The Number That Tells the Story
3,521%.
The tariff that erased
an entire industry overnight.
On April 21, 2025, the US Department of Commerce finalised anti-dumping and countervailing duties on solar panels imported from four Southeast Asian countries: Cambodia, Vietnam, Thailand, and Malaysia. The tariff rates varied significantly by country and by company. Malaysian manufacturer Jinko Solar received the lowest — 41.56%. Thai company Trina Solar received 375.19%. Vietnamese manufacturers faced rates up to 395.9%. Cambodia received 3,521.14%.
The rate for Cambodia was the highest because Cambodia's manufacturers had stopped cooperating with the US investigation — leaving trade officials to apply "adverse inference" rates based on the facts available, which assumed the worst. But the cooperation problem was symptomatic of a deeper reality: the companies running solar factories in Cambodia were Chinese companies, the factories were owned and operated by Chinese interests, the components were Chinese, the technology was Chinese, and the only thing Cambodian about the product was the return address on the shipping container. When the US Commerce Department came asking how the factories worked and where the parts came from, there was nothing the Cambodian government could credibly say in defence — because the answer would have confirmed exactly what the investigation was looking for.
$2B+
Cambodia solar exports to US in 2023 — peak of the boom
3,521%
Final US tariff rate on Cambodian solar panels — highest in modern US trade history
~$10M
Cambodia solar exports by 2025 — the collapse was total
How It Happened — The Anatomy of a Policy Arbitrage
A loophole opened.
Chinese capital flooded in.
It looked like industrialisation.
It wasn't.
To understand Cambodia's solar collapse, you need to understand what created the boom in the first place. The United States has had tariffs on Chinese-made solar panels since the Obama administration — initially in response to complaints that Chinese manufacturers were dumping panels below cost and receiving unfair government subsidies. These tariffs have been raised, extended, and reinforced by every administration since 2012. By 2022, Chinese solar panels faced tariffs that made them effectively uncompetitive in the US market.
Chinese solar manufacturers responded with a straightforward workaround: move the final assembly step to countries not covered by the China-specific tariffs. Malaysia, Vietnam, Thailand, and Cambodia all saw rapid investment in solar panel assembly facilities from Chinese companies. The formula was simple — ship Chinese-manufactured solar cells and components to a factory in Southeast Asia, assemble them into finished panels, attach a "Made in Cambodia" (or Vietnam, or Thailand, or Malaysia) label, and export to the US at the China-tariff-free rate. The panels were not meaningfully more Cambodian than a letter forwarded through a Cambodian post office is Cambodian.
Cambodia's government saw the investment statistics and the export numbers and read them as success. Solar panels became Cambodia's single largest export to the United States. Headlines described Cambodia's industrial transformation. Investment promotion materials cited the solar sector as evidence of economic diversification. None of the people writing those materials asked the question that the US Commerce Department eventually asked: what, exactly, is Cambodia adding to these panels? The answer — minimal assembly labour, a building, and a location — was not sufficient to constitute a Cambodian product under US trade law.
2012
US Imposes Tariffs on Chinese Solar
Obama administration begins imposing anti-dumping and countervailing duties on Chinese-manufactured solar panels. Chinese manufacturers begin looking for alternative export routes.
2018–22
Chinese Investment Floods Southeast Asia
Chinese solar manufacturers build assembly facilities in Cambodia, Vietnam, Thailand, and Malaysia. Cambodia's solar exports to the US begin growing rapidly. Investment statistics look spectacular. The underlying model is tariff arbitrage.
2023
The Peak — $2 Billion in Exports
Cambodian solar panels become the country's top export to the United States. Over $2 billion in annual exports. Government officials cite the sector as evidence of economic modernisation. It is the high-water mark. The investigation has already begun.
2024
The Investigation
US Commerce Department begins anti-dumping and countervailing duty investigation. Preliminary findings that Chinese companies are circumventing existing China tariffs through Southeast Asian operations. Cambodian manufacturers stop cooperating with the investigation. Preliminary tariffs announced.
Apr 2025
3,521% — The Collapse
US Commerce Department finalises 3,521.14% tariff on Cambodian solar panels. The rate set because Cambodia's manufacturers refused to cooperate — adverse inference applied. Imports from Cambodia collapse immediately. The factories close or relocate. Within months, Cambodia's solar exports approach zero.
2025
The Aftermath — Nothing Remains
Solar exports collapse to under $10 million. Workers laid off. No skills transferred. No domestic supply chain. No technology retained. Chinese companies relocate assembly operations to Laos, Indonesia, and other countries not yet covered by the tariffs. Cambodia is left with empty factories. The boom left no trace except economic statistics that no longer exist.
The 3,521% Tariff — What It Actually Means
The mechanics of how a tariff kills an industry overnight
A 3,521% tariff does not mean prices rise 35 times. It means the product becomes commercially impossible to import. A solar panel module that costs $100 to manufacture faces a $3,521 tariff — making the delivered cost to a US customer $3,621 for a $100 product. No buyer pays that. The trade stops immediately and completely.
The 3,521% rate is made up of two components: a 117.18% anti-dumping duty (for selling below fair market value) and a 3,403.96% countervailing duty (for receiving Chinese government subsidies that flowed through to the Cambodian operations). The countervailing duty rate of 3,403% is, as the lead US trade attorney described it, "among the highest rates I've ever seen in any kind of countervailing duty investigation." It reflects the magnitude of Chinese government subsidies that were found to be flowing through Cambodian facilities — subsidies so large that they made the panels commercially impossible to manufacture at comparable cost elsewhere.
The contrast with Malaysia tells the whole story. Jinko Solar's Malaysian operations received a tariff of 41.56% — high, but survivable with price adjustments. Malaysian operations had more genuine local content, more cooperation with the investigation, and were seen as more authentically Malaysian than Cambodian operations were Cambodian. The tariff rate is, in this sense, a quantitative measure of how real the industrialisation actually was.
The Framework — Three Types of Industrialisation
Fake. Real but borrowed.
Sovereign.
Letters 72, 73, and 74 together constitute a case study series in what developing country industrialisation actually looks like at different levels of depth. The Cambodia solar collapse, India's frozen fry revolution, and Indonesia's nickel gambit represent three fundamentally different approaches — and three fundamentally different levels of durability.
Type 1 — Policy Arbitrage
Fake industrialisation
🇰🇭 Cambodia Solar
- Mechanism: Exploit a policy gap — a tariff loophole, a trade preference, a regulatory exemption — to route foreign-made goods through a domestic address.
- What it looks like: Investment statistics surge. Export figures appear. Employment is created in assembly. Government officials declare industrial success.
- What it is: A forwarding address. The value, the technology, the supply chain, and the profits belong to the foreign company exploiting the loophole.
- Durability: Zero. Policy gaps close. Trade investigators catch up. When the loophole closes, the industry disappears instantly because nothing real was ever built.
- Legacy: Empty factories. Unemployed workers with assembly-line skills that don't transfer. Export statistics that don't repeat. No technology, no supply chain, no industry.
Type 2 — Ecosystem Integration
Real but borrowed industrialisation
🇮🇳 India Fries · 🇻🇳 Vietnam Electronics
- Mechanism: Leverage a genuine comparative advantage — cost, geography, agricultural conditions, skilled labour — to attract foreign investment into industries that require real local capability.
- What it looks like: Contract farming networks. Technology transfer. Local supplier development. Skills accumulation. Infrastructure investment. Cold chains, processing plants, logistics.
- What it is: Real production of real goods using real local inputs and real local capabilities. The foreign partner captures the brand and the margin, but genuine local economic value is created.
- Durability: Medium to high. Dependent on cost competitiveness and on building local capability faster than costs rise. Vietnam's challenge: moving up the value chain before labour costs close the gap with China.
- Legacy: Skills, supply chains, infrastructure, accumulated knowledge. When the foreign partner leaves or the model evolves, something real remains that can be built upon.
Type 3 — Resource Leverage
Sovereign industrialisation
🇮🇩 Indonesia Nickel
- Mechanism: Control a scarce resource that the world needs and use that control to force the value-adding processing to happen domestically.
- What it looks like: Export bans. WTO challenges. Industrial parks in remote regions. Massive foreign investment in processing facilities. Battery supply chain integration.
- What it is: Genuine leverage based on irreplaceable endowment. The world cannot substitute the nickel. It must come to Indonesia for the processing or do without.
- Durability: High while the resource is scarce and the processing advantage holds — but vulnerable to technology shifts (LFP batteries not needing nickel) and to the question of who captures the processing margin (so far, China).
- Legacy: Industrial infrastructure, technical workforce, processing capacity, supply chain position. The question is whether Indonesia can convert processing leverage into true manufacturing sovereignty.
The Policy Gap Test — How to Tell the Difference
Before celebrating an industrial boom,
ask five questions.
The Cambodia solar collapse was predictable — not because trade investigators are particularly clever, but because the model failed every test of genuine industrialisation. The questions that would have identified it as policy arbitrage rather than real industry are not sophisticated. They are the questions any honest analyst should ask when a developing country reports unexpected export growth in a sector where it has no prior history.
Where do the inputs come from?
Chinese cells, Chinese components, Chinese equipment. Nothing Cambodian except the building and the labour.
Potatoes grown by Gujarati contract farmers on Indian soil using Indian-developed seed varieties. Genuinely Indian inputs.
What skills does the workforce acquire?
Panel assembly. Skill that transfers nowhere — not to other industries, not to higher-value roles in the same industry.
Agronomy, processing, cold chain logistics, quality control. Skills that build a food processing industry capability.
What happens if the policy changes?
Everything stops. The Chinese companies move to the next loophole country. Cambodia is left with empty factories.
The farming network, the processing plants, the cold chain infrastructure, and the agricultural knowledge all remain and can serve other markets.
Does local productivity underpin the competitiveness?
No. Competitiveness comes from Chinese subsidies routed through Cambodia — not from any Cambodian production advantage.
Yes. Gujarat's soil conditions, lower labour costs, shorter shipping to Asia, and contract farming efficiency all create genuine cost advantages.
Who captures the profit?
Chinese companies. Cambodia captures labour wages and some tax revenue. The margin, the technology, and the brand all leave the country.
HyFun Foods is an Indian company. The processing margin stays in India. Farmers earn premium prices. Local employment is created at multiple levels.
The Honest Read — Cambodia Isn't Alone
Cambodia is the most extreme example, but it is not the only one. Vietnam's electronics sector — which represents 30%+ of Vietnamese exports — passes more of the industrialisation tests than Cambodia's solar did, but still faces the fundamental question of whether the value is being created by Vietnamese companies or simply processed by Vietnamese workers in foreign-owned factories. The distinction matters for long-term development, even if both models produce export statistics and employment in the short term.
The harder truth is that distinguishing real from fake industrialisation requires asking uncomfortable questions about investment flows that governments receiving investment are not incentivised to ask. When Chinese companies invest in Cambodia, the Cambodian government sees foreign investment, employment, and export revenues. Asking whether the industry is real requires a kind of analytical honesty about what is being built that conflicts with the political desire to present economic success stories. Cambodia was not uniquely naive. It was the most extreme case of a common pattern.
The deeper lesson is about what policy gaps actually are. A policy gap — a tariff loophole, a trade preference, a regulatory exemption — can attract foreign investment. But the investment attracted by a policy gap is investment that will leave when the gap closes. Durable industrialisation requires a comparative advantage that exists independently of the policy environment. Gujarat's soil produces better processing potatoes than most of the world regardless of US tariff policy. Indonesian nickel exists in the ground regardless of WTO rulings. Cambodian solar assembly existed only because of a tariff structure that was always going to be challenged.
The Series — Three Letters, One Framework
What durable growth actually looks like.
The NGE View
The verdict.
What We Believe
✓
Policy gaps are not competitive advantages. The moment a tariff loophole closes, a trade preference expires, or an anti-circumvention investigation concludes, any industry built on that gap disappears. Cambodia's solar boom was not an industrial transformation — it was a temporary redirection of Chinese export capacity through a Cambodian address. The trade statistics looked real. The export revenues were real. The underlying industry was not. No government building economic strategy on a policy gap has built durable prosperity. The gap always closes.
✓
The five questions matter more than the export statistics. Where do the inputs come from? What skills does the workforce acquire? What happens if the policy changes? Does local productivity underpin the competitiveness? Who captures the profit? If the honest answers to those questions point to foreign inputs, transferable assembly skills, policy-dependent competitiveness, and foreign profit capture — the industry is not real, regardless of what the export figures say. Cambodia passed none of these tests. India's frozen fry industry passes most of them. Indonesia's nickel industry passes some. The tests are not complicated. They are just uncomfortable to apply to investment your government is celebrating.
✓
The most durable growth comes from ecosystem-driven models that integrate infrastructure, skills, and local capabilities. This is not a new insight — it is the lesson of every successful industrial policy from South Korea's chaebols to Taiwan's semiconductor industry to Germany's Mittelstand. What makes India's fry industry resilient is not the export statistics but the accumulated agricultural knowledge in Gujarat, the cold chain infrastructure, the tissue culture seed systems, the HyFun Foods company that is genuinely Indian. What makes Indonesia's nickel position resilient is not the processing facilities but the irreplaceable resource underneath them. What made Cambodia's solar position fragile was the absence of any of these things. The forwarding address is not an industry. The ecosystem is.
NGE · A Futuristic Investment Letter
Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Written from first principles. Not consensus. Not noise.
— Pawan Bhatia · NextGen Economics · Bangalore, India