|
Field
Trial Title
|
Before
Career Incentives and Turnover: Evidence from Ethiopian Manufacturing Workers
|
After
Misperceptions of Career Incentives and Turnover: Evidence from Ethiopian Manufacturing Workers
|
|
Field
Abstract
|
Before
High turnover rates are a common issue in manufacturing firms in developing countries. In one of the main industrial parks in Ethiopia, the government has speculated that misperceptions regarding job characteristics among new hires contribute to high turnover rates. We propose to implement a new information treatment, with particular focus on long-run promotion incentives, and analyze the effect on belief update and turnover decisions. We will sample 2,000 new female young workers before they start working, and randomly provide information on entry-level operators, medium position, and high position. We plan to follow up workers within 1 week and after 6 months to measure their update in perceptions of salary and promotion within the industrial park and outside, whether workers have already left the industrial park, welfare outcomes, productivity, and skills.
|
After
High turnover rates are a major issue in industrialization of developing countries. In one of the major industrial parks in Ethiopia, we identify a key determinant of turnover from our pilot study – career incentives. Despite the importance of career incentives during job search, workers have substantial misperceptions of career incentives within the industrial park, and it is difficult for workers to learn the true career incentives from on-the-job experience. We propose an intervention that provides accurate information of promotion likelihood and salary after promotion to understand how the intervention might help workers learn about the career incentives, and to causally estimate the effect of misperceptions of career incentives on workers’ turnover. We plan to sample 1,200 new female young workers, randomly provide the information of the career path on 400 workers, and follow up workers after 45 days and 6 months to understand how misperceptions prevent workers from optimal job search, and the causal effect on long-run employment and welfare outcomes.
|
|
Field
Trial Start Date
|
Before
May 10, 2021
|
After
March 01, 2022
|
|
Field
Trial End Date
|
Before
April 30, 2022
|
After
December 31, 2022
|
|
Field
Last Published
|
Before
May 10, 2021 11:40 AM
|
After
February 15, 2022 10:55 PM
|
|
Field
Intervention Start Date
|
Before
May 10, 2021
|
After
March 14, 2022
|
|
Field
Intervention End Date
|
Before
August 31, 2021
|
After
June 30, 2022
|
|
Field
Primary Outcomes (End Points)
|
Before
Updated perceptions of job characteristics of the industrial park, workers' turnover, workers' productivity
|
After
Updated perceptions of job characteristics of the industrial park, whether workers stay in the firm in the follow-up surveys
|
|
Field
Experimental Design (Public)
|
Before
We will implement a cross-randomized experiment on 2000 female young workers. For the first treatment arm (entry-level operator), we will provide information of salary as an entry-level operator. For the second treatment arm (promotion), we will provide salary and promotion likelihood of a higher position after 6 months. Within the second treatment arm, for half of the workers, we will provide information of the medium-level position; for another half, we will provide information of the high-level position.
|
After
We will implement a clustered randomized experiment on 1,200 female young workers. Workers will be sampled over the course of 50 days. We will randomly select 25 days, randomly select 2/3 of the workers on those days, and provide them with true promotion likelihood ("Out of 100 entry-level workers, how many of them will be promoted to upper level?") and salary after promotion ("What's the average salary of upper-level positions?").
|
|
Field
Randomization Method
|
Before
Randomization done in office by a computer
|
After
We will generate a randomization file before the intervention that pre-determines workers with certain IDs on certain days will be treated.
|
|
Field
Randomization Unit
|
Before
Individual
|
After
The randomization is clustered by days of hires. Within the treated clusters, individuals will be randomly treated.
|
|
Field
Planned Number of Clusters
|
Before
22 firms and 60 days of hiring
|
After
50 days of hiring
|
|
Field
Planned Number of Observations
|
Before
2,000
|
After
1,200
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
500 control, 500 only provided with entry-level operator information, 500 only provided with promotion information (250 with medium-level position, 250 with high-level position), 500 both treatment (250 with medium-level position, 250 with high-level position)
|
After
25 treated clusters (600 workers), 25 control clusters (600 workers). Within treated clusters, 400 out of 600 workers are treated.
|
|
Field
Additional Keyword(s)
|
Before
|
After
Turnover
|
|
Field
Secondary Outcomes (End Points)
|
Before
Workers' welfare outcomes (income, consumption, health), plans in the future (education, marriage, migration), skills (general and specific), work attitude
|
After
Worker's effort (absent days, overtime, performance pay collected from firms' records), welfare outcomes (income, consumption, health), plans in the future (education, marriage, migration), skills (cognitive and dexterity)
|
|
Field
Pi as first author
|
Before
No
|
After
Yes
|
|
Field
Building on Existing Work
|
Before
|
After
No
|