Back to History Current Version

Contractual features and the appointment of high-level bureaucrats

Last registered on January 22, 2026

Pre-Trial

Trial Information

General Information

Title
Contractual features and the appointment of high-level bureaucrats
RCT ID
AEARCTR-0017586
Initial registration date
January 19, 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
January 22, 2026, 1:57 PM EST

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

Affiliation
Institute for Fiscal Studies

Other Primary Investigator(s)

PI Affiliation
University of British Columbia
PI Affiliation
Institute for Fiscal Studies

Additional Trial Information

Status
On going
Start date
2025-11-26
End date
2026-08-31
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This study investigates preferences in the appointment of high-level bureaucrats in local authorities--who are arguably more consequential for public service delivery than their counterparts in central and federal governments, the recent focus of most studies on bureaucrats. We study the contractual process from both the demand side –i.e., mayors’ screening– and the supply side –i.e., bureaucrats’ selection into the role. We survey mayors and appointed high-level bureaucrats across municipalities in Peru. In this survey, we introduce discrete choice experiments to elicit average preferences underlying mayors’ appointments of high-level managers within their office, as well as bureaucrats’ valuation over alternative contract arrangements. From the bureaucrats’ side, we elicit their valuation over alternative contract arrangements that a hypothetical individual, similar to the respondent, is facing for a similar position held by the respondent in a neighbouring municipality. Looking at wage differentials, we can flexibly back out a willingness to pay distribution for different contractual features. From the mayor’s side, we estimate mayors’ preferences over characteristics of bureaucrats to be selected into office. Mayors are presented with hypothetical profiles of potential job candidates for a role in their office. Using their stated preference for profiles, we can estimate how mayors trade off various characteristics of candidates. The aim is to shed light on optimal contracts for appointed bureaucrats.
External Link(s)

Registration Citation

Citation
Bancalari, Antonella, Claudio Ferraz and Julia Loh. 2026. "Contractual features and the appointment of high-level bureaucrats." AEA RCT Registry. January 22. https://doi.org/10.1257/rct.17586-1.0
Sponsors & Partners

Sponsors

Partner

Experimental Details

Interventions

Intervention(s)
See experimental design.
Intervention Start Date
2025-11-26
Intervention End Date
2026-08-31

Primary Outcomes

Primary Outcomes (end points)
1. Bureaucrats: Willingness-to-pay for non-wage contractual attributes.
2. Mayors: Choices about alternative work arrangements.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
Our study in based in Peru, where we leverage the opportunity to collaborate with the Ministry of Economy and Finance who introduced a program-- named "Punche Gerentes"-- aimed at attracting and retaining high-quality senior bureaucrats. Our experiment is embedded in a nationwide survey of mayors and bureaucrats targeting municipalities in Peru. The online survey will be collected from late November 2025 onwards--the month when the evaluation for contract renewals start. Although the Ministry will circulate the link to the online survey, the email and information sheet will clearly indicate that this survey is conducted for research purposes, specifying the affiliation of the researchers, and highlighting that no personal data will be shared with the Ministry. Survey responses are not personally identifiable by the State. The participants are the municipal mayors, as well as the appointed managing directors of the four offices charged with public investment within a municipality: (1) Logistics, (2) Administration, (3) Budgeting, and (4) Infrastructure.

The discrete choice experiment is randomly allocated across participants using the Bayesian Adaptive Choice Experiments (BACE) developed by Drake et al. (2025). This method allows us to elicit preferences quickly and efficiently using a dynamic experimental framework. BACE generates an adaptive sequence of menus from which subjects will make choices, with each menu optimally chosen using the information provided by the subjects’ previous choices. While we use a Bayesian approach to randomly generate menus, we will also estimate the willingness to pay–or willingness to hire–associated with each attribute following frequentist inference.

Bureaucrats experiment
Participants are shown two descriptions of contractual arrangements for a hypothetical bureaucrat similar to them (age, marital status, education, public sector experience) and for a similar position to theirs in a neighboring municipality. The two contracts differ in their features and their monthly wage. The features and wages are assigned to participants at random. Each participant faces different contract choices throughout 10 rounds. The description of the hypothetical bureaucrat remains the same across rounds. We tell participants that the type of job is the same and ask them which contract they would choose if both were available. The options are:
1. Probability that the mayor will terminate the contract before its end date, regardless of performance: {5%, 50%}
2. Contract renewal subject to public investment goals: {Not monitored with goals, set and monitored by mayor, set and monitored by SERVIR-MEF}.
3. Corruption: {The mayor is being investigated for corruption, The mayor has never been investigated for corruption}.
4. Salary: {−10%(10%)150%} from reported current salary.

Mayors experiment
Mayors are shown two different profiles of hypothetical job candidates. The candidate profiles differ in their characteristics. Each participant faces different contract choices throughout 10 rounds. The description of the hypothetical mayor remains the same across rounds. Participants are asked to make a recommendation on which of the two candidates to select as a generic technical bureaucrat for a hypothetical mayor similar to themselves. The dimensions we will randomly vary are:
1. Networks: {Without recommendation from mayor’s family or friends - applied to job ad, Recommended by mayor’s family or friends}
2. Helped mayor in the campaign of the last elections: {Helped mayor in last elections, Did not help mayor in last elections}
3. Highest educational attainment: {bachelors, postgrad}
4. Managerial experience: {3 , 9} years.

We anticipate that some applicants will be inattentive and not read the position descriptions carefully. To measure inattention, we will compare the rate of people choosing the lower paid job when the same descriptions are posted but with different wages. Furthermore, towards the end, we introduce an additional round where we ask participants to select the dominated option as an attention check.

Survey: we will collect detailed data that will help us characterize each respondent, including socio-demographics, education, experience, ability through an IQ test and personality traits including the Big 5, public sector motivation, integrity, political ideology and economic preferences (i.e., risk appetite and altruism). We will also collect data on their current and last contract arrangement, including salary and contract features, and their satisfaction with these features.
Experimental Design Details
Not available
Randomization Method
Randomization done by a computer on a remote server.
Randomization Unit
Individual
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
1. Bureaucrats: approx. 1,000
2. Mayors: approx. 200
Sample size: planned number of observations
1. Bureaucrats: approx. 1,000 by 10 choices 2. Mayors: approx. 200 by 10 choices
Sample size (or number of clusters) by treatment arms
Not applicable. Randomization is at the attribute-level within respondents via BACE. Effective observations are respondent by profile.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Not applicable. Our plan is to estimate willingness-to-pay magnitudes for non-wage work attributes for bureaucrats and preference parameters for mayors.
IRB

Institutional Review Boards (IRBs)

IRB Name
Behavioural Research Ethics Board, Office of Research Ethics, The University of British Columbia
IRB Approval Date
2025-09-11
IRB Approval Number
H25-02443