Does the demand for redistributive policy increase in response to intensifying income inequality within a society? If so, as price is adjusted to eliminate excess supply and demand and return the market to equilibrium, there exists a form of control device that withstands deepening income inequality. Unfortunately, there is no such conspicuous correlation between income inequality and redistributive policy Some researchers, including Alesina and Angeletos(2005), have shown that the decision-making on redistributive policy is not affected by income inequality itself but is directly influenced by the judgment on income inequality by members of society; specifically, the social and normative perceptions regarding the fairness of the fundamental sources of income inequality. This study departs from the awareness of such an issue. The problem with determining the optimal level of redistribution in a society is figuring out which part is to be subjected to the redistribution with regards to income inequality. The criteria should be based on people’s normative perceptions of fairness within their society. The definition of social norm regarding fairness in this paper is neither given as already known(‘a priori’) nor deductively inferred but is embedded in an individual’s judgment of what is considered fair. From this perspective, the World Values Survey is used to empirically examine Koreans’ perceptions about fairness and preferences for redistribution. Of course, fairness norms covered in this study are restricted to income inequality and distributional justice. Distributional justice, in this context, refers to the criteria for fair distribution of physical resources such as income and wealth amongst the members of a society. In addition, preference of the norms of fairness and redistribution can differ on a person-to-person basis, and social beliefs have to be defined in such a way that can best represent and describe the preferences of each individual. Therefore, the distribution, not the average, is the core subject of analysis. Additionally, the study uses kernel density estimation as an appropriate method and scrutinizes some of the factors that determine people’s spectrum of fairness perceptions.