When Paul Reville ended his five-year tenure as the Massachusetts Secretary of Education in 2013, the state was widely acknowledged as a nationwide leader in K-12 education. Massachusetts students consistently ranked at or near the top of the National Assessment of Educational Progress (NAEP) and other national rankings of student achievement, and Massachusetts students also outperformed those of many other countries on the Programme for International Student Assessment (PISA) exams. Despite these successes, Reville could not shake an uncomfortable truth: the overall achievements of Massachusetts students masked large gaps on every measure of performance that correlated strongly with socioeconomic status, race/ethnicity, disability, and English language learner status.
These gaps are not unique to Massachusetts; they exist in every state across the country. Poverty in particular has been a persistent and overwhelming predictor of poor outcomes on all measures of child achievement and wellbeing across the United States, which several decades of school reform have not been able to change. Conversations about education reform have generally avoided or minimized the impact of poverty on student success, either because of the belief that poverty is too difficult a challenge to address directly or out of concern that poverty will be used as an excuse for poor performance.
Despite this reluctance to discuss the effects of poverty directly, its impact on children is clear and pervasive. Children living in poverty experience disadvantages at every life stage relative to their better-resourced peers. By age three, children in low-income families have heard 30 million fewer words than children in better-off families. Only about 46 percent of children aged three through six in families below the federal poverty line are enrolled in center-based early childhood programming, compared to 72 percent of children in families above the federal poverty line.1 Poor children are about 25 percent less likely to be ready for school at age five than children who are not poor.2 Once in school, these children lag behind their better-off peers in reading and math, are less likely to be enrolled in college preparatory coursework, less likely to graduate, and over 10 percent more likely to require remediation if they attend a four-year post-secondary institution.3 All of these issues compound one another to create a cycle of low opportunity: children in poverty are less likely to achieve high educational attainment, and low educational attainment leads to lower median weekly earnings and higher rates of unemployment.
Children in poverty are also more likely than other children to experience multiple adverse childhood experiences (ACE)—e.g., witnessing violence, having substance abuse in the home, or having an incarcerated parent—which have been shown to hinder academic and social-emotional growth. Children in families below the poverty level are more than three times as likely to experience two or more adverse child and family events as children in families at 400 percent or more of the poverty level.4 These ACEs in turn contribute to disproportionately high adverse health and mental health outcomes for children in low-income households. This multitude of non-academic and academic challenges traps low-income children in a self-reinforcing cycle that causes and perpetuates gaps between them and children with access to more resources.
Socioeconomic status is not the only category in which disparities persist. Substantial opportunity and outcome gaps are also prevalent between white students and students of color, students with disabilities and students without, and English learners and those proficient in English. While 20 years of education reform has yielded some progress for America's students, it has failed to achieve the central goals of American public education: excellence and equity for all.
Recognizing the need for a new children’s opportunity agenda, Reville founded the Education Redesign Lab at the Harvard Graduate School of Education in 2014. The Lab’s goal is to create systemic, silo-breaking approaches to addressing the comprehensive needs of children, especially those living in poverty, by developing personalized systems of support and opportunity both within and outside of school. The core ideas underpinning the Lab’s work are that schools alone cannot address all the factors that lead to negative outcomes for children and that it will take a coordinated, system-wide approach to make real change.
To overcome widespread inequity in supports, opportunities, and outcomes for children, the Lab is working to ensure that all children have access to personalized systems of support and opportunity starting in early childhood and throughout a developmental pathway that builds student engagement and agency while preparing them for success in higher education and careers. Features of these personalized systems include:
Systems of whole-child support: No matter how much schools improve, children need more than academic supports or improved schools to thrive; they must also be physically and emotionally healthy to be prepared to learn throughout their K-12 years. In a true support model, early childhood, health, mental health, and social service supports will be comprehensive, braided with educational services, and designed to address critical barriers to learning and development.
Student-centered, tailored learning: Meet children where they are and give them what they need to succeed. Through well-executed, student-centered learning, each student is able to achieve academically to the best of his or her abilities. This definition of learning broadens the typical “personalization” field to include a whole-child personalized approach, so that cities create learning environments where students are at the center.
Systems of opportunity: Systems of opportunity enable each student to chart his or her own path toward a rewarding career. Low-income children are much less likely to participate in preschool or out-of-school enrichment activities than their more affluent peers, and when they do, the programs are generally inconsistent in quality and availability. To address these gaps, a coordinated system of programs and services needs to be crafted so that every child has access to preschool, expanded learning, summer, and work-based opportunities that enrich them as learners and help them build the important skills, networks, and social capital that will serve them in the future.
The Lab leverages three key strategies to define and advance this vision for whole-child personalized systems:
- Field Work: Supporting city leaders and teams to accelerate the implementation of effective practices, enabling conditions, and systems integration;
- Research: Disseminating curated and original actionable research on the critical components and examples of redesigned education systems;
- Movement Building: Promoting this vision, in collaboration with key partners, through strategic communication, capacity building, and advocacy around critical state and local policies.
In February of 2016, the Lab launched the By All Means initiative to test and refine its theory that meeting the complex array of children’s needs and developing their interests and talents requires a system-wide approach. Six small- to mid-sized cities joined the initiative to work in partnership with the Lab over a two-and-a-half-year period: Louisville, Kentucky; Oakland, California; Providence, Rhode Island; and Salem, Somerville, and Newton in Massachusetts. The goal was for the cities to begin designing and implementing new, personalized systems for serving children while the Lab supported and documented this process in order to identify enablers of and barriers to progress that could help other cities take on this work.
By All Means (BAM) is informed by a mayor-driven “collective impact” approach to addressing social challenges through collaborative action.5 Mayors have a unique ability to drive this work. According to Oakland mayor Libby Schaaf, “As mayors, we have the opportunity to change how public systems work, as opposed to just starting another program…. We can change the expectations and beliefs of an entire generation of children. There is no nonprofit organization that can deliver that kind of promise, and we should really be aware of that opportunity that only government has.”
Collective impact encompasses a number of core practices, including having one or more strong champions, identifying shared goals and common metrics, and creating a “backbone” organization to support the work.6 A second defining frame for this work is the central role of leadership in effecting change. Ronald Heifetz’ theory of “adaptive leadership” acknowledges the complexity of many leadership challenges and the need for new mindsets and ways of leading rather than simply technical solutions to many leadership challenges.7
The Lab chose to work with cities as the unit of change because they are increasingly emerging as sites of innovation and leadership on issues that are struggling to gain traction nationally. Cities also represent relatively discrete systems that already provide a number of direct services to children. Four of the six BAM cities are in New England, one is in the South, and one on the West Coast. The cities range in population from 760,000 (Louisville, KY) to 43,000 (Salem, MA). They represent the complex array of governance relationships between mayors, school superintendents, and school boards across the country, from mayors who directly select superintendents, to those with have no formal influence on the schools at all. Finally, they vary demographically, with differing levels of poverty, different racial and ethnic makeups, and different histories. The variations among the cities have afforded the Lab opportunities to compare how a city-wide initiative could be implemented in different contexts.
The Lab identified cities for potential inclusion in BAM through a combination of word of mouth, research, and outreach, which included a presentation at an October 2015 meeting of the United States Conference of Mayors. Using the criteria below, the Lab then invited the cities to submit applications for membership.
Mayoral Commitment and Leadership: Mayoral commitment and willingness to lead a collective action approach was the first essential element for participation. The rationale for this is that addressing all the factors affecting children’s wellbeing requires a city-wide effort, and that this starts with strong leadership from the top. The Lab believes that mayors have the political capital and convening power to make BAM a high-visibility, high-impact effort.
City-School District Partnership: While mayoral leadership is central to the Lab’s theory of action, schools are and will continue to be the hub of service provision for children. The aim of BAM is to create a broad coalition to share the responsibility for children with the schools. For this to happen, it is essential that the mayor and school superintendent have a strong working relationship and a shared commitment to the BAM work.
Existing Work: By design, each of the cities chosen for the consortium had already begun to take action toward a more comprehensive approach to serving children. The expectation was that BAM would build on and accelerate these actions while knitting together different initiatives under a single framework, rather than helping cities build something completely new. In some cases, cities already had a substantial number of initiatives underway.
Stability: Although there is no way to guarantee this—as shown by the unexpected leadership departures in several BAM cities—the Lab looked for cities that appeared to have some degree of stability in their key leadership positions. It also looked for sites that were relatively stable financially, while recognizing that cities and school districts face constant financial pressures.
Size: The Lab intentionally targeted small- and mid-sized cities for BAM, since these offered the greatest chance of success and learning in the early stages of the work.
Because BAM is intentionally designed to be experimental, with variation across cities, the Lab took a “light touch” approach to core city commitments and Lab supports. Each city agreed to participate in the following required elements of BAM’s model, and the Lab provided resources on best practices rather than prescriptive requirements on how to best implement or take advantage of them:
Children’s Cabinet: Children’s Cabinets are the governance structure for each city’s BAM work. These cabinets create a high-level mechanism to coordinate services for children across city and non-governmental organizations. Each city in the consortium has formed a cabinet that is chaired by the mayor, is co-chaired by the superintendent or another city leader, and includes representation from health and social services and other government and community organizations.
Consultant: To ensure the work of the cabinet moves forward between cabinet meetings and has a designated facilitator, the Lab supports a part-time consultant in each city. The Lab and the cities worked together to identify candidates for these positions; in some cases, the consultants already had deep local experience, while in others they were newcomers to the city contexts. The role is envisioned as a process facilitator rather than a content expert.
Twice-Yearly Convenings: To further support the cities’ work, the Lab sponsors a semi-annual series of convenings at the Harvard Graduate School of Education. Starting with the first convening in May 2016, city teams have come together with the Lab staff and outside experts as a way to deepen and accelerate the work and to build opportunities for cross-city sharing of information and resources. Each convening has included a mix of presentations and panels by top education and policy experts, “team time” for city teams to work together, and opportunities for cross-city sharing of progress. The Lab also arranges for individual meetings between cities and experts in particular areas of interest, such as financing, equity, or early childhood education.
Documentation and Evaluation: Cities agreed to participate in an ongoing documentation and evaluation process, which serves multiple purposes: to share lessons with a broader audience; to assist cities in tracking their progress on a range of process, opportunity, and outcome measures; and to inform the Lab’s iterative approach to supporting the cities in this work.
Additional Supports: The Lab has provided a range of additional supports tailored to the needs of each city that has included Reville and others’ participation in key city events, helping cities identify and connect with program partners and potential funders, and assistance with data use and outcome measures.
As the cities began their work, there was an obvious need to create a way of measuring each city’s progress toward its goals. For an initiative initially designed to last only two-and-a-half years, this presented a challenge: it would take time to design and implement the elements of the work that could lead to improvements in student outcomes, with the first potential child outcomes likely not coming until near the end of the time period or even later. The early work in each of the cities focused on creating the Children’s Cabinets, and the variation in initiatives from city to city meant that using common measures across all six cities was unfeasible.
To address these issues, the Lab developed a framework through an iterative process that acknowledged the three phases of the work in each city:
- Creating cross-agency governance structures through the cabinets;
- Providing increased programming and services to children; and
- Improving outcomes for children.
The Lab shared an early version of the framework with city teams prior to the October 2016 convening and met with the cities’ data or evaluation representatives during the convening to review it and its use. The current framework, derived from collective action approaches, contains three process-focused categories that can be measured qualitatively through evidence, and two outcome-focused categories that can be measured quantitatively through data.
This Measures of Success framework serves several functions: it tells a story about the systems-level work to better serve children and youth across multiple initiatives, it provides accessible information that can be used to inform cabinet-level conversations and discussions with potential partners and funders, and it can trigger conversations and collaborative action to improve capacity to use data for effective decision-making. The Lab also hoped to identify a small set of common quantitative measures that could provide some cross-city comparison of progress, but this proved impossible given the range of initiatives and focus on different age groups. Instead, the Lab has included chronic absenteeism as the single measure to be tracked in each city. While this measure will not make it possible to compare progress across city initiatives in any meaningful way, it does provide a common indicator that can serve as a proxy for a number of factors that influence child wellbeing, including health, mental health, and family stability.
The Lab asked each city to define its outcomes and obtain cabinet approval for them prior to the May 2017 convening. Though this was accomplished, the measures carry different levels of weight in different cities: while several cabinets have begun using the measures framework during meetings to track their progress, others have not yet incorporated them into ongoing discussions about the work. In addition, some teams have seen the work shift in a way that necessitates a change in the outcomes. Since the spring of 2017, the Lab has been working with the consultants and city data leads to identify and collect the specific data that cities will use to track their outcomes.
Documenting and learning from the work of the By All Means consortium is a central component of the Lab’s iterative approach. This research is designed to distill the enablers of and barriers to this cross-sector work, in order to inform the ongoing work of the consortium as well as to provide actionable research for other cities interested in undertaking a similar comprehensive agenda for children. This report documents the first 18 months of BAM’s initial two-and-a-half-year duration and focuses on the following elements: leadership, including cabinet effectiveness; partnerships and relationships; external factors; data; and funding. The final report, to be released in 2019, will focus on sustainability, moving to implementation, the experiences of participants in the new programs and services, and trends in outcomes for children across a range of metrics.
Data for this overview include multiple, in-depth interviews with each city’s mayor, superintendent, and other key participants in the change process; observations of Children’s Cabinet meetings; reviews of minutes from additional cabinet meetings; and anonymous surveys given to participants at the conclusion of each of the Lab’s convenings. To date, Lab researchers have visited each of the six cities twice, at approximately six-month intervals.
During the early stages of By All Means, the six cities in the consortium created Children’s Cabinets if they did not already have one, identified specific areas of focus for their work, and began, in most places, implementing initiatives directly benefiting children. Among the initiatives the cabinets have undertaken are increasing access to preschool, improving behavioral health services, expanding access to personalized learning and summer programming, and implementing individualized supports and enrichment plans. Several cities have also focused on creating data-sharing agreements between different agencies. In each case, the goal is to move toward creating systems that integrate services and supports in order to make it possible for every child to come to school ready to learn every day. In addition, each city has identified an initial set of measures by which it will track its progress, and most have either secured funding or developed a funding strategy.
The following overview provides initial findings of the enabling factors that have been important in determining how the work has unfolded and some of the challenges cities have faced in their efforts to create new systems. It features brief illustrations of these enabling factors and challenges where possible. The city-specific summaries that accompany this overview provide greater detail on how BAM has evolved in each city.
Collective impact approaches require strong individual relationships and organizational partnerships to succeed. In the context of BAM, the most important partnership is between the mayor and superintendent; while mayors lead the effort to create a new city-wide responsibility for children, superintendents govern the largest and most important child-serving agency. All BAM superintendents welcomed the recognition that schools alone cannot address all the challenges of children living in poverty and the shift to a more collaborative responsibility for providing practical supports.
The role of the BAM-sponsored consultants in each city is to facilitate the work of the cabinets, both during and between meetings, and to ensure that the work continues to move forward. Consultants are also the primary point of contact between the cities and the Lab, facilitating a range of tasks that includes overseeing the adoption of Measures of Success indicators, developing city presentations for the convenings, and reviewing city-specific materials produced by the Lab.
The Convenings Provided Support and Planning Time
The convenings have proven essential to the work of the consortium: all cities have highlighted the value of having focused, extended team time away from their cities as an important contributor to their progress. Of the many benefits articulated by participants, the most common related to the power of bringing everybody together in a neutral location, which facilitated a high level of focus on the work and relationship-building among the cabinet members; the value of opportunities to learn from other cities and from national experts on topics relevant to their cabinet work; and the importance of the friendly “pressure” and natural accountability that emerges when convening with other cities, since cabinets are eager to demonstrate progress and, simultaneously, do not want to feel as if they’ve fallen behind their peers.
Data is essential both to identify high priority needs in the community and to evaluate the effectiveness of the cabinet’s interventions to address those needs. During BAM’s early months, cities began asking for support on data use and integration. Some cities already had sophisticated data operations, but wanted to take the next steps toward integrating data across different agencies to support collective impact; other cities had data they collected but did not fully utilize and wanted to develop stronger practices for using data as a tool for decision-making and change. To address this, Lab staff began supporting cities individually on the development of, and data collection for, their Measures of Success. The Lab is also supporting one of the cities in creating a broader strategy for data use.
Several other cities have been developing additional data capacity in support of this work, including creating MOUs to share data across agencies, surveying students about their school connectedness, asking families about the barriers they face in accessing preschool, and making use of data collected through a programmatic partnership to improve their understanding of community and individual student needs. For example, one city in the cohort doubled the number of summer learning slots available to children in 2017, and did so using the following data-based strategies: analyzing student data to identify the need for summer programming, requiring providers to agree to track certain data about participants, and implementing pre- and post-tests to determine how effective these programs were in reducing summer learning loss. The city’s contract with summer program providers will be re-negotiated annually according to these results, thereby making effective use of funding as well. Another member of the cohort has addressed data gaps in particular areas of interest by moving the school registration processes online, allowing the city—for the first time—to gather data on the community’s childcare needs in real time.
Two cities in particular have excelled at incorporating data into their decision-making, with one poised to become a national leader in this work thanks to a significant external grant. This grant is supporting the city in the creation of a data dashboard that aligns with collectively determined goals across the areas of education, health, economic security, housing, and safety. In the other city, the mayor’s data lead is a member of the cabinet, enabling that culture of data to permeate the cabinet’s work. This city is pioneering innovative systems strategies to address the needs of children, including the possibility of assigning an ID at birth to track children’s progress from infancy to early childhood to K-12. Doing so would allow agencies to coordinate interventions, identify children’s individual needs sooner, and address them faster and more effectively. Due to FRPA and HIPAA laws, this effort has hit a number of roadblocks, but the cabinet has overcome challenges with persistence, knowing that effective rollout of such a project is central to their shared vision.
It is important to note, too, that cities have moved ahead on projects that benefit children and increase access to services even where they lack nuanced data, knowing the importance of meeting children’s needs sooner rather than later while also iterating and refining the work.
Finding startup and sustained funding is a perennial challenge of collective action work as well as a key measure of its success and sustainability. The Lab has provided indirect financial support to the cities through the part-time consultants, the convenings, and strategy and fundraising contacts, but it does not provide direct grants to carry out the work. The cities have found success utilizing several different strategies for funding, including reallocating or raising public dollars, identifying private funding, or securing in-kind support. In some cases, this support was provided by a cabinet member’s organization; in other cases, the cabinet successfully sought outside funding.
The first 18 months of the By All Means initiative have demonstrated both the power of a comprehensive agenda to support children and the challenges in implementing new systems to further this agenda. Leadership is essential for the work to be sustainable long term, but so are structural supports, staffing capacity, and broad stakeholder buy-in. As much as anything else, persistence matters: our strongest cities proceeded with the work amidst changes in leadership, cabinet membership, focus areas, structures, and processes. They forged innovative new relationships and connections while overcoming complex political histories and gaps in data.
As the Lab anticipated, By All Means has progressed at different rates in different cities for a variety of reasons. For example, the size of the larger cities presented challenges—more complex governance structures, greater needs for funding, more children to serve—but also advantages, such as more robust staffing and more community partners. In these larger cities, the cabinets needed to strategically align BAM with the complex landscape of existing initiatives, in a way that generated coherence and broad buy-in, to ensure the sustainability and success of the work. The progress of each city’s work was influenced by each of the themes reviewed in this document: leadership, partnerships and relationships, external factors, data, and funding. The Education Redesign Lab has, in its capacity as a partner and convener, worked to support the individual needs of each city as leaders there have encountered obstacles and celebrated successes.
The accompanying Keys to Success are distilled from the Lab’s observations of the first 18 months of this initiative. To learn more about the experiences of each city, visit the individual case studies. As the work continues to unfold and cities forge ahead on implementing new and innovative ways of working across traditional sectors, the Lab will be looking for the barriers to and enablers for sustaining the work on behalf of children after this initial startup phase.
1. Child Trends Data Bank. Early Childhood Program Enrollment. (August 2014). Retrieved from https://www.childtrends.org/wp-content/uploads/2014/08/08_Early_Childhoo...
2. Isaacs, J.B. Starting School at a Disadvantage: The School Readiness of Poor Children. (March 2012). Brookings Institution. Retrieved from https://www.brookings.edu/wp-content/uploads/2016/06/0319_school_disadva...
3. U.S. Department of Education, Institute of Education Science, National Center for Education Statistics; http://nces.ed.gov/nationsreportcard/naepdata; Pathways to College Network (2004); U.S. Department of Commerce, Census Bureau, Current Population Survey (CPS), October 1990 through 2013. See Digest of Education Statistics 2014, table 219.75; Adams, P. et al. (April 2012). Remediation: Higher Education’s Bridge to Nowhere. Complete College America. Retrieved from https://www.insidehighered.com/sites/default/server_files/files/CCA%20Remediation%20ES%20FINAL.pdf
4. Child and Adolescent Health Measurement Initiative (2013). Overview of Adverse Child and Family Experiences among US Children. Data Resource Center, supported by Cooperative Agreement 1-U59-MC06980-01 from the U.S. Department of Health and Human Services, Health Resources and Services Administration (HRSA), Maternal and Child Health Bureau (MCHB). Retrieved from http://www.childhealthdata.org/docs/drc/aces-data-brief_version-1-0.pdf?Status=Master
5. Edmondson, J., and Zimpher, N.L. Striving Together: Early Lessons in Achieving Collective Impact in Education. Albany: SUNY Press, 2014; Hanleybrown, F., Kania, J., and Kramer, M. “Channeling Change: Making Collective Impact Work.” Stanford Social Innovation Review, January 26, 2012.
6. Kania, J., and Kramer, M. “Collective Impact.” Stanford Social Innovation Review 9, no. 1 (Winter 2011): 36–41.
7. Heifetz, R.A. Leadership Without Easy Answers. Cambridge, Mass: Belknap Press of Harvard University Press, 1994.