A Promise in the Name of God

How to cite

Neshat, Navid; Kim, Yongseok. 2025. "A Promise in the Name of God." AEA RCT Registry. AEARCTR-0015040. https://doi.org/10.1257/rct.15040-2.0

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

Navid Neshat* , Yongseok Kim

* Primary investigator

Registry ID
AEARCTR-0015040
DOI
https://doi.org/10.1257/rct.15040-2.0
First registered
Last updated
Status
In development
Trial dates
2024-12-21 – 2025-07-01
Countries
Private
JEL codes
D8, C93
Registry record
http://www.socialscienceregistry.org/trials/15040

Abstract

In environments with limited contract enforcement, sellers struggle to credibly signal product quality. By analyzing audio recordings of hundreds of trade negotiations in Afghanistan, we aim to examine how frequently sellers invoke God to assure buyers of their products' quality. We further evaluate the effectiveness of these religious oaths by conducting a field experiment.

Intervention

The intervention involves instructing sellers to avoid making religious vows during their conversations with customers in a given session. The treatment is randomly assigned at the seller-date-session level, while the control group receives no such instructions.

Experimental design

To evaluate the effectiveness of religious oaths on sales rates, we partner with a local wholesale company in Afghanistan. The experimental design is as follows: A team of six professional sellers, each with at least some experience as street vendors, will operate handcarts to sell goods daily, which is a common practice in the country.

Each seller offers five items from their handcarts: black tea, candy, pasta, dried dates, and turmeric. These items were selected for their non-perishable nature and widespread availability in local markets. They are sold in two formats: factory-packaged and loose (unpackaged). Factory-packaged items are sealed, similar to those in developed markets, while loose items are displayed in food-grade storage containers.

Each cart carries 15 kg of each item, and for a given cart, all items are either in factory-packaged format or in a loose format. Over 30 days, at the beginning of each day, the partner company randomly assigns six carts to the six sellers. The neighborhood where each seller operates is also randomly selected from a list of areas in Kabul where mobile vendors are commonly found. Sellers work in two shifts: morning (8:00 AM to 12:00 PM) and afternoon (2:00 PM to 6:00 PM). In the afternoon, they relocate to a nearby street of their choice.

The intervention involves instructing sellers to refrain from making religious vows in their conversations with customers. The treatment is randomly selected at the seller-date-session level. The control group receives no such guidelines.

Each day, sellers receive a randomly assigned cart with 15 kg of each item at no cost. We provide goods for free at first, and we set the prices, and we take the money of what is sold. Seller compensation consists of a fixed payment of AFN 10,000 (approximately 140 USD), regardless of sales performance, plus a 2 percent commission on total sales. The fixed payment exceeds the minimum monthly wage for a permanent government worker (AFN 6,000) by over 60 percent. The commission is included to incentivize effort.

Item prices are set at the median price quoted by at least five local stores for the corresponding format (factory-packaged or loose). At the end of each session (morning and afternoon), we record the quantity of each item sold. These data will be analyzed to assess the impact of religious oaths and product formats on consumer purchasing behavior.

Design details

Randomization method
The randomization was conducted by computer.
Randomization unit
seller-date-session
Clusters
6 sellers
Observations
6 sellers×30 days×2 sessions per day
Arms
900 sales amount control, 900 sales amount treatment
Primary outcomes
For each item in a given format (i.e., factory-packaged or loose), the outcome variable is the sales amount, measured at the seller-session-date level.

Cited by

No citation data available for this registration yet.

Keywords