Changing Social Norms: the Importance of “Organized Diffusion” for Scaling Up Community Health Promotion and Women Empowerment Interventions

Social norms—the unwritten rules of acceptable behavior shared by members in a group—can contribute strongly to group members’ choices and actions. Scholars and practitioners working to improve global health and promote equitable development are currently investigating how social norms theory can inform the design of prevention interventions in low- and mid-income countries (LMIC) (Cislaghi and Heise 2017). In the last few years, non-governmental organization (NGO) practitioners have implemented programmatic strategies to influence social norms related to a wide range of gender and health-related behaviors in LMIC, such as contraceptive use (Costenbader et al. 2017); child marriage (Lee-Rife et al. 2012); female genital cutting (FGC) (Mackie and LeJeune 2009); and intimate partner violence (Shakya et al. 2017). An effective tool to change social norms are “community discussions,” where members of the same group identify local harmful practices and the norms that sustain them, eventually renegotiating both to achieve greater health, well-being, and empowerment for themselves and others in their group (Linos et al. 2013). Due to the time required, however, community discussions are often criticized by those who claim they are limited in reach. Practitioners are thus left with the dilemma of trying to engage both the greatest possible number of people in each given group of their interventions, and the largest possible number of groups in the region or country where they are working.

The aim of this paper is to examine how “organized diffusion”—the sharing of knowledge encouraged by practitioners and led by program participants—can be a fruitful strategy to increase the reach of community discussions, ultimately helping interventions achieve effective social norms change. Little empirical work has been done to examine how intervention-led diffusion processes can result in normative change, health promotion, and empowerment in LMICs. Historically, interest in diffusion in development studies and health promotion mostly focused, respectively, on the adoption of new technologies to improve rural agrarian practices and on the spread of new health-related knowledge (Greenhalgh et al. 2005). However, interest in how diffusion of information shapes culture and behavior is not new in the social sciences. Since Rogers’s (1962) foundational book (The Diffusion of Innovations); theorists in communication science; marketing (Robertson 1967); political science (Mintrom 1997); sociology (Katz et al. 1963); and business studies (Nonaka and Takeuchi 1995), to cite a few examples, have shown interest in understanding how new ideas gradually take hold in groups of people (for a review, see Greenhalgh et al. 2005).

We ground our analysis of how organized diffusion can contribute to social norms change on data from three case studies, respectively from three different interventions: (1) the Community Empowerment Program (CEP), implemented in Mali; 2) change starts at home (change), implemented in Nepal; and 3) Voices for Change (V4C), implemented in Nigeria. To do so, we first review key theoretical concepts. Next, for each intervention, we explain its structure, look at the methods used for data collection, and present key results. In the discussion session, we draw theoretical and programmatic implications. Concluding remarks summarize key messages emerging from our comparative analysis.

Great attention is being paid to the role that “social norms” play in influencing behaviors that shape people’s ability to protect their health and achieve their life potential. Recent attention to using social norms theory to achieve change emerged, in part, from the realization that changing harmful practices through factual information and economic inducements alone is not effective (Gelfand and Jackson 2016; Kumar et al. 2015).

Many theories of what social norms are and how they influence behavior exist (Bell and Cox 2015). Most of the literature refers to Cialdini’s theory, that identifies two types of normative beliefs: (1) one’s belief about what others in one’s group do (called descriptive norms) and (2) one’s belief about what others in the group approve and disapprove of (called injunctive norms) (Chung and Rimal 2016; Mackie et al. 2015; Miller and Prentice 2016). People comply with social norms for various reasons, including, for instance, because they are uncertain about what is the best behavior to achieve something in a given situation, they want to express membership in a group; they anticipate a social reward, or because they are forced to by those who have power over them (Bell and Cox 2015). Theorists suggest that to change a social norm it is critical to reach out to people’s “reference group,” that is, it is critical to engage the entire network of those who share the norm in question (Mackie et al. 2015; Miller and Prentice 2016; Saxena 1971).

While in high-income countries “social marketing” approaches that aim to change social norms by correcting people’s misperceptions about what others do are often chosen to achieve social norms change (Berkowitz 2010; Gidycz et al. 2011; Miller and Prentice 2016; Stock et al. 2014). In LMICs, two (sometimes intersecting) main intervention strategies are most commonly found: wide-reaching media campaigns that often incorporate social marketing strategies (Tankard and Paluck 2016) and participatory discussions between members of the same reference group (Vaitla et al. 2017). Both approaches have possible shortcomings. Community-based discussions might have limited reach due to the relatively elaborate and resource-intense nature of the intervention. While media campaigns can fail to reach the intended audience or to spark the public dialog needed for people to change their perceptions about what others in their group do and approve of due to the fairly unidirectional nature of media broadcasts.

To overcome the former challenge, some practitioners have tested a participant-led method to share information with non-participating members of their group. This method is commonly known as “organized diffusion.” Studying how the CEP facilitated change in social norms supporting FGC in West Africa, Mackie and Lejeune identified six phases in the process of diffusion of the knowledge (Mackie and LeJeune 2009). Phase one includes the discussions happening before the program, as rumors about the intervention generate curiosity. Phase two refers to the creation of the new knowledge with a selected group of participants. In phase three, participants share their knowledge with one “adopted” member in their community: usually a family member with whom they discuss what was interesting to them during project activities that day. Then, in phases four to six, information spreads out from the intervention community to new communities, eventually reaching people across the entire larger group (ethnic group, region, or country).

To our knowledge, until today, the potential of organized diffusion for normative change in LMICs was mostly theoretical. The literature on programs that facilitate normative shifts does not explicitly examine organized diffusion, a process different from but related to “community mobilization.” Community mobilization is the final piece of organized diffusion, where participants raise awareness and generate community action in theirs and other communities. Very few programs formally integrate participants’ continuous and sustained sharing of new understandings with their immediate and larger social networks in their strategy, and, as a consequence, few studies have looked at the potential of this piece of an organized diffusion process. One notable exception is the evaluation of the SASA! intervention in East Africa that achieved change in the norms supporting domestic violence through community mobilization (Abramsky et al. 2014). When Abramsky and colleagues evaluated the SASA! program, however, they did not tease out specifically the contribution of the mobilization component. Later, Starmann et al. (2018) conducted a secondary data analysis to look at how the communication materials specifically contributed to SASA!’s success, finding that radio programs and interpersonal communication contributed to the change. The importance of organized diffusion also emerged in other iterations of the program. Studying an adaptation of SASA! in Rwanda, Stern and colleagues (Stern et al. 2017) found that visibility of change helped increase organized diffusion, eventually changing the behavior of nonparticipating members of the community, even when they personally disagreed with the new behavior. Their finding that organized diffusion can first change either people’s attitudes or practices echo those from a qualitative study of the CEP (Cislaghi et al. 2016).

Despite this evidence suggesting that organized diffusion does work, we have limited evidence (especially from interventions other than SASA!) that can help elucidate to what extent organized diffusion—among other strategies—is contributing to social norms change. This paper is, to our knowledge, one of the very few that specifically examines organized diffusion across multiple programs to assess its importance. Using a comparative case method (Goodrick 2014), we examine the effectiveness of organized diffusion to achieve social norms change across three different norms change interventions in three different contexts (The CEP in Mali, Change in Nepal, and V4C in Nigeria). We provide relevant background information, explain our data collection and analysis methods, and present results on organized diffusion by case. This information is also summarized in Table 1 below, as an introductory overview to the three case studies.

The NGO Tostan is a veteran of community-led social norms change in West Africa (Cislaghi 2018; Gillespie and Melching 2010; Kuenzi 2005). Their CEP, implemented since the 1990s in thousands of villages in rural West and East Africa, aims to support communities in achieving self-identified goals, which include reducing child marriage (Cislaghi et al. 2017; Jewkes et al. 2015; Michau, Horn, Bank, Dutt,, and Zimmerman 2015; Warburton 2014). Despite achieving positive results across a range outcomes including governance, education, health, environment, and economy (Cisse et al. 2018), the NGO gained considerable attention for its impact on FGC (Gillespie and Melching 2010) documented through a randomized trial (UNICEF 2008) and qualitative studies (Cislaghi 2017; Cislaghi et al. 2015).

The CEP has three components. The first is a 30-month curriculum on democracy, human rights, problem solving, hygiene and health, literacy, and numeracy. Approximately, 40 men and women take part in an informal education program that makes use of participatory pedagogical strategies (Bajaj et al. 2016; Cislaghi et al. 2017). The educational classes furnish a space where individual capacities, skills, and aspirations can grow, where group member norms can be renegotiated, and where economic constraints can be addressed, through the creation of group strategies for revenue generation. The second key component of the CEP is the Community Management Committee, a 17-person community group whose task is to implement the vision emerging from the classes in collaboration with the whole community. The third component is organized diffusion. Class participants share their learning with peers and family members, and committee members raise awareness throughout their locality. Together, participants and members of the committee also organize community mobilization activities in their villages and, eventually, in neighboring villages, motivating people living in those villages to join in the process to change local harmful norms. Intervention staff report that the engagement of local politicians and religious leaders also greatly contributed to the success of their program.

Change is a multicomponent social behavior change communication and community engagement strategy designed to prevent male-perpetrated intimate partner violence (IPV) against women in Nepal (Clark et al. 2017). The intervention was designed and implemented by Equal Access International (EAI) as part of a randomized controlled trial designed and run by Emory University. Change seeks to shift attitudes, norms and behaviors that underpin IPV perpetration in Nepal, positing that “changes in social norms can successfully promote sustainable change that protects women from IPV” (Heise 2011).

Change’s strategy has four core components: (1) a 39-week, “edutainment” radio program involving drama and discussion elements; (2) a 40-week “couples” curriculum delivered to 360 married couples via weekly facilitated Listening and Discussion Groups (LDGs); (3) wider community engagement largely through LDG organized activities; and (4) trainings for religious and community leaders (Clark et al. 2017). Informed by both the radio content and the curriculum and guided by EAI trained local facilitators, the LDGs provide a safe space for couples to critically reflect on existing harmful gender norms and renegotiate more mutually respectful relationships.

The pathway to change envisioned by the intervention follows three distinct phases: (1) a critical reflection phase; (2) a skill-building phase (where couples are exposed to and learn new life skills); and (3) a community mobilization phase, which encourages organized diffusion through community actions delivered by the LDGs. Throughout the three phases, LDG members are encouraged to share what they learn in the group meetings with family members, who are also invited to join the group sessions at three time points over the course of the 9-month intervention.

V4C was an innovative 5-year program (2012–2017) implemented in four of Nigeria’s states (Enugu, Lagos, Kaduna, and Kano) and focused on changing young people’s attitudes and practices in three main behavioral areas: violence against girls and women, support for women’s role in household decision-making, and support for women’s political leadership. The program sought to create change at three levels: (1) among individuals, through “Safe Space” gender courses offered to young women and men online and in person in selected higher education institutions in the focal states. In these Safe Spaces, young people were encouraged to reflect on and debate gender issues and take actions to promote gender justice; (2) in wider society, through a branded communications campaign targeting young people through radio discussions and dramas, TV, social media, and billboards; and (3) within formal institutions, through the passage of legislation enhancing women’s rights and supporting women’s participation in political structures. In such a way, V4C intended to catalyze agents of change, ready and able to diffuse their new attitudes and behaviors among peer networks and create the societal space for them, and others, to adopt more equitable gender behaviors. By virtue of their post-secondary education, the students were well-positioned to influence others in their home communities (Welsh et al. 2017).

Results from the three case studies show the potential for organized diffusion. The three sets of results suggest that facilitating a process through which participants share their knowledge with others can help achieve change in existing social norms, ultimately contributing to change in their practices. Existing theory can assist us to interpret study findings.

Neo-diffusionism (Kashima 2009, 2014) suggests that through communication, ideas are passed from one cultural agent to others. For this communication to affect the listener’s opinions, the speaker needs to tailor an appropriate message. Through the conversation, the speaker and the listener agree on the understanding of what they talked about, strengthening both the listener’s and the speaker’s beliefs relative to their conversation. Across the three case studies, our findings corroborate Kashima’s central tenet, suggesting that intracultural processes of diffusion can be facilitated when the listener and the speaker know each other well (for experimental evidence, see Lau et al. 2001). In the CEP case study, we showed that diffusion first happened in the family, where the speaker’s capacity to assess the listener’s knowledge and to anticipate their reactions is, on average, likely higher than with others in the rest of the community. In the Change study, the most prominent persons with whom the participant discussed anti-VAWG messaging was a neighbor, family member or husband. In the V4C study, the peers of directly exposed participants exhibited significant change, just on a lesser scale than those directly exposed. Neo-diffusionism also purports the more frequent the communication, the more the new information will spread across the social network, evolving into a new reality. As the new understandings become meaningful, not only to individuals but also to communities as a whole, they embody more than new emerging beliefs: they become standard acceptable ways of making sense of the world shared within one’s social group, eventually changing how members of that group think and act. In our data, we found that organized diffusion increased the positive changes in behaviors that were sustained by harmful social norms, suggesting that the new understanding and knowledge were indeed becoming part of a new shared social narrative of acceptable actions.

In addition to the theoretical literature on diffusion of knowledge within complex cultural systems, the literature on social movements can be of further assistance in interpreting the significance of the findings emerging from our case studies, as it incorporates greater awareness of the resistances faced by those who attempt to change an unequitable status quo. Christiansen (2009) argued that social movements have four key stages in their life cycle: (1) emergence (widespread discontent); (2) coalescence (population collective aware of widespread discontent); (3) bureaucratization (the formalization of the movement into an organization); and (4) decline (the end of the movement either because it succeeded or was repressed). Our findings shed greater light on how social movement move from phase 1 to phase 2. We found that a core group of motivated activists is needed to and effective at increasing individual and collective awareness of a widespread discontent with the current status quo. Then, as those motivated activists reach out to others in their community, their new understandings of how things could be different can facilitate coalescence of intents. At this point, a general sense of individual unease becomes more concrete: it gets discussed in conversations that generate new visions, strengthening people’s collective intentions to address what is causing it. Together, diffusion theory and the four stages of social movement model offer a theoretical explanation to why we found very similar patterns across three cases studies, with important implications for practice.

Collectively, the key message emerging from our findings is that integrating organized diffusion strategies within social norms interventions has the potential to achieve greater and more diffuse impact reaching others than those who were immediately and more intensively exposed to the intervention. If future community-based interventions intend to achieve social norms change within participants’ communities, they should equip participants with knowledge and skills to engage others in their network in transformative conversations. Intervention strategies that request participants to “adopt” other participants (as in the case of the CEP) can be of assistance here. The three interventions above offer three models of how this can be done, and a description of their programs exist in the literature (Cislaghi 2017; Clark et al. 2017; Denny and Hughes 2017a). While these interventions were on potentially controversial topics (such as FGC, IPV, or gender equality), we hypothesize that the organized diffusion strategy might be adapted and used for other outcomes, such as, for instance, parenting practices (Weber et al. 2017).

Our study has some limitations. The first is that the three case studies use different outcome measures to examine diffusion, the most direct being that of CEP. However, results from the three case studies converge supporting a robust finding of the presence of organized diffusion that is not confined to any one intervention type or outcome measure. The second limitation is related to the potential for participant bias. The diffusion effect might have only changed participants’ capacity to respond according to what they thought was expected of them. Data collection aimed to address this limitation by asking participants about their own practices and, in the case of the CEP case study, by triangulating information about injunctive family norms across family members. However, these limitations should be taken into account when interpreting study findings.

Social norms can sustain harmful practices. Interventions being carried out in LMICs often integrate community-based dialogs to achieve social norms change but are criticized for not reaching change at scale. In this paper, we have analyzed data from three interventions, uncovering the potential of participant-led organized diffusion of knowledge and understandings across participants’ social network. Future research and practice shall increase our understanding of the most effective and ethical ways in which organized diffusion can help achieve greater social norms change to improve global health, well-being, and empowerment. Doing so has the potential to generate additional impact with little additional investment.