Learning versus Signaling in the Workplace: Evidence from the Adoption of Commitment Savings Accounts

How to cite

Kanz, Martin; Klapper, Leora; Breza, Emily. 2015. "Learning versus Signaling in the Workplace: Evidence from the Adoption of Commitment Savings Accounts." AEA RCT Registry. AEARCTR-0000913. https://doi.org/10.1257/rct.913-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Martin Kanz* , Leora Klapper , Emily Breza

* Primary investigator

Registry ID
AEARCTR-0000913
DOI
https://doi.org/10.1257/rct.913-1.0
First registered
Last updated
Status
In development
Trial dates
2015-10-12 – 2016-06-30
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/913

Abstract

While an increasing number of low-income households in developing countries has access to a formal bank account, there are many barriers that prevent households from accessing a wider range of welfare-enhancing financial products and services. Although technically “banked”, these individuals often find the formal financial sector daunting and complicated and fail to actively engage with it. As a result, they are not able to fully utilize the facilities it offers and more importantly, they are unable to use the formal financial sector as a mechanism to meet their financial goals. In this project, we examine how workers learn about a new financial product being offered as an add-on to a basic salary account and examine the mechanisms that drive the adoption decision. We partner with a leading bank in Bangladesh and offer a simple "Deposit Pension Savings" (DPS) product to salaried workers in the garment sector in Bangladesh who already have an electronic payroll account. Workers are given the option to enroll in the DPS plan and commit to having a fixed savings amount debited from their monthly salary. We vary the conditions under which the product is presented to eligible workers and study to what extent the adoption decision is driven by learning about the product versus the desire on the worker's part to convey information to the employer.

Intervention

We work with a leading bank in Bangladesh to offer a DPS commitment savings product to garment workers that already receive their monthly salary into an electronic payroll account. Workers are given detailed information about the product and have the opportunity to sign up for a DPS account. If they do so, they can choose an amount to be debited from their monthly salary and deposited into their savings account.

The intervention varies the conditions under which workers are offered the product. All workers in the treatment groups receive a detailed presentation of the product terms and characteristics and complete a short survey at the end of the information sessions. The main intervention consists of two additional features that are added to the product presentations for randomly chosen subgroups of workers: (1) an endorsement of the product from the employer, delivered by a floor manager (2) an announcement that feedback about the sign-up decision will be provided to the employer.

The intervention is set-up as a simple 2x2 design with the following treatment groups:

1. Control [No offer]

2. T1: No employer endorsement + no feedback to employer [Basic offer]
3. T2: No employer endorsement + feedback to employer
4. T3: Employer endorsement + no feedback to employer
5. T4: Employer endorsement + feedback to employer

We compare take-up rates under these experimental conditions to examine to what extent the take-up decision is driven by (i) information about the product versus (ii) employer endorsement (information from a credible source) versus (iii) feedback to the employer i.e. the ability to signal to the employer. Using data on personal characteristics, we will also examine whether the take-up decision varies systematically by worker characteristics across the treatment groups.

Experimental design

The intervention is set-up as a simple 2x2 design with the following treatment groups:

1. Control [No offer]

2. T1: No employer endorsement + no feedback to employer [Basic offer]
3. T2: No employer endorsement + feedback to employer
4. T3: Employer endorsement + no feedback to employer
5. T4: Employer endorsement + feedback to employer

Design details

Randomization method
The randomization was conducted using a computer random number generator.
Randomization unit
We randomize at the level of the individual worker and stratify by location in the factory, income category and gender.
Clusters
NA
Observations
2,000 workers to receive offer (conditional on employer permission)
Arms
400*5 (conditional on employer permission)
Primary outcomes
The main outcome of interest is the take-up decision. All workers in the treatment groups will be asked after the product presentation whether they wish to sign up for the product ("yes" or "no"). For consumer protection reasons, workers will also be given the option to make a decision within two days of the presentation, in which case we record both the initial response and the final take-up decision. We will also examine heterogeneity in take-up by personal characteristics, collected in a short baseline survey and from proprietary data on worker productivity supplied by the factories participating in the experiment. The baseline survey also collects phone numbers of respondents so that (conditional on worker and employer consent) we can collect follow-up information on actual amounts saved and impacts on worker outcomes 3-6 months after the intervention.

Cited by

No citation data available for this registration yet.

Keywords