Balancing Progressivity and Efficiency: A Survey Experiment to elicit Preferences

How to cite

Sorribas-Navarro, Pilar; Arenas, Andreu; Foremny, Dirk. 2024. "Balancing Progressivity and Efficiency: A Survey Experiment to elicit Preferences." AEA RCT Registry. AEARCTR-0013839. https://doi.org/10.1257/rct.13839-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Pilar Sorribas-Navarro* , Andreu Arenas , Dirk Foremny

* Primary investigator

Registry ID
AEARCTR-0013839
DOI
https://doi.org/10.1257/rct.13839-1.0
First registered
Last updated
Status
In development
Trial dates
2024-09-16 – 2024-09-21
Countries
Brazil (), Spain (), United States of America ()
JEL codes
H11; H21; H24; H31; H50
Registry record
http://www.socialscienceregistry.org/trials/13839

Abstract

This study aims to estimate individuals' preferences regarding the progressivity and efficiency of the Personal Income Tax (PIT). Variations in PIT designs result in differing impacts on redistribution and efficiency. While more progressive taxes achieve higher redistribution, their effects on tax revenues and economic growth remain inconclusive. To address this, we employ an experimental conjoint survey design to elicit individual preferences for regarding this trade-off.

In this experiment, participants will be presented with pairs of proposals for the design of the personal income tax and will choose their preferred option. Each proposal will provide details on different dimensions of the proposal: progressivity (marginal tax rates across six income brackets), tax revenues generated, the budgetary implications, and impact on economic growth.

This research is conducted across three countries—Spain, the USA, and Brazil—each with distinct levels of inequality, development, and importance of the welfare state.

Experimental design

This design is a conjoint analysis of preferences for tax policy. Alternatives of policy proposals are randomly chosen in pairs to participants, who rate them, and choose the most preferred one.

Design details

Randomization method
The data collection is done by Netquest, a market-research company with ample evidence in conjoint designs. They will randomize the conjoint profiles shown to the survey participants.
Randomization unit
Randomization is at the individual level so that alternatives are orthogonal to individual characteristics.
Clusters
2000 individuals per country (Brazil, Spain, US)
Observations
2000 x 8 x 2 = 32000 for each country.
Arms
Each alternative is randomly chosen over all possible combinations of the attributes with only one constraint: the exclusion of regressive income tax alternatives. The change in fiscal revenues approximately corresponds to the change in marginal tax rates.
Primary outcomes
1-Marginal Means 2-AMCEs 3-WTPs 4-Weights of the attributes And variations of these outcomes, as explained in the pre-analysis plan.

Cited by

No citation data available for this registration yet.

Keywords