First-hand experience of short-term incarceration

Last registered on March 28, 2022


Trial Information

General Information

First-hand experience of short-term incarceration
Initial registration date
March 26, 2022

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
March 28, 2022, 7:15 AM EDT

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


Primary Investigator

University of Zurich

Other Primary Investigator(s)

PI Affiliation
ETH Zurich, KOF Swiss Economic Institute
PI Affiliation
ETH Zurich, KOF Swiss Economic Institute
PI Affiliation
University of Zurich, Department of Economics

Additional Trial Information

In development
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
We conduct a randomized controlled trial with the general population and public sector employees to study the effect of first-hand experience of short-term incarceration on perceptions of pains of imprisonment and attitudes towards law and order. In collaboration with the Department of Justice and Home Affairs Zurich, we randomly invite interested citizens to take part in a test run of a newly built prison in Switzerland and measure their attitudes towards punishment.
External Link(s)

Registration Citation

Arman, Arto et al. 2022. "First-hand experience of short-term incarceration." AEA RCT Registry. March 28.
Sponsors & Partners

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

Request Information
Experimental Details


Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Perception of pains of imprisonment: We will compute an index by extracting the first principal component using Principal Component Analysis (PCA) from (1) an incentivized guessing task regarding the average well-being of inmates in a prison in the canton of Zurich and (2) a survey instrument capturing people’s beliefs about their own well-being during imprisonment. In the guessing task, individuals will guess the average answer inmates gave when asked about their well-being in prison on a scale from zero to ten, and are paid for accurate guesses.

Preference for law and order: We will compute an index by extracting the first principal component using PCA from (1) a donation task and (2) a survey question capturing people’s attitudes towards law and order. In the donation task, individuals will decide whether and how much the experimenter should donate to an organization that supports a tougher penal system or to an organization that supports a more moderate penal system (participants do not know the actual identities of the organizations).
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Furthermore, we will explore the effects of the treatment on questionnaire measures of trust in institutions and perceptions of procedural fairness.
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
To participate in a test run conducted in the jail cells of the prison “Gefängnis Zürich West” (GZW), adult individuals with a clean criminal record living in the canton Zurich could applied and indicated their preferred dates to participate in the test run and if they would like to stay overnight. Those applicants were screened regarding the conditions of admission by the members of GZW and then an anonymized list of the eligible participants was handed to the research team.

We invited all participants in that list to an online baseline survey, where we measured participants’ beliefs about – and preferences regarding – the justice system and correctional institutions. We will use these pre-treatment outcomes as control variables in our analysis to increase the precision of our estimated treatment effects. We also collected a variety of other questionnaire measures such as their risk attitudes, criminal attitudes, educational level, and previous jail experience.

After the invitation to the baseline survey, we randomized a share of the applicants to be eligible to participate in the test run (treatment group), while the other participants were not (control group). The head of the GZW then filled the cells for their test run based on the list of participants that were randomly assigned into the treatment group, not exhausting the pool of individuals in the treatment group (for more details, see experimental design details).

Subsequently, the test run will be conducted by the Department of Justice and Home Affairs, starting at noon on the 24th of March and ending at noon on the 27th of March. Participants in the treatment group, that were assigned to participate by GZW and show up for their appointment, will spend a pre-defined amount of time in their assigned cell.

In the second survey (main survey), for which we will invite participants in the afternoon of the 27th of March, we will measure the participants’ perceptions of pain of imprisonment and attitudes towards the justice system. Furthermore, we will elicit their beliefs about the well-being of inmates in a prison in Zurich through an incentivized guessing task and measure their attitude towards law and order through a donation task.
Experimental Design Details
Applicants were requested to specify their preferences regarding weekday and duration of their stay during the application process. Some places in the test run were reserved in advance for political figures, members of the court and members of the media. For the remaining available cells we randomly selected 185 individuals to be eligible to participate in the test run (treatment group) and allocated the rest to the control group. Members of the control group were not allowed to participate in the test run. The number of people allocated to the treatment group was larger than the actual test run capacity such that the Department of Justice and Home Affairs had more flexibility in allocating participants given the participant’s preferences for dates and duration. From this group, the administrative staff at GZW selected individuals according to the applicants’ participation preferences and their characteristics (gender, smokers, weekday and duration) as well as the available cells and invited them to participate in the test run. Due to this process, some people in the treatment group cannot participate and some more might not show up on their assigned day. Thus, we will need to factor in non-compliance in our analysis.
Randomization Method
We implemented a stratified randomization with two strata based on whether the applicants are employed by the canton of Zurich or not. This was done due to the assumption that applicants working for the canton of Zurich were significantly different from other participants in some of their attitudes and beliefs regarding the correctional settings.
Randomization Unit
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
0. Participants were randomized into treatment and control at the individual level. However, as some participants will share cells for some amount of time, we will take into account dependencies at the cell level in the statistical analysis.
Sample size: planned number of observations
After announcing the test run in the media, individuals from the general population, political figures and members of the court applied for a place in the test run in an online application process. Members of the media had to apply through a separate channel. For the baseline survey, we invited the 802 applicants that applied through the standard online application via individualized survey links. Of these 802 invitations roughly fifty were sent to the participating political figures and members of the courts, which are excluded from the randomization and our main analysis. 726 participants successfully participated in the first survey wave (response rate of around 90 percent). We were informed later on that four members of the media had applied a second time through the standard online application. Therefore, we had invited them for the baseline survey and included them in the randomization. As the Department of Justice and Home Affair forbid members of the media to participate in the survey, we excluded them from the invitation to the main survey and the analysis. Therefore, we will invite all remaining 798 participants for the second survey wave, expecting a lower response rate (e.g., 80 percent). Even though we exclude the political figures and members of the court from the main analysis, we might use this non-experimental sample for explorative purposes using before-after comparison. We will run further exploratory analysis, to evaluate differences in the treatment effects for employees of the canton of Zurich and for individuals that had experienced a prison from the inside prior to our experiment (e.g., visited an inmate).
Sample size (or number of clusters) by treatment arms
We expect 185 participants in the treatment group and 560 participants in the control group to participate.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
Human Subjects Committee of the Faculty of Economics, Business Administration, and Information Technology
IRB Approval Date
IRB Approval Number
OEC IRB #2022-013


Post Trial Information

Study Withdrawal

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

Request Information


Is the intervention completed?
Data Collection Complete
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials