Productivity, tax, and social preferences

How to cite

Seres, Gyula. 2023. "Productivity, tax, and social preferences." AEA RCT Registry. AEARCTR-0011197. https://doi.org/10.1257/rct.11197-1.2

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

Gyula Seres*

* Primary investigator

Registry ID
AEARCTR-0011197
DOI
https://doi.org/10.1257/rct.11197-1.2
First registered
Last updated
Status
In development
Trial dates
2023-04-13 – 2023-12-31
Countries
Singapore ()
Registry record
http://www.socialscienceregistry.org/trials/11197

Abstract

Prior real-effort labor market experiments find that the tax rate does not always have a negative effect on effort. This is not consistent with expected payoff-maximizing behavior as taxation reduces the marginal utility of effort. We address the puzzle by coontroling for social preferences. Using a stylized model, we show that stronger social preferences increase the marginal benefit of effort. We test this in a randomized laboratory experiment in which the subjects are alotted in groups. They complete paid typing tasks which income is taxed and the tax is redistributed within their group. We randomize the tax rate and estimate its effect between people of different social preferences.

Intervention

Before the main part, the usbjects complete a stage eliciting their social preferences. Then, the main part starts. The subjects participate in a computerized repeated real-effort experiment with a mundain typing task. There are one unincentivized trial round and five incentivized rounds. Their performance is measured by the number of finished tasks in a 4-minute-long round. For this, they receive a flat rate payment of 0.20 SGD per task. That is, their gross payment in a round is equal to 0.20 SGD x tasks. From Round 2, they are subjected to a tax rate of X percent. The subtracted amount is redistributed evenly within their randomly matched fixed groups of 3 participants. The session concludes with a demographic questionnaire. The treatments differ in the tax rate X: (10, 50 or 70%) and the provision of feedback between rounds (feedback vs no feedback).

Experimental design

The computerized experiment is done using zTree. The social preference elicitation stage will use the protocol of Andreoni and Miller (2002). In the typing task, the subjects recieve a random sequence of 15 letters and digits. The task is to type it in reverse using their keyboard.

Design details

Randomization method
The treatment randomization is done at the session level. All other randomizations are done by computer.
Randomization unit
Treatment is randomized at the session level. Group matching and the task are done at the individual level.
Clusters
150 groups
Observations
450 subjects
Arms
100 subjects in the feedback treatments (10%, 50%, and 70%); 50 subjects in the no feedback treatments (10%, 50% and 70%)
Primary outcomes
effort (score), social preferences

Cited by

No citation data available for this registration yet.

Keywords