A recruitment fee is the factory's debt, not the worker's

Almost every manufacturer in Malaysia that sells to multinational customers now has a policy against recruitment fees. The policy is usually well drafted. It says that workers must not pay for their jobs, that the employer bears the cost of recruitment and that agencies breaking the rule will be dropped. Kelvorne had such a policy from 2019. In 2021 we discovered how little it had changed.

The reason was not bad faith by anyone in our company. Our policy bound the Malaysian agencies we contracted with. Those agencies relied on partners in Nepal, Bangladesh and Myanmar, who in turn relied on sub-agents in villages and district towns. The fee was charged two or three links down that chain, usually in cash, months before a worker reached our gate. When we interviewed 300 of our migrant workers that year, 78 per cent said they had paid, a median of about RM 10,900, often borrowed from relatives or moneylenders at interest. Our policy had made the fee invisible to us without making it go away.

How a fee paid in Kathmandu becomes a cost in Johor

It is tempting to treat a fee paid abroad as a matter between a worker and an agent. Our chief financial officer made that case to me, and it was not a frivolous one. The money was never ours. We had no contract with the sub-agents. No court was likely to find us liable for what they charged.

My answer was operational rather than legal. A worker who arrives owing a year's wages is a different employee from one who does not. They will ask for every hour of overtime available, and the code most of our electronics customers use caps the working week, overtime included, at sixty hours. They will hesitate to raise a safety concern or to refuse a shift. If a better-paid job appears across the industrial estate, they may leave to repay the lender sooner, taking their training with them. And if a customer's auditor or a foreign customs authority concludes that debt is binding people to the job, the consequences fall on our shipments, not on the agent. US customs detained goods from several Malaysian manufacturers over forced-labour indicators in 2020 and 2021, and debt bondage is one of those indicators.

The fee, in other words, does not stay where it was paid. It returns to the factory as overtime pressure, silence, attrition and shipment risk. A cost that comes back to the factory belongs in the factory's accounts.

Repaying what we never received

Once we accepted that, a forward-looking policy was not enough. Banning fees for new hires would have protected people not yet recruited and left 1,400 people on our lines carrying debts we had allowed to be created. We decided to repay every worker still employed who had paid a fee, and every worker who had left in the previous two years whom we could trace.

The method mattered as much as the decision. An independent social-compliance firm interviewed workers in Nepali, Bengali and Burmese, away from their supervisors. We set each amount from the worker's own account, tested against what was known about fee levels in each country, and paid in three instalments into workers' own bank accounts rather than through any agent or supervisor. By March 2022 we had repaid RM 15.8 million, about S$5.1 million, to 1,412 people.

We made mistakes. We calculated amounts before inviting workers to elect representatives to check the method, and 160 workers disputed their figures; some of them were right. We spent weeks perfecting the payment timetable, and rumour filled the silence. Hiring slowed for five months while we replaced two agencies, and about S$14 million of revenue slipped into the following year.

What we feared most did not happen. Our Johor plant director expected repaid workers to resign. Instead, voluntary attrition among migrant workers fell from 19 per cent to 11 per cent within a year. Someone who no longer owes a moneylender has less reason to chase a slightly better wage, and more reason to stay where they were treated fairly.

Running recruitment like a supply chain

The more lasting change was organisational. Recruitment had sat with human resources and compliance, functions that write rules and check them. We moved accountability for the recruitment chain into operations, where I manage it as I manage castings or tubing. Our Malaysian agencies are treated as first-tier suppliers. Their home-country partners are the tier beneath, and we audit them in the same way, in their own offices, with our own staff making pre-departure visits. We now recruit directly in Nepal. Every cost of bringing a worker to us, including medical checks, permits, the levy and the flight, sits in our labour standard cost, at about RM 6,800 a hire.

That figure is the point of the exercise. Once recruitment cost is in the standard cost, it competes honestly with automation, overtime and the location of the next plant. When we weighed robotic welding cells in Johor, the full cost of recruitment, rather than a fee hidden in a worker's debt, was part of the comparison. Decisions about where to make a part became decisions about what a worker really costs.

Some peers tell me that repaying fees puts a company at a disadvantage against competitors who do not. For a period it does. But customers increasingly ask for evidence rather than policy, and a manufacturer that can show repayments, direct recruitment and audited partners is easier to award work to. I also think the disadvantage is smaller than it looks, because the costs of indebted labour are real even when nobody has added them up.

For a board, I would suggest three questions. Do you know what your migrant workers paid to reach you, from their own accounts rather than from your agencies? Who in management owns the recruitment chain below your first-tier agencies? And is the full cost of recruitment in the cost of your product, or in somebody's loan?