Trump’s Polysilicon Tariffs Are Making Things Worse

The administration recently made its homegrown solar panel base look bigger by doing two contradictory things at once. On August 6, it added a 15% tariff and price floors to solar feedstock called polysilicon and its products — wafers, cells and modules. At the same time, the homegrown manufacturing idea is losing its last 30% federal tax credit when the Inflation Reduction Act (IRA) 45X advanced manufacturing credit phased out on July 4, 2025.

The first order of business is to clarify the real question at stake. This is not about whether machines will replace workers. It is about who captures the gains if they do. The second order is to find out the trade-offs. Those who benefit and those who lose from trade protectionism. To be sure, the case for supply-chain diversification has merit. China’s dominance of the solar industry is a product of decades of state-backed overcapacity and below-cost pricing, not merely superior efficiency. An American capacity in critical materials is defensible on national-security grounds, especially where polysilicon feeds into the semiconductor supply chain as well. The Supreme Court’s recent invalidation of earlier levies imposed under the International Emergency Economic Powers Act has also narrowed the tools available to the executive, making Section 232 tariffs the legally available fallback. Yet tariffs are a blunt and self-defeating way to address the problem. The historic record is dismaying. A study by the Solar Energy Industries Association found that the 2018 to 2021 safeguard tariffs imposed 62,000 American jobs, $19 billion in investment and just 2,000 manufacturing positions. The administration is simultaneously raising the cost of solar through tariffs and removing the financial incentive to install it.

The deeper incoherence is political. The One Big Beautiful Bill Act signed into law on July 4, 2025, eliminated the 30% federal tax credit for residential solar and accelerated the phase-out of credits for commercial and utility-scale projects. The residential segment, forecast to contract by 19% in 2026, is expected to fall to its lowest installation level since 2020, says Canary Media. BloombergNEF analysts estimate it will take more than a decade for the industry to match the installation records set in 2023. Total solar installations in 2025 fell 14% to 43.2GW. The utility-scale sector shrank by 16%, with nearly all the decline crammed into the fourth quarter as developers delayed operations to preserve remaining tax-credit safe harbours. For all the rhetoric about energy independence, the American solar manufacturing base remains small, incomplete and unprofitable without trade support. The country now produces every major component in the solar supply chain — polysilicon, ingots, wafers, cells and modules - for the first time since SolarWorld closed its Oregon plant in 2013. Domestic module capacity has reached approximately 65GW, enough to cover the Solar Energy Industries Association’s (SEIA) expectation of 44GW of installations in 2026. But cell capacity lags at a mere 3.2GW. Solar glass is not produced domestically at all. First Solar, the only firm with a fully homegrown supply chain, uses thin-film cadmium-telluride technology rather than crystalline silicon, which accounts for most utility-scale installations. Even where factories exist, solar-grade polysilicon production in America is not profitable at current prices. The tariffs are not protecting a thriving industry. They are keeping an uncompetitive one alive.

The incentive structure is telling. Tariffs concentrate benefits - they protect a small number of factories and their workers. They diffuse costs - they raise prices for every installer, developer, utility and homeowner across the country. The politics may prove nastier than the economics. A handful of well-placed manufacturers gain a rent; millions of consumers pay a little more and do not notice, or blame their local installer rather than the White House. This is not the way to build industrial competitiveness. It is the way to build a permanent constituency for inefficiency.

To be sure, the case for supply-chain diversification has merit. China’s dominance of the solar industry is a product of decades of state-backed overcapacity and below-cost pricing, not merely superior efficiency. An American capacity in critical materials is defensible on national-security grounds, especially where polysilicon feeds into the semiconductor supply chain as well. The Supreme Court’s recent invalidation of earlier levies imposed under the International Emergency Economic Powers Act has also narrowed the tools available to the executive, making Section 232 tariffs the legally available fallback. Yet tariffs are a blunt and self-defeating way to address the problem. The historical record is dismaying. A study by the Solar Energy Industries Association found that the 2018 to 2021 safeguard tariffs imposed $19 billion in private investment and just 2,000 manufacturing positions. FTI Consulting estimates that new antidumping duties on Southeast Asian solar imports, imposed in April 2025, could raise utility-scale project costs by nearly 15% and put 14GW of capacity at risk. The academic evidence is worse: research by Todd Gerarden and colleagues at the Kennedy School of Government found that consumer costs from past solar tariffs totalled $5.5bn, vastly exceeding the benefits to domestic producers and tariff revenues combined.

The administration’s goal is to build a domestic solar industry that can compete without protection. The first step is to stop subsidising the idea that it can. The problem is that it is unlikely for American solar to become commercially viable at current prices. Instead, the administration has simultaneously raised the cost of solar through tariffs and removed the financial incentive to install it. The administration’s preferred instruments would have been the Inflation Reduction Act’s 45X advanced manufacturing credit, which paid up to $0.04 per watt for domestic cell production. That was a better instrument than tariffs because it rewarded output rather than excluding competition. Its phase-out, combined with the residential credit’s elimination, removes precisely the demand-side pull that would make domestic supply economically meaningful. The administration could have retained deployment credits, tightened the rules on foreign-entity ownership of manufacturing capacity and let the market sort out which American factories were efficient enough to survive. Instead, it has chosen the path that is easiest politically and worst economically: raise prices, protect producers and blame China.

The tariffs will not end Chinese dominance. They will simply make American solar more expensive, slow deployment further and funnel billions in consumer transfers to a handful of well-connected manufacturers. If the aim is to build a solar industry that can compete without protection, the first step is to stop subsidising the idea that it can. The problem is that it is unlikely for American solar to become commercially viable at current prices.