25 May 2016
‘Even in the area of the world that - one would assume - has the lowest rate of corruption, even here corruption costs us a lot’, says Shaul Shalvi, associate professor in experimental economics at UvA’s Center for Research in Experimental Economics and political Decision-making (CREED) as he points out the estimated €120 billion a year cost of corruption in the European Union, only a little less than the EU’s annual budget.
To Shalvi, his main research focus - identifying the mechanisms at the roots of corruption – speaks for itself. ‘I pay taxes… I want to know: how do these mechanisms work, what is driving corruption? What are the dynamics of every-day corruption?’
The Israeli national has been studying the topic for years. Since obtaining a PhD in psychology at the UvA in 2011, Shalvi has held a number of positions including Associate Professor in the Psychology Department of Ben-Gurion University in Be'er Sheva, Israel. Last August he returned to his alma mater as an Associate Professor in experimental economics and in psychology. In September 2015, the European Research Council ERC granted Shalvi 1.5 million euros for his study At the roots of corruption: a behavioural ethics approach, which will run until September 2020.
For his initial experiments, Shalvi and his team have 20 to 30 people playing mostly interactive games behind a computer. ‘This is experimental research, in a controlled setting. But we can use the resulting insights for society. They enable us to design an environment that allows people to behave morally.’
In one example, participants are paired and person A is asked to roll a die and tell person B what number he rolled. Then person B may roll a die and say what he rolled. If both participants benefit when they rolled the same number, person B is more likely to lie than if they don’t, Shalvi and his team found in 2014: ‘People want to be honest, but they also want to collaborate. It turns out that in settings that are mutually beneficial and where collaboration is possible, people prefer collaboration to honesty.’ In such circumstances, the number of lies rises 50% when people are in a group, compared to when they are not, the researchers found. ‘Our experiments also show that things go wrong if an honest person is paired with a liar.’ The psychological mechanisms behind this are still unknown – does a feeling of shared responsibility dampen remorse? – but Shalvi did find that paired participants lie less if they don’t both get the same benefit from it.
Lab outcomes such as these are relevant for organisations that use bonus schemes, Shalvi argues. ‘Collaboration is often encouraged blindly. But though an all for one, one for all structure in which everyone benefits equally from higher results might promote collaboration, it also makes people more inclined to unethical behaviour.’ If an organisation wants to limit corruption, having more diverse groups and diversity in incentives might be wise, he adds. ‘It would be interesting to study the trade-off between the cost and the benefit of collaboration.’
With the ERC funding, one of the things Shalvi intends to study is the formation of corrupt groups. Earlier research on obligatory staff rotation found that requiring people to change jobs regularly led to a lower incidence of lying compared to forcing people to stay in the same position – hardly surprising since forced rotation means staff get exposed to different standards and there is less possibility to develop bad routines. Shalvi: ‘But what happens when staff is not obliged to rotate but may if they want to? Will staff move until they find someone with whom they can act corruptly?’
Shalvi added such a setting to his research. This study is still in its infancy but preliminary findings indicate that allowing staff to choose whether to stay or to move, the way EU institutions are organised, is not as effective at reducing corruption as forcing them to rotate, as is the case within the United Nations. ‘If you give people the option but not the obligation to switch, I would suggest you have some monitoring.’
Several other research topics related to corruption fascinate Shalvi, including ethical blind spots – when come people simply don’t see their own corrupt behaviour - and the way factors such as temptation and time-pressure interact. But to seriously combat the problem, academic research into the mechanisms of corruption alone is not enough, he finds. ‘I want to create a setting in which people who study corruption using an experimental approach, people who experience it in their working environment, and policy makers can meet to discuss and exchange ideas. This platform is still a tentative plan but it would start off with a conference next year: What What leads to corruption, and how can we tackle and control it?’
More information? Please email S.Shalvi@uva.nl.