The paper stresses that domestic anti-corruption strategies almost always include such line of action as anti-corruption awareness-raising. In the first place, it implies raising awareness of corruption as a phenomenon, of its forms and consequences among the general public. The key expectation of these activities is to increase the number of persons reporting corruption offences to competent authorities, and rejection of corruption through the change of social norms also by promoting the principles of non-tolerance of corruption, transparency, integrity etc.

However, recent studies point out that in certain conditions anti-corruption awareness-raising cannot produce the intended results or can even backfire by increasing corruption.

Since 2016, nine such studies focused on the phenomenon under consideration have been released in seven countries, namely in Costa Rica, Indonesia (Jakarta), Papua New Guinea (Port Moresby), South Africa (Manguzi), Nigeria (Lagos), Mexico and Albania. The studies assessed the impact of anti-corruption awareness-raising in the form of short messages on the fight against corruption at the domestic level on its public perception and readiness to participate in corruption interactions. All in all, 19 such messages were tested:

• Increasing rate of bribery in country – Costa Rica;
• Grand corruption is endemic – Indonesia;
• Petty corruption is endemic – Indonesia;
• Government successes in anti-corruption – Indonesia and Nigeria;
• Citizens can get involved in anti-corruption – Indonesia;
• Corruption is endemic – Papua New Guinea, Nigeria, Albania;
• Corruption is illegal – Papua New Guinea;
• Corruption is against religious teachings – Papua New Guinea, Nigeria;
• Corruption is a “local” issue – Papua New Guinea, Nigeria;
• Bribery declined recently in region – South Africa;
• Corruption steals tax money – Nigeria;
• Citizens strongly condemn corruption – Mexico, Albania;
• Wealth is lost to other countries – Albania.

The participants of the survey were randomly assigned to one of the two groups, treatment or control: those who belonged to the first group received one of the above-listed messages, whereas the members of the second group were not influenced in any way. Then, all participants underwent a survey or were invited to play a real money game where they had a chance to bribe an official to do better than the rival. After that, the results of the treatment and control groups were analysed and compared to assess the impact of the messages received.

In ten cases, the preliminary circulation of anti-corruption messages had at least one undesirable effect. For example, the information on the increasing rate of bribery in Costa Rica led to the increase in the number of respondents indicating that they were disposed to take a bribe; the message on the government’s successes in anti-corruption decreased public confidence in the real effectiveness of those measures etc. In seven other cases, the messages produced no effect, and only in two cases they had a positive outcome (“Citizens strongly condemn corruption” in Mexico and “Corruption is a “local” issue” in Papua New Guinea).

The U4 experts believe that this effect can be caused by the use of descriptive data in some messages suggesting that other members of society are involved in corruption, benefit from it and/or contribute to its spreading (for example,” Petty corruption is endemic”, ”Corruption is endemic” and the like). As a result, instead of understanding the need to counter corruption, those who receive these messages can consider the involvement in corruption as socially acceptable.

However, the studies also unveiled that not only descriptive but other forms of messages turned out to be ineffective: if the former almost always produced negative effects, the latter either had negative effect on the beliefs of the participants or had no effect at all (even if the message was positive, for instance, concern the government’s successes in anti-corruption). The researchers believe that in this case the cause was the lack of truly tangible validity of respective statements for the citizens.

Additionally, the authors underline that, based on the studies where the messages had a positive impact, it is impossible to make definite general conclusions, as in both cases repeated studies did not produce the same effect.

The studies described had certain methodological flaws. In particular, they did not take into account the effect of the following factors:

• who broadcasted the message, specifically, what media were involved;
• the frequency and the intensity of circulation of relevant messages;
• whether the broadcasting of the messages was combined with other anti-corruption measures etc.

But even taking into account the shortcomings of the studies reviewed, their findings indicate that anti-corruption awareness-raising in the form of short messages rarely produces expected results, which raises doubts about the expediency of this kind of measures to effectively counter corruption. In assessing this expediency, the U4 experts put forward the following recommendations:

• pre-test the messages on a sample of the target audience and use experimental techniques rather than rely exclusively on focus groups and interviews where the participants are believed to be able to give expected answers and even unconsciously distort the results due to the pressure exercised by the researchers;
• be prepared that even the most promising campaign to broadcast anti-corruption messages can turn out to be unsuitable for changing certain behavioral models, for example, in an authoritarian state or where corruption is functional;
• combine anti-corruption awareness-raising with other anti-corruption measures;
• ideally, refuse to widely use short anti-corruption messages until they are sufficiently elaborated and prove to be consistently effective.