Casual Income Inequality: The Reference Group Model

One prominent explanation for the discrepancy between the perceived and actual income inequality is the reference group theory (Hyman, 1960). Reference groups usually refer to relatively small groups such as family, neighborhoods, or regions where the individuals live. Individuals will usually compare themselves to others when evaluating their own positions. If they do not properly take into account the selection process and fail to apply the Bayes’ rule to deduce the income distribution for the entire population from sub-sample, their conclusions can result in biased perceptions. The task to infer accurate assessment of income distribution is rather complex. Those individuals that have limited abilities to form expectations and properly process
However, after discovering their true position in the income distribution, a self-interest may motivate agents to change attitudes in favour of more redistributive policies. Owing to the unique testing procedure, Cruces, Perez-Truglia, and Tetaz (2013) were able to test this hypothesis. Contradictions between the subjective assessment of agent’s position and the actual position were pointed out to a randomly assigned treatment group. Confronting the agents with an accurate information had a significant effect on their stated attitudes towards redistribution. Those agents, who underestimated their position on the scale of income distribution, did not change their preferences for redistribution. On the other hand, those who underestimated their relative position and received accurate information demanded more redistribution than those in the control
The geographic reference group bias refers to a generalization of reference group as if it represents the whole country. Researchers used the respondents’ area of residence as a geographical proxy for reference groups. The fully naïve agent will report his position within the reference group as if it was his position in the overall distribution of income. For “sophisticated” agent with an ability to form rational expectations and to apply Bayes’ Rule, relative income within a reference group should not have a further effect on the perceived position on a national scale of income distribution after controlling for other explanatory variables. Cruces, Perez-Truglia, and Tetaz (2013) showed that survey respondents’ income rank within their locality shaped their perception of inequality at the country