The report stresses that the impact of social norms on corrupt behaviours of public officials has been increasingly discussed by the research community, but a comprehensive review of studies on this topic has never been made. 

The authors state that the report is designed to fill this gap. The paper explores two key issues:

1. What do we know about the social norms that drive corrupt behaviors among civil servants in contexts of endemic corruption*?
2. What do we know about how to change the social norms that drive corruption among civil servants?

Social norms

Studies on the effectiveness of anti-corruption measures carried out in the last decade, show few examples of success and sustainable results, while conventional instruments targeted at the formal structures and rules in the public sector considered as corruption-prone often do not work. The authors believe that the only anti-corruption strategies targeting civil servants that have rigorous evidence of their effectiveness are audits and e-governance. At the same time, the widespread arguments around effective corruption prevention measures such as sufficient wages of servants turn out to be insufficient to counter corruption as the impact of social norms is not taken into consideration.

In Uganda for example, despite higher wages and greater autonomy of the Revenue Authority aimed at enhancing its independence from political interference, over time corruption became its major issue. The researchers believe that this was due to the great importance of family networks in the country: increased pay rates may imply more extensive social obligations, and in some cases actually result in a net loss to the individual. This state of affairs can develop into a vicious circle with higher wages leading to more corruption because the tax officer has to make up for the loss caused by such obligations.

Besides social (including gendered) norms, literature identifies formal and practical norms as well that can determine the effectiveness of anti-corruption strategies.

1) Formal norms refer to how civil servants are expected to behave based on the official rules and regulations of the civil service and codes of conduct.

In most countries these norms are modeled on Weber’s ideal bureaucracy, defined by a strict rule-bound mode of operation, and impartial in carrying out its duties. Other bureaucratic models include colonial, communist, international organizations and New Public Management bureaucracies.

At the same time even the countries that seem to have similar bureaucratic models have the bureaucracies unique in terms of their adaptation and evolution over time, which is not often taken into account in anti-corruption strategies. The authors highlight that it is important to understand that changes in bureaucratic structures or formal rules can happen at once, for instance, when the policy of “zero tolerance” for corruption is enforced. However, this does not mean that the way bureaucracy actually functions will change at once as well.

2) Practical norms can be understood as the expectations of what a civil servant needs to do to accomplish his/her duties, what will be “practical” in a certain situation defined by quickly adopted actions/decisions. For example, servants can ignore some legal acts, because they are considered as paralyzing bureaucracy; or if there are no formal procedures/instructions, the servant will improvise to perform a specific task. All in all, formal norms determine how bureaucracy should function, while practical norms mean how bureaucracy actually functions.

3) Social norms are the mutually held expectations of how a civil servant should behave in a given situation (for instance, guanxi in China or “blat” in Russia). Social norms are often divided into descriptive or injunctive: descriptive norm is a belief about what others do, whereas an injunctive norm is a belief about what others think one should do. Norms are maintained by social influence such as “approval, including positive sanctions, or disapproval, including negative sanctions; or by one’s belief in the legitimacy of others’ expectations; among enough members of the reference groups” and may vary based on time and context.

Social norms generally depend on the context, but some forms of “social logics” in a way or another can be detected in any country:

1. Negotiation: the logic is that everything, including the rules themselves, is negotiable. This means that “there is no consensus on many of the rules of this “game”, which are selected, arranged, modified and reinvented along the way”.
2. Gift-giving: gift-giving as thanks for service rendered is a practice that can be found around the globe. In each context, the practice has its own logic and reason it developed. For example, in the Soviet Union, blat, meaning “ways of getting things done through personal contacts, pulling strings and exchanging favours”, arose to help overcome shortages or cumbersome bureaucratic processes. However, these practices have continued in contemporary social practices despite the fact that Russia no longer experiences the same shortages it did during the Soviet Union era. In the countries where this social norm is embedded, refusing a gift in return for helping someone may create shame not only for the giver, but also for the recipient.
3. Solidarity networks are social networks of relationships between family, friends, neighbors, and peer groups. Phrased differently, these are the various reference groups an individual is part of and shares social norms with. Anyone who fails to respect his/her obligations to a member of one of the networks to which he/she belongs suffers reproach, and becomes the object of pressure from all members of the network.
4. Predatory authority (prebendalism) is the perceived “right” that people in power have to use the benefits of the public office for their own benefit and the benefit of those close to them.
5. Redistributive accumulation is the expectation that those in power will use that power “to eat and feed family and friends”. To refuse to use public office for personal gain is seen as “a simultaneous show of ingratitude, egoism, pride, naiveté, and even stupidity”, especially in fragile states where time horizons - both political and everyday - are short: implying that, because you do not know what will happen tomorrow, you must ensure you eat enough today and store some for later.

The listed social logics help justify the actions that are actually corrupt, but are not necessarily considered as socially unacceptable. The knowledge of these logics is crucial for understanding how to counter corruption among civil servants.

4) Gendered norms are social norms defining acceptable and appropriate actions for women and men in a given group or society. For instance, in the Democratic Republic of the Congo, women experience more severe consequences of involvement in corruption. These norms are embedded in formal and informal institutions, nested in the mind, and produced and reproduced through social interaction. They play a role in shaping women and men’s (often unequal) access to resources and freedoms, thus affecting their voice, power and sense of self.

Accountability and pressure

On the one hand, civil servants, defining administrative and professional legitimacy of their behaviours, proceed from the principles of ideal bureaucracy; on the other hand, they base their social legitimacy, i.e. social acceptability of their actions, on whether their behaviours are in line with the social and cultural logics of a wider public. This can lead to tensions in the event that the official has to decide whose expectations prevail in every specific moment, those of the state or those of society. In practice, social expectations often prevail over the need to comply with formal rules.

The types of reference groups and expectations that they link to civil servants differ a lot: it may seem that civil servants constantly navigate “multiple lines of accountability” adapting themselves to them:

[Civil servants] reminded me of chameleons, apparently slow, inscrutable, their eyes always in motion, scanning independently from each other in all directions, changing their colours according to the environment in order to make them invisible to their enemies, yet never losing sight of their aims.

In order to better understand these reference groups, it is useful to differentiate between pressures that officials experience:

1. Pressures within the civil service:
   1. Hierarchical (or vertical) pressures are pressures on a civil servant from their superiors and those higher in the civil service hierarchy.

Out of vertical pressures come two different types of accountability relationships: bureaucratic accountability (internal, downward, and highly controlled accountability of civil servants to their superiors) and patron-client accountability (a type of informal relationship in which “the client expects protection and aid from his patron, who in turn has a claim to loyalty, work or submission”);

1. Peer (or horizontal) accountability comes from colleagues or other civil service professionals working in the same area;
2. Representational (or democratic) accountability means the relationship between elected politicians and their constituents: officials experience pressures from politicians in decision-making and can take the side of those who “take care of their electorate” of which they are actually a part even if they disagree with politicians;

1. Pressures from broader society: in contexts where government goods and services are limited, exclusive or almost non-existent in the average person’s daily life - all defining characteristics of fragile states - civil servants become important brokers in their social with their members constantly reminding them of social obligations to help them:
   1. Kinship pressures are those originating from social obligations to one’s family and friends. In many parts of sub-Saharan Africa, a person’s social status is defined by the number of individuals they are able to support and the amount of wealth they are able to redistribute among both kin and clients. The success of the civil servant achieving that position or a higher one is seen as a collective success. Higher positions within the civil service also lead to growing requests for help from family members;
   2. Sociability/vicinal pressures are similar to but broader than kinship pressures and include other types of networks of close social belonging, such as ethnic groups, villages, schools, clubs, political parties etc. in which one feels an obligation to help others.

All these types of pressure and accountability may overlap, compete or reinforce each other. For example, a patron higher up in the civil service asks a lower-level civil servant to ignore a request that came to his desk. The officer formally satisfies the logic of bureaucratic accountability with regard to one criterion, respect of hierarchical order, in contradiction to the second requirement of bureaucratic accountability, the implementation of the law. Another example: a civil servant does not have a personal vehicle, but wants to help his housekeeper by providing vehicles for a funeral procession and uses public vehicles. He is acting on his social obligation to his housekeeper, whom he considers to be like family (kinship pressures), but follows the official procedures to request the vehicles, even though their use for the funeral is against the civil service code.

Anti-corruption strategies

While anti-corruption strategies that integrate changes in social norms are drafted in theory and are an emerging practice, this area remains new and insufficiently explored with no reliable and unequivocal evidence that these strategies work in practice. The data on effectiveness of these strategies consist primarily of laboratory-based experiments, some field-based experiments, historical analysis, and contemporary case studies.

For example, a study was based on laboratory experiments conducted on undergraduate and graduate students at the University of Oxford to test whether the social norms in a specific country context influence how individuals respond to similar situations related to corruption in other contexts. The student assigned the role of a private citizen had to decide whether and how much to offer the public servant as a bribe for an official service, whereas the individual playing the civil servant had to decide whether and how much to accept as a bribe. The impact of being caught and facing legal consequences was simulated with a mandatory cost deducted from the amount received from the bribe. The authors found that they could predict whether the undergraduate student assigned the role of private citizen would pay/accept the bribe based on the individual’s home country, but that these predictions were not as accurate for graduate students. They speculated that those graduate students were likely to have spent more time in the United Kingdom (and out of their home country) indicating that their values may have changed following the change in context.

Depending on social norms, the following main strategies can be defined:

1. Pressures within the civil service – hirachical:

• Change the top-down norm;
• Disrupt networks;

2. Pressures within the civil service – peer:

• Introduce organisational culture change;
• Address pluralistic ignorance;

3. Pressures within society – kinship:

• Collectively establish pro-integrity norms;
• Create space for horizontal negotiation on norms;
• Instigate collective action;

4. Pressures within society – societal/vicinal:

• Strengthen alternative norms;
• Coordination signals;
• Foster images of integrity that trump images of reciprocity.

Drafting of anti-corruption strategies integrating changes in social norms has certain limitations that should be taken into consideration:

1) Social norms are only one of many factors that affect corruption. Corrupt behaviours and related social norms are often based on the need to overcome bureaucratic obstacles and get access to other unavailable goods or services. The possibility to affect social norms to counter corrupt behaviours is particularly evident in the cases where the relevant norm is not defined by the need to survive: for instance, there is a clear distinction between civil servants soliciting bribes because so are their expectations and civil servants soliciting bribes where they are necessary to compensate low or unpaid wages;

2) It remains unclear whether anti-corruption strategies, including those that integrate changes in social norms, can address grand corruption.

Additionally, in selecting a strategy, it is necessary to carry out thorough preliminary analysis of existing social norms to define potential entry points for their change.

For instance, a team of researchers in Nigeria conducted a survey of more than 4,000 Nigerian households to understand how social norms influence corrupt behaviors in law enforcement and the public health sector. The analysis showed a divergence between injunctive and descriptive norms. Many respondents, for instance, had an ‘inflated’ view (also known as ‘pluralistic ignorance’) of the acceptability of bribery in traffic stops. Based on this finding, the authors provided a series of recommendations to challenge the common belief that specific corrupt behaviors are acceptable. This led to tangible programmatic recommendations, such as media programs and engaging trendsetters working on social norms.

Effectiveness of anti-corruption strategies

Some anti-corruption strategies integrating changes in social norms have proved their effectiveness in practice:

1) Aligning state-society norms. This strategy implies the use of existing social norms to counter corruption. For example, Rwanda has successfully conducted a large-scale reform of the public sector, based, in particular, on: a) formalizing informal dynamics: with the arrival of a new leader the informal practice of imihigo (a cultural practice where leaders publicly commit to achieving certain things and the pride associated with doing so, as a mechanism to reward them) was reinvented and now forms the backbone of a central performance monitoring scheme that covers all public institutions and officials in Rwanda, and imihgo contracts are now signed directly with the President, often at public events or in front of the Rwandan Parliament; and 2) supporting activities that foster attitudes and beliefs countering corruption: for example, the creation of ingando, i.e. camps where citizens went to learn about the history of Rwanda and the work of the government, and itorero that was a program targeting select individuals from the elite and was seen as a more advanced version of ingando aimed at indoctrinating individuals to defend state policies.

2) Nudgesany aspects of the choice architecture that alter people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives. In other words, these are interventions attempt to encourage people to make certain choices by encouraging them to make those choices, not by changing the options available. The effectiveness of this strategy has been confirmed primarily by lab-based research. However, one of these studies was conducted on the ground and produced positive results as well: the impact of posters presenting social nudges attempting to reduce bribery was tested in Manguzi, South Africa. The posters had messages that said fewer and fewer people in the community were paying bribes. The game was designed to determine if the posters influenced whether or not the participants would choose to engage in the bribe, whether they thought they ought to (injunctive norm) and what they thought other participants were doing (descriptive norm). The results indicate that the posters changed the perception of the prevalence of bribery - from the perception that it was common to the perception that it was decreasing (i.e. descriptive norm) - and that participants were less likely to engage in bribery afterwards.

3) Whistleblowers and guerilla bureaucrats (individuals who fight back against unprincipled principals). In general, the findings demonstrate that these individuals are motivated by a number of factors, including personal convictions, with the strength of these factors depending on the context.

The use of other anti-corruption strategies still demonstrates mixed results (which most likely is due to their employment in inappropriate conditions):

1) Principled principals, codes of conduct, and ethics training. Evidence suggests that each of these strategies requires sustained top-down leadership over an extended period to be successful. For example, a Public Ethics Commission established to control proper conduct of high-level officials in Brazil received strong backing from the President and, based on interviewee responses, is seen to have changed the social norms among the upper echelons of the civil service from 1999–2004; however, its effectiveness decreased dramatically after the President left office, it lost its political backing and struggled to enforce the code of conduct. Another study conducted in the form of a survey in Poland found that, by themselves, codes of ethics (formal norm setting) and disciplinary codes (punitive measures) are ineffective but, when combined and enforced, they may reduce the perception of kickbacks among civil servants.

2) E-governance. Strategies based on the implementation of e-governance use information technology to provide and automate specific government services and are one of the most promising areas also in terms of countering corruption, as they create the opportunity for citizens to directly access services without having to go through a civil servant, which reduces opportunities for corrupt behavior by removing the choices available to civil servants. Theoretically, they limit a civil servant’s ability to influence the process and, therefore, may mitigate the impact of social norms. However, it is necessary to take into account the fact that the use of ICTs is not immune from co-optation: in Indonesia, for instance, an e-tender system for procurement bids was not a panacea for procurement corruption, as civil servants and actors outside the bureaucracy still colluded to inflate the price of goods and services needed by local government. In contrast, the online platform in Rwanda has proved to be an effective instrument; Rwanda, however, has a strong enforcement of their policy of zero-tolerance for corruption among civil servants suggesting that the effectiveness of e-governance platforms may also be influenced by the broader reform environment.

3) Local social contracts. This strategy suggests shifting the focus from what a well-functioning state “should look like” to “what is working” and scaling that. For example, if improved governance is the goal, local social contracts may lead to more accountable governance, though officially they could be considered a form of corruption and are likely to marginalize “out-group” individuals. This strategy can be particularly relevant for the countries with systemic corruption, as they are often characterized by multiple local social contracts in the absence of a nation-wide one. However, the impact of e-governance on reducing the level of corruption by changing social norms remains largely unexplored.

4) Public recognition. This approach flips traditional accountability approaches by seeking to “name and fame” “good-doers”, instead of focusing on “naming and shaming” wrongdoers. This type of strategy may be attractive because it takes a proactive instead of a reactive approach and attempts to establish behaving with integrity as a social norm. However, there is no objective evidence of effectiveness of this strategy.

There are some other strategies highlighted by the authors of the review that generate some takeaways, but have no positive evidence:

1) “Best practices”. Based on a study of 6,500 public servants across Chile, Ghana, Kosovo, and Uganda, authors of a study find that public service reform “best practices” may actually be “worst practices” in contexts where corruption is endemic. Using Chile as an example of a country with strong rule of law and anti-corruption social norms, they compared the responses of civil servants from Ghana, Kosovo and Uganda in scenarios in which mangers are given greater discretion over hiring, firing and pay. Respondents in Ghana, Kosovo and Uganda soundly stated that this type of increased managerial discretion would lead to greater corruption and “unprincipled principals”. While in other countries such as New Zealand the practice produced a positive effect.

2) Public awareness campaigns. These campaigns aimed at either discouraging corruption or addressing pluralistic ignorance are common anti-corruption strategies. However, this messaging may have a negative impact or no impact at all if it does not consider the context in a proper manner (we wrote about it in detail here).

•  

Based on the analysis conducted, the authors of the review make several important conclusions in how to develop and implement anti-corruption strategies integrating changes in social norms:

1) Context matters. Both the context of the civil service and the broader political settlement in a country should be taken into consideration in selecting an anti-corruption strategy - specific social norms at play and how they do or do not influence a corrupt behavior will depend on the context. These social logics can tell us broadly what to look for, and lines of accountability can help to focus our analysis on where to look, but the interrelated factors for one context will differ from the next.

2) Social norm-based strategies are not “silver bullets”. While social norms may be a factor driving corruption, there are also environmental factors that may be more influential.

3) Understand how the intervention fits into the broader system. This makes it possible to understand, in the first place, which type and level of change the program may be able to achieve, and, in the second place, what unintended negative consequences may be.

4) Interventions should be based on input from civil servants. To understand the social norms at play, as well as the potential opportunities and challenges in shifting them, practitioners should seek the input of civil servants in their analysis as well as their program design.

5) Strategies that do not work may still teach us something and help understand why an intervention may not have achieved the desired results.

---

*Endemic corruption (or systemic corruption) is a situation where corruption is a part of functioning of the state at all levels, is actually recognised as a norm and encouraged, whereas integrity can be punished.