As communities strive to meet the needs of children and families in an increasingly challenging economic environment, there is a clear need to leverage existing resources to the fullest extent possible and to identify creative ways to expand on these assets. One of the most valuable potential resources to be marshaled in this effort is the large amount of information available across community agencies and organizations serving children, youth, and families. Data that represent combined knowledge across agencies will allow researchers to conduct more accurate and more extensive investigations of community issues, including underlying causes and the identification of effective responses. However, the isolated and disparate nature of human services data typically presents a significant challenge to both agencies and researchers. Although cities, counties, schools, social service agencies, and community non-profit organizations each have their own internal data, they typically have neither ready access to data from the other entities nor the ability to combine their own data efficiently. The result is that many important questions related to the impact and effectiveness of community services remain unanswered. Communities often have little or no understanding of the iterative, cumulative, or longterm effects of programs and services, either alone or in combination. An inability to combine data from multiple sources, each of which may have a different and unique “piece of the puzzle,” also means that communities lack a sophisticated understanding of how multiple factors interact to determine child and family outcomes.

As funders are increasingly calling on agencies to demonstrate the impact of their programs or services or to risk losing financial support, the need for strategies that facilitate outcomes research and evaluation are greater than ever. Over the past decade, a diverse group of stakeholders in Charlotte, North Carolina, including policymakers, city and county agencies, non-profit organizations, foundations, and social and behavioral scientists, have increasingly recognized the importance and difficulty of measuring the impact of programs designed to improve the lives of children and their families. In response to this need, The University of North Carolina at Charlotte Institute for Social Capital, Inc. (ISC) was created to serve as a link among agencies, with the broader goal of fostering research and data-based community decision-making. In addition, ISC was designed to link the community with university researchers with expertise in a wide range of areas relevant to children, youth, and families. The stated mission of ISC is to foster university social and human capital research and to increase the community’s capacity for data-based planning and evaluation.

In this article, lessons learned over the first five years of implementing ISC will be used to examine key considerations and successful strategies for sharing data across agencies for purposes of research. In order to illustrate the value of an entity such as ISC, the results of a pilot project examining academic achievement for localAfrican American youth will be presented. The potential benefits of this and similar studies will be discussed, including details about the use of their findings to support community initiatives.

1 Building an Institute for Social Capital

A collaboration “brings together two or more agencies, groups, or organizations at the local, state, or national level, to achieve some common purpose of systems change” (Backer 2003, p. 4). However, true collaboration is much more than simply working together toward a common goal. The fundamental notion behind collaboration is that by joining together, collaborative partners have a greater chance of achieving their goals than if each acts independently. Perhaps most importantly, collaboration is designed to be mutually beneficial to all participants. One important area in which community partners can collaborate to create value for all is through research and datasharing. Data obtained through collaborative activities with community partners can serve as a powerful tool not only to support requests for resources at the local, state, and national level (e.g., by providing information on [1] the specific population to be served, including relevant characteristics and need and/or [2] the impact of programs and services), but also to sustain and to increase attention and commitment to project goals. Wandersman (2003) encouraged the development of such community-centered models, which represent best practices in regard to collaborative efforts that complement other interventions to promote well-being. These community-centered models “begin with the community and ask what it needs in terms of scientific information and capacity building to produce effective interventions” (p. 230). However, as noted by Lerner et al. (1998), most existing efforts fail to build on community assets and to create the capacity to sustain effective programs.

In the context of datasharing, notions of collaboration closely parallel the fundamental principles of social capital as described by Putnam (2000). More specifically, collaboration, like social capital, should work by (1) allowing individuals and organizations to resolve collective problems; (2) permitting communities to operate more efficiently (i.e., trust and trustworthiness improve the efficiency with which tasks and broader goals are accomplished); (3) expanding awareness of how organizations’ fates are linked; (4) establishing networks and other structures that facilitate the flow of information required to facilitate the accomplishment of goals; and (5) producing a positive impact on individuals’ lives.

ISC grew from several converging local, national, and academic research initiatives. At the local level, the Mecklenburg County Children’s Alliance, a network of more than 40 child and youth-serving organizations, worked together for more than 2 years to identify a way of sharing data across agencies. During the same time period, the Foundation for the Carolinas took the lead in generating community conversations around how best to build social capital in the Charlotte-Mecklenburg (County) community. Local funders were also exploring ways in which they could obtain higher-quality evaluation data from the agencies they supported. They further believed that having more standardized data across agencies could better inform funding priorities. These local conversations were set in the context of broader national discussions, in service agencies, funding agencies, and academia, around the practical importance of scientific research and the need for an infrastructure to support research within the social and behavioral sciences.

In Fall, 2003, The University of North Carolina at Charlotte sponsored a feasibility study to identify the structures necessary to respond to the growing academic and community need for ready access to human services data, to determine how best to build social capital within the Charlotte-Mecklenburg area, and to examine the viability of an entity to advance datasharing. As a result of this study, in April 2004, the Foundation of UNC Charlotte founded The University of North Carolina at Charlotte Institute for Social Capital, Inc. (ISC) to support the improved use of data and research across human service agencies. As part of its charge to build a research foundation capable of compiling, validating, and analyzing community data in order to inform community decisionmaking, ISC was tasked with constructing a unique database that would combine data across multiple sources, linking variables at the individual level. Because of the paramount need to protect the confidentiality of all data housed in the ISC Community Database, the Institute was structured as an independent entity and operates as a 501(c)(3) nonprofit corporation. While closely affiliated with the University through a cooperative service agreement, ISC nonetheless operates as a separate, private legal entity, allowing it to provide a level of data security that the university, which is subject to State open records laws, could not ensure. ISC is governed by a board of directors comprised of community leaders across the domains of education, child welfare, health care, mental health, juvenile justice, law enforcement, and corporate and nonprofit management.

Along with The University of North Carolina at Charlotte, a doctoral/research intensive institution and the largest institution of higher education in the Charlotte region, ISC’s other key partners include Charlotte-Mecklenburg Schools, the Mecklenburg County Department of Social Services, the Mecklenburg County Sheriff’s Office, Communities in Schools, the Charlotte-Mecklenburg Police Department, the Mecklenburg County Children’s Alliance, the United Way of the Carolinas, and a host of other community agencies and organizations. Through its partnership with UNC Charlotte, ISC is able to leverage faculty, students, and other university resources to provide the community with valuable analytic support, including technical support, to help organizations define research efforts and required data; to assist organizations in conducting data analyses; and to support organizations in interpreting the results. At the same time, through its partnership with various other child, youth and family serving agencies, ISC is working to integrate data on clients and services across the community that can serve as the basis for research and evaluation. Moreover, by uniting the efforts of its numerous partners, ISC can engage in and facilitate work that benefits from the input of multiple expert sources, representing both science and practice.

By making shared data available for research, ISC allows collaborative work to be conducted more efficiently and economically. A project utilizing shared data would typically require the parties to expend significant time and other resources to negotiate the complex datasharing agreements needed to ensure that all of the appropriate legal protections are in place; to transfer and to translate data into a compatible format; to merge and to match the data in a reliable manner, making appropriate provisions for the resolution of contradictions in the data (e.g., reconciling different racial/ethnic variables; determining that Thomas J. Smith, Tom Smith, and Tommy Smith are, in fact, the same individual). However, ISC has already addressed each of these challenges and can now extend the benefits of this collaborative work to university and community researchers, providing them with the de-identified individual-level data they need to analyze questions of interest.

2 The ISC Community Database

In order to make high-quality data available for purposes of informing program planning and evaluation, ISC is in the early stages of building and housing a comprehensive database of local social service data. Software Solutions Lab (SSL) serves as the technical arm of ISC and has responsibility for database development and maintenance. Created in 1999 as a principal component of the UNC Charlotte College of Computing and Informatics, SSL serves as an in-house software research and development consulting group, providing the university and the Charlotte business community with world-class information technology services. SSL is composed of full-time, senior staff who are professional software developers, each with more than 10 years of experience in commercial research and development.

By merging data from different sources, the ISC Community Database will be a valuable resource for data that can be used to improve our understanding of how programs perform, with whom programs work, and in what combinations programs work. For example, by looking at its own data cross-referenced with child welfare data, an early childhood school-based program can examine the extent to which it serves children who have been the subject of a report of child abuse or neglect, as well as whether program outcomes vary based on participating children’s history of maltreatment. The ISC Community Database also has the ability to generate “virtual control groups;” for example, the same early childhood program can compare the subsequent school performance of its participants to that of a group of students matched on school, grade, and gender who did not participate in the program. As also illustrated by the previous example, the ISC Community Database offers the additional benefit of facilitating longitudinal research as children’s data can be tracked across time.

Central to ISC’s function is ensuring the security and confidentiality of data. In fact, ISC is, in part, such a valuable resource because it serves the function of de-identifying data and making it available for research and evaluation purposes without compromising the privacy of individual-level data. The infrastructure and security procedures developed by ISC and the Software Solutions Lab (SSL) represent multiple layers of security, constituting what practitioners refer to as “defense in depth.” This layered approach takes into consideration the legal, physical, technical, and procedural components of the data security structure. As previously discussed, the legal status of ISC as an independent nonprofit corporation ensures that, unlike a State entity, it cannot be compelled by the courts to release any data, nor even to report whether a particular individual is represented in the ISC Community Database. In regard to the physical security structure, ISC essentially maintains four distinctly separate physical locations: the server, where the unencrypted data reside; two workstations where all processing instructions are issued; the dictionary that relates transaction numbers to agency data sets; and the secure storage for exchange media. The ISC server is located in leased space within a data center on the UNC Charlotte campus. This machine contains the ISC Community Database, where all identified information are stored and joined together. Professional systems staff ensure that the machine is secure and patched at all times. Access to the data center is controlled by card-key security; only vetted systems staff—and in the case of ISC, the SSL contractors escorted by these staff to the ISC server—are ever allowed in the area. The only computers allowed to communicate with this server are two thin-client machines located in a secure office near the ISC suite; this office is locked at all times. The thin-client traffic with the server is entirely encrypted, and all local storage devices have been disabled. These client machines are password-protected and do not have the ability to print or to transfer data to a disc or other storage device.

Before new data arrive at ISC, staff have prepared for the exchange by issuing a transaction number for the exchange. This transaction number is the sole identifier for the data that will be deposited. Before the data leave the depositing agency, they are encrypted using standard PKI techniques,Footnote 1 coordinated by the Institute. When data arrive, ISC’s SSL contractors deliver them to the server by hand. The physical media are then locked in a fireproof safe. Afterwards, the SSL contractors return to the thin-client workstations, where they decrypt the deposited data, and load them into the database. Eventually, the deposited data records are joined with the information already in the database. When a researcher is granted a limited data license, the SSL contractors burn the de-identified report to a physical medium (typically CD-ROM), that is delivered directly to the approved party.

As emphasized, it is ISC’s policy never to disclose personally identifiable information. All requests for (de-identified) data are carefully reviewed by ISC’s Data Research and Oversight Committee (DAROC). DAROC is a standing committee of ISC, and the Board of Directors has adopted written policies and procedures to govern its operations. The Board appoints DAROC members, who represent various ISC stakeholders, including the data depositors, other community agencies and nonprofit organizations, and the University. Approval by the University’s Social Sciences Institutional Review Board is a prerequisite for DAROC review and approval; accordingly, DAROC provides an added layer of protection designed to focus on ensuring that even de-identified data do not pose a threat to confidentiality. To inform the DAROC review process, ISC’s SSL contractors run a preliminary version of the report and provide a summary describing the contents of the report, including the number of rows (records) and the number of distinct values per column. In this manner, DAROC can determine whether a sample or subset of the sample is so small as to risk a breach of confidentiality

The pilot version of the ISC Community Database was launched in 2008, with data from four initial depositors: Charlotte-Mecklenburg Schools (which is comprised of 172 schools across Mecklenburg County, including the City of Charlotte), the Mecklenburg County Department of Social Services, the Mecklenburg County Sheriff’s Office (which operates the local jail and juvenile detention facilities), and Communities in Schools, a non-profit organization focused on preventing school dropout. A total of 61,615 unique individuals were represented in this pilot database, thousands of whom had been served by two or more of the depositor agencies. To illustrate further the scope of the data contained in the pilot ISC Community Database, CMS deposited 201,839 student-years worth of data covering pre-school through grade 12 for students enrolled in the seven high schools identified as the most challenged in the district from 2000 through 2008. This deposit was significantly expanded in March, 2009, such that the updated version of the ISC Community Database now under construction will include data for all students enrolled in CMS from 2000 through the current academic year for the full course of their time in the district. In short, researchers will have access to more than 1.2 million student-years worth of data.

3 Sample Project: The Charlotte Post Initiative

The potential benefits of ISC include not only better quality and availability of data through the ISC Community Database, but also the ability for researchers trained in scientific methods to work with community researchers and agency personnel to ensure that research is driven by sound questions, informed by both science and practice, and that rigorous methods are employed. Through this collaborative model, ISC seeks to unite researchers with community members as early in the project planning process as possible. One case that illustrates the benefits of this approach involves an initiative by The Post Foundation to assist African American students in achieving school success. Founded in 1878, The Charlotte Post is the oldest minority-owned, locally published communication directed at the Charlotte African American community. Building on its work of honoring and providing scholarships to outstanding African American high school seniors, in 2007, the Post Foundation approached ISC to partner in its efforts to enhance school success for all African American youth in the community. Beginning with the inception of this project, ISC was able to leverage its resources, including University researchers and the ISC Community Database, to assist the Post Foundation in taking a scientific data-based approach to identifying a strategic plan of action. Accordingly, even though it is still in its early phases, this project illustrates the benefits of ISC as a resource for the community

The achievement gap between African American and other racial groups has been a perplexing problem for American educators for more than 40 years (Herrnstein and Murray 1994; Loehlin et al. 1975; Lynn 1996; Neisser et al. 1996; Williams and Ceci 1997). Recent educational reform (i.e., No Child Left Behind) has increased attention to this issue. In the past few decades, numerous interventions have been put in place with the intention of “leveling the playing field” for all students. Although these measures have had some success in narrowing the gap, it has not closed. During this same period of time, a number of methodological issues have been raised regarding how educational researchers have studied and intervened on the issue of academic disparities (Reynolds and Brown 1984; Schmitt 1989). Two particular concerns have been the definition of the problem and the general methodology used to measure differences between groups.

In reporting on the academic achievement gap, researchers have often differed as to the metrics used for making comparisons. Focusing on and assessing differences based on students’ academic performance (i.e., student grades), standardized grade level assessments (i.e., end of grade assessments), or national assessment results (i.e., SAT, ACT) will all produce differing results (Hedges and Nowell 1999; Perie et al. 2005; Phillips et al. 1998). This inconsistency raises the question of what is being measured and whether it is comparable to what other research has measured, a factor that is particularly important when researchers are seeking to generalize their findings to the larger body of research on the achievement gap.

A related issue is whether, or which, within-group differences are considered. Researchers have often approached African American students as a homogeneous group for which a singular intervention or “magic bullet” might appropriately address the academic performance of the entire group. However, such a conceptualization ignores the considerable diversity within the African American population. In fact, a number of subgroups within the African American population have been shown to perform at or above the mean of other groups (Akom 2003; Horvat and Lewis 2003; O’Connor 1997; Tyson et al. 2005). Thus, a failure to consider within group differences may mask the effects of important characteristics of this diverse population. Similarly, although national comparisons may be informative about general trends, they have limited utility for designing and evaluating interventions aimed educational disparities within a local school system, which are shaped by the unique and specific attributes of the local context.

At the outset, the Post Foundation recognized that their goal of improving school performance among African American youth in the Charlotte community was immense. In order to determine where the Foundation might most effectively target their efforts, ISC researchers examined longitudinal data from African American students in those schools identified by CMS as the most challenged in the district, along with a group of high schools with similar demographic patterns. Limited by their ability to obtain data across time, investigations into the academic achievement gap for African American students have frequently relied on cross-sectional data, examining the performance of a specific group at only one point in time. Although this methodology allows for expedient and convenient sampling, it fails to take into account the developmental trajectory of the population and thus fails to reflect developmental change. Often, the longitudinal methodologies that can enable researchers to consider change over time are too difficult and costly to conduct. However, using the ISC Community Database, these researchers were able to examine the full trajectory of students’ academic performance during their tenure in CMS.

The initial study relied on data drawn from the ISC Community Database for all African-American students enrolled in ninth grade at any time during the 2000–2001 through the 2006–2007 school year in one of the seven CMS high schools represented in the database. Data were first analyzed to determine how students’ performance in high school correlated with their performance in earlier grades. The primary variable of interest was students’ end-of-grade (EOG) test scores. The North Carolina End-of-Grade Tests are standardized tests that are administered in public schools statewide to measure student performance on the goals, objectives, and grade-level competencies articulated in the North Carolina Standard Course of Study. Students performing at Levels I and II on the EOG tests have not reached proficiency and are deemed to have insufficient or minimally sufficient knowledge and skills necessary to succeed at the next grade level, respectively. Students performing at Level III are deemed proficient, whereas scoring at Level IV indicates consistently superior performance in the subject area.

The first set of analyses indicated that only 39% (n = 1,424) of the female African American students represented (n = 3,644) were proficient in math in grade 3. Proficiency levels for this group improved to 59% (n = 2,419 of a total 4,091 female students for whom scores were available) in grade 5 and to 63% (n = 2,976 of a total 4,693) in grade 8. A similar pattern was observed for reading. Less than half (44%, n = 1,580) of African-American girls (n = 3,632) were proficient in reading at grade 3, but they improved over time; 53% (n = 2,147 of a total 4,069) were proficient in grade 5, and 70% (n = 3,303 of a total 4,690) achieved proficiency in grade 8.

These results were mirrored in analyses of boys’ EOG performance. Only 36% (n  =1,146) of the African-American male students examined (n = 3,186) were proficient in math in grade 3. However, as with the female students, boys’ scores improved over time, with 56% (n = 1,961 of a total 3,527) scoring as proficient by grade 5, and 59% (n = 2,395 of a total 4,054) scoring as proficient in grade 8. In reading, only 34% (n = 1,084) of male students (n = 3,156) were proficient in grade 3. Boys’ proficiency level in reading improved slightly to 43% (n = 1,472 of a total 3,453) in grade 5 and 60% (n = 2,431 of a total 4,042) in grade 8.

Taken together, these results suggest that many local African American students are struggling academically as early as third grade. Although a number of them are able to improve their performance to achieve proficiency in reading and/or math over time, particularly in regard to African-American boys, almost half (41%) are continuing to struggle as they enter high school.

Given that reading is essential for students’ academic success, a second set of analyses were conducted to explore further the relationship between those lowest scoring students’ reading performance in grade 3 and their performance in high school. These analyses used a subsample of students who had reached ninth grade, and thus had not dropped out prior to high school, for whom third grade EOG reading scores were available. In addition, only those students who had been enrolled for a sufficient period of time to allow them to complete the courses of interest were included in these analyses. Results demonstrated that among students scoring at the lowest EOG level (Level I) in reading in third grade (n = 1,022), 77% (n = 787) later failed to achieve proficiency in Algebra I and 82% (n = 889) were not proficient in English I. Notably, both Algebra I and English I are “gate-keeping” courses in CMS, in that a student must pass these courses in order to graduate from high school. Although a student may pass English I and Algebra I despite having achieved a low-level EOG score in these areas, the fact that so many of these students failed to meet state standards for proficiency is nonetheless troubling.

The next set of analyses were designed to investigate the association between these same students’ EOG reading levels in grade 3 and their later enrollment and performance in college preparatory classes. More specifically, these analyses examined whether students who did poorly in their early school years enrolled and performed at or above proficiency in Algebra II and Biology. It should be noted that successfully passing Algebra I allows a student to enroll in Algebra II, even though he or she may not have received a score of proficient in Algebra I. In addition, enrolling in Biology and then Chemistry can be viewed as a proxy for taking college preparatory classes. Overall, both measures afforded some ability to gauge how many students who struggled with reading in grade 3 were prepared to complete college preparatory classes. Of those students who scored at Level I in reading in grade 3 (n = 1,022), only 268 (30%) ever enrolled in Algebra II, and 72% of these students (n = 194) failed to reach proficiency. Stated differently, only 28% (n = 74) of those students who were the lowest performers in reading in third grade eventually achieved proficiency in Algebra II, a course regularly required for college admission. Likewise, of those ninth grade students who were not proficient in reading in grade 3 who later enrolled in Biology (n = 727, 71%), 93% (n = 678) did not score as proficient. Of the only 237 (23%) students who were not proficient in reading in grade 3 who enrolled in Chemistry, 94% (n = 223) failed to reach proficiency.

The final set of analyses examined the reading proficiency of those African-American students who repeated ninth grade at least once. Results showed that of the 786 students who repeated ninth grade, 78% (n = 613) had scored at either Level I or Level II in third grade reading.

As a whole, these findings evidenced a critical need for the Post Foundation to focus its efforts at a time point earlier than students’ entry into grade 3. A substantial percentage of those African American students in challenged schools in Charlotte = Mecklenburg who were identified as performing below proficiency levels in third grade continued to exhibit a pattern of poor academic performance through high school that severely limited their educational opportunities. Although this pattern may begin earlier than grade 3, this is the first time point at which standardized assessments are available across all students.

Notably, these introductory findings, which are essential to informing data = based programming are often not available to community groups such as the Charlotte Post Foundation. Although the type of data utilized for these initial analyses are regularly maintained by local school districts, typically, these data are not easily accessible—particularly to nonprofit groups. Not only do such groups lack the resources and expertise necessary to negotiate the many legal and data-security issues attendant to interagency datasharing agreements, but these organizations need to obtain data relatively quickly in order to ensure (1) that the initiative is timely and responsive to current community need and (2) that they can leverage existing enthusiasm and momentum in the community. Moreover, most community organizations lack quick and easy access to the research skills that ISC was able to bring together in the Post research team.

In response to the project data, which the Post Foundation presented at a 2008 communitywide event, the Foundation has identified a target group of second grade African American students whom they will link to existing community programs and supports designed to bolster academic performance beginning in 2009–2010. ISC continues its work with the Post Foundation and will be evaluating the impact of these programs and services. By depositing data from the participating agencies and programs into the ISC Community Database, the partners are ensuring that the research team will be able to track the impact of programs, both alone and in combination, on school performance longitudinally, as well as to compare the performance of participating students to that of a matched comparison group. The researchers will also be able to use data from other community agencies to examine how programs impact children and youth receiving other services. For example, the pilot version of ISC Community Database contains child welfare data for 19.4% (n = 1,697) of the students included in the initial project analyses, such that the school performance of students who have been involved in child welfare cases as a victim and/or perpetrator of abuse may be examined separately. Similarly, there are Sheriff’s Office data from 1.5% (n = 131) of these students, such that additional analyses can be conducted for this group of students, who have been jailed as juvenile offenders.

4 Conclusions and Directions for the Future

Reliable data are necessary to develop a complete and accurate understanding of the effects of human and social capital initiatives to inform service planning, program evaluation, public policy, and academic research. Community-centered collaborative models, such as the Institute for Social Capital, can be a powerful tool for addressing this need. Datasharing efforts in particular can benefit communities by identifying areas of community need, informing the development of appropriate and effective response strategies (i.e., programs and policies), providing information that can be used to enhance services and service delivery, and, importantly, increasing the efficiency and effectiveness with which all of these activities can be accomplished.

Although the scientific research ISC will facilitate has the potential to contribute significantly to advancing knowledge on child, youth, and family services and policy, ISC has been designed to be much more than a vehicle for academic research. Its primary goal is to strengthen the Charlotte—Mecklenburg community by building the research and evaluation capacity of governmental agencies, nonprofits, and other organizations serving children, youth, and families. During the next three years, ISC anticipates adding no fewer than ten new data depositors from the ranks of local agencies. Our goal is not only to expand the ISC Community Database substantially through new depositors, but also to work with agencies to conduct projects that will provide direct benefit to them and to the community at large. Each of these projects will address a specific need identified by a community agency serving children, youth, and families in Mecklenburg County.

In order to reach its full potential and to ensure that it serves as a valuable community resource, ISC will need to work closely with community agencies to enhance their awareness of and ability to access the services available through the Institute, including the ISC Community Database. Through planned social marketing and community engagement initiatives, ISC will not only disseminate its work to but also include community stakeholders in the development and implementation of effective strategies for ensuring that ISC becomes an established community resource.

There is much yet to be learned about successful collaboration. For example, little is known about what motivates partners and/or gives them the ability to collaborate, as well as the processes involved in establishing and sustaining successful initiatives. As noted by Bazzoli and colleagues (1997), the many theories on collaboration that have been generated across a variety of disciplines can be summarized in three factors: the perceived need for collaboration, a willingness to collaborate, and the ability of organizations to collaborate. Each of these three factors is affected by the environment, which is the product of not only community, but also organization and individual members’ characteristics. Additional work to identify key features of these factors, as well as how they interact, can provide valuable guidance on how collaborative initiatives can succeed. Moving forward, ISC plans to identify ways to investigate empirically those strategies that contribute to success—and sustainability. Perhaps most importantly, though, ISC will continue to help the Charlotte-Mecklenburg community to build its social capital by bringing together resources to find new, effective, and economical ways to enhance the lives of children, youth, and families.