Bloomberg Opinion’s climate columnist visited Michigan, the former heart of the solar industry, and China to learn how good, old-fashioned capitalism won out.
It all starts with a crystal.
To make the solar cells that are projected to become the world’s biggest source of electricity by 2031, you first melt down sand until it looks like chunks of graphite. Next, you refine it until impurities have been reduced to just one atom out of every 100 million — a form of elemental silicon known as polysilicon. It’s so vital to the production of solar panels that it can be likened to crude oil’s role in making gasoline. The polysilicon is then drawn out into a vast crystal, resembling a Jeff Koons steel sculpture of a sausage, before being sliced into salami-thin wafers. These are then treated, printed with electrodes, and finally sandwiched between glass.
The basic process has changed little since the first cell was invented in 1954 by scientists at Bell Laboratories in New Jersey exploring whether silicon could be used to power computer processors. “It may mark the beginning of a new era,” The New York Times wrote at the time in a front-page article announcing the discovery, “leading eventually to the realization of one of mankind’s most cherished dreams — the harnessing of the almost limitless energy of the sun for the uses of civilization.”
The seven decades since tell the remarkable story of how America squandered its invention of solar photovoltaics, or PV, to the point where it will never recover. As recently as 2010, a small town in central Michigan was the world’s biggest producer of solar polysilicon. Nowadays, the US is barely in the game, and more than 90% of the total comes from China. That country’s clean-technology exports “threaten to significantly harm American workers, businesses and communities,” President Joe Biden said May 14, announcing 50% tariffs on Chinese solar cells.
Washington blames China’s dominance of the solar industry on what are routinely dubbed “unfair trade practices.” But that’s just a comforting myth. China’s edge doesn’t come from a conspiratorial plot hatched by an authoritarian government. It hasn’t been driven by state-owned manufacturers, subsidized loans to factories, tariffs on imported modules or theft of foreign technological expertise. Instead, it’s come from private businesses convinced of a bright future, investing aggressively and luring global talent to a booming industry — exactly the entrepreneurial mix that made the US an industrial powerhouse.
The fall of America as a solar superpower is a tragedy of errors where myopic corporate leadership, timid financing, oligopolistic complacency and policy chaos allowed the US and Europe to neglect their own clean-tech industries. That left a yawning gap that was filled by Chinese start-ups, sprouting like saplings in a forest clearing. If rich democracies are playing to win the clean technology revolution, they need to learn the lessons of what went wrong, rather than just comfort themselves with fairy tales.
To understand what happened, I visited two places: Hemlock, Michigan, a tiny community of 1,408 people that used to produce about one-quarter of the world’s PV-grade polysilicon, and Leshan, China, which is now home to some of the world’s biggest polysilicon factories. The similarities and differences between the towns tell the story of how the US won the 20th century’s technological battle — and how it risks losing its way in the decades ahead.
Visiting the City in the Woods
If you own a mobile phone, a computer, a car, or a home appliance, it’s likely there’s a little bit of Hemlock in your home right now. Hemlock Semiconductor Corp. produces about one-third of the world’s chip-grade polysilicon, which finds its way into almost every electronic device on the planet. Solar polysilicon is simply the poor cousin of the stuff computer chips are made from: While impurities of one part in 100 million are considered acceptable for solar panels, microprocessors need to be pure to as much as one part in 10 trillion.
The constant stream of chemical tanker trucks going to and from the plant is the only sign that a vital node of the global economy is hidden among Hemlock’s soybean, corn and blueberry fields, dotted with red barns, clapboard homes and flagpoles. Two hours north of Detroit and inland from Lake Huron, its main street hosts a discount store, a coin-op laundry, a Ford dealership, a vet clinic, a handful of liquor stores and chain restaurants, a tavern and very little else.
It’s about as close to Middle America as you can get. Former President Donald Trump won surrounding Saginaw County, long a Democratic stronghold, by 1,074 votes in the 2016 presidential election. Four years later, Biden took it back on a margin of 303. Hemlock is part of Michigan’s 8th Congressional District, the pivot point of the entire US House of Representatives. Currently, exactly 217 seats are more Democrat-leaning, and 217 more Republican-leaning.
Hemlock Semiconductor keeps a low profile in this rural area. Driving past stands of the hemlock pines that give the town its name, you’re almost at the gate of the roughly 800-acre (three-square-kilometer) site before you notice its distillation towers, warehouses and a low industrial hum. The company turned down Bloomberg’s requests to visit the facility or interview executives.
People in town appreciate the semiconductor factory as a provider of more than 1,300 jobs, alongside funding for the town fair and school board — even if it’s always been a bit opaque. Katherine Ellison, a local historian, grew up in the 1980s two roads across from the plant, nicknamed “The City in the Woods” by her friends. “You’d kind of stumble upon this huge, lit-up structure at night,” she told me. “People who weren’t from here would ask, “What is that?”
Operations started long before solar power was taken seriously. The year was 1961, a time when the founders of Intel Corp. were looking at using polysilicon to build the first integrated circuits for use in the Apollo space program. That seemed an ideal business for Dow Corning, a long-standing joint venture between Dow Chemical Co. and Corning Inc. specializing in silicon-based chemicals such as glues, sealants and breast implants.
Dow had established itself in the 1890s in nearby Midland to take advantage of rich underground deposits of brine that could be refined into useful chemicals. The trains rattling back and forth there upset the delicate polysilicon purification process, so a new plant was established on isolated farmland in Hemlock, 14 miles (22.5 kilometers) to the south.
It’s never been an easy business. Moore’s Law — the celebrated rule of innovation that turned computers from room-sized, costly devices in the 1950s into affordable microchips that can sit on a fingertip — also required constant reductions in the costs and volumes of polysilicon in use, making it hard for companies to make consistent profits. “The purer it is, the less you need,” Denise Beachy, Hemlock’s president from 2014 to 2016, told me.
As far back as 1984, a study for the US Department of Energy noted that Hemlock was “an old, high-cost plant” where Dow Corning was “reluctant to invest.” With the venture unwilling to spend money, fresh capital was brought in the same year by selling about one-third of the factory’s equity to Japan’s Shin-Etsu Handotai Co. and Mitsubishi Materials Corp.
Things started to change around the year 2000, as rising concerns about climate change coincided with a surge in oil prices and the prospect of subsidies for renewables. Solar panels were traditionally so costly they were only used for highly specialized applications such as space probes, as well as watches and pocket calculators that only sip power. Suddenly in the early 2000s, solar started to look like a competitive way of producing energy.
As a result, PV-grade polysilicon — made until then from material rejected by chipmakers — seemed like it might become a valuable commodity in its own right. Almost overnight, it went from a backwater to a boom industry. The growth has yet to stop. Since 2005, annual installations of solar panels have increased at an average annual rate of about 44%. This year, the capacity of new modules installed globally every three days is roughly equivalent to what existed in the entire world at the end of 2005.
Hemlock initially surfed this wave. In 2005, it announced a $400 million to $500 million plan to increase production at the plant by half. Eighteen months later, it promised $1 billion more to add a further 90%. One more billion was announced amid the 2008 financial crisis, along with yet another $1.2 billion for a separate plant in Clarksville, Tennessee.
Those numbers sound big — but they were insufficient to keep up with demand.
There are a few reasons for that. First, Hemlock was owned by a joint venture between two American and two Japanese chemical companies, which between them produce everything from fiber optic cables to smartphone glass, plastics to insecticides, pill casings to machine tools and gold bullion. Such setups are notorious for complexity, which can undermine their ability to adapt quickly to changing conditions. Any fresh spending needed to get signed off by four corporate boards, none of whom saw solar poly as a priority.
Making matters worse was the fact that, when solar power started to take off in the late 1990s, Hemlock’s main shareholder Dow Corning was in the middle of a decade of bankruptcy protection — the result of lawsuits from women who claimed to have been harmed by its silicone breast implants.
Another factor was energy. As much of 40% of the cost of producing polysilicon is power, and the Hemlock factory is the biggest single-site consumer of electricity in Michigan — a remarkable statistic, when you consider the state also includes the immense General Motors Co. and Ford Motor Co. factories in Detroit.
Local electricity costs are relatively high. The 2008 expansion in Hemlock only went ahead after the state’s governor Jennifer Granholm — now President Biden’s secretary of energy — signed a bill giving the facility tax credits to protect it from electricity price hikes. Clarksville was proposed for a new plant because of its access to cheap power from the Tennessee Valley Authority, a New Deal-era hydroelectric project.
Beyond all that, though, the reluctance to invest more aggressively was driven by the conviction that polysilicon was, and would always be, a cozy oligopoly. Until the mid-2000s, the raw material for all the chips and solar panels on the planet was produced at just 10 facilities in the US, Europe and Japan. They were under the control of seven companies, of which Hemlock was comfortably the largest. This provided the kind of cartel-like price security enjoyed by the Organization of the Petroleum Exporting Countries (OPEC).
The solar panel-makers who depended on the seven companies for their polysilicon hated the situation — as did anyone who wanted to see the costs of solar power fall and its scale increase. “The supply of feedstock has lagged far behind the demands of the industry,” lamented one 2007 report on the sector, which blamed the shortage on oligopolistic behavior by big producers. As a result, all hope that solar power might help the world avert disastrous global warming seemed futile. In an influential 2006 report for the UK government, the economist Nicholas Stern predicted it would take decades for renewable power to become competitive with fossil fuels.
This didn’t bother the polysilicon producers, who were running at full capacity and enjoying scarcity that allowed them to drive up prices for consumers. The result was a fatal complacency. “Hemlock is extraordinarily profitable,” Jim Flaws, Corning’s chief financial officer, boasted in a 2009 call with investors.
Things were about to change drastically.
From Selling Fish to Fuel
In China’s southwestern Sichuan province in the mid-2000s, Liu Hanyuan was looking for new investment opportunities. He was born to a peasant family in Meishan, a small city on the broad banks of the Min river. After leaving school in the early 1980s — at a time when Deng Xiaoping’s reforms were starting to allow entrepreneurship in the state-controlled agriculture sector — he invented a mesh cage that could be suspended in fast-flowing water to farm fish and borrowed $69 (500 yuan) from his father to commercialize it.
The technique was a success. Looking to the next opportunity, Liu started producing fish food in a kitchen hand-grinder to sell to other farmers. That did even better: There was a huge market for such pellets as aquaculture production grew at double-digit rates throughout the 1980s and 1990s to meet China’s seemingly insatiable appetite for seafood. By 2002, Forbes magazine named Liu the country’s ninth-richest person, and he became a regular delegate to the Chinese People’s Political Consultative Conference, a quasi-parliament composed of the politically-connected great and good. His company, Tongwei Co., posted $1.5 billion (10 billion yuan) in revenue in 2008.
But a slowdown was coming. By the mid-2000s, almost all the water usable for fish farms had been exploited, and the newly-minted tycoon had to find alternative sources of growth. One happened to be downstream from Liu’s hometown, in Leshan.
Like Hemlock, Leshan lies atop a prehistoric sea, making it rich in brine and a natural home for a chemicals industry. Unlike Hemlock, it is blessed with a superabundance of cheap electricity. Sichuan is where the tributaries of the Yangtze tumble down from the foothills of the Himalayas in a precipitous drop, before starting a meander to the sea. Leshan itself, whose population of 1.5 million makes it a small town in Chinese terms, is home to one of the world’s biggest Buddha statues. As tall as a 17-storey building, the seated figure was carved out of a cliff in the 8th century by monks, hoping it would protect river craft from the treacherous rapids where the Min and Dadu rivers mingle. The Three Gorges Dam, the world’s biggest hydroelectric project, is further downstream. It’s a lively but ramshackle city, steaming in Sichuan’s humid, rainy summer heat, with scarcely a glimpse of the sun on which Leshan’s solar industry depends.
In Germany, the most prominent solar tycoon had other ideas. In just over a decade, Frank Asbeck had capitalized on Berlin’s renewable subsidies to build his manufacturer, SolarWorld AG, into a giant. That success turned him into a billionaire with a talent for self-promotion. He signed up Larry Hagman, the archetype of a Texas oilman in the 1980s soap opera Dallas, as a spokesman advocating for clean power. At one point, as GM circled bankruptcy in 2008, Asbeck even offered to buy its European Opel brand.
The chaos in the industry put Asbeck’s empire into jeopardy. SolarWorld was manufacturing its own panels at plants in Germany, South Korea, California and Oregon, at prices well above rival products coming out of China. But his US-based facilities gave Asbeck a card to play.
That allowed him to approach the Department of Commerce in October 2011 with a claim that China was not just selling cheap solar panels but dumping them on overseas markets at costs below what they were charging domestically. The US found in favor of SolarWorld six months later and imposed duties on China-made panels ranging as high as 250%. It would be the first of several waves of trade restrictions imposed against Chinese photovoltaics.
Claims of dumping are contentious and hugely consequential. They’re often brought by oligopolists who have had their comfortable hegemony disrupted by cheaper foreign rivals. If they win, they will gain government protection against their own inability to compete. Their domestic customers will typically take the other side of the debate and can be equally cynical: Buyers of solar panels simply want the cheapest modules they can get, regardless of the effect on local jobs and supply chains.
It’s conventional for insiders to act as if such decisions are made on sound objective criteria, but in truth it’s almost always a political mess built on a shaky foundation of low-quality data. When reviewing antidumping cases, “economists overwhelmingly reach the conclusion that the investigations are distorted and biased” in favor of manufacturers, according to one 2016 study of the solar dispute.
Meanwhile, the core questions are often almost impossible to answer. Is Tongwei’s cheap electricity from a state-owned utility a form of government subsidy? What about Hemlock’s tax credits protecting it from high power prices? Chinese businesses can often get cheap land in industrial parks, something that’s often considered a subsidy. But does zoning US land for industrial usage count as a subsidy too? Most countries have tax credits for research and development and compete to lower their corporate tax rates to encourage investment. The factor that determines whether such initiatives are considered statist industrial policy (bad), or building a business-friendly environment (good), is usually whether they’re being done by a foreign government, or our own.
What’s clear in retrospect, however, is that Asbeck’s claims had little solid basis. The telltale sign of a subsidized industry is that prices spring back up once competitors have been squeezed out and the government withdraws support — but the opposite has happened with solar panels, which now sell for roughly 5% of what they cost in 2011.
At the World Trade Organization (WTO), which has a more rigorous definition of subsidy than US and European governments, there has only ever been a single case alleging subsidy against the Chinese solar industry — and that has been dormant since 2011. A separate WTO panel in 2014 found the US antidumping decision resulting from Asbeck’s complaint went against the trade body’s own rules.
Still, one might have expected American solar-panel manufacturers to have responded with jubilation to Washington’s trade broadside against Beijing. In fact, the effect was much more like fear. The US in 2011 was making more money selling polysilicon and solar machinery to China than it was spending buying completed panels. That meant it was highly vulnerable to retaliation. In July 2012, two months after Washington found in favor of Asbeck, the counterattack began: China’s Ministry of Commerce announced an investigation into whether the US was dumping polysilicon into the mainland market.
Chinese panel producers didn’t wait for their government’s ruling to act. With spot prices slumping well below the long-term contracts preferred by incumbent polysilicon producers, they canceled purchases en masse. By the end of the year, the decline in Hemlock’s sales into China had “reached dire levels,” Corning’s CFO Jim Flaws told a January 2013 earnings call. “The market for solar grade polysilicon is almost non-existent now.”
The European Union, where Asbeck had been at work bringing more antidumping cases, was in a similar situation, but officials there struck a compromise agreement in 2013 that helped the main local polysilicon producer, Wacker Chemie AG, maintain access to China. Despite concerted lobbying, the US failed to do the same. The result was a direct hit on Hemlock, with tariffs of 57% on American polysilicon imports.
That was just what China’s nascent polysilicon industry needed. “At that time, Chinese polysilicon producers were not cost competitive,” says Johannes Bernreuter, an analyst who’s studied the solar poly market since the early 2000s. “That gave them a protection wall to develop. It was no coincidence that six Chinese manufacturers returned from a dormant state to production in the course of 2013 when the antidumping duties were introduced.”
The opposite happened in the US. Taking fright, Mitsubishi sold out of the Hemlock venture in 2013. The following year, about six months after Beachy took over as Hemlock’s chairman, she had to announce the closure of the promised Tennessee plant. “It was pretty painful to be honest,” she recalls.
In Pasadena, Texas, a polysilicon factory owned by solar developer SunEdison Inc. was shuttered in 2016 when the company went bankrupt, blaming Beijing’s retaliatory tariffs. Its innovative technology was bought by GCL Technology Holdings Ltd., a Chinese rival.
In Washington state, REC Silicon ASA shut its plant in Moses Lake in 2019, again citing the tariffs imposed by China. A second REC factory in Butte, Montana limped on until its closure was announced this February. Sustained by ongoing demand from its original customers in the chip industry, Hemlock kept going — but it quit making PV-grade polysilicon altogether in 2019 and 2020. At a time when the solar industry was scaling new heights, America’s manufacturing sector had quit the field.
The East Is Red, the Sun Is Rising
Today, Tongwei has expanded beyond recognition.
It now has facilities dotted across China’s renewables-rich outlying regions, where power is cheap: in Sichuan; in tropical, hydro-fueled Yunnan; and in Inner Mongolia, rich in sunshine and wind, as well as dirty coal. Even its Leshan operations have long outgrown their original site. The new plant sprawls over a vast campus downstream from the city. Silvery pipes carry chemicals to the distillation towers where silicon is purified, resembling a monumental church organ. A phalanx of pylons brings in electricity. A second Tongwei plant on a similar scale sits just across the road, while GCL Technology has a third hulking over a neighboring plot.
The site is sparkling. The roughly 2,000 people who work here mostly toil away from sight. They have dormitories, canteens, and an on-site gymnasium. A brand-new blue glass building on an artificial hill above a pond is a museum for visitors, illustrating the polysilicon production process and the history of the business. On the grass slope, beds of orange flowers have been pruned to spell out the Chinese characters for “lucid waters and lush mountains” — a quote from Xi Jinping seen as emblematic of the president’s support for the environment.
Inside, a framed certificate confirms that the factory’s 2022 power supply was 100% renewable — a total of 2.38 gigawatt-hours, nearly enough to power Ireland for a month. Nearby, an illuminated panel displays a bar chart of the world’s polysilicon centers. On the far right of the chart sits a squat bar representing Hemlock, with a modest 30,000 metric tons a year. The Leshan site alone can produce about 120,000 tons, and Tongwei as a whole will have a capacity of 480,000 tons this year.
Those numbers are astounding when you consider the amount of energy they represent: 480,000 tons is enough to generate sufficient solar electricity to power Mexico for a year — or Indonesia, or the UK and Ireland put together. Over their lifetime, those solar panels will provide nearly five times as much useful energy to the world economy as all the oil and gas in Exxon Mobil Corp.’s underground petroleum reserves.
Tongwei may be little-known outside China, but it’s by far the world’s biggest producer of polysilicon and should already be considered one of the most important energy companies globally. That dominance will only grow if plans announced in December to nearly double output go ahead.
Over a lunch of chicken-feet skewers, steamed fish, braised tofu, and orange juice in the factory’s office suite, the site’s strategic development director, Ding Xiaoke, speaking through a translator, describes a business strikingly different to the popular image of state-subsidized behemoths intent on undermining rivals in Europe and the US. The tariffs on Chinese solar products recently announced by Biden don’t concern him, because the plant doesn’t have any customers in the US. He’s interested in setting up production bases overseas but worried they won’t match Sichuan’s low costs.
“For Tongwei, everything is about the market,” he said. Political issues such as trade barriers “might determine how fast we put investment in a specific region, but it won’t stop us growing.”
Above all, he describes a business not only lacking state support, but largely left to its own devices by its corporate parent. (The Leshan facility is technically run by Sichuan Yongxiang Co., a subsidiary of the fish food-to-solar Tongwei Co. parent.) By Ding’s account, it’s as likely to compete as collaborate with Tongwei’s polysilicon facilities in other Chinese provinces.
That image of pure independence isn’t 100% accurate. Tongwei Co.’s own accounts list a total of 2.19 billion yuan ($301 million) in government grants and tax concessions to the parent company since 2009, with more than half the total accrued last year as its capacity expansion went into overdrive. At the same time, the financials provide precious little evidence of the sort of comprehensive backing that’s usually assumed to explain the low cost of Chinese photovoltaic panels, especially when benchmarked against First Solar, the only US rival with comparable accounts that’s remained in operation throughout the past decade.
Far from benefiting from cheap land, tax discounts and below-market loans, the Chinese company, if anything, is notable for how closely its business resembles its smaller US rival. The value of its land rights as a share of property, plant and equipment is about 4.9%, compared with a 0.8% share for the land on First Solar’s balance sheet. That suggests that — far from receiving benefits — Tongwei is spending more on land. Since the start of 2009, taxes on its income have amounted to about 30% of the pre-tax total; First Solar managed a dramatically lower 12.8%. Tongwei’s weighted average cost of capital — a proxy for any advantage it’s getting from low-cost loans — was 11.9%, almost identical to the 11.8% at First Solar.
As for subsidies, First Solar’s since 2009 have amounted to about three times the amount reported by Tongwei — $967 million in grants, tax credits, loan guarantees — according to a database compiled by corporate accountability lobby Good Jobs First. About 90% of the total was development and export financing for projects in Chile, Canada and India. (Thanks to its complex ownership structure, Hemlock Semiconductor doesn’t provide comparable financial data, although data from the company and the Good Jobs First database indicates $618 million in subsidies since 2008.)
The real support that Tongwei has received has been something far more indirect: the certainty of robust government backing for renewable power. Long after Germany and Spain canceled the subsidies that encouraged the renewables industry to grow so fast during the 2000s, China’s program was still active. It didn’t provide any direct support for manufacturers but ensured a level of demand from utilities that allowed solar factories to grow beyond their troubled infancy to their current profitable status. (The Chinese subsidy program was closed at the end of 2021.)
By providing policy certainty and an investment-friendly environment — two things that businesses lobby for in every country on the planet — China has built a solar industry whose lead is by now likely to be unassailable. Polysilicon is the bedrock of the solar supply chain. If it can’t be produced at competitive prices, the domestic industry is at best going to be assembling photovoltaic products manufactured elsewhere. That’s now the only plausible path for the rest of the world, according to Bernreuter.
“I don’t think there will be a renaissance for the US, Europe and Japan. I cannot imagine that,” he says. “They cannot compete with the Chinese players any more.”
Learning the Lessons of History
If you want to imagine an alternative path for the global solar industry, you need only look to the history of Hemlock’s home state.
When Henry Ford was laying the foundations of the modern automotive industry in Detroit, one of his key innovations was building on a scale that shocked the competition into submission. His Highland Park plant was the largest factory the world had ever seen when it opened in 1910. The River Rouge facility, built less than a decade later, was nearly 10 times larger. It accommodated its own power plant, docks and steel mill, in an area greater than London’s entire financial district.
The secret? Economies of scale. A small factory in a stable market is unlikely to reduce costs very much. However, when building to epic proportions in a market growing at heroic rates, relatively minor tweaks to the manufacturing process will compound over time to drive ever-shrinking prices. In the 1960s, Intel Corp. founder Gordon Moore predicted that the number of transistors on computer chips would double every two years, a prediction that has held true up to the present. The same process explains why the similar semiconductors manufactured into solar panels cost about 4% now of what they did in 2009. China’s excellence in such cheap, large-scale manufacturing has turned it into the production base for 95% of the world’s iPhones. It’s hardly surprising that the same expertise has given it a lead in the solar industry, too.
For that process to work, however, manufacturers have to be convinced that their bold investments will pay off — either because they’re more efficient than their rivals, or because they’re confident that long-term demand will be unstoppable.
Businesses and nations commit aggressively to the projects that they think present opportunities for them. China’s support for solar developers is so unwavering, in part, because — unlike the US (which is currently pumping more oil and gas than any nation in history) — it’s desperately short of domestic energy sources, other than coal reserves whose costs are ever-rising and whose fumes threaten to choke its cities.
The amount of coal, petroleum or hydroelectric power that a country can produce is an ineluctable fact of its geography, so China’s ability to transform such energy into economic growth is heavily dependent on imports from other countries. That’s a concern for Beijing — but solar and wind power are different. The key to their development isn’t geological happenstance but manufacturing prowess, one field in which China has few peers.
An economist would tell you that the ideal way to structure global trade is for countries to specialize in the products where they have the greatest comparative advantage. If China can produce cheaper solar panels than anyone else, then other nations should buy them and send back in return whatever they can produce at record-low prices. The corn and soybean fields that surround the Hemlock polysilicon plant are a case in point. Last year, US exports of the two crops amounted to $28 billion and $13.7 billion respectively, a total greater than the $22 billion it spent on imported solar panels.
The Covid-19 pandemic made that view look impossibly unfashionable, as it triggered a worldwide rethinking of supply chains. Alarm in the US and EU grew further when Russia weaponized its gas exports amid the invasion of Ukraine. If authoritarian powers have too much control over a source of energy, they might be able to bend even democracies to their will.
That analogy, however, doesn’t really make any sense in the case of solar. Russia’s gas companies sell fuel, but China’s solar businesses sell machines for producing energy from daylight. The distinction is crucial: Moscow can turn off Europe’s gas taps, but Beijing can’t turn off the sun. Even so, the mere fact of China’s dominance, largely shrugged off in 2019, has assumed the significance of a global emergency in 2024.
The trouble is, it’s too late now to wind back the knock-on effects. If US and European subsidies to solar hadn’t been cut because of the bad political smell from Solyndra and the wave of austerity that followed the 2008 financial crisis, then local renewable developers would have been more active and manufacturers would have seen more demand for their products. If a charismatic German hadn’t bounced the US into an accidental tariff war with China in the early 2010s, those proposed American polysilicon plants might have been built. The process innovation that’s happened over the past decade in Leshan might have occurred in Hemlock, Clarksville, Moses Lake and Pasadena instead.
Installers see no reason to prefer local products. “Over the last year or so we have not seen a huge disparity between Chinese- and American-made,” says Randy French, founder of Independent Solar, a company that’s been installing residential systems for 21 years in California, Arizona, Nevada and Texas. “They’re all really great products right now,” he says. “Chinese or anything.”
The US solar industry that’s left is moribund at best. SunPower Corp., a once-venerable name in the US solar industry which was worth at much as $12.5 billion at its peak in 2007, filed for bankruptcy in August. Three weeks later, Switzerland’s Meyer Burger Technology AG canceled plans to build a 2 GW solar cell plant in Colorado, saying the site was no longer financially viable and it would manufacture in Germany instead.
The following week, Maxeon Solar Technologies Ltd., SunPower’s manufacturing arm before a 2019 spin-off, dropped its earnings guidance, after customs checks at the Mexican border halted imports to the US, its largest market. Start-up CubicPV Inc. in February announced it was abandoning plans to build a US plant capable of slicing polysilicon into enough thin wafers to make 10 gigawatts per year of panels.
There are winners, to be sure, but not on a scale to meet the coming demand. First Solar, well-protected by tariffs, has seen its shares gain by about a quarter since the Biden administration’s latest round of levies in May.
South Korea’s Hanwha Solutions Corp., which produces panels under the Qcells brand, has also done well. Democratic presidential nominee Kamala Harris visited its plant in Dalton, Georgia last year and joked of her “maybe unique passion for solar energy”. In August, Qcells received a $1.45 billion government loan guarantee for a nearby factory that will produce 3.3 GW of wafers annually. That’s roughly one-tenth of the solar panels the US installed last year, and about one-twentieth of what will be needed annually if the country is to meet Biden’s promise to decarbonize America’s grid by 2035.
In Hemlock, too, activity is picking up. A US law passed in 2022 banned the import of products produced in China’s Xinjiang region unless there’s clear evidence that there was no forced labor in their supply chain. In theory, that shouldn’t be an insurmountable issue for Tongwei, which doesn’t operate in Xinjiang and appears to be “the safest bet in the Chinese polysilicon market,” according to a 2021 study by researchers at Sheffield Hallam University. In practice, however, the ban has acted as a de facto block on all polysilicon produced in China.
That law inspired Hemlock Semiconductor to restart sales and production for the solar industry in 2021. Just to the west of the existing plant, work has already begun on a $375 million expansion. Residential plots have roughly doubled in price from $7,000 per acre to as much as $15,000 in anticipation of the influx of workers, according to Abbey Miller, a part-time real estate agent and server at a local diner.
“The U.S. solar industry is at an important inflection point and the market is demanding U.S.-made polysilicon because of its high quality and traceability,” a spokeswoman for Hemlock said in an e-mailed statement. “The momentum in the market will only accelerate the build out of a domestic supply chain and help us explore opportunities to expand capacity to keep pace with growing demand.”
It’s not clear how much of Hemlock’s new production will go to the solar industry, and how much to semiconductors — but if you were making a bet, you still wouldn’t put it on PV. What the US really cares about is the ultra-pure polysilicon for microchips that’s central to America’s desire to make computing power — not solar power — a key national security advantage against China. The CHIPS Act, signed by Biden in 2022, provides around $52 billion in subsidies to the US microprocessor industry, sums beyond the imagination of what the solar sector has enjoyed anywhere. The Loan Programs Office, a government agency with authority to issue hundreds of billions in loans to promising energy projects and headed up by former solar entrepreneur Jigar Shah, didn’t lend money to a PV manufacturer between the collapse of Solyndra in 2011 and August’s loan to Qcells.
The purported rationale of US tariffs on imported solar is one founding father Alexander Hamilton would have recognized: to protect a fledgling industry until it is strong enough to stand on its own feet. No one I have spoken to in the solar industry, however, sees that happening. To have a functioning PV sector you need every piece of the supply chain — from polysilicon, ingot production and wafer slicing to cell manufacturing and finally module assembly. There’s precious little sign that’s going to emerge on a sufficient scale in the US, not least because China’s production efficiencies are so much greater.
Successive waves of tariffs have done little more than create a Potemkin solar industry, while putting a tax on clean power as the climate crisis festers. The US installed about half as many solar panels in 2023 as the European Union, despite a far richer natural endowment of clear skies and bright sunlight.
If America wants a sort of small-batch, artisan solar sector to give the impression it’s doing the work of staving off climate change, while going all-in on expanding petroleum production instead, then it’s hit on the right policy. US manufacturers can survive in the walled garden of the domestic market, but the protectionism that sustains them means they’ll never grow efficient and cheap enough to survive the savage competition of the global market. That’s where Chinese companies reign supreme. “China really wanted the solar industry,” says Hemlock’s former president, Beachy. “We gave that away as a country.”
It’s a tragic failure of vision and ambition. In Detroit a century ago, US automobile entrepreneurs created an industry that irrevocably transformed cities, countries and economies. This time around, China’s innovators are the ones changing the world. With assistance from Taylor Tyson
Graphics edited by Elaine He
Images edited by Yuki Tanaka
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the author of this story: David Fickling at dfickling@bloomberg.net
To contact the editor responsible for this story: Howard Chua-Eoan at hchuaeoan@bloomberg.net
