About the Project

"Skills for Growth" project aimed to comprehensively study the manufacturing sector in the Dhaka metropolitan area. A representative sample was drawn from 200 establishments, comprising 2,000 workers and 500 managers. The questionnaire was modeled on US O*NET skills questionnaire, measuring the importance and level of 35 distinct skill dimensions as relevant to the job performance of workers.

Project History

Survey of Skills and Employers Recruiting Behavior (SSERB)

A previous study conducted in Peru explored human capital differences between poor and rich countries, specifically examining job skill requirements. Researchers found that Peruvian jobs have significantly more uniform skill profiles than those in the U.S. They developed a theoretical model in the O-ring tradition, suggesting that when there is uncertainty about labor availability, firms may prefer hiring unspecialized workers. This preference leads to a labor market where workers, known as "toderos" (jack-of-all-trades), are expected to perform a broad range of tasks.

• SSERB Survey Period & Sample: Conducted in Peru (2017-2018) and the U.S. (2020-2021), covering 1,000 firms and 5,000 workers in Peru, and 308 employers in the U.S.
• Sampling Strategy: In Peru, half of the firms were selected based on employment distribution, while the rest were reported by surveyed workers. Occupational classification matched the U.S. Standard Occupational Classification (SOC) at the four-digit level.
• Skill Measurement: Ten skill categories were evaluated, including cognitive, social, writing, customer service, project management, and computer skills. Both surveys used the same skill taxonomy and five-point rating scale, following the O*NET methodology in the U.S.
• Key Findings: Peruvian jobs required a broader range of skills than U.S. jobs (mean skill importance: 3.3 vs. 2.7), with workers performing multiple tasks rather than specializing. Despite using informal hiring methods, vacancies in Peru were filled much faster (9.06 vs. 28.5 days). The study also found that hiring generalists, or "toderos," was more optimal for firms than hiring specialists.

Objectives

To analyze skill gaps and labor market inefficiencies in 13 major sectors to inform workforce development strategies in Bangladesh.

• To Identify & Quantify Skills: To assess 35 skill dimensions using the US O*NET model to understand job-specific skill demands.
• Evaluating Skill Levels: Comparing existing worker competencies with employer expectations to pinpoint skill gaps.
• Analyzing Labor Market Frictions: Investigating factors like educational gaps, training accessibility, and technological shifts affecting employment and productivity.
• Developing Actionable Recommendations: Providing insights for policymakers, industries, and educational institutions to enhance skill development.
• Promoting Economic Growth: Ensuring workforce alignment with industry needs to boost productivity, lower unemployment, and support sustainable development.

Implementation

To achieve our goal, we conducted a comprehensive survey across Dhaka, Gazipur and Narayanganj, using two sampling methods to identify firms. Our data collection involved listing survey, pilot survey and manager-employee survey, gathering insights from 2,000 employees and 500 managers.

Skill Assessment Using the US O*NET Questionnaire

We used the US O*NET Skill Questionnaire to evaluate 35 skill categories across job roles, measuring both importance and proficiency using a seven-point scale. Participants rated each skill from "Not Important" to "Extremely Important" and assessed their own competency based on predefined benchmarks.

This structured approach provided quantifiable insights into cognitive abilities (e.g., critical thinking, reading comprehension), technical skills (e.g., programming, troubleshooting), and social skills (e.g., negotiation, coordination). The data helped identify skill gaps and training needs, ensuring workforce capabilities align with industry demands for improved job performance and readiness.

Learn more about the US O*NET Questionnaire here.

Survey Execution

Sampling Methods

A comprehensive data collection strategy was implemented with two sampling methods:

One was geo-sampling, which identifies firms using satellite images and the other one was industry-area-firm size combination which selects firms based on census data.

Survey Location

The study was conducted in Dhaka, Gazipur, and Narayanganj, selecting establishments based on industry representation, geographic distribution, firm size, and managerial insights to ensure diverse employment conditions and job roles.

Figure: Location Map of the Study Area

Data Collection Phases

The data collection process is structured into three key phases: a listing survey to document the firms, a pilot survey to refine the methodologies, and a manager and employee survey to capture the skill assessments.

Listing Survey

The listing survey aimed to find and categorize manufacturing firms while collecting important data for future research. In just seven days, in the first week of July 2024, 42 trained surveyors visited around 10,000 firms across Dhaka's industrial areas. This helped to understand where workers are, what industries they work in, and the challenges businesses face. Careful planning, test surveys, and industry classification improved the process. To keep data accurate, surveyors followed a step-by-step method and held daily review meetings.

Pilot Survey

Two pilot surveys were conducted before the manager and employee survey to refine the questionnaire. Pilot-1 (September 26, 2024, Dhaka) focused on evaluating survey clarity for both managers and employees, revealing issues like difficulty measuring intangible capital and confusion about bonuses and skill-related questions. Adjustments were made to improve question phrasing and accessibility. Pilot-2 (December 23, 2024, Mirpur, Dhaka) tested the revised questionnaire with ARCED team members across multiple industries. It highlighted new challenges, such as managers needing detailed explanations, employees being too busy, and income reporting requiring decimal inputs. Solutions included better scheduling, clearer question preambles, and alternative incentive methods like mobile recharge.

Manager and Employee Survey

The manager and employee survey began in January 2025. Before starting, we conducted a three-day training session for enumerators. Five teams, each with one supervisor and a total of 24 enumerators, carried out the survey. Data collection ran from January 22 to February 25, with over 100 responses submitted daily. By the end of the survey, we successfully gathered data from 2,000 employees and 500 managers, matching our initial target.

The survey results will provide valuable insights into skill gaps and employment conditions across various industries, helping inform policies and strategies for workforce development. The data collected will be analyzed to support evidence-based decision-making in enhancing labor market dynamics.

Engagements

Dive into the research paper that reveals the job skills requirements and human capital differences between the rich and poor countries, Survey of Skills and Employers Recruiting Behavior (SSERB)