Offline Reinforcement of Household Financial Planning among Displaced Persons in Armenia: A Three-Arm Cluster Randomized Trial

Last registered on October 07, 2026

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0019887

Initial registration date

October 06, 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

October 07, 2026, 11:12 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

Lesman Ghazaryan

Affiliation

DeVinci Higher Education

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Syuzanna Martirosyan

PI Affiliation

Grenoble Ecole de Management

Contact Investigator

PI Name

Jonathan Karver

PI Affiliation

World Bank Group

Contact Investigator

Additional Trial Information

Status

On going

Start date

2026-10-05

End date

2026-12-28

Keywords

Behavior, Finance & Microfinance, Post-conflict, Welfare

Additional Keywords

household budgeting, financial planning, financial capability, displaced persons, refugees, behavioral intervention, social and behavior change communication, financial resilience, cluster randomized trial, Armenia

JEL code(s)

C93, D14, D91, O12

Secondary IDs

Prior work

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

Abstract

This study evaluates whether offline behavioral interventions can improve household budgeting and financial planning among people displaced from Nagorno-Karabakh and currently living in Armenia. The evaluation uses a three-arm cluster randomized controlled trial across 18 communities in Shirak, Lori, Tavush and Gegharkunik.
All study communities receive a social media campaign promoting household budgeting and forward financial planning. Communities are then randomly assigned to one of three conditions: social media only; social media plus printed materials and billboards; or social media plus community training sessions. The design therefore estimates the additional effect of each offline component relative to social media alone.
The primary outcome is a financial planning behavior index measuring concrete household practices, including recording income and expenses, allocating money to essential spending before other expenditures, saving toward a specific purpose, and discussing household spending plans before money is spent. Baseline and endline surveys are conducted with the same respondents.
The study also examines mechanisms and secondary outcomes including planning horizon, financial self-efficacy, awareness and use of financial services, income predictability, and exposure to the intervention. The primary analysis will estimate intention-to-treat effects using the randomized community assignment.

External Link(s)

Registration Citation

Citation

Ghazaryan, Lesman, Jonathan Karver and Syuzanna Martirosyan. 2026. "Offline Reinforcement of Household Financial Planning among Displaced Persons in Armenia: A Three-Arm Cluster Randomized Trial." AEA RCT Registry. October 07. https://doi.org/10.1257/rct.19887-1.0

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)

All study communities receive a social media campaign promoting household budgeting and forward financial planning. Communities are randomly assigned to one of three conditions: (1) social media only; (2) social media plus printed materials and billboards; or (3) social media plus community training sessions. The evaluation estimates the incremental effect of each offline component relative to social media alone.

Intervention Start Date

2026-10-20

Intervention End Date

2026-12-20

Primary Outcomes

Primary Outcomes (end points)

Financial planning behaviour index at endline.

Primary Outcomes (explanation)

The primary outcome is a continuous financial planning behaviour index ranging from 0 to 4. Respondents receive one point for each of four behaviours reported during the previous 30 days: (1) recording household income and expenses; (2) setting aside money required for essential expenditures before other spending; (3) saving money toward a specific purpose; and (4) discussing with another household member how money would be spent before spending it. The same items are measured at baseline and endline. The endline index is the primary outcome, with the baseline index included as a covariate in the primary ANCOVA specification.

Secondary Outcomes

Secondary Outcomes (end points)

Planning horizon; financial self-efficacy; awareness and use of financial products and services; goal-directed saving; current financial goal; perceived prevalence of financial planning within the community.

Secondary Outcomes (explanation)

Secondary outcomes capture financial-planning mechanisms and related behaviours. These include the respondent's financial-planning horizon, confidence that their actions can improve the household's financial situation, awareness and recent use of banking or financial services, whether the household is currently working toward a specific financial goal, and perceptions of financial-planning behaviour among peers. These outcomes will be analysed separately from the primary financial-planning behaviour index and will not replace the pre-specified primary outcome.

Experimental Design

This study is a three-arm cluster randomized controlled trial conducted in 18 communities in Armenia. The community is the unit of randomization and the individual displaced person is the unit of analysis. Six communities are assigned to each arm: social media only, social media plus printed materials and billboards, or social media plus community training.
Baseline data are collected before randomization and before any intervention exposure. Randomization is conducted after baseline data collection. The same respondents are followed at endline after the intervention period. The primary analysis compares each treatment arm with the control arm using intention-to-treat estimates.

Experimental Design Details

Not available

Randomization Method

Randomization conducted in office by computer after completion of baseline data collection, using stratified/constrained randomization based on marz and pre-specified community characteristics.

Randomization Unit

Community (cluster); individuals within communities are surveyed but are not individually randomized.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

18 communities

Sample size: planned number of observations

1,078 individuals at baseline; approximately 856 individuals expected at endline

Sample size (or number of clusters) by treatment arms

Control: 6 communities; Treatment 1: 6 communities; Treatment 2: 6 communities. Approximately 360 baseline respondents and approximately 285 endline respondents per arm on average. Exact individual sample sizes vary slightly because three communities have smaller eligible populations; constrained randomization assigns one smaller community to each arm.

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

Power calculations assume 18 communities allocated equally across three arms, with 6 communities per arm, approximately 50 completed endline observations per standard community, 80% power, and a 5% two-sided significance level. Because the intra-community correlation and baseline outcome prevalence are not yet known, sensitivity analyses are reported across ICC values of 0.05, 0.10, 0.15 and 0.20 and baseline binary-outcome prevalences of 10%, 15%, 20% and 25%. Under the central planning scenario of 15% baseline prevalence and ICC = 0.10, the minimum detectable effect for a pairwise treatment-control comparison on a binary budgeting outcome is approximately 21.7 percentage points at the 5% two-sided level. Using a conservative 2.5% two-sided threshold for the two planned treatment-control comparisons, the corresponding MDE is approximately 24.6 percentage points. For the continuous financial-planning behaviour index, under ICC = 0.10, the pairwise MDE is approximately 0.61 standard deviations without baseline precision gains. If baseline outcomes and pre-specified covariates explain 25% of endline variation, the MDE falls to approximately 0.53 standard deviations. These are analytical planning calculations. Final power calculations will be updated using observed baseline prevalence, observed outcome variance, observed intra-community correlation, achieved sample sizes and the executed randomization.

Supporting Documents and Materials

IRB

Institutional Review Boards (IRBs)

IRB Name

IRB Approval Date

IRB Approval Number

Analysis Plan