Anti-Corruption Programs — Know Your Crowd!

By Cheyanne Scharbatke-Church & Diana Chigas

Social norms exist within a group. They represent mutual expectations, not just common beliefs, within the group about what is the right way to behave in a particular situation. And it is the approval, disapproval or other social sanction from the members of the group that helps ensure compliance with the norm. Therefore, understanding the group — who is in and who is out — matters for programming. If we believe that social norms should be part of multi-faceted anti-corruption programming, then we need to know how to identify and change social norms, which includes determining who is in the group.  But what do we know about identifying the group? 

Social norms literature provides surprisingly little concrete guidance on what constitutes a group or how to identify who is in the group, although references to the role of the group (or “reference group”) abound. Yet how can one effectively target an intervention if one doesn’t know what the boundaries of the group are? With this challenge in mind we have mined the theory and practice literature in an effort to provide more clarity.

What is not a characteristic of a reference group?

The literature offers more on what is not a characteristic of a reference group than it says about what is necessary to be deemed a group, so we will start there. 

There are no geographic, size or formality requirements for a group. Group boundaries are not necessarily defined by:

Formal Nature

Groups do not have to have formal structures. While a group can exist formally (e.g. a committee or staff of a non-profit), it is also possible for a group to be a loose set of people with a common affiliation.


Groups are not necessarily based on physical proximity. Groups may overlap with geographic boundaries, like the residents of a village, yet it is also possible that the group is physically spread out, like people in a diaspora.


Groups do not have to be limited in number. Social norms can exist among a handful of people who self-identify as important to each other or among millions as in the case of a social norm that is held by the majority of citizens in a country.

What is important in determining who is in the group? 

The literature coalesces around one key characteristic of reference groups for identifying, understanding and influencing social norms:

The group is important to the individual and vice versa

A key characteristic is that the people in the group matter to the individual. In other words, the individual needs to care about the opinion or perspective of the people in the group in order for the potential positive or negative sanction to have an impact on their behavior. 

We believe that is not sufficient to define a group. In addition, the individual also has to matter to the group. While this is not emphasized in the literature, we believe that it is not enough that the person care about the group; the group also has to see the person as one of theirs. Given that social norms are mutual expectations there must be a reciprocity in the relationship; the other members of the group also need to care how the individual behaves.  Therefore, we suggest that the sense of belonging — both by the individual toward the group and the group toward the individual — is important to the mutual aspect of expectations that characterize social norms. 

Once a member of a group, not always a member: the importance of a group to a person can change, resulting in the degree of influence of the group’s norms increasing or decreasing. Different groups may be more or less important to people at different stages of life, for instance. For anti-corruption practitioners, this means monitoring group influence could be an important part of understanding the context.

People belong to many reference groups

People are members of many different groups (with different social norms, sometimes overlapping) at the same time. The question for someone trying to change social norms is which group matters for the behavior that we wish to change.

As an example, consider a colleague of ours who is a judge in Kampala. A few years ago she was on maternity leave, and during that time she was part of many groups, all with different and overlapping social norms. She could see herself as part of:

  • a group of mothers with babies in her neighborhood (no formal structure, geographically based, importance of group will likely diminish over time),
  • a fundraising committee for a nonprofit that provides menstrual products to schools (formal structure, not geographically dependent, influence will remain as long as participating in committee)
  • as a parent at her older child’s school (no formal structure, somewhat geographically dependent, influence will likely diminish when child leaves school)
  • her university alumni association from her LLM in the United States (formal structure, not geographically based and will continue in influence as long as she feels connected)
  • her profession (judges) (formal and informal structure, not geographically based, and continuing influence), as well as her courthouse (geographically based, defined boundaries, influence as long as she is in the particular district)
  • as a female judge (no formal structure, not geographically based, importance lasting as long as she identifies with other female judges)
  • her family or clan (informal structure, not geographically dependent, long lasting importance)

The group is salient in the particular situation where the behavior happens

If we belong to many reference groups, and if those groups have different internal mutual expectations, then which group impacts one’s behavior in a particular situation? The literature suggests that in order to influence individuals’ behavior through social norms, the group needs to be salient to the situation where the behavior occurs. Salience is frequently referenced, but without much clarity as to what it means. 

Our interpretation is that the group’s norms will be influential when the group itself is relevant to the situation and behavioral choices at hand. In a professional context, for instance, the mutual expectations of your ‘friends back home’ (i.e. a reference group amongst many) may not be seen by an individual to be relevant to the context.

For instance, consider an economics professor who has been a member of the same small university department for 20 years. She lives in a place where bribery is common, particularly among the police service. She is considering offering a bribe to have her son released from police detention on what seems to be a fair charge. If this professor’s academic department had a strong social justice lens and her colleagues advocated for fairness in the rule of law, this could influence her away from offering a bribe. The bribing behavior would be contrary to the mutual expectations of what was the right behavior in this group. Due to the culture within the economics department around these issues, they become salient to this situation.

While the physical presence of members of the group can matter, in that if they are physically present in the situation it could strengthen the influence of the norms on behavior, presence is not required. It is asserted in the literature that the likelihood of the group learning of the behavior can be sufficient to influence behavior.

We feel there must be more known about what constitutes salience. When reference groups have competing mutual expectations and both could be seen as relevant to a particular situation, which one influences the behavior? How do we know which group’s norms play the stronger influence?

Assistance appreciated

Did this description help or hinder your understanding? Are there ways we could make it easier to follow? How about more useful?

Do you have materials that offer guidance on what constitutes “the group” when it comes to social norms? If so, please put them in the comments section so that everyone can benefit. 

About the authors

Cheyanne Scharbatke-Church is Principal at Besa: Catalyzing Strategic Change, a social enterprise committed to catalysing significant change on strategic issues in places experiencing conflict and structural or overt physical violence. She has significant experience working on anti-corruption and state legitimacy in the DRC and Uganda. As a Professor of Practice, she teaches and consults on program design, monitoring, evaluation and learning. Cheyanne is also the Corruption in Fragile States blog series editor.

Diana Chigas, JD, is the Senior International Officer and Associate Provost at Tufts Universityand a Professor of the Practice of International Negotiation and Conflict Resolution at the Fletcher School of Law and Diplomacy. Diana has worked with governmental and non-governmental organizations on systemic conflict analysis, strategic planning, and reflection and evaluation to improve the impact of peace programming. Diana has over 25 years’ experience as a facilitator and consultant in negotiation and conflict resolution, as well as supporting and evaluating social change programming in conflict-affected countries. Her interest in corruption emerged from her experience supporting peacebuilding programming, where corruption has consistently been a key factor hindering effectiveness and driving conflict.

Related Posts

6 thoughts on “Anti-Corruption Programs — Know Your Crowd!

  • March 6, 2019 at 2:35 pm

    Hi both – great post and I’ve enjoyed reading the others on social norms too.

    One question, on: “The group is important to the individual and vice versa”. You propose that the other members of the group also need to care how the individual behaves. Great point. However, I wonder if this is still, ultimately, in the perception of the individual?

    ie, is it really a question of whether the individual believes they belong to the group, or believes that the group cares about them? (Technically they might believe they belong, even though the rest of the group doesn’t actually care about them. Perhaps unusual but it could happen?)

    As per the earlier post, when we talk about social norms as beliefs about what others do, and beliefs about what others think we should do, this seems to mean that the mutual expectations are still in the “eye of the beholder”. I wonder if the same applies here.


  • March 6, 2019 at 3:25 pm

    Hi Derek

    Thanks for the comment — it is thought provoking. On one hand it would make sense, that if the individual feels like they are part of the group then the informal rules that dictate behavior would be adopted by that individual because they self-identify in this way. But what about the sanction (positive or negative) component of social norms as they are the key to maintaining the norms? In the short term the individual could be self-sanctioning e.g. I better do this or they will think less of me, but what about over time? If the group doesn’t see the individual as belonging they don’t expect them to comply and therefore do not provide the reinforcing signals e.g. a smile/nod or a frown; does the norm maintain its ability to influence behavior over time? It wouldn’t mean they don’t do it but it could mean the strength of the norm is quite weak. That is my initial reaction, but like all good questions, I will keep thinking on it.

  • March 6, 2019 at 5:49 pm

    Dear Cheyanne / Dear Diana,

    Interesting post. I think there is the need for some small N studies of bribe givers and takers to generate some evidence to the hypotheses here of group affiliation role in behavior regarding bribe giving and taking. The general hypothesis seems to be ‘bird of a feather flock together’ and where they do more mutually enforced flocking (socialization) , they are more likely to mimic the group behaviors.

    A worker employed by an NGO involved in anti-corruption activities and regularly involved in their Church which also actively promotes anti-corruption would be LESS inclined to bribe than a worker whose employer had no involvement in anti-corruption and who either did not affiliate with a Church or affiliated with a Church which had no involvement in anti-corruption efforts. My anecdote, the first worker who paid the traffic cops bribes 5 times over a 2 years period. But maybe we need to dig deeper on my other group affiliations.

    Regards / Stephen

  • March 8, 2019 at 4:47 pm

    Hi Stephen

    Interesting comment. I agree that there needs to be far more study into how the social norm literature applies to corrupt behavior and applies to corrupt behavior in fragile contexts in particular. One of our overarching questions is around the transferability of the literature as the vast majority of the research, particularly lab style experiments, comes from Western contexts.

    Thanks Cheyanne

  • March 15, 2019 at 3:32 am

    People are more prone to break rules when they are in a home environment than in a foreign environment. This is best explained by a saying,”you are what you do when nobody is watching”. If this is true, is not group norm also a temporary solution to the problems of corruption?

    • March 20, 2019 at 2:44 pm

      Hi Narayan

      Thanks for the interesting reflection. The social norms research actually speaks a bit to this issue from the perspective of the strength of a social norm. It is hypothesized that one characteristic of norm strength is the degree of detectability of the behavior itself. We would argue for corruption that it is whether the behavior can be observed as well as if it is likely that a member of the relevant reference group will find out about it. So we may not have witnessed the exchange of money, for instance, but know that someone who is a midlevel civil servant should not have the means at their disposal legally to obtain a collection of expensive brand-name handbags. The norm behind a behavior that can not be detected still has influence but maybe not as much. There is also some work that says that norms can be integrated so much into one’s personal identity that the sanction for breaching e.g guilt can be self-imposed.


Leave a Reply to Narayan Manandhar Cancel reply

Your email address will not be published. Required fields are marked *