Data Mining Reveals a Global Link Between Corruption and Wealth

Emerging Technology From the arXiv: “Social scientists have never understood why some countries are more corrupt than others. But the first study that links corruption with wealth could help change that…One question that social scientists and economists have long puzzled over is how corruption arises in different cultures and why it is more prevalent in some countries than others. But it has always been difficult to find correlations between corruption and other measures of economic or social activity.
Michal Paulus and Ladislav Kristoufek at Charles University in Prague, Czech Republic, have for the first time found a correlation between the perception of corruption in different countries and their economic development.
The data they use comes from Transparency International, a nonprofit campaigning organisation based in Berlin, Germany, and which defines corruption as the misuse of public power for private benefit. Each year, this organization publishes a global list of countries ranked according to their perceived levels of corruption. The list is compiled using at least three sources of information but does not directly measure corruption, because of the difficulties in gathering such data.
Instead, it gathers information from a wide range of sources such as the African Development Bank and the Economist Intelligence Unit. But it also places significant weight on the opinions of experts who are asked to assess corruption levels.
The result is the Corruption Perceptions Index ranking countries between 0 (highly corrupt) to 100 (very clean). In 2014, Denmark occupied of the top spot as the world’s least corrupt nation while Somalia and North Korea prop up the table in an unenviable tie for the most corrupt countries on the planet.
Paulus and Kristoufek use this data to search for find clusters of countries that share similar properties using a new generation of cluster-searching algorithms. And they say that the 134 countries they study fall neatly into four groups which are clearly correlated with the wealth of the nations within them….Ref: arxiv.org/abs/1502.00104  Worldwide Clustering Of The Corruption Perception”