Malaysia wants to cut its reliance on foreign labour to 10% of the workforce by 2030, down from the current 13% cap. The stated route is automation.
The Star ran a piece on Sunday where the people who'd actually have to buy the machines explained why that's harder than it sounds.
1. The number that ends most conversations
Automating a small or medium manufacturer costs between RM500,000 and RM1 million. That's the figure quoted in the piece.
For an MSME running on thin margins, that isn't a capital expenditure decision. It's the whole business. You're asking a company that might turn over a few million a year to spend the equivalent of several years of profit on equipment, on the promise that labour policy stays put long enough for it to pay back.
2. Nobody in the story disagreed with automation
This is the part worth noticing. Every single person quoted was pro-automation. The argument wasn't whether. It was how fast, and who pays.
Malaysian Employers Federation president Datuk Seri Syed Hussain Syed Husman said automation "should be viewed as a gradual transition rather than an immediate substitute." He also said Malaysia shouldn't frame this as "foreign workers versus automation."
SME Association of Malaysia president Chin Chee Seong: "The objective should not simply be to replace foreign workers with machines."
Deputy Investment, Trade and Industry Minister Sim Tze Tzin was pushing the other way: "We want businesses to understand the importance of automation, the importance of moving up."
Everyone agrees on the destination. The disagreement is entirely about the timetable, and timetables are where policy either works or quietly doesn't.
3. Why "replace the worker with a machine" is the wrong frame
Syed Hussain's point about not making this "foreign workers versus automation" deserves more airtime than it got.
A machine doesn't replace a worker one for one. It changes what the job is. Someone has to run it, maintain it, program it, and fix it at 2am when a production line is down. Those are better jobs. They're also jobs that need people who've been trained for them.
Which loops straight back to the underemployment problem. Malaysia has 1.96 million skill-underemployed workers and a graduate underemployment rate of 35.5%. There's a genuine argument that the technical workforce for an automation push already exists and is currently doing something else.
4. The thing that would actually move this
If the government wants 10% by 2030, the binding constraint isn't willingness. It's financing.
An MSME can't self-fund RM500,000 to RM1 million. It can service a loan against it if the payback is credible, and the payback is only credible if the foreign labour cap actually holds. Every time the cap gets relaxed, the business case for the machine gets worse, and the firm that bought one gets punished for moving early.
That's the real policy problem. Not persuasion. Predictability.
5. What to watch
Budget 2027, tabled 9 October. Finance Minister II Datuk Seri Amir Hamzah Azizan has named "raising the ceiling," including MSMEs, as one of three priorities.
If that turns into automation financing with a long tenor and a stable labour cap alongside it, 2030 is arguable. If it turns into another grant scheme with a RM50,000 ceiling, the maths doesn't reach.
Source: 'Automation not a quick fix', The Star, 23 August 2026, by Charles Ramendran.