Defining a Community is Not a Simple Process
Over the past few years, the Transparency for Development (T4D) program team has been sharing stories and lessons from our mixed methods research into whether citizen engagement at the community level can improve maternal and newborn health outcomes.
Engaging with and empowering communities is an integral component of transparency, accountability and participation programs. But communities are not homogeneous, and the best approach to working with a community is likely to differ depending on what we actually mean when we use the term “community.” For example, do we define the term geographically, such as people who live in the same neighborhood, village or county? Do we mean all people, or a particular subgroup, such as only village leaders, or a small but representative mix of average citizens?
How does the T4D program define community? We get asked this question all the time, and it’s one we often asked ourselves during the design process. Here are a few insights into what we mean by “community” in the context of the T4D program.
A ‘lived’ community
Traditional community scorecard interventions often include participants drawn from all villages accessing a particular service, such as a health facility or school. For example, participants in the CARE International community scorecard intervention in Malawi were drawn from the beneficiary community of a health center, which encompassed more than one village. This makes sense if the intervention is focused primarily on service delivery. But the T4D program focuses on the health problem of maternal, newborn and child health (MNCH), which encompasses issues related to service delivery, but extends beyond to include problems that can be addressed by the community itself (e.g., access issues or lack of knowledge of healthy MNCH practices).
We chose to focus the program around the “lived” community of a village. The idea is that all of the villages served by a particular facility might not have the same community level issues, even if they do share a similar service delivery experience at the facility. Furthermore, participants from the same village are likely to have pre-existing relationships with each other and with local leaders, making it more likely that they would be able to work together to improve MNCH in their community.
This decision was not self-evident. Others have argued that in order to improve the performance of frontline service providers, or to successfully interact with government officials who have authority over the health facility, it is necessary for many villages to work together — more people, more power. This was one of many tradeoffs the team faced in designing the intervention.
Who to include
Once we decided to work at the village level, we needed to determine the size of the T4D community meetings. Some transparency and accountability interventions keep community involvement completely open to all, rather than focusing on specific community representatives, and we initially considered this approach.
We quickly learned through an iterative piloting process that too many participants made it difficult for the facilitator to manage meetings, and led to domination of the meeting by a few vocal participants. We also learned that, by keeping meetings open, we were inviting curious, but uncommitted, onlookers, which diluted the engagement of the community group as a whole. As a result, we decided on a closed group of 15-16 community participants, known as Community Activists, or CAs.
Another important decision, especially once it was determined we would work with a small group of invited participants, was who to invite. We wanted to ensure the community activist group was representative of the community as a whole. With this mantra in mind, we strove to include a mix of participants from different education backgrounds, genders, ages, and, where applicable, from the various ethnic or religious groups in each village. We also attempted geographic diversity; where a village included more than one subvillage, we tried to recruit at least one participant from each.
Beyond demographic characteristics, we looked for people with certain characteristics. Indonesia program partner PATTIRO suggested a list of attributes such as “doers” (those who had previous experience working for the benefit of the community) and those who were “vocal” (CAs who would be comfortable speaking up at meetings and approaching others for help), and suggested ways to identify and recruit participants with these characteristics. We also looked for enthusiasm and deep personal interest in the topic of maternal, newborn and child health.
The final characteristic, and perhaps the most important of all, was recruiting people with the time to participate. This is another instance where piloting came in handy: we had an experience in one pilot community where the CAs were enthusiastic and exhibited leadership qualities, but were simply too busy to schedule meetings. In this community, the program never got off the ground.
Who to exclude
Village leadership and health service providers
The village head and the midwife can be allies in this type of program. They may have connections, power, or specialized knowledge helpful in creating change. In fact, many transparency, accountability, and participation programs purposefully include these actors in the mix. However, village leadership and health providers can also be the problem: in one pilot community in Indonesia, the village head was using the village’s only health facility as his private home.
In the end, we decided not to invite village leadership and health service providers to be community activists in the T4D program. This didn’t mean they couldn’t be involved — just that the people who did participate had to decide whether or not to engage with them. In many communities, participants chose to do just that. We saw a mix of actions that included community activists asking the village head to broker a meeting with their health workers (examples of which are described here) or working with health facility leadership to install and manage a complaint box at the health facility. We also saw communities where the CAs chose a different path, such as one village in Tanzania that organized a chicken co-op to raise funds to defray costs associated with MNCH services. In this example they did not work with the health workers or village authorities at all. This flexible approach of “non-prescriptive” social actions — another of the T4D design principles — was intended to enable community members themselves to determine who to work with to improve their MNCH.
Traditional birth attendants
Traditional birth attendants (TBAs) are generally not part of the formal health system. They are also not considered “skilled” health professionals in terms of delivery care. But TBAs are part of the community ecosystem and, in many places, have played a role in prenatal, postnatal and delivery care for generations of women.
TBAs can serve as important allies for community members in this type of program. In fact, we’ve heard of examples of midwives and TBAs setting up partnerships so that both are present at the birth of a child. However, TBAs can also be part of the problem. For example, in Indonesia, dukun bayi (the local term for TBA) provide a number of services to expectant and recent mothers, including prenatal massages and help with child care and housework after the birth, making them a tempting alternative to skilled care. There is also an element of tradition here, where members of the same family have received care from the same dukun bayi for generations. Just as with village leadership and frontline service providers, not including TBAs as part of the core group of community activists leaves greater freedom for the group to decide whether and how to engage with TBAs in the way that is most productive for their communities.
We weighed many pros and cons here, including that TBAs might feel excluded if not invited, or that they might become scapegoats in meetings if they were. After piloting, and in close consultation with PATTIRO and our Tanzania partner, Clinton Health Access Initiative, we came out in two different places in Indonesia and Tanzania: in Indonesia, the T4D program did not explicitly include TBAs, but in Tanzania, where we determined that excluding them might negatively impact the community dynamic, TBAs were invited to participate.
Defining a community is not a simple process, and different definitions can lead to different conclusions in terms of participation, accountability and health actions. We defined the community to enable participation and representation by a diverse mix of people and to create a space for dialogue, encouragement and action. Approaching how we defined community in a very deliberate way and iterating on this definition helped us increase citizen participation by people whose voices are frequently unheard. And the question of who was included (and who was not) played an important role in the problems and the actions that became the focus of the intervention.
Jessica Creighton is assistant director for the Transparency for Development (T4D) project at the Harvard Kennedy School’s Ash Center for Democratic Governance and Innovation.