Social rewards fail to address corrupt practices in government - Uganda study

By Si Ying Thian

Social recognition rewards failed to change attitudes and behaviours related to corruption, although local leaders strongly believed in its effectiveness, according to a management study in Uganda.

Conventional anti-corruption approaches tend to be punitive, so the researchers wanted to test if positive reinforcement would work in addressing corrupt behaviours. Image: Canva

An experiment to reduce corruption through non-monetary means found that elected officials became demotivated due to corruption happening at other levels of government. 

 

In the study funded by the UK Foreign, Commonwealth & Development Office, the researchers rewarded the officials managing Bwindi National Park in Uganda with social recognition, hoping to improve performance and discourage corrupt behaviours among these leaders.  

 

However, the leaders slipped back into corruption as they felt that, despite their own best efforts, the outcomes of projects under their charge were ultimately “beyond their control” due to other corrupted officials.

 

The study was published in the Journal of Policy Analysis and Management this month.

 

The East African nation of Uganda is well-known for its high levels of corruption, where much of society does not consider it problematic that government officials use public offices to benefit themselves or the communities around them, according to a paper published in a social sciences journal in 2013.

 

Conventional anti-corruption approaches, such as the use of penalties, tend to be punitive in nature and often prove to be ineffective when social norms do not support reporting in the first place, pointed out another study published in 2010.

 

This is why researchers from the United States and Uganda wanted to test if positive reinforcement – by offering rewards for good behaviour – would work in addressing corrupt behaviours.

Carrot approach didn’t improve performance and integrity

 

The research found that the rewards had no effect on elected leaders completing significantly more checklist items or spending more time on committee work for the national park. 

 

Social recognition was used as a reward in this study as previous research had shown the effectiveness of social status and prestige to motivate committee members in deeply-embedded cultures.

 

This is especially so for smaller communities with dense social networks.

 

The research employed social recognition in the form of radio announcements and signs to publicly recognise elected leaders if they oversee a local project according to national guidelines. 

 

Although elected leaders claimed that they were more likely to perform better if they are recognised publicly, their self-reported status showed no evidence of them delivering on their planned projects even when rewarded. 

 

Through a survey and experimental games designed to reveal underlying norms, the study found that elected leaders continued keeping funds for themselves despite having the reward system in place.

 

The researchers also found that residents continued to lack confidence in their leaders’ ability to perform, and still gave in to corrupt leaders even when they had a chance to give a monetary reward to an effective leader. 

 

Explaining this, the researchers suggested that the long-term experience of corruption could demoralise the public due to apathy, powerlessness and resignation.

Roots of corruption 

 

Previous research could provide an insight into why corruption can be deep-rooted, and why such incentives are not robust enough to fight corruption.


A study by the same researchers in 2020 found that people are less likely to report corrupt acts when the responsibility of doing so is spread across many levels of government, making it difficult to hold any one person accountable.

 

Corruption can also function as a form of cooperation between parties, due to the psychological tendency to reciprocate and collaborate, according to another study in 2015.

 

A growing body of research questions the efficacy of anti-corruption approaches based on monitoring and enforcement without first building supportive norms, according to a 2012 study.

 

The researchers pointed to the role of governments in establishing an anti-corruption culture, and suggested that it could cascade down to societal norms. 

 

Additionally, the researchers developed three scenarios where rewards could function as a source of motivation.

 

Leaders could implement rewards for public officials who have a clear motivation for professional or political advancement, or those who do not already enjoy high social or professional status, such as frontline bureaucrats or public service providers. 

 

Social rewards might also work better in societies or organisations which already have strong, existing anti-corruption norms, likely where corruption is less endemic.

Study design

 

The researchers conducted 60 focus group discussions and 48 key informant interviews among both residents and elected leaders from both control and treatment villages.

 

For the treatment group, villages would offer elected leaders with symbolic, positive recognition if they successfully oversaw a local project according to national guidelines. 

 

The tasks included documenting the rationale for selecting certain contractors, regularly overseeing project implementation, and releasing funds to contractors only upon the completion of projects. 

 

The researchers concluded by suggesting there be “more structural reforms,” and then complementing these macro-level reforms with positive recognition, to maximise the effectiveness of the interventions.

 

“This study suggests that recognition schemes should not be considered in isolation, but must be seen within the context of more systematic and structural approaches to anti-corruption,” they stated. 

 

Source: Journal of Policy Analysis and Management

“The limits of awards for anti-corruption: Experimental and ethnographic evidence from Uganda”

Authors: Buntaine, M.T., et al.

doi: 10.1002/pam.22540