Inequality, Life Expectancy, and the Intragenerational Redistribution Puzzle

How to cite

Meemann, Christine; Traub, Stefan; Krieger, Tim. 2023. "Inequality, Life Expectancy, and the Intragenerational Redistribution Puzzle." AEA RCT Registry. AEARCTR-0011510. https://doi.org/10.1257/rct.11510-2.0

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

Christine Meemann* , Stefan Traub , Tim Krieger

* Primary investigator

Registry ID
AEARCTR-0011510
DOI
https://doi.org/10.1257/rct.11510-2.0
First registered
Last updated
Status
In development
Trial dates
2023-07-01 – 2024-06-30
Countries
Private
JEL codes
D630, D810, H550, I140
Registry record
http://www.socialscienceregistry.org/trials/11510

Abstract

In most OECD countries, pension reform policy has decreased the level of intragenerational redistribution over the last three decades, that is, redistribution among members of the same generation with high and low pension entitlements. This trend has occurred despite heterogeneity in life expectancy linked to socioeconomic status having a regressive impact on outcomes. This study contributes to solving this puzzle by means of a controlled laboratory experiment. We study the causal relationship between inequality of entitlements, mortality risk, and the size of redistribution in a stylized social security system.

Intervention

The aim of the laboratory experiment is to study the causal relationship between heterogeneity in life expectancy due to economic inequality and the amount of intragenerational redistribution in a stylized social security system. In the experiment, subjects are assigned the role of an involved social planner. In the framework of a hypothetical social security system, the social planner is asked to redistribute benefits among a group of persons who differ in their contribution-based entitlements. Depending on the treatment, the group members may also differ in their 'mortality risk', which is implemented in the experiment as the probability of default in terms of receiving a zero payoff instead of the benefit. The decision of the social planner is made from behind the veil of ignorance (VOI), that is, she learns her position in the group in terms of her own entitlement, mortality risk, and (expected) benefit only after having made her choice about the level of redistribution.

As a new feature compared to a former version of this experiment, positions in the group are still randomly assigned, but the social planner knows prior to her choice that she has to earn the entilement associated with her position ex-post. In the experiment, entitlements will be associated with a waiting time, whereby a higher waiting time is attached to a higher entitlement (as this results from a higher effort). Subjects are fully-informed and have to fulfill the waiting time of their position at the end of the experiment.

In a new version, entitlements will be associated with a real-effort task instead of a waiting time. Subjects have to position a certain number of sliders at the end of the experiment. The number of sliders corresponds to the entitlement of their position. Still, subjects are fully-informed that they have to complete the slider-task at the end of the experiment.

Experimental design

The redistribution task, which is repeated 4 times with a different parametrization, is to choose the preferred degree of redistribution of the initial social security contributions from among the five positions of a group. The five positions vary with respect to their contribution-based entitlements and associated waiting time. In the new version, waiting time will be replaced by a slider-taks. Depending on the treatment, the group members may also differ in their 'mortality risk', which is implemented in the experiment as the probability of default in terms of receiving a zero payoff instead of the benefit.

Design details

Randomization method
Randomization by a computer.
Randomization unit
Session level and group level randomization.
Clusters
4 treatments between-subjects
Observations
4 treatments between-subjects 2 sessions per treatment 30 participants per session 4 observations per participant = 240 observations per treatment = 960 observations in total New survey: 4 treatments between-subjects, 2 sessions per treatment, 25 participants per session, 4 observations per participant = 200 observations per treatment = 800 observations in total
Arms
30 participants per session * 2 sessions per treatment = 60 participants per treatment New survey: 25 participants per session * 2 sessions per treatment = 50 participants per treatment
Primary outcomes
Specifically, we investigate whether and how the social planner reacts to differentials in the distribution of mortality risk. The experiment involves four risk treatments (between-subjects): a control treatment without mortality risk; a treatment with equal mortality risk; and two treatments with skewed mortality risk, where income positions with lower entitlements exhibit a higher mortality risk. In one of the two skewed-mortality-risk treatments, redistribution results in a loss of efficiency. In the other treatment, subjects receive a compensation for the expected efficiency loss. Our key outcome variable is thus the preferred level of redistribution between treatments.
Secondary outcomes
Furthermore, we are interested if the preferred level of redistribution changes with the replacement ratio and the distribution of entitlements (within-subjects). We therefore repeat the decision task four times (rounds): With a high and low replacement ratio and with low and high variance.

Cited by

No citation data available for this registration yet.

Keywords