What do External and Internal Corporate Social Responsibility do? Evidence from a Firm-level Field Experiment

How to cite

Wu, Hugh Xiaolong. 2020. "What do External and Internal Corporate Social Responsibility do? Evidence from a Firm-level Field Experiment." AEA RCT Registry. AEARCTR-0005959. https://doi.org/10.1257/rct.5959-1.0

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

Hugh Xiaolong Wu*

* Primary investigator

Registry ID
AEARCTR-0005959
DOI
https://doi.org/10.1257/rct.5959-1.0
First registered
Last updated
Status
On going
Trial dates
2020-04-01 – 2021-02-01
Countries
Private
JEL codes
J0, L2, M1, M5
Registry record
http://www.socialscienceregistry.org/trials/5959

Abstract

This study provides answers to two research questions. First, what are the impacts of corporate social responsibility (CSR) messages on employee attrition, productivity, and well-being? Second, we compare two types of CSR information-external CSR and internal CSR- to examine which type of CSR is more effective.

We conduct a 3-month field experiment on more than 8,000 workers in 145 stores of a large multi-national spa chain home-based in China. The spa worker's job is physically exhausting. Workers have long working hours and the attrition rate is very high. Thus, the strategic provision of corporate social responsibility messages could be an effective managerial practice to improve employee attrition, productivity, and worker well-being.

We assign stores into two treatment groups and one control group. In both treatment groups, workers receive a corporate social responsibility message every other day through the company's information system. Our first treatment, labeled as "external CSR", is to send workers information about the firm's external CSR events in history. Examples include but are unlimited to the firm's subsidizing the socially disadvantaged groups, building schools, and donating during natural disasters. Our second treatment, labeled as "internal CSR", is to provide employees with information on the firm's internal CSR events, for example, setting up relief funding, providing resources for, and helping out its own workers during difficult times. The difference between the two treatments is that external CSR is not directed related to worker's welfare, while internal CSR is closely related.

The field experiment starts on June 1st, 2020, and lasts 14 weeks (3 months).

Intervention

We provide workers with two types of corporate social responsibility messages (external and internal CSR) every other day.

Experimental design

We assign stores into two treatment groups and one control group. In both treatment groups, workers receive a corporate social responsibility message every other day through the company's information system. Our first treatment, labeled as "external CSR", is to send workers information about the firm's external CSR events in history. Examples include but are unlimited to the firm's subsidizing the socially disadvantaged groups, building schools, and donating during natural disasters. Our second treatment, labeled as "internal CSR", is to provide employees with information on the firm's internal CSR events, for example, setting up relief funding, providing resources for, and helping out its own workers during difficult times. The difference between the two treatments is that external CSR is not directed related to worker's welfare, while internal CSR is closely related.

Design details

Randomization method
Randomization was done in office by a computer: We select employees into treatments or control groups by store using the stratified randomization method. In addition to the average attrition rate (store-level), we stratify on store sales and size, since these characteristics are correlated with turnovers, productivity, and well-being.
Randomization unit
Store
Clusters
145 stores
Observations
8,000 workers
Arms
48-49 stores per arm
Primary outcomes
Personnel Turnover Employee productivity (sales, repeated customer, compensation, attendance, etc.) Store-level performances (e.g. revenues, customer visits) Employee well-being
Secondary outcomes
Pro-social behavior (this outcome is for another separate study, measured by sales one helps coworkers achieve)

Cited by

No citation data available for this registration yet.

Keywords