The median Malaysian citizen's pay rose to RM2,793 in 2024. It buys what RM2,103 bought in 2010.

A viewer asked this under a video about the cost of living, and 24 people upvoted it:

So what's the real problem that cause inflation then?

We went to check, and the first thing worth saying is that we cannot answer that question, and neither can any table the Department of Statistics publishes.

What we can do is the thing the question is really reaching for: put the pay rise and the price rise side by side, from the government's own numbers, and see which one won. That turns out to have a specific answer, and it is not the answer most people expect.

Prices rose 34.6% in fifteen years, and fell in one of them

Malaysia's Consumer Price Index is published monthly, as open data, with no gaps. This is the easy half.

Source: DOSM, Consumer Price Index (open data), dataset cpi_headline, division 'overall' · zero-based axis

Taking the average of the twelve monthly readings in each year, the index went from 100.02 in 2010 to 134.62 in 2025 — a rise of 34.6%. The steepest single year was 2017, at 3.8%. The line falls exactly once, in 2020, by 1.1%. That is the only year since 2010 in which the average Malaysian price level was lower than the year before.

Put the other way round: a ringgit earned in 2024 buys what 75.3 sen bought in 2010.

The median citizen earned RM2,793 a month, and it bought what RM2,103 bought in 2010

The pay half is harder to get at, and the difficulty is worth knowing about. Wages are not in DOSM's open-data API at all — we probed eighteen plausible dataset names and every one returned a 404. They are published once a year, as a release document, and each year has to be read on its own.

The most recent is the Salaries & Wages Survey Report 2024, released on 29 September 2025. It says this:

The median monthly salaries & wages for Malaysian citizens rose to RM2,793, up from RM2,602 in 2023, while the mean monthly salaries & wages increased to RM3,652 from RM3,441 in the previous year.

And then, two paragraphs later, it says something that almost no coverage of it repeats:

The real median monthly salaries & wages for Malaysian citizens recorded a higher growth of 5.4 per cent, rising from RM1,995 to RM2,103.

Source: DOSM, Salaries & Wages Survey Report 2024 · zero-based axis

Those are four numbers for two years, and the gap between the pairs is the whole subject. RM2,793 is what the median citizen was paid. RM2,103 is what it was worth. The difference is RM690 a month, and it is not a deduction, a tax or a fee. It is fifteen years of price rises.

DOSM does not say what "real" means. We worked it out, and it checks to the ringgit.

The release prints real wages without stating the deflator or the base year, which makes RM2,103 an unusable number — you cannot compare it to anything if you do not know what it is measured in.

So we tested it. Dividing the nominal figures by our own annual-average CPI on the 2010 = 100 base:

• RM2,793 ÷ 1.3279 = RM2,103.3, against DOSM's published RM2,103
• RM2,602 ÷ 1.3040 = RM1,995.4, against DOSM's published RM1,995
• RM3,652 ÷ 1.3279 = RM2,750.2, against DOSM's published RM2,750
• RM3,441 ÷ 1.3040 = RM2,638.8, against DOSM's published RM2,639

Four for four, to the ringgit. So DOSM's "real" wages are in 2010 ringgit, deflated by the headline CPI — established by reproduction rather than by citation, because the release never says so. It also means the CPI arithmetic underneath this article is confirmed against an independent DOSM publication.

One consequence matters if you are doing this at home: RM2,103 has already been deflated. Adjusting it for inflation a second time is the specific error this structure invites.

In the one year we can properly compare, pay beat prices

Between 2023 and 2024, median nominal pay rose 7.3% and the price level rose 1.8%. That is why the real figure rose 5.4%. Mean pay rose 6.1% nominal, 4.2% real. The fastest-rising group was low-skilled workers, up 7.4% to a mean of RM2,128, from RM1,982.

This cuts directly against the familiar line that wages have been flat for years. In 2024, on DOSM's own measure, they were not flat — they moved up faster than prices, and they moved up fastest at the bottom.

It is one year. We are comparing 2024 with 2023 because those are the two years we hold, and a two-point comparison is not a trend. Anyone who tells you what Malaysian real wages have done since 2010 is working from a series we could not find published, and we are not going to invent one to make a better-looking chart.

The average is not the middle, and the breakdowns use the average

Two averages run through this release and they are not interchangeable. The median — the person in the middle — is RM2,793. The mean is RM3,652, which sits RM859, or 31%, above it.

The distinction is not pedantry, because the breakdowns by state, by sector and by skill are all reported against the mean, and the mean is the higher number.

Source: DOSM, Salaries & Wages Survey Report 2024 · zero-based axis

Five places in Malaysia have mean pay above the national mean of RM3,652: Putrajaya, Kuala Lumpur, Selangor, Labuan and Pulau Pinang. Everywhere else in the country is below it. Read that chart as a chart of means and it is informative; read it as "what people there earn" and it is RM859 too generous before you start.

One more label that should not be dropped: every wage figure here is for Malaysian citizens. Malaysia's formal workforce includes a large non-citizen component that these numbers do not describe.

One division of the price index rose 58%. We are not going to tell you which one.

The headline index hides an enormous spread, and this is the closest the data gets to the question actually asked.

DOSM's CPI breaks into thirteen divisions. Between 2010 and 2025, while the headline rose 34.6%, one division rose 58.2% and another rose only 22.2% — and that second one actually fell in 2015, 2016 and 2020. Two households with different spending patterns did not experience the same inflation. They were not close.

We can tell you that. We cannot tell you which divisions they are, because DOSM's API returns them as bare numeric codes, 01 to 13, with no lookup table alongside them. We probed four candidate dataset names for the mapping and all four returned 404. The standard international classification would let us guess, and the behaviour of each series fits that guess neatly. A behavioural fit is not a label, and a guess printed in bold becomes a fact by the third time it is quoted. So the codes stay codes until we have the document.

The one change of documented size

If you want a candidate for "the real problem", the largest deliberate change to Malaysian household prices in the current period is in the federal budget, and it is not hidden.

The Ministry of Finance's own table puts subsidies and social assistance at RM67,360 million in 2024 (actual), RM57,062 million in 2025 (revised estimate) and RM49,004 million in 2026 (budget estimate) — a fall of RM18.4 billion, or 27%, in two years. The Ministry attributes it to "the implementation of fuel subsidy rationalisation and lower global oil prices."

The line this rests on

“Subsidies and social assistance: 67,360 → 57,062 → 49,004. RM18.4 billion off the line in two years.”

Retrieved 29 Jul 2026 · check it yourself

That is a real, large, documented shift in what the government pays toward the price of things people buy. What it is not is an explanation of the CPI. The price level rose 1.4% in 2025; nothing in these documents attributes any part of that to the subsidy change, and neither do we. Who receives the fuel subsidy, in ringgit, by income group, is a separate question with household data behind it.

The honest answer

The question asks for a single cause. The documents do not contain one, and the honest reason is that identifying a cause requires a model — of pass-through, supply shocks, policy and money — where every figure above is a measurement. No DOSM table supplies that, and we acquired no Bank Negara material to stand on instead.

What the measurements say is more specific and more useful:

Prices are 34.6% higher than in 2010. The median citizen's RM2,793 is worth RM2,103 in that money. In the one year we can compare, pay grew faster than prices, and fastest for the lowest-paid. And the average that gets quoted is RM859 above the middle.

The premise that Malaysian pay is simply losing to Malaysian prices does not survive contact with the 2024 release. What is true is narrower and harder to argue with: the level of pay is low, the erosion since 2010 is real and equal to RM690 a month at the median, and one year of catching up is one year.

The next thing to get is the missing wage years. Each one exists, in its own release, and each has to be read. Until they are, this article compares two years and says so.

---

Notes on the figures. CPI figures are annual averages we computed as the arithmetic mean of the twelve monthly index values in each calendar year, on DOSM's published base of 2010 = 100; that construction is confirmed by the deflator reproduction above. 2026 is excluded throughout because the published series covers January to June only, and a six-month average plotted as a year overstates it. All percentage changes, the 75 sen purchasing-power figure, the RM690 and RM859 gaps and the RM18.4 billion two-year fall in subsidies are our own arithmetic on the cited values. Wage figures are for Malaysian citizens and come from DOSM's annual Salaries & Wages Survey Report, not from its separate quarterly Employee Wages Statistics (Formal Sector), which has a different scope and definitions; the two are never spliced here. Real wage figures are in 2010 ringgit and have already been deflated. The state comparison is of mean pay, as the release reports it, and would not hold as stated of the median. Subsidy figures are 2024 actual, 2025 revised estimate and 2026 budget estimate, and are federal operating expenditure, not household prices.

Not verified. Every wage year before 2023 — DOSM publishes no wage series in its open-data API (eighteen dataset ids probed, all 404), so each earlier year must be read from its own release document, and we have not yet done that. The names of the CPI divisions — DOSM's API returns COICOP codes without a lookup table (four candidate dataset ids probed, all 404), so no division is named here. The CPI base year — 2010 = 100 is established in this article by reproduction, twice over, and not by a citation from a DOSM CPI release; we have also not established whether the published 1980-onward series is spliced across earlier rebasings, which is why no claim here reaches back before 2010. And the cause of Malaysian inflation, which is not in any document cited above.