Abstract:
Zipf's law states that the population size of a city is inversely proportional to its population rank of the city. This paper examines the applicability of the Zipf's law to the world rank of corruption. The relationship between corruption and its rank is found to be approximately log-linear but less than perfect for Zipf's law. Due to a slight concavity of the relation, either a piecewise regression or a non-linear model provides an extremely convenient tool for predicting the degree of corruption across countries. Although limited number of observations, an alternative characterization of the corruption ranks appears to obey the Zipf's law more closely.
