Vietnam's growth model is rented hours, and the price of an hour is falling

Vietnam's growth model has been simple to describe. It rents out hours.

Foreign companies bring the capital, the technology, the brand and the customer. Vietnam supplies the factory floor, the port, the stability and the labour. Everything the country earns from that arrangement is, in the end, a price for time.

The wage ladder

The numbers are not flattering in the way they are usually presented. According to JETRO's 2024 survey of Japanese companies operating in Asia, the average basic monthly wage for a manufacturing worker was:

Vietnam sits in the middle of that ladder: cheaper than Malaysia, Thailand and Indonesia, more expensive than the Philippines. Within Vietnam the spread is wider still, from about $242 a month in Hai Duong to roughly$399 in Ho Chi Minh City. The country is not the region's bargain basement any more, and it has been raising wages faster than most of its neighbours.

That matters because the selling point is no longer absolute cheapness. It is cheapness relative to output, delivered at scale, on time, inside a supply chain that has been assembled over fifteen years.

One customer, one channel

Concentration is the second half of the story. Vietnam exported about $153.2 billion to the United States in 2025, up 28.2% on the year and roughly 32% of total exports. Over 77% of all Vietnamese exports are produced by foreign-invested enterprises.

So the chain runs: foreign capital builds the plant, foreign demand buys the output, and the domestic contribution is mostly labour and land. That is not a criticism of the model. It is a description of where the value sits, and of how few decisions Vietnam controls inside it.

Why this time is different

Automation has been promised before. Factories still moved to Asia through the 1990s and 2000s because robots were rigid: excellent at repeating one motion, useless at a task that changed every week, and expensive to reprogram.

Two things have changed. First, the learning. A system that can be taught a new task by demonstration, rather than by an engineer rewriting every step, removes the cost that kept flexible assembly in human hands. Second, the cost of the machine itself, which has fallen while wages have risen.

The visible evidence is already mundane. Xiaomi runs a phone plant outside Beijing that barely turns the lights on. Driverless taxis carry passengers daily in several American cities. Humanoid robots are being trialled on car lines. None of that is a finished transition; all of it points the same way.

If the price of an hour of machine work falls below the price of an hour of Vietnamese work — for tasks that can be specified — then the reason to ship a factory across the Pacific weakens. Not overnight. Ten to fifteen years is the plausible window, which is one school generation.

What stays scarce

There is a version of this argument that ends in resignation. It should not.

The useful question is not how to keep every factory. It is what remains expensive when machines can make almost anything. The answer is the things people pay for because a person is doing them: care for the old, health, education, and the parts of tourism and culture that depend on welcome rather than throughput.

Those sectors share a property that manufacturing does not. Their value is not in the unit cost. It is in trust, language, patience and presence — the parts that are hardest to automate and easiest to underprice.

Rich countries are ageing quickly. Australia's intergenerational report projects its population aged 85 and over to triple by 2066; Japan and South Korea are further along. Robots can remind someone to take a pill. They do not hold a hand.

The bet to stop making

The last fifteen years were won by betting on factory relocation, and the bet paid. The risk is treating a winning bet as a permanent one. An economy with a single channel — foreign capital, foreign demand, domestic hours — is exposed to a single decision it does not make.

Diversifying does not mean abandoning industry. It means not asking one sector to carry the whole country's growth, and building the parts of the economy whose value survives the moment when an hour of machine time becomes cheaper than an hour of anyone's attention.

The question is no longer how to keep the factories. It is what is still scarce.

Source: JETRO FY2024 Survey on Business Conditions of Japanese Companies Operating Overseas; Vietnam Customs and Ministry of Industry and Trade export data for 2025.

Contact: [email protected]