Over the years, more and more funders, program directors, public officials, and scholars have pushed for collaborative approaches to achieve significant social change. The impetus for this has largely been the realization that despite multitudes of initiatives dedicated to issues such as improving educational outcomes for young students, eliminating hunger and diseases, and providing better employment opportunities for disadvantaged people, wide-ranging progress remains elusive.

In their seminal article on collective impact, authors John Kania and Mark Kramer argue that such efforts can benefit from a backbone organization—a superstructure that provides coordination, planning, facilitation, reporting, and administrative skills to participating organizations. Examples of backbone organizations in the education sector include StrivePartnership in Cincinnati, Commit! in Dallas, Alignment Nashville, and Partners in Progress in Oakland and Brooklyn.

In our own research on food insecurity in Houston, however, we have observed that the backbone organization is neither the only feasible nor even the best structure for grounding multi-organization collaborative initiatives. Instead, we see it is one of four possible structures on which collaborative efforts can hang. Each of the four structures has strengths and weaknesses related to what we see as five critical elements of collaboration: user focus, a common agenda, self-reinforcement, the potential for scale, and a broad scope. Determining which structure will work best for partnering organizations requires that they examine the context and goals for their collaboration.

Food Insecurity and Interventions in Houston

While Houston is one of the largest, most diverse, and economically robust cities in America, it also features a lack of social infrastructure that has exacerbated certain poverty and health issues. The State of Health of Houston and Harris County 2015-2016 reported that 22.4 percent of people in Houston lived below the poverty level—that’s 6.6 percent higher than the national average and 4.9 percent higher than the state average. The number of adults without health insurance and persons with disabilities living in poverty in Houston is also greater than the national average. In addition, about two-thirds of Houston’s population is overweight or obese, and more than a tenth of the people in the metropolitan area are expected to develop diabetes.

Many of these health concerns stem from chronic inaccessibility to affordable, nutrient-rich foods, particularly fresh fruits and vegetables. Houston has fewer supermarkets per capita than other large cities. A 2011 map produced by the University of Houston’s Community Design Resource Center showed that more than 250,000 low-income Houstonians lived more than a mile from the closest grocery store, and nearly a quarter of those did not have access to a vehicle. In a city like Houston, where there is urban sprawl and an uneven public transportation network, this means healthy food choices are out of reach for many low-income people. The State of Health report also states that more than 25 percent of children in Houston live in food insecure households.

In our fieldwork, we learned about many supply- and demand-side interventions combatting the lack of healthy food access in areas like these (often called food deserts). Some are collaborative, and some are not. Supply-side interventions concentrate on creating new sources of affordable and nutritious foods for people who lack access. They include interventions such as governmental subsidies to entice private-sector grocers to build stores at undeveloped (or “greenfield”) sites, or bring fresh fruits and vegetables into corner stores. Demand-side organizations aim to increase users’ desire for existing sources. The nonprofit organization Recipe for Success’s Seed to Plate program, for example, aims to make school-aged children enthusiastic about gardening, food preparation, and healthy eating, and the nonprofit Urban Harvest teaches gardening to both youngsters and adults. In addition, a number of nonprofit organizations, such as Children at Risk and Hunger Free Texans, do not offer interventions, but support these programs through their community building and political advocacy activities.

But while efforts like these are worthwhile, their disjointed nature tends to limit and isolate their impact. Initiatives often differ in how they define the problem. Some focus on nutrition and health, some on economic development, others on social justice. This makes collaboration challenging, because organizational and program priorities do not seem to intersect. Project timelines also differ. Some initiatives promise a “quick fix,” where politicians and local community leaders make vivid appeals to the public, while others evolve slowly and may struggle to attract donors. Finally, initiatives have different levels of community interaction. We have seen community gardens appear without much input from the community members expected to maintain them, while certain food pantries regularly survey their service area to see what people most need. These differing missions, resources, and protocols often lead organizations with similar outputs in the same neighborhood to “compete” for customers, volunteers, and funders. As a result, their impact is limited.

However, we have witnessed several collaborative structures that hold more promise for spreading social impact. These approaches differ in how they pursue narrow versus systemic goals. They place different requirements on their beneficiaries, and they vary in their ability to scale.

Collaborative Structures

To start, it is useful to define five elements we see as imperative for collaboration:

  • User focus. Collaboration impact is best achieved when the people an initiative aims to serve as co-creators, rather than just recipients or beneficiaries.
  • A common agenda. The more alignment there is between the definition of the problem and the goals of the collaboration, the more likely it is that participating organizations will achieve impact.
  • Self-reinforcement. Collaborative activities and outcomes that produce rewards for each partner make progress more sustainable without external support.
  • The potential for scale. Collaborations that can grow easily and quickly are more likely to produce impact.
  • A broad scope. Pursuing a broader scope of activities tends to provide more holistic solutions to social problems and address more root causes.

These elements can help us evaluate the strengths and weaknesses of the four collaborative structures food insecurity organizations in Houston are using, and understand how they work together.

Structure 1: Dominant player supply chain

In the greater Houston Area, Houston Food Bank (HFB) dominates hunger relief distribution and logistics. It is the largest food bank in America, serving 800,000 people in 18 counties throughout Texas and delivering more than $100 million worth of food. It is a distribution hub to more than 600 organizations, and while 44 percent of the food it provides comes from the national Feeding America program, HFB also procures donations from farms and vendors such as Sysco.

Since 2012, HFB has leveraged its position to partner with organizations such as the nonprofit Brighter Bites (BB) to pursue demand-side initiatives. BB promotes nutrition education for elementary-aged children enrolled in eligible schools, teaching kids how to prepare nutritious meals with fresh fruits and vegetables, encouraging them to share their learning with parents, and sending each child home with about 40 pounds of fresh fruits and vegetables every two weeks. BB began in Houston in 2012 with eight schools and 900 students. Today, it serves 78 schools and more than 9,800 students, and includes sites in San Antonio, Austin, and Dallas. In Houston, HFB keeps this excellent and award-winning intervention together. Its relationship with a wide array of vendors allows it to provide a variety of healthy foods, and its infrastructure network—trucking, warehousing, sorting, and distributing—helps ensure that BB can offer reliable, high-quality services.

A common agenda is at the heart of the dominant player supply chain structure. In this case, HFB and BB define the social problem as a lack of access to fresh fruits and vegetables, and agree that providing these foods alongside education is the best way to help children and families achieve healthier eating. HFB supports the collaboration logistically to increase its impact. The model is also self-reinforcing. Since these two organizations are linked in the supply chain, they benefit from each other’s effectiveness and efficiencies, as do their other partners. BB’s partner schools, for example, can count on timely delivery of high-quality produce. Finally, because HFB has already created a capacity for expansion, and BB, although small, has leveraged relationships with its partners—schools and school districts, and the parents of students in those schools—the collaboration has the ability to scale fairly quickly and easily.

The primary disadvantage of this model is that it is standardized, and thus ignores potential users who do not fit neatly within the target population. One of the shortcomings of the BB program, for example, is that it does not reach students whose parent or guardian cannot pick up their bag of food after school—a common circumstance at low-income schools. While BB recognizes this and has pilot-tested after-hours and off-school site deliveries (such as at a local YMCA), it is a challenge for HFB, which excels at large-scale logistics and is one step removed from the families. A second disadvantage is that the scope is narrow. The dominant player supply chain focuses only on one segment of the food insecure population. Thus, while the program serves many people, many others are neglected.

Structure 2: Neighborhood wraparound

The second collaborative structure we observed features food pantries and other service providers. On check-in, food pantry staff assess whether visitors need additional services beyond food—for example, job assistance, transportation, counseling services, or a health screening. Because staff is attuned to these interconnected needs—even employing social services support staff for such referrals—visitors receive the knowledge and assistance they need to access those services. At times, the pantries also invite other service providers to do monthly or quarterly skin cancer screenings or flu shots on site.

The strength of this structure is its high level of attention to the needs of users. The directors of two Houston pantries, Target Hunger and Community Family Centers, are keenly attuned to the needs of the people they serve and have trusting relationships with them. Another strength is the model’s broad scope: People’s needs are defined more holistically and include underlying problems that can lead to hunger. Furthermore, because referrals to wraparound services are local and accessible, users are generally familiar with and thus more likely to use them.

A downside to this approach is that it lacks common agenda; the collaborating partners may disagree on the primary social issue facing users and what to prioritize. It’s also generally not self-reinforcing. The personal conviction and dedication of food pantry directors drives them to link their organizations to others, but other service providers may not feel reciprocally obligated. The biggest deficiency, however, is scale. Efforts like these touch only individuals who actively come to a certain location for food. They miss users who live far-away from the pantries or do not seek help. Finally, the services provided by the pantry are somewhat restricted by the local services present in the neighborhood.

Structure 3: Systemic (or “umbrella”)

Perhaps the most ambitious collaborative effort we have seen in this space is the Harris County BUILD Health Partnership, which operates in nearby Pasadena, Texas. Originating in 2015, the organization partners with sponsoring organizations to improve the entire food system. Initiatives within the partnership address local food production, distribution, and consumption, and have a common goal to improve community health by increasing the quality and access to food.

BUILD Health Partnership created a committee of several prestigious and well-resourced local organizations, including county government, Pasadena city government, University of Texas MD Anderson Cancer Center, University of Texas Health Science Center at Houston, and HFB. The partnership was jumpstarted by a $250,000 grant from the BUILD Health Challenge, a national awards program aimed at improving community health. The partnership, which stewards one of seven implementation grants funded nationally, emerged from participating groups’ prior work on Healthy Living Matters, a county-wide collaboration to curb childhood obesity. The organizing committee, in turn, recruited other organizations around production, distribution, and consumption.

We call this type of arrangement an umbrella structure, under which organizers assemble and build a production and delivery ecosystem. They start by identifying organizations that can support and supply services, such as governmental agencies that have access to funding, staffing and physical assets (such as health workers in a public clinic), and mandates (such as laws requiring health screenings). From there, the organizations agree on a mission, identify important services areas, and attempt to recruit experts in those areas. For example, the founding BUILD partners recruited health clinics to conduct food insecurity screenings and refer those in need to another recruit, a local food pantry. One advantage here is that suppliers get engaged early in the process.

The umbrella collaboration contributes to impact in three ways. There is a common agenda among the organizers. The common agenda of improving access to healthy foods allowed the organizers to recruit other organizations to join. The second is scope. In the case of the BUILD Health Partnership, the scope was the entire food system and access to healthy foods within the community was the target. The third is scale. Because organizers recruit other providers to create the system, no single organization needs to add a new service.

But there are also three deficiencies. The umbrella is not self-reinforcing. For example, a community college student qualifying for a food scholarship, one of the partnership’s programs, would not be able to use that scholarship at neighborhood corner stores that stock fresh fruits and vegetables, another program within the partnership. Second, while the common agenda may be strong at the top, not all organizations in the collaboration may share it. Most of the activities within the collaborations are continuations of existing services. The umbrella structure is also top-down instead of bottom-up; people who use services are treated as beneficiaries with little input into the creation of the system. As a result, they are less likely to engage and more likely to remain food insecure.

Structure 4: The backbone (or “drain”)

Although it’s more tangentially connected to food security, one of the most interesting collaborations we have seen in Houston uses a backbone organization. Founded in 2015, ProUnitas set out to increase K-12 public school student access to local nonprofit organizations by placing an obsessive focus on student behaviors. Working with schools in low-resources neighborhoods that have historically displayed some of the worst learning outcomes in the Houston area, ProUnitas created a metric for measuring attendance, in-school behavior, and satisfactory homework performance on a weekly basis. If student behaviors deviate from “good” in any category, a ProUnitas counselor embedded in the school immediately visits the student. The counselors, who maintain rapport with the students, aim to discover the cause of the behavior change and provide a solution. If a student’s attendance becomes an issue because, for example, a parent lost a local job, was forced to move out of town, and therefore couldn’t prepare meals at home, the counselor might connect the student with food social services. (Note: We researched this effort because of its inclusion of food services.)

We call this structure “the drain,” because it is set up to catch people who, so to speak, fall to the floor and are in danger of getting flushed away. ProUnitas aims to decrease the information asymmetries and transaction costs of the students it serves by directing them to the proper service providers. The advantage of this collaborative structure is that it can put laser focus on the welfare of users. It uses student behaviors at school as indicators of other, perhaps more primary problems within their lives, such as food insecurity. The scope of the effort is also relatively broad. As a backbone structure, ProUnitas helps recruit partner organizations that are useful to students, assists them with grants, and maintains strong relationships with local governments and other important partners. Furthermore, unlike the neighborhood wraparound structure led by food pantries, ProUnitas does not offer any direct services to its users. Thus, its fundraising and organizational efforts are directed toward supporting a network of other services providers that serve people directly.

The disadvantage of this structure is that coordination across organizations is difficult to do and sustain. The drain organization must be able to find high-quality providers across a wide range of services. The screening process to partner with ProUnitas is lengthy, but even so, it might recruit providers that do not necessarily share a common agenda. This can create some difficulties, particularly since the metrics of success are not determined collaboratively. Second, the collaboration is also not necessarily self-reinforcing. For example, while ProUnitas demands that a student in trouble immediately receive ancillary services, partner organizations might prioritize other people they serve. Counselors refer students to the services and follow-up with them, but the partner organizations use their own resources and logistics to handle students who come their way. Additionally, no single partner organization is responsible for a student’s success. While these drain organizations are beginning to gain national traction, they remain limited in scope and still face challenges in receiving widespread financial support from donors who traditionally favor direct service delivery.

Building Blocks for Significant and Sustained Impact

In any given community, these collaboration structures can co-exist and even work together to reinforce a range of efforts. Both the dominant player supply chain and neighborhood wraparound structures implement, while the umbrella and drain structures coordinate (at the planning level and implementing level, respectively). The dominant player supply chain and neighborhood wraparound structures implement agreed-on interventions, providing standardized and efficient services where possible and customized services when necessary. An umbrella organization like BUILD Health Partnership brings together diverse organizations to build a common agenda, determine the nature of interventions, develop metrics, and allocate resources. And the drain model can draw on agreed-on metrics to ensure that no one an initiative aims to serve falls through the cracks.

Using a range of structures addresses the limitations of any single collaborative form. For instance, creating a common agenda in the backbone collaboration is difficult because the metrics are imposed by the backbone organization on the service providers rather than emerging from them through a more collaborative process. In addition, the initial evidence of backbone organizations’ success in the educational space in other cities can be partly attributed to the homogenous nature of their target group demographic—school aged children. Further, the government mandates school attendance providing the backbone organization a relatively easy path to access children. However, the users, institutions, and organizations surrounding the food insecurity space differ significantly from those of the education space. Food users are heterogeneous. People who are food insecture are young and old, varied in their access to food sources, and might procure food from private sector conventional grocery stores, corner stores, urban farmers’ markets, fast food, as well as from nonprofits such as local food pantries. Thus, it is going to be exceedingly difficult to design a single organization or even a single collaborative structure to effectively serve such users.

There is no perfect, one-size-fits-all structure for effective collaboration. Different partnerships require different resources and support systems, and as the food space in Houston shows, it takes all kinds to create significant change.