A community-based, anti-ageist approach would start from the place of collaboration with a community to discuss needs, collectively research strategies to address these needs, and develop agreed upon measures for intervention. This process would need to include young people and their families as equal partners. This may be happening at some organizations, but my research did not locate one where this is fully taking place. Given that so many organizations have already begun and have significant investments, a better approach might be to attempt to incorporate this approach into existing activities. Based on the perspectives explored here, this might include four major changes. First, organizations might invite youth and parents to be part of governance structures. This includes the board of directors and any other governance and advisory committees. It might also include having young people and family members on staff at the organization. Further, it could include both a challenge and necessary support to involve constituent agencies in the PN initiative to invite youth and parental participation in new ways. My experience attempting to do this work indicates that it is a challenge—both for young people/family members and for boards of directors—to attempt this kind of integration. Boards, often comprised primarily of professionals who have often served on other boards, tend to see young people as “not-yet-adults” and therefore to not take them as seriously as traditional board members. Young people and family members, in turn, often do not have previous board experience and therefore do not know how to navigate board issues, politics, or all of their responsibilities. Boards can consider special orientation processes for these board members. They might also consider changing their approach to meetings as well, adopting youth work approaches to meetings, such as experiential activities (which often make meetings more engaging for adult members as well).

Once young people and family members are invited into decision-making roles, organizations could start to reconsider elements of their work with youth/families present. What metrics do board members now feel matter to them? They might also survey the community, conduct interviews, and host public meetings, to try to learn about neighborhood priorities. Rather than treat goal metrics as static, they might attempt to regularly engage with the community learn about goals of importance to them. It may also be that metrics themselves have to change—that summative metrics turn out to be less important than individual or family goals and metrics. Organizations could open their thinking to possibilities like these that treat the purpose of metrics and data collection differently. Perhaps most importantly here, organizations may need to recognize that the metrics they choose and data they collect influence their funding models—and not be afraid of changes this process might necessitate. In this model, the funding should follow, rather than guide, the community’s goals and shared metrics for success.

A structured, yet accessible model to begin to integrate changes like these in Promise Neighborhood efforts is presented in the work of Douglas Schuler (2008) on pattern languages. A pattern language defines a set of process-based patterns that can be applied to particular situations. In this case, Schuler’s pattern language system could be applied to transform PN efforts toward narrow goods into an exemplar of collective intelligence for the common good.

Specifically, many of the patterns he identified can be used to generate a collective intelligence—shared knowledge, data, and thinking—that drives toward a set of goods shared by a broader community. Patterns like Collective Decision Making, Participatory Design, Community Networks, Transparency, and Community Inquiry are strategies that can be employed by organizations and community members to generate and harvest their collective knowledge, and then to transform that knowledge into actionable, collective activities. Patterns like these are particularly useful because they capture the salient parts of complex social processes in simple terms and can be easily applied across contexts. A PN interested in re-working itself from the standpoint of the community might apply a pattern like Participatory Design in considering how to create a PN, a pattern like The Power of Story to understand community needs, one like Community Inquiry to identify shared goals, and a pattern like Collective Decision Making to determine priorities and begin to take shared action. They might keep the Transparency pattern as a part of decision-making processes and regular activities in order to maintain a strong connection between the organizations involved and the community they operate within (Schuler 2008).

While utilizing these processes might add complexity, and thus slow down the work initially, it is likely that over time the intense cooperation of the community will lead to efforts that more closely align with community needs and are strongly supported by the community at-large, assisting in the success of these efforts.

6.1 Limitations

This research draws upon publicly available information to explore, analyze, and assess the ways that Promise Neighborhoods make an effort to further work toward the common good through collective intelligence. The documents available to analyze paint a picture of organizations that, despite good intentions, do not seem to be making strong, explicit efforts to harness grassroots, community-based collective intelligence. It is likely, however, that at least some of these organizations are making efforts to include community perspectives in their work and decision making. For example, I found a few organizations that used community surveys to develop an understanding of the people they were meant to serve. I suspect others may have community advisory boards that are not listed in their publicly available documentation. Further research should attempt to more closely examine the PNs to try to understand the ways that they might currently be including community perspectives, which could serve as a platform to start developing a richer, deeper, and more inclusive collective intelligence.

Despite these research limitations, it seems clear that there are serious absences in the governance, decision-making processes, and publicly stated efforts of these organizations that should be addressed if these organizations hope to harness the collective intelligence of the communities they are working in.

7 Conclusion

It is my contention that these agencies—intentionally or no—are an example of a broader trend in social service administration to use very specific types of data to shape an increasingly narrow field of choice and possibility for young people. Given that young people and their families have very little role to play in either the definition of the shared outcomes/metrics of these initiatives or the choices they make about what they do within the offerings available to them, we might start to wonder about exactly who the “village” is that is raising our children. The village may in fact be more a computer and its makers than the neighborhood and its young people.

These changes form the basis of an ongoing research project that I am conducting on collective impact strategies and the changes they are making in the political economy of data for social workers and service users. The degree to which these changes are affecting clients’ lives and will affect their lives in the future is still unknown. However, these significant infrastructural (Star and Ruhleder 1996) changes in service provision open the door for new forms of collective intelligence that both have positive potential and raise significant concerns.

The next stage of my project will work with social workers and young people to investigate possible responses and alternatives to these CI strategies. What, for example, might collective intelligence look like if it were based on the experiences of community members and on the shape they desire their community to take—a community-based, anti-ageist approach (parallel to Adam Greenfield’s proposals for smart cities, 2013)? What might happen if the community took charge of determining metrics, collecting data, and analyzing it (Lennie et al. 2011; Sabo-Flores 2008)? The opportunities seem great as well as challenging, given the immense resources and ideological pressure of existing approaches to CI. Regardless, there seem to be worthwhile avenues for community practitioners, community members, researchers, and activists to pursue.