Perceptions of Relative Income Positions and Preferences for Redistribution
How to cite
Fukuda, Rui; Ando, Rintaro. 2025. "Perceptions of Relative Income Positions and Preferences for Redistribution." AEA RCT Registry. AEARCTR-0017134. https://doi.org/10.1257/rct.17134-1.2
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
As demographic ageing intensifies fiscal pressures, understanding public preferences for redistribution is vital to sustaining social security systems and anticipating inequality—both in Japan and beyond.
This study investigates how individuals’ awareness of their relative income position influences their views on fairness, inequality, and redistribution. To address this question, we implement a randomised information intervention survey in collaboration with a Japanese municipality. The survey is linked to administrative tax records, allowing respondents to be informed of their actual income rank within local reference groups. Treated individuals receive information on their rank within demographic cohorts. This design builds on and extends the findings of Hvidberg et al. (2023), shifting the analytical focus from fairness perceptions to redistribution preferences.
The study has three main objectives. First, it examines the causal impact of positional information on attitudes towards redistribution, contributing to a re-evaluation of the Prospect of Upward Mobility (POUM) hypothesis by incorporating the role of misperceptions. Second, it documents the accuracy of income-related beliefs in Japan, including perceptions of (i) median income within reference groups, (ii) one’s own position within those groups, and (iii) expectations about future income. Third, a follow-up survey conducted several months later assesses the persistence of the intervention’s effects over time. Future income expectations are validated by linking survey responses to tax records in the subsequent year.
For details on the survey design, please refer to the Supporting Documents and Materials, available via the corresponding button. These include the Survey Design and Methodology document as well as participant-facing survey materials.
Intervention
We conduct an online survey targeting residents of a Japanese municipality. As part of the survey, participants are randomly assigned to either a treatment or control group. The treatment group receives personalised information about their income rank within local reference groups, based upon their self-reported income. These ranks are calculated using administrative tax records. The control group does not receive any rank-related feedback.
Experimental design
We conduct an online survey targeting residents of a Japanese municipality to examine income-related beliefs and preferences for redistribution. The survey embeds a randomised information intervention and is designed to document both causal effects and baseline misperceptions.
The target population comprises approximately 5,000 residents who (i) possess a 2025 municipal tax record that includes income data for the year 2024, (ii) were born between 1976 and 1980, and (iii) hold Japanese nationality. Each individual receives a personalised invitation letter containing a unique QR code, which links to a web-based questionnaire. Responses are collected exclusively online via a self-administered format.
Respondents are randomly assigned to either a control or treatment group. The treatment group receives personalised feedback on their income rank within the municipal income distribution, based on self-reported income and administrative tax records. Randomisation is implemented automatically via the survey platform.
The first wave of the survey begins with the dispatch of invitation letters on 30 October 2025. Respondents are initially asked to complete the questionnaire by 21 November. A reminder postcard is sent to non-respondents, and the final deadline for participation is 10 December 2025. A follow-up survey will be conducted approximately six to twelve months after the initial wave to assess the persistence of treatment effects. This second wave will target only those who completed the first survey.
For details on the randomisation procedure, survey instruments, and implementation protocols, please refer to the Survey Design document.
Design details
Cited by
No citation data available for this registration yet.