Research question

Skills training by private firms in Bangladesh is not necessarily resulting in higher rates of employment, better paid employees or more productivity, and levels of fraud can be high. Private training providers (PTPs) in Bangladesh are usually incentivized to provide skills that generate employment: one-third of their payment per trainee is released when the trainee finds employment. In practice, there is significant fraud in overstating employment. Up to a third of public resources may be wasted in this way, and the true cost is much higher. How can an anti-corruption strategy improve skills training outcomes, reduce corruption and achieve better productivity outcomes for employers?

Project summary

The garments sector is a major exporter and employer in Bangladesh, and around 80% of employees are low-income women. Skills training programmes should enable workers to increase their skills and make firms more productive, enabling faster employment and wage growth. However, while complaining about the lack of skills, Bangladeshi garments factories are curiously reluctant to pay for training, and often prefer to employ slightly cheaper unskilled workers at the factory gates. Fraud occurs when private training providers (PTPs) whose payments are linked to results overstate employment success of their trainees, sometimes in collusion with agencies charged with releasing payments. Our initial informal investigations suggested between ten to thirty percent of resources may be wasted due to fraud, and the true cost is much higher because this fraud creates further incentives for poor quality training.

To solve this puzzle we use insights from research on the determinants of productivity. The productivity of a worker in a factory depends jointly on the skill of the worker and the organizational capability of the factory. A badly organized factory is not likely to observe any significant improvement in productivity by putting a skilled worker onto a poorly functioning production line. In contrast, in a factory with high organizational capabilities, an unskilled worker slows down the production line and skill deficits are easily detected.

The research suggests that corruption and fraud in skills training programmes may be a result of an inappropriate design of incentives to improve the marketability of skills, while monitoring and enforcement does not work well, even in very well-managed programmes. A feasible and effective anti-corruption strategy would be to jointly support skills training together with support for the organizational upgrading of firms. A joined up policy may create support for more effective monitoring and enforcement because a significant number of insiders including both PTPs and employing firms will support the implementation of the strategy and help to identify and exclude the minority who may continue to engage in fraud.

Listen to Mushtaq Khan discussing this project in our ACE podcast:


Mushtaq Khan and Joseph Ajefu (SOAS University of London), Sumaiya Kabir Talukder and Mehnaz Rabbani (BRAC Institute of Governance and Development – BRAC University).

Read the project summary

This project is now completed.