The paper entitled Strategic Anti-Corruption Communications: Guidance for Behaviour Change Interventions contains recommendations on how to promote anti-corruption awareness-raising based on the theory of change (ToC). The studies proving the success of its application have already been conducted by the BIG experts: according to the findings of a pilot project in Tanzania that we have already written about, anti-corruption awareness-raising lead to a considerable reduction (by 14 to 44 per cent) in the intent to offer gifts to medical workers.

Impact of anti-corruption messages

The paper stresses that in certain conditions anti-corruption awareness-raising can lead to an unexpected result and even have an opposite effect increasing corruption. The same conclusion was previously made by the U4 experts in their analytical publication Message Misunderstood: Why Raising Awareness of Corruption Can Backfire.

The BIG experts make some additional conclusions in their guidance regarding the impact of different types of anti-corruption messages on general public:

• Messaging of conventional wisdoms on the pervasiveness and harm of corruption that are already known to the citizens and confirming the beliefs that already exist generally produces opposite effect, provoking “fatigue from corruption”;
• Employment of “politically correct” language appealing to the illegal, immoral or unethical character of corruption does not produce tangible effect in the public attitude to corruption: even if corruption is widespread, citizens tend to blame the corrupt, recognising the harmful effect of corruption, which means that these messages contain no new information;
• Selection of a specific (narrow) target audience and adjustment of messages to it can produce positive effect.

Anti-corruption attitudes, behaviours and social norms

The paper highlights that the resilience of specific patterns of corruption in society can be explained by existing social norms, including:

• Descriptive social norms, based on the perceived dominant patterns of behaviour within a given group (what one believes most people are doing) in response to a particular situation;
• Injunctive social norms, based on the acceptability of a particular behaviour (what one believes most people think is considered the right course of action in response to a particular situation).

The influence of certain social norms can lead to collective action dilemmas when people find themselves socially pressured to engage in corruption despite their personal preferences, or when perceptions about social norms are influenced by stereotypes or biased information and thus not aligned with true collective preferences.

The researchers suggest using different approaches to formulating suitable anti-corruption messages taking account of potential impact of social norms:

• Revealing true injunctive norms (real attitude of citizens to corruption cannot coincide with their behavioral models while messaging responsive to their internal attitudes can create the conditions for changing their behaviour);
• Revealing true descriptive norms (it can be useful to try to challenge the existing wide-spread narratives, for instance, that corruption is pervasive, by claiming the opposite);
• Challenging both injunctive and descriptive norms. This approach was used in the pilot project in Tanzania: devising anti-corruption messages appealing to health workers by both authoritative figures (management) and some hospital staff recruited as “champions” to disseminate the messages through their social networks was aimed at indicating a shift in the injunctive social norm (the acceptance of “gifts” is no longer socially acceptable), while placing posters and desk signs in hospitals targeting the facility users was meant to signal a change in the descriptive norm (i.e. most people are no longer offering “gifts” to health workers).

Anti-corruption awareness-raising mechanisms

Anti-corruption messages can be conveyed through different mechanisms: priming, learning and persuasion. For these mechanisms to produce positive effect, it is needed that:

• They take into account specific corrupt practices that should be addressed, peculiarities of their implementation and existing conditions, and be underpinned by reliable evidence;
• Messages incorporate the vocabulary used by the target groups to indicate corrupt interactions, have the right tone to attract attention and elicit a reflection conducive to updating beliefs rather than cause “corruption fatigue”;
• Messages are communicated to the target groups by reputable and trustworthy individuals, and potential need to employ different dissemination channels to reach different audiences.

Use of behavioural science

The paper stresses, however, that even if all above-mentioned conditions are met, messages cannot produce the desired effect. In this case, the authors suggest turning to behavioural science.

Firstly, it should be taken into account that people can be open to change their beliefs when their environment changes. This is why besides the language and the content of anti-corruption messages, the timing of their dissemination can be important: for instance, they can be conveyed immediately after elections or in a situation where a significant event changes corruption risk perceptions (for example, a large-scale corruption scandal is unveiled).

Secondly, the efficiency of anti-corruption interventions can be enhanced by incentivizing citizens to independently search for additional information: research demonstrates that when people are less confident in their beliefs, it is more likely they will search for new information; consequently, challenging conventional wisdoms can be an important first step triggering the desired reaction.

Theory of change

As regards the use of the ToC for anti-corruption interventions, the authors of the guidance recommend taking into consideration the following elements in preparing the intervention:

1. Goals. Understanding the desired result is the basis for anti-corruption awareness-raising.

Relevant examples may include:

• Change in corruption perceptions, impact of corruption on society, anti-corruption action of the government, citizen participation in the fight against corruption, beliefs of other people;
• Promotion of participation of society in the fight against corruption by engaging citizens in monitoring and control, reporting of corruption offences, campaigning and voting;
• Convictions to abstain from corrupt actions, which will allow decreasing bribery, favouritism, vote-buying and enhance law enforcement.

2. Problem and its context. Yet another important element is the definition of a specific problem to address through anti-corruption awareness-raising and its context.

An example may be the following: “Bribes are frequently exchanged between hospital staff and users in hospital x, impeding fair access to healthcare and endangering patients’ health.”

It is important to note that this problem in most cases will be impossible to solve, especially where corruption is functional, without taking into consideration the context. As a result, the effort aimed at changing behaviour should be supported by the measures to mitigate the factor favouring corruption such as amendment of legal acts, wage increase, provision of sufficient amount of medicines etc., as well as take into account potential consequences.

3. Preconditions. After the goal is set and the problems to be solved by the intervention are defined, it is necessary to clarify the necessary preconditions and initial outcomes that need to be achieved in order to reach the desired ultimate outcome.

To this end, it is necessary to answer the following questions:

• Which outcomes (ultimate outcome and preconditions) are feasible considering the local context and possible entry points?
• Who is the target group?
• Are we formulating concrete changes, not just project deliverables?
• On what level (i.e., system level, local level, individual level) are we aiming for changes?
• What type of behavioural change must take place to achieve the ultimate outcome?
• Are the ultimate outcome and the preconditions specific, measurable, achievable, realistic, and time-bound?
• Can the outcomes be solely achieved by the programme? What is in the power of influence of the programme?
• What resources are needed as preconditions?

The answers to these questions should be followed by the elimination of all obstacles detected to effectively change citizens’ behaviours.

4. Behavioural categories. In preparing for anti-corruption interventions, it is also advisable to consider the underlying assumptions and expectations (“behavioural categories”): descriptive norms, injunctive norms, personal attitudes and behaviours.

Hence, quantitative baseline and endline data on the four behavioural categories should be collected and analysed to track how changes are being elicited by the intervention.

5. Power of influence. The authors of the guidance remind of the need to consider the power of influence of the awareness-raising programme and non-programmatic preconditions that require the adjustment of the programme. For instance, a programmatic precondition would be to increase awareness of corruption implications and thus increase public participation in anti-corruption efforts. For example, it is not in our power of influence to improve the transparency and accountability of government institutions, but it is a precondition to improve trust and confidence in the government, which in turn would be a precondition to increase public participation in anti-corruption efforts.

6. Messengers. Finally, the authors of the guidance highlight the importance of the choice of messenger, i.e. peers, persons of authority (managers, politicians), agents with certain expertise etc. The prerequisite is that the target audience places trust in these individuals.

Final recommendations

The BIG experts put forward the following additional recommendations on how to organise anti-corruption awareness-raising:

• Invest in a strong monitoring and evaluation framework that would allow for monitoring the achievement of results throughout the project;
• Engage stakeholders in intervention design, observe real-world behaviour rather than merely survey experiment results and use messages about the concrete impact of corruption, especially on the inner circle of the target groups;
• Develop a long-term strategic approach envisaging not only anti-corruption awareness-raising but also changes in other factors of the choice architecture (for instance, the regulatory framework), community meetings, cultivating networks of champions, engaging opinion leaders and credible monitoring.