Impact Evaluation of Providing Assistance to Public Officials Who Serve Rental Subsidy Beneficiaries on Subsidy Take-Up and Staff Performance

Last registered on September 28, 2026

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0019788

Initial registration date

September 21, 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

September 28, 2026, 8:51 AM EDT

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

Last updated

September 28, 2026, 9:46 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

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

Request Information

Primary Investigator

Name

Pablo Celhay

Affiliation

Pontificia Universidad Catolica de Chile

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Javiera Selman

PI Affiliation

Universidad de Chile

Contact Investigator

Additional Trial Information

Status

In development

Start date

2026-01-01

End date

2027-01-31

Keywords

Behavior, Governance, Welfare

Additional Keywords

Bureaucratic performance; Public administration; Field experiment; Performance information; Task allocation; Housing vouchers; Neighborhood quality.

JEL code(s)

C93; D73; I38

Secondary IDs

Prior work

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

Abstract

This document presents the pre-analysis plan for the research project “Impact Evaluation of an Online Counseling Program Providing Neighborhood Information to Rental Voucher Recipients in Chile.” The Chilean Ministry of Housing and Urbanism (MINVU) introduced the first rental voucher program in Latin America in 2013. Yet take-up remains limited: each year, approximately six out of ten awarded vouchers are not used, and many households that do use their vouchers relocate to neighborhoods with relatively poor characteristics.

This project evaluates strategies to improve the performance of public agents who assist beneficiaries of the Chilean Rental Voucher Program. Public workers are randomly assigned to a control group or one of five treatment groups. All groups use Aprendo y Arriendo (AyA), an online platform that provides standardized counseling tools and neighborhood-level information to support interactions with beneficiaries. The control group receives basic training on the platform and standard lists of beneficiaries.

The treatments vary along two experimental dimensions. First, agents may receive performance feedback in addition to the basic platform training. Second, beneficiary lists may be prioritized according to the expected cost of successfully completing a rental placement, distinguishing between relatively high- and low-cost cases. Some treatment groups receive performance feedback only, others receive prioritized beneficiary lists only, and two groups receive both interventions. This design allows us to separately and jointly estimate the effects of performance feedback and case prioritization on agent behavior and performance. We study whether these interventions increase voucher utilization, improve the efficiency of agents’ efforts, and affect the characteristics of the neighborhoods into which beneficiary households relocate.

External Link(s)

Registration Citation

Citation

Celhay, Pablo and Javiera Selman. 2026. "Impact Evaluation of Providing Assistance to Public Officials Who Serve Rental Subsidy Beneficiaries on Subsidy Take-Up and Staff Performance." AEA RCT Registry. September 28. https://doi.org/10.1257/rct.19788-1.1

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

The intervention is implemented through Aprendo y Arriendo (AyA), an online platform used by public agents who support beneficiaries of Chile’s Rental Voucher Program. The platform provides standardized counseling tools, beneficiary information, and neighborhood-level information to assist agents in helping households use their vouchers and search for rental housing.

All study groups receive access to the AyA platform and basic training on its use. The experimental interventions vary along two dimensions. First, some agents receive performance feedback based on metrics related to their work with beneficiaries. Second, some agents receive prioritized beneficiary lists that rank cases according to their expected difficulty or cost of achieving a successful rental placement. The prioritization can emphasize either relatively high-cost or relatively low-cost cases.

The five treatment arms combine these interventions in different ways: performance feedback only, high-cost prioritization only, low-cost prioritization only, performance feedback plus high-cost prioritization, and performance feedback plus low-cost prioritization. The control group receives basic platform training and standard, non-prioritized beneficiary lists without performance feedback.

Intervention Start Date

2026-05-01

Intervention End Date

2026-12-31

Primary Outcomes

Primary Outcomes (end points)

Voucher utilization: whether the beneficiary's rental contract is activated
within the exposure window. Activation is the final administrative step and
the point at which the subsidy is actually used. Source: MINVU contract
administrative records.

Contract creation: beneficiary has at least one rental contract created in
the system.
Contract validation: the beneficiary's contract went through the validation process. Note: contracts declared invalid count as validated, because the indicator measures whether the process was completed, not whether its
outcome was positive.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Public agents and rental voucher beneficiaries are randomly assigned to experimental groups. The design varies two dimensions: whether agencies receive performance information in addition to basic training, and whether beneficiaries are assigned to agent lists prioritized toward relatively high-cost or low-cost cases. This generates five treatment groups and one control group. The design allows us to estimate the separate and joint effects of performance feedback and case prioritization on agent performance and voucher utilization.

Experimental Design Details

Not available

Randomization Method

Randomization will be conducted by computer. Assignment will be stratified using pre-specified characteristics of public agents, agencies, and beneficiaries.

Randomization Unit

Randomization occurs at multiple levels. Agencies/offices are randomized to receive basic training or training with performance metrics. Beneficiaries are randomized to control, high-cost-first, or low-cost-first list assignments. Public agents are also randomized within the experimental design.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

51 SERVIU local housing offices (agencies)

Sample size: planned number of observations

16,252 voucher holders (beneficiaries), of whom 13,849 are eligible for the beneficiary-level arm; 466 public officials across the 51 offices

Sample size (or number of clusters) by treatment arms

Office-level arm (51 offices, 16,252 beneficiaries):
26 offices treatment (8,424 beneficiaries) — platform with performance
metrics and interactive case management
25 offices control (7,828 beneficiaries) — platform with standardized
program information and AI chatbot only, plus a static case list
Beneficiary-level arm (13,849 eligible):
6,927 treatment — access to beneficiary information site
6,922 control — no access
2,403 beneficiaries ineligible for this arm (retain office assignment)

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

Office-level arm. Main outcome: share of assigned beneficiaries with a rental contract created (percentage points, office level, 51 offices). Baseline mean 51.8 pp, SD across offices 10.2 pp. Estimated with stratum fixed effects and baseline adjustment; observed standard error 0.89 pp. MDE (80% power, 5% two-sided) = 2.5 pp = 0.23 SD of the cross-office outcome level Beneficiary-level arm. Individual randomization, no clustering, 6,927 vs 6,922. For a binary outcome with baseline rate near 0.5, MDE (80% power, 5%) = 2.4 pp = 0.05 SD. Calculation ignores precision gains from stratification and baseline covariates, so it is conservative.

Supporting Documents and Materials

IRB

Institutional Review Boards (IRBs)

IRB Name

Facultad de Economía y Negocios Universidad de Chile

IRB Approval Date

2024-07-10

IRB Approval Number

Acta 13 - 2024

Analysis Plan

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

Request Information