Beliefs in Merit, Trickle-Down Economics, and Preferences for Redistribution: An Experiment with the Top and Bottom 20%

Authors – Roberto Brunetti , Gianluca Grimalda , Maria Marino

Abstract:

We conduct an online survey-experiment in the US to study how people’s beliefs about individual merit and trickle-down economics influence their preferences for redistribution. Participants (N = 2646) from the top and bottom quintiles of the income distribution, referred to as spectators, decide how much money to redistribute between two people, referred to as stakeholders. The first stakeholder, named Person 1, is an entrepreneur earning more than $100,000 in real life and is assigned $50 within the experiment for a task he/she has performed for us. The second stakeholder, named Person 2, earns less than $10,000 in real life, does not perform any task for us and is assigned only $1.

The spectators are informed that Person 1 earned the initial endowment by carrying out a task for us while Person 2 carried out no task. Before choosing how much to redistribute from Person 1 to Person 2, spectators read their profiles. We manipulate several real-life characteristics of the two stakeholders using a conjoint table analysis. First, we manipulate their effort expressed in hours worked per day. Second, we manipulate Person 1’s talent by describing him/her as either having founded the firm he owns or having inherited it. Third, in order to test the different mechanisms underlying the trickle-down hypothesis, in three different treatments we manipulate, i) the amount of money donated to charity in the last year by Person 1 (Philanthropic channel), ii) the number of employees working in Person 1’s firm (Employment channel), and (iii) the number of patents obtained by Person 1’s firm (Innovation channel). Stakeholders’ characteristics are combined randomly for a total of 16 conjoint tables administered in a within-subject design.

We find that the effort of Person 2, the low-income stakeholder, has by far the greatest impact on spectators’ redistributive choices: Spectators redistribute significantly more when Person 2 is described as a hard worker. Furthermore, the effort and talent of Person 1, the high-income stakeholder, impacts negatively the redistributive choices, but in all cases to a lesser degree than the effort of the low-income stakeholder. Finally, we find that the trickle-down attribute has an impact on redistributive choices only when it is framed in terms of donation to charity and it impacts only the redistributive choice of low-income spectators.