Perception of Income Inequality: A Multidimensional Scaling Study

Existing research shows that popular income inequality measures fail to reflect their respondents’ perceptions of income inequality. However, most of the current literature focuses on what is negative, telling us what individuals do not perceive. This paper presents an alternative methodology to help uncover actual perceptions of inequality, how people perceive inequality instead of how they don’t. Multidimensional scaling, a statistical tool for visualizing dissimilarity data as a low-dimensional map, is used on results of a simple grouping task with a given distribution set. The outcome is a perception map that presents respondents’ answers spatially, which enables additional insight into respondents’ thinking. The map created by the respondents’ replies, presented in this paper, indicates that their decisions are driven by two factors: what the biggest gap in incomes of a given distribution is and whether some groups have equal incomes. The result additionally validates multidimensional scaling as a tool for measuring income inequality perception and opens new ways of improving inequality perception questionnaires.