The Political Economy of Pension Reforms

Last registered on August 20, 2026

View Trial History

Pre-Trial

Trial Information

General Information

Title

The Political Economy of Pension Reforms

RCT ID

AEARCTR-0017544

Initial registration date

August 18, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

August 20, 2026, 9:30 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Name

Giulia Giupponi

Affiliation

Bocconi University

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Amalie Jensen

PI Affiliation

University of Copenhagen

Contact Investigator

PI Name

Vincent Pons

PI Affiliation

Harvard Business School

Contact Investigator

Additional Trial Information

Status

In development

Start date

2026-03-31

End date

2030-06-30

Keywords

Electoral, Welfare

Additional Keywords

social security, pension, pension reform, policy preferences, political preferences

JEL code(s)

H55, P00

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

We study what respondents in Denmark know and think about the functioning of the public pension system. Thanks to the match of survey responses to administrative data from Statistics Denmark, we can test whether respondents have accurate information about their own pension situation. We randomly provide a subset of respondents with information about their pension rights and features of the pension system, with a focus on fiscal sustainability, redistribution within and across generations, or pocketbook motives. We then study how information on these dimensions changes respondents’ views on pension policy and the role of government, and their stated political behavior.

In a parallel, non-experimental component, we exploit quasi-random variation in individuals' normal retirement age generated by Danish pension legislation to estimate the effect of actual policy exposure on the same outcomes, using a regression discontinuity design. This part of the analysis is pre-specified in the pre-analysis plan accompanying this registration.

External Link(s)

Registration Citation

Citation

Giupponi, Giulia, Amalie Jensen and Vincent Pons. 2026. "The Political Economy of Pension Reforms." AEA RCT Registry. August 20. https://doi.org/10.1257/rct.17544-1.0

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

The intervention consists of randomized information treatments embedded in an online survey administered to respondents in Denmark. The treatments provide factual information about the Danish public pension system, focusing on fiscal sustainability, redistribution across generations, redistribution within generations, or individual pension entitlements. The intervention examines how information changes respondents’ policy views and political preferences.

Intervention Start Date

2026-09-16

Intervention End Date

2026-10-11

Primary Outcomes

Primary Outcomes (end points)

Primary outcomes are survey measures collected immediately after treatment exposure.
We pre-specify three general primary outcomes: (i) satisfaction with the current Danish public pension system; (ii) support for the planned regular increases in the normal retirement age; and (iii) perceived overall fairness of the pension system.
We additionally pre-specify treatment-specific primary outcomes, each targeting the dimension a given treatment highlights: confidence in the state's ability to pay the respondent's own benefits, for the fiscal sustainability treatment; and, for the remaining treatments, how fairly the groups each treatment is about are perceived to be treated by the system. All respondents receive all of these questions.

Primary Outcomes (explanation)

Prior beliefs about the facts conveyed by each treatment are elicited from all respondents, including the control group, before treatment exposure. We construct measures of misperception by comparing these priors to the true values (as reported in the information treatments), and use the distance between prior and truth as a pre-specified moderator in heterogeneity analyses.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary outcomes are: (i) stated preferences over reforms to the retirement age, pension benefit levels and early retirement options, together with an index summarizing their fiscal-consolidation content; (ii) views of government (perceived efficiency, trust, preferred scope and satisfaction); (iii) democratic attitudes; (iv) stated political behavior and party preference.

Secondary Outcomes (explanation)

Experimental Design

The experiment is a survey-based randomized controlled trial. Individual respondents are randomly assigned with equal probability to one of five information treatments or a control group after baseline questions and before outcome measurement. Randomization is implemented by the survey software and is not stratified.

Experimental Design Details

Not available

Randomization Method

Randomization done in office by a computer

Randomization Unit

The randomization is done at the level of the individual respondent. There is a unique level of randomization.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Planned sample size of 7,200 individuals. We will send out invitations to 90,000 respondents hoping for an 8% response rate.

Sample size: planned number of observations

7,200 individuals

Sample size (or number of clusters) by treatment arms

1,200 individuals assigned to each of the five treatment arms, and 1,200 individuals assigned to the control group

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

We compute minimum detectable effects (MDEs) for a pairwise comparison of a single treatment arm against the control group, with two-sided significance of 0.05, power of 0.80, and equal arm sizes. Treatment is assigned at the individual level and is not clustered, so no design effect applies. Under our baseline assumption of an 8% response rate (1,200 responses per arm, 2,400 observations per pairwise comparison), the MDE is 0.114 standard deviations. It is 0.145 standard deviations under a pessimistic 5% response rate (750 per arm) and 0.102 standard deviations under an optimistic 10% response rate (1,500 per arm). Even in the pessimistic case the MDE falls below the 0.15 standard deviation benchmark suggested for information-provision experiments (Haaland, Roth and Wohlfart, 2023). Comparisons that pool several treatment arms will be better powered than single-arm comparisons.

Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

IRB