Fact-Checking 'Taiwan Is Filthy Rich Yet Its People Feel Nothing': How Data Overthrew My Assumptions
While scrolling through videos, a title caught my eye: "Why Taiwan is filthy rich, yet its people feel nothing."
My initial reaction was an instinctive nod. TSMC is printing money and GDP keeps hitting record highs, yet nobody around me feels life is getting better—isn't that obvious? But then I stopped myself. The phrase "it's obvious" is usually the cue to go check actual data. So I spent an afternoon digging through raw figures from the Directorate-General of Budget, Accounting and Statistics (DGBAS), the Ministry of the Interior (MOI), and the National Audit Office (NAO).
The conclusion: The premise of the video holds true, but the mechanism behind why it holds true is completely different from what I had assumed. In fact, two core assumptions I held were thoroughly disproven by the data. Here is the breakdown.
1. How Rich Is Taiwan? Even More Exaggerated Than I Thought¶
First, let's verify the first half of the statement. Taiwan's economic growth reached 8.63% in 2025, a 15-year high. I knew that. What I didn't know was that on August 14, 2026, DGBAS revised the growth rate for that year up to 11.05%—a 39-year high.
In 2024, Taiwan's per capita GDP surpassed Japan's for the first time, and in 2025 it overtook South Korea—a reversal after South Korea had pulled ahead in 2003, 22 years ago. DGBAS projects per capita GDP will reach $45,332 in 2026.
What best demonstrates "having more money than can be spent" is excess savings—the surplus national savings left uninvested domestically (based on DGBAS forecasts from August 14, 2026):
| Year | Excess Savings | Share of GDP |
|---|---|---|
| 2020–2023 | ~NT$3 Trillion | — |
| 2025 | NT$5.70 Trillion | 19.84% |
| 2026 (Est.) | NT$8.47 Trillion | 25.13% |
| 2027 (Est.) | NT$11.00 Trillion | 29.49% |
Sailing from 3 trillion to 11 trillion within three years means "filthy rich" isn't hyperbole—it is literal.
Three caveats apply:
1. Exchange rate effects: The New Taiwan Dollar appreciated toward 28 against the USD in 2025, inflating USD-denominated figures (DGBAS cited exchange rates as a major factor pushing per capita GDP past $40,000).
2. Population decline: A smaller denominator elevates per capita figures.
3. High industrial concentration: Analysts note that excluding semiconductors drops the growth rate below 2%.
Even after applying these discounts, the first half of the claim stands firm.
2. Disproven Assumption #1: "Wages Haven't Risen"¶
I originally believed the explanation was "twenty years of real wage stagnation." I had heard this narrative for years and accepted it as fact.
The numbers revealed otherwise—at least for the past two years. DGBAS statistics published in February 2026 for the full year 2025 show:
| Indicator | 2025 | YoY Change |
|---|---|---|
| Avg. Regular Wage | NT$47,884 | +3.09% (Largest gain in 26 years) |
| Median Regular Wage | NT$38,406 | +3.03% |
| Real Regular Wage | — | +1.40% (5-year high) |
| Real Total Wage | — | +2.21% (7-year high) |
Adjusted for inflation, wages recorded genuine positive growth at rates rarely seen in recent years. Claiming that "wages haven't moved at all" is factually incorrect for 2025.
However, the same report contains another revealing metric: 69.74% of employed persons earned below the average wage, the highest proportion since tracking began in 2020. DGBAS explicitly attributed this to tech sector compensation pulling up the mathematical mean. Industry disparities are stark—Financial & Insurance services averaged NT$70,997 and Info-Tech averaged NT$69,595, while Accommodation & Food Services stood at just NT$34,961.
Lesson one: "Average wage gains" and "widespread wage increases" are two completely different things. In hyper-concentrated economies, they can even diverge. A rising mean can itself be evidence that the majority is falling behind.
3. The Real Answer: Who Received the Created Value?¶
If wages rose, where does the widespread sense of indifference come from? The answer lies in income distribution.
Viewed through distribution, GDP breaks down into four components: Compensation of Employees, Operating Surplus, Consumption of Fixed Capital (Depreciation), and Net Taxes on Production and Imports. An audit report by the National Audit Office analyzing incremental GDP growth from 2017 to 2021 provides the clearest picture:
| Distribution Component | Increment (NT$ 100M) | Share of Increment |
|---|---|---|
| Operating Surplus | 16,732 | 44.96% |
| Compensation of Employees | 13,454 | 36.15% |
| Consumption of Fixed Capital | 6,594 | 17.72% |
| Net Taxes on Production & Imports | 434 | 1.17% |
| Total GDP Increment | 37,214 | 100% |
Out of the NT$3.72 trillion created during those five years, wage earners captured 36.15%, while corporate profits captured 44.96%.
The long-term structural trend is even clearer: Employee compensation as a share of GDP declined from 51.06% in 1992 to 43.03% in 2021—its lowest level since 1992. Meanwhile, private-sector employees as a proportion of total employed workers climbed from 56.23% to 71.14%. More people are sharing the pie, but the labor slice of the pie has shrunk. This is the National Audit Office's own official explanation for popular indifference.
International comparisons underscore how unusual this is. In 2020, Taiwan's GDP grew +3.39% while the US contracted -2.77% and Japan contracted -4.28%. Yet in that year:
| 2020 Metric | Taiwan | United States | Japan |
|---|---|---|---|
| Employee Compensation / GDP | 44.7% | 54.5% | 52.7% |
Taiwan expanded while others contracted, yet Taiwanese workers received 8 to 10 percentage points less of total economic output.
4. Disproven Assumption #2: "Tech Workers Took All the Money"¶
Following the logic above, my next hypothesis was: High-earning tech employees must be hoarding all the wage growth.
This assumption was also wrong, and in a fascinating way.
While the ICT sector contributed 35.8% of incremental GDP growth, with ICT manufacturing accounting for 70.36% of industrial value-add, breaking down ICT's internal value distribution reveals:
| ICT Sector Distribution Component | Share of ICT Increment |
|---|---|
| Operating Surplus | 48.33% |
| Compensation of Employees | 27.66% |
| Consumption of Fixed Capital | 24.22% |
| Net Taxes on Production & Imports | -0.22% |
The key lies in that 24.22%. Nearly a quarter of ICT's newly generated output converts directly into equipment depreciation—reflecting the capital-intensive nature of advanced semiconductor fabrication plants. Combined with 48.3% flowing into corporate operating surplus, only 27.66% remains for employee compensation. This is noticeably lower than the all-industry average of 36.15% and the service sector's 38.41%.
So the correct conclusion isn't "tech employees took all the money," but rather: The structural value distribution of the tech industry directs less than 30% to wages, and Taiwan's economic growth engine happens to be the exact industry type that distributes the lowest share to labor.
This distinction matters. The former would imply an income tax distribution issue; the latter is an industrial structure reality that tax adjustments alone will not fundamentally reshape.
Where did that 45% corporate surplus go?
1. Retained corporate earnings (driving excess national savings past NT$11 trillion).
2. Dividend payouts (foreign institutional ownership of TSMC stands near ~69%).
3. Statistical accounting nuances (earnings from 1.27 million self-employed individuals and 470,000 business owners are classified as operating surplus under standard national accounts framework).
Additionally, money is not fleeing overseas. Taiwan's GNI has consistently exceeded GDP in recent years (2025 GNI reached NT$29.53T vs GDP NT$28.72T). Net foreign factor income remains positive (+NT$813.7 billion). Wealth generated abroad exceeds foreign income extracted from Taiwan. Capital stays in Taiwan, primarily held as corporate equity and financial assets.
5. The Unexpected Finding: Wealth Inequality in Taiwan Isn't as Extreme as Believed¶
This was the single most counterintuitive discovery of my investigation.
I expected the data to conclude with "Taiwan suffers from catastrophic wealth inequality." Surface numbers appear to support this: DGBAS's 2024 Family Income and Expenditure Survey shows the disposable income gap between the top and bottom 20% of households widened to 6.17x—a 14-year high. Household wealth disparities reached 66.9x (though that wealth survey reflects 2021 data released in April 2024, using different methodologies).
However, introducing global context alters the picture significantly. Taiwan's household wealth Gini coefficient stands at 0.606, lower than Australia (0.611), the UK (0.620), France (0.676), Japan (0.678), and Germany (0.727). A 66.9x wealth gap ratio sounds shocking until compared to Australia (93.1x), the UK (109.5x), South Korea (140.1x), or France (627.4x).
Income Gini figures tell a similar story. Taiwan's unadjusted domestic household Gini of 0.341 drops to 0.270 when converted to standard OECD equivalized scales. Comparing unadjusted 0.341 against the UK (0.387) or US (0.465) overstates Taiwan's inequality.
Furthermore, lower-income brackets fare relatively better in terms of asset safety: Taiwan's bottom 20% hold 0.94% of national wealth, compared to 0.11%–0.67% across South Korea, Australia, France, and the UK, driven by high household savings habits and fewer negative-net-worth households.
Thus, the proper diagnosis is not "Taiwan is exceptionally unequal," but rather: Taiwan's growth flows heavily toward capital over labor, leaving median households unable to keep pace with macroeconomic expansion.
6. Housing Prices: Consuming Wage Gains¶
Housing represents a major driver of public indifference.
In Q1 2026, Taiwan's national mortgage burden ratio reached 41.46%, with a price-to-income ratio of 9.47x. The Ministry of the Interior classifies ratios below 30% as reasonably affordable, while 40%–50% represents a heavy burden. Taipei City recorded a 64.41% mortgage burden and a 14.72x price-to-income ratio.
When regular wages rise 3% while housing prices remain pegged above 9x annual income, real purchasing power gains get largely absorbed.
Yet my initial framing required adjustment here as well: Homeownership in Taiwan stands at 83.95%, while renting households account for only 7.62%. On paper, most Taiwanese households are asset-side beneficiaries of property appreciation rather than victims. Housing inflation acts as an intergenerational wealth transfer from non-owners to existing owners, rather than a total loss of national wealth.
The real macroeconomic cost lies elsewhere: Capital is locked up in land rather than productive investments. Private consumption accounts for only ~43% of Taiwan's GDP, compared to Japan (53.1%), South Korea (48.5%), the US (67.9%), and the OECD average (~60%).
In short: Taiwan has NT$11 trillion in excess savings sitting idle, while private consumption shares linger near global lows. Money is abundant, but velocity is constrained.
7. Key Takeaways¶
If I were to rewrite the video title based on empirical data, it would read: "Taiwan is remarkably wealthy, but growth channels fail to reach median households."
Four principles worth remembering:
1. Mean vs. Median: Means and medians tell opposing stories when growth is concentrated. Always check which metric is being cited.
2. Incremental vs. Stock Distribution: Incremental value distribution explains sentiment far better than stock distribution. People experience rate-of-change, not static levels.
3. Cross-Country Context & Standards: Raw inequality metrics are virtually meaningless without international benchmarks.
4. Public Data Is Readily Available: All numbers cited come directly from public DGBAS, MOI, and NAO reports. The answers were always there; they just required digging.
Comments
Loading comments…
Leave a Comment