Business Practices from Foreign Cultures: The Effects of Chinese and American Trainings on Ethiopian Workers

How to cite

Wu, David; Ortiz, Miguel. 2026. "Business Practices from Foreign Cultures: The Effects of Chinese and American Trainings on Ethiopian Workers." AEA RCT Registry. AEARCTR-0018302. https://doi.org/10.1257/rct.18302-1.0

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

David Wu* , Miguel Ortiz

* Primary investigator

Registry ID
AEARCTR-0018302
DOI
https://doi.org/10.1257/rct.18302-1.0
First registered
Last updated
Status
In development
Trial dates
2026-04-13 – 2026-12-31
Countries
Ethiopia ()
Registry record
http://www.socialscienceregistry.org/trials/18302

Abstract

This project studies whether organizational practices developed in advanced economies can be successfully transplanted to improve firm performance in the developing world. It focuses on two contrasting coordination models commonly used by multinational firms operating in Ethiopia: a Chinese-style model that emphasizes hierarchical discipline, rule-following, and centralized authority, and a US-style model that emphasizes horizontal teamwork, open communication, and shared responsibility. While both approaches are widely adopted in practice, there is limited causal evidence on their effectiveness when implemented in local firms operating in different cultural and institutional environments. The study asks whether these foreign organizational practices improve worker performance, and through which mechanisms of coordination such improvements occur. To answer this question, the project implements a randomized controlled trial with 150 manufacturing firms in Addis Ababa. Firms are randomly assigned to either a control group or to one of the two training interventions, each delivered through a three-day workshop to five workers per firm. Outcomes are measured using three complementary sources: supervisor evaluations of both treated and untreated workers, detailed worker surveys capturing attitudes, well-being, and behavioral traits, and a standardized performance task designed to measure productivity in a controlled setting. Data are collected at baseline, two weeks after the intervention, and three months later, allowing the study to assess both short- and medium-term effects of alternative organizational models.

Intervention

We will first obtain a list of manufacturing firms registered at the Addis Ababa city administration. From this frame, we will randomly select 150 firms, stratified by garment sector and non-garment sectors (metal work, woodworking, leather production). Because garment sector mainly employs female workers and non-garment sectors mainly employ male workers, we will stratify our treatment by whether a firm is in garment sector, and exclusively select female workers from garment sector and male workers from non-garment sectors.

We will collect baseline characteristics one week prior to the intervention. This will include three surveys: (i) Firms’ survey, to measure firm characteristics, including size, organizational structure, and performance evaluation and reward systems. (ii) Workers’ survey, to measure workers' valuation of their job and work more generally, attitudes toward teamwork and discipline, mental and physical health, and behavioral measures of discipline and creativity, among other outcomes. (iii) Supervisors’ survey: Supervisors from each firm will evaluate 10 workers. Five of these workers will be selected to participate in the training (for treated firms), while the other five will serve as within-firm comparisons (allowing us to study spillovers and improve precision). As a result, we will obtain supervisor evaluations for 1,500 workers in total.

Two weeks after the intervention, we will conduct a second round of data collection. This will include the workers' survey and the supervisors' evaluation survey to capture short-run effects. In addition, we will administer a performance task designed to resemble a manufacturing process and provide a standardized measure of performance across firms. Three months after the intervention, we will measure longer-term effects. We will administer the firms' survey, the workers' survey and the supervisors' evaluation survey.

Experimental design

Both workshops will be held at training facilities provided by the Technical and Vocational Training Institute and will last three days.

Chinese-style discipline workshop: This workshop resembles practices used by prominent Chinese firms operating in Ethiopia. It emphasizes hierarchical discipline and structured coordination. Activities include physical, team-based exercises—such as coordinated marching—where following instructions and synchronizing with the group are essential. Teams compete in these activities in an engaging format designed to foster collective discipline.

US-style teamwork workshop: This workshop resembles training programs offered by leading soft-skills providers in the United States. It focuses on trust, open communication, and egalitarian collaboration. Activities include problem-solving games, trust exercises, and communication challenges aimed at strengthening interpersonal relationships and fostering a cohesive work environment where ideas can be shared.

Design details

Randomization method
Randomization is determined by a computer program in Stata after baseline survey.
Randomization unit
Firm-level
Clusters
150 firms
Observations
1500 workers (500 workshop participants, 500 spillover workers in the treated firms, 500 control workers)
Arms
50 firms will be assigned to Chinese-style workshop; 50 firms will be assigned to US-style workshop. For each randomly selected firm, all five workers selected for the workers' survey at baseline will participate in the assigned workshop. In total, 250 workers will participate in Chinese-style workshop; 250 workers will participate in US-style workshop.
Primary outcomes
Supervisors' evaluations for each worker: - A normalized index of overall performance - A normalized index of teamwork ability - A normalized index of discipline (ability of following rules and regulations) Workers’ midline and endline measures: - Number of questions correctly answered in the discipline test (discipline measure) - Number of objects listed in creativity test (creativity measure) - A normalized index of attitudes on discipline - A normalized index of attitudes on teamwork - Whether a worker quits the job (turnover measure) Performance task: - Worker’s individual performance (5-point Likert scale) evaluated by enumerators - Team-level outcomes on quality, time efficiency, enumerators’ subjective evaluations (coordination, task allocation, planning, communication frequency, communication effectiveness, creativity, engagement)
Secondary outcomes
Worker’s midline and endline measures: - A normalized index of depression - A normalized index of anxiety - A normalized index of physical health - Willingness to accept a hypothetical offer with a special work session in the current firm as a measure for value of work - Willingness to accept a hypothetical offer from elsewhere as a measure of value of outside options - Job satisfaction and appreciation - Altruism and trust level measured from a dictator game - Value for both Chinese-style and US-style trainings - Views on China and US economic models

Cited by

No citation data available for this registration yet.

Keywords