On 5 January 2026, at the briefing on Vietnam's 2025 economic and social figures, the statistics office under the Ministry of Finance released three numbers every HR department should write down.
Labour productivity across the economy rose 6.83% at constant prices. At current prices it reached an estimated 245.0 million dong per worker, equivalent to US$9,809, up $626 on 20241.
Average worker income was 8.4 million dong a month, up 8.9%, an increase of 685,000 dong on the year before1.
Average consumer price inflation for 2025 was 3.31%1.
All three sit in the same release. Putting them side by side is the reader's job.
The first subtraction, and the trap inside it
At a glance, income grew 8.9% and productivity 6.83%, a gap of two points, and the conclusion seems to write itself: pay is running ahead of productivity.
That conclusion fails on a technical point. The 6.83% productivity figure is at constant prices, with price effects stripped out. The 8.9% income figure is nominal, with inflation still in it.
Take the 3.31% CPI out of 8.9% and real income growth is about 5.4%. Against real productivity growth of 6.83%, the picture reverses: real productivity grew faster than real pay.
It is a simple calculation, and because it is simple it gets skipped, in news coverage and in budget proposals alike. Whoever has this detail in hand at the board meeting holds the stronger position, whether they are defending a pay rise or arguing against one.
Where the productivity came from, when 70% of workers hold no qualification
The statistics office attributes part of the gain to a better-skilled workforce: the share of workers holding a degree or vocational certificate reached an estimated 29.2% in 2025, 0.8 percentage points higher than in 20241.
Read that from the other end. About 70.8% of Vietnam's labour force holds no formal qualification or certificate. And the improvement over a year is 0.8 points.
So most of the 6.83% did not come from new certificates. It came from elsewhere: workers shifting into higher value-added sectors, investment in equipment, and skill learned on the job that no document records.
That has a direct consequence for anyone running training. If the real capability of the workforce is built at work, what system captures and transfers it? In most Vietnamese companies the answer is that no system does. It lives in the heads of team leaders, and it leaves when they do.
A tighter labour market than you think
The rest of the release describes a market with little slack.
Employment in 2025 stood at 52.4 million, up 578,300. Unemployment among the working-age population was 2.22% for the year, down 0.02 points, at 2.51% in urban areas and 2.02% in rural ones. Underemployment was 1.65%, down 0.20 points, at 1.25% urban and 1.93% rural1.
One figure usually overlooked: 3.8 million people did subsistence work in 2025, 122,500 more than in 2024. They are not wage employees, and they are also a potential supply if formal employment becomes attractive enough.
With unemployment around 2%, a workforce plan resting on the assumption that you can hire when you need to is betting on a market that no longer exists. The urban-rural gap in both measures also says where supply remains and where it has run out.
Four numbers for the 2027 salary budget paper
National data does not set any single company's pay rise. It gives you benchmarks. These four are enough:
One, inflation. CPI averaged 3.31%. This is the level that keeps real income flat. Anything below it is a real pay cut, however much larger the payroll number looks.
Two, market income growth. Up 8.9% nominal. This is the competitive benchmark. Rise materially below it and the gap to market opens up, and employees will see it.
Three, your own company's productivity. Revenue or value added divided by average headcount, compared across two years. This is the only number that actually answers how much more you can afford to pay.
Four, national productivity. Up 6.83%. Use it to see whether you are moving faster or slower than the economy.
If number three is below number one, you have a bigger problem than pay.
What to do this quarter
Plenty of Vietnamese companies have never calculated their own productivity, because it appears in no standard report. The method is not complicated: take the year's net revenue or value added, divide by average headcount for that year, then do the same for the prior year and compare.
With that figure in hand, the pay conversation at board level changes. Instead of debating impressions of the market, you present your company's productivity growth next to three national benchmarks. A four-line table.
One caveat worth stating in the paper itself: the 2025 figures were estimates at the time of publication, not final official numbers. Cite the source and the publication date so a later reader knows how much weight to put on them.
