Social Connections Can Predict a Town’s Corruption Risk

Towns with fragmented social networks have a higher risk of corruption compared to towns with more diverse social networks, according to the paper “Social capital predicts corruption risk in towns” recently published in Royal Society Open Science.

“We determined that we could predict the risk of public contract corruption in towns across Hungary by analyzing the online social connections formed by residents of those towns,” said Johannes Wachs, a PhD candidate in CEU’s Department of Network and Data Science and lead author of the paper. 

When looking at corruption scores based on competition for public contracts in towns across Hungary, Wachs and his co-authors noticed wide variations between towns that could not be completely explained by average income, unemployment rate, educational attainment levels or other socio-economic variables.

“In terms of corruption, we see a lot of heterogeneity across Hungarian towns, which is surprising because Hungary is not such a heterogeneous country,” explained Wachs. “We wanted to understand why corruption was so persistent in certain places, but nearly absent in others, and we had a sense from looking at the sociological literature that corruption is embedded in the norms of a place and how people in a community interact with each other.” 

The researchers used anonymized personal data from the Hungarian online social network iWiW, which was active from 2002 to 2012 and had upwards of 3.5 million users in Hungary, to map the social networks of individual towns. In some towns, social networks were very densely connected, and there were no identifiable isolated groups. But in other towns, the population was fragmented—there were multiple distinct groups of people that had little interaction with people from other groups.

“We found that towns with fragmentation had a significantly higher corruption risk in their public contracts,” said Wachs. “We believe the social and cultural norms in these fragmented towns make individuals think in terms of ‘insider versus outsider.’ This is a perfect breeding ground for corruption, which itself is based on excluding certain people from public contracts and public services in order to favor the people in another group.”

The researchers also considered whether residents’ social connections outside of their own town could predict local corruption. They found lower levels of corruption in towns with residents who had diverse social connections to people all around Hungary. Wachs argued that diversity of social connections made people in these towns used to dealing with more impartial norms and thus less accepting of corruption, in contrast to towns with more socially isolated groups.

So what can these results tell us about how to address corruption?

“Our results show, on a very local level, why seemingly similar towns experience such a difference in public sector corruption. Current top-down interventions for combating corruption aren’t sufficient, because corruption is embedded in the mores and norms of a place. If you want to understand how corruption thrives—whether in small towns or whole countries—you have to look at how people interact with each other,” said Wachs.

“Social capital predicts corruption risk in towns” was co-authored by Wachs, who is also a post-doctoral researcher at RWTH Aachen; Taha Yasseri of Oxford; Balazs Lengyel of the Hungarian Academy of Sciences; and Professor Janos Kertesz, head of CEU’s Department of Network and Data Science.