FEATURECHINA
A Chinese worker assembles solar panels in Jiangxi province. The U.S. cannot preference its domestic solar manufacturers without rebuffing the World Trade Organization.
Solar has been the fastest-growing source of electric power in the United States for several years. The cost of energy produced by solar photovoltaic panels declined by 82 percent between 2010 and 2019 and became cheaper than coal and natural gas in 2018. Solar now comprises 3 percent of total electricity generation, totaling 43 percent of new generating capacity in 2020. But almost all of the panels and their component parts are made in China. That has to change.
We need to re-create a solar panel manufacturing industry for national-security, economic, environmental, and ethical reasons. Building a solar manufacturing industry will require changes in our current trade policy, as well as coordinated procurement to require a domestic solar supply chain, and R&D support for accelerating the development of more efficient panels. These interconnected elements also need to be linked to job training and placement for people with employment barriers and those losing their jobs in the transition from renewable to fossil fuel industry jobs.
The United States employs about 250,000 workers in the solar industry—about 162,000 in installation and 34,423 in manufacturing. The remainder are in operations and maintenance and other supportive jobs. If we combine a strategy to transition to renewables with one to build a domestic solar industry and supply chain, employment could increase dramatically. A ten-gigawatt increase in production would add 62,500 direct and 75,000 total manufacturing jobs and would increase installation employment as well.
But China currently dominates in manufacturing all components of solar modules. A solar module begins with refined and melted polysilicon. A hair-thin wafer of the polysilicon forms the basis of a solar cell, which converts light energy to electricity. There are several types of cell structures, the most common being single cell (monocrystalline) and polycrystalline, which is used in what are called thin-film solar panels because they are built from flexible materials. China dominates in the production of each element—the refined silicon, the wafers, the cells, and the completed module enclosed in glass and metal or all glass (bifacial panels).
China’s Shady Solar Strategy
China has a long-term strategic vision for dominating essential industries—and solar is one of them. In 2016, I wrote in the Prospect about how China used a combination of subsidies and free land to attract U.S. solar companies. They stipulated, however, that they couldn’t sell their products in China, which violated free-trade principles and the promises China made to other WTO countries in its WTO accession protocols, but the U.S. government did not make an issue of it and many U.S. companies found it to be an offer they couldn’t refuse.
China started developing its own solar industry and dumped products on the world market at below-cost prices—something that also violates trade law. By 2011, prices began falling dramatically and many U.S. and German producers couldn’t compete. Now, China produces 75 percent of solar modules globally.
It’s a similar story with polysilicon. The U.S. was the world’s leader in polysilicon production until China used similar policies to gain dominance. A polysilicon plant uses an energy-intensive process that refines metallurgical-grade silicon to the high purity levels needed for solar cells and semiconductors. Although it took China a while to develop the manufacturing knowledge to be competitive, once it did the industry took off. In 2017, China imposed stiff tariffs on American and South Korean polysilicon, which meant the growing Chinese solar panel production sector had to use domestic polysilicon.
Investment and tariffs have been effective—China’s polysilicon industry, all of it in western China, has grown 27-fold in the past ten years. Since 2017, 91 percent of new capacity worldwide has been in China. Two-thirds of the world’s polysilicon market will be controlled by five companies in China and Hong Kong by 2021.
And it’s the same story with wafers and cells. The U.S. currently has almost no production capacity for cells, while China produces 80 percent of the world’s output. In 2020, China accounted for 99 percent of global wafer production, 80 percent of global cell production, and 75 percent of global module production—all substantial increases over its market share in 2010.
A March 2020 U.S. International Trade Commission report estimates that U.S. cell production dropped 75 percent between 2016 and 2018. U.S. cell imports jumped from 308 megawatts from January–June 2018 to 951 megawatts during that same period in 2019, an increase that coincided with the ramping up of new domestic module manufacturing.
Many U.S. states, cities, companies, and individuals are installing solar to avoid carbon emissions. They would be surprised to know that the majority of the solar panels installed in the United States come from dirty coal–fired factories and that some components of their panels are made with forced labor.
SOURCE: REC SILICON ASA
About 45 percent of the world’s solar-quality polysilicon is produced in the Xinjiang region of China. Transforming silica into silicon for solar panels is an energy-intensive process that relies on electricity produced from dirty coal. Further, some manufacturers in western China have been accused of dumping toxic wastewater produced in the process into rivers. These practices produce significant air and water pollution that affects agricultural production and human health in the central and western regions of China where it is commonly produced.
The Australian Strategic Policy Institute has documented detention, re-education, and forced labor of Muslims in the Xinjiang region, where many inputs to solar panels are produced. A 2021 Horizon Advisory report names Chinese solar giants Daqo New Energy, East Hope Group, GCL-Poly, and Jinko Solar among the companies in the Xinjiang region using forced labor.
These companies are still exporting to the United States. American law bans importation of goods produced with forced labor, and there is considerable evidence that many of these goods make their way into the solar supply chain. The U.S. has banned cotton and tomato imports from the region and placed restrictions on hair products and computer inputs from Xinjiang, but to date, not polysilicon or other components of solar panels produced in the region.
The China Photovoltaic Industry Association not only denies the use of forced labor in the industry but uses threatening language to warn against restricting imports: “We also solemnly tell some American association and companies that if they intend to use this as an excuse to restrict and suppress relevant parties and companies in China, interfere with normal business cooperation and competition, and seek personal gain from it, they will not only violate international trade rules and market economic principles, but also destroy them.”
U.S. Solar Policy: Fits and Starts
In dramatic contrast with China’s policy for supporting every element of the solar production industry, U.S. policy has been scattershot. Rather than engage in long-term industrial planning, the main federal approach to support the solar industry has been tax credits, offered intermittently, while the Department of Energy has funded research and the government intermittently has subsidized some production and installation.
At the state and municipal level, we have a hodgepodge of renewable mandates and subsidies focused on installation. The main policy instrument is the renewable portfolio standard (RPS), which requires utilities to purchase a set percentage of their power from renewable sources by a set date. Currently, 38 states plus Washington, D.C., have an RPS (or similar program). While state subsidies have supported the record-breaking growth of solar energy deployment, they have done nothing to support manufacturing of solar components.
The American Recovery and Reinvestment Act (ARRA) of 2009 demonstrates the effectiveness of a direct federal subsidy of the clean-energy sector. Of its total investment of $840 billion, ARRA spent $92 billion on clean-energy technologies, including clean-energy generation, grid modernization, electric vehicles, transit, energy efficiency, and workforce training to support these industries.
To jump-start projects, the $25.7 billion invested for clean-energy generation paid developers 30 percent of their project costs in cash rather than as a tax credit between 2009 and 2015. Another $4.6 billion was provided for guaranteed loans to companies investing in renewable energy. Although Solyndra became the poster child for those arguing that the government can’t pick winners, total interest payments to the government from the loans exceeded losses from loans by $30 million. The program invested in 183 projects, which leveraged private investment of nearly $5.4 billion. Of the 183 clean-energy projects, 58 went to solar equipment manufacturers, totaling $1.1 billion. Before the stimulus, solar provided less than 1 percent of the nation’s electricity. As a result of the stimulus, solar more than doubled to 2.3 percent.
We need that scale of investment, and more, to expand domestic solar capacity. Currently, there are 21 solar panel manufacturers in the United States, 15 of which are American-based.
China has a long-term vision for dominating strategic industries—and solar is one of them.
Some existing U.S. companies demonstrate how to reclaim domestic supply chains by building on established regional competence. When Tempe, Arizona–based First Solar opened its first factory in Perrysburg Township, Ohio, in 2001, it was the country’s largest solar panel manufacturer. It was part of a solar boom in the Toledo area, building off research at the University of Toledo and historical strength in glass technology and manufacturing. Eighteen years later, the company opened a second factory nearby to produce its new Series 6 line. Representing a cumulative investment of over $1 billion, the new factory expanded First Solar’s domestic capacity to 1.9 gigawatts. This larger line of panels is designed for commercial, industrial, and utility-scale uses.
The new plant added about 500 jobs, with the company now employing approximately 1,450 employees at both Ohio factories. The factory was in production 18 months after breaking ground, illustrating how quickly the country could ramp up domestic production. The company has said that it will exit 2021 with 2.6 gigawatts of capacity in Ohio as a result of process improvements.
First Solar sources all of its inputs for its production in Ohio from a diverse range of suppliers that, notably, does not include companies in China. “We operate a diversified supply chain and, wherever possible, source domestically produced content,” said Mike Koralewski, First Solar’s chief manufacturing officer. “Solar manufacturing at scale is a significant job creator, but its real impact often gets lost in the narrow focus on jobs directly created on the factory floor. Our manufacturing operations in Ohio indirectly support as many as 7,000 supply-chain jobs across America.”
First Solar’s expansion motivated the NSG Group to build a $265 million specialty coated-glass plant nearby with about 70 percent of its production going to First Solar. The 511,000-square-foot facility employs 110 hourly workers and 40 in salaried positions. It’s the first glass furnace to be built in the United States in four decades. NSG, in turn, sources materials such as soda ash from Wyoming and sand from Michigan.
The cadmium telluride (CdTe) based thin-film panels First Solar makes are the second-most prevalent type of solar panel after crystalline. They are comparable or better than crystalline panels, and because they don’t use silicon they are easier and cheaper to produce and have a lower carbon footprint. The company also leads the industry internationally in providing recycling of its modules, which it has been doing for the past decade. Recycling recovers more than 90 percent of the CdTe from every module, which can be used repeatedly. First Solar is the only solar manufacturer to achieve the coveted EPEAT ecolabel of the Green Electronics Council for energy and water efficiency of the production process and use of recycled material.
First Solar is a member of the Ultra Low-Carbon Solar Alliance, launched in October 2020 with the goal of increasing market demand for solar panels manufactured with low embedded carbon. Its members include ten solar companies representing the full spectrum of the solar value chain.
The Alliance hopes to have a third-party-verified embodied-carbon-specific ecolabel intended to complement existing sustainable-product labels. To be certified, a solar module company and its suppliers have to submit to lifecycle analysis to disclose their embedded carbon emissions.
France already has a version of this certification in place, and the European Commission is considering a requirement that all PV modules include carbon footprint information.
DAVID DAYEN
A solar array in Palm Springs, California. China produces around three-quarters of solar modules globally.
Alliance Executive Director Michael Parr is talking to federal and state agencies and other large purchasers of solar to encourage a preference for modules with the ecolabel. The strategy does not violate trade agreement rules against preferencing domestic suppliers because any company is eligible to seek certification. An ecolabel will initially privilege U.S. and European producers since there’s roughly twice as much embodied energy in Chinese modules, according to Parr. And it wouldn’t raise the price of solar projects—more than 60 percent of the embodied carbon in a given project is in the polysilicon and wafer, which is only about 6 to 8 percent of its cost.
While preferencing certified ecolabel producers is one approach to building a clean domestic solar supply chain, we also need a combination of trade policy, procurement policy under Buy American, manufacturing tax credits, and continued research and development.
The Trade Policy Conundrum
Members of the World Trade Organization (WTO) agree not to discriminate among trading partners and to treat imported and locally produced goods equally. That means the U.S. can’t give preference to domestically made renewable-energy products and also follow WTO rules. One or the other has to give.
Other nations have used targeted industrial policies, and have been the subject of complaints by—of all nations—China. (Do as I say, not as I do.) China actually won a major case in which the WTO ruled in its favor. In 2013, Japan and the European Union mostly won a case against Ontario for paying higher prices for solar energy from locally produced equipment. In 2016, the U.S. brought a case against India before the WTO for its buy-local provisions for cells and modules. India retaliated by bringing a case against eight U.S. states with solar subsidies and buy-local policies. The WTO ruled against the U.S. and required changes in the policies. But with Trump having refused to appoint WTO appellate judges, enforcement of rulings is effectively impossible.
Several manufacturing advocates told me that the U.S. should give preference to domestic solar despite the fact that it violates the WTO. It took two years after Japan and the EU brought the case against Ontario for the WTO disputes settlement mechanism to issue its first ruling. The party can then appeal the decision, and if still found in violation, the party can protest the sanctions. If sanctions are applied, they don’t start retroactively from when the rule was broken, but only after the final decision has been rendered. All this assumes that the WTO has appellate-system judges to hear appeals. Last week, the Biden administration indicated that it would not appoint a judge, putting all trade cases on hold. At some point, the president will need to decide whether domestic industrial goals take priority over a badly flawed trading system that his predecessors of both parties have long promoted, at the expense of U.S. manufacturing.
Our dependence on China for solar deployment is no less dire than our dependence on its semiconductors.
While deliberately violating WTO rules may seem extreme, our dependence on China for solar deployment is no less dire than our dependence on semiconductors, on which the Biden administration is beginning to act. Due to increased demand for computers and other tech products during the pandemic, along with Trump’s banning imports from Chinese companies suspected of using semiconductor technology for spying, the nation is experiencing a shortage of semiconductors that are in many consumer goods. Because the industry is so vital to the U.S. economy, Biden is expected to sign an executive order soon to offer financial support to the industry. This same sense of urgency should apply to the solar supply chain.
The tariffs that the Trump administration placed on foreign solar modules, under Sec. 201 of the Trade Act, which allows retaliation against dumped imports, motivated three foreign producers (Hanwha Q Cells, Jinko, and LG) to open U.S. module plants in response to the tariffs. While the Solar Energy Industries Association (SEIA) claims the U.S. lost 62,000 jobs and $19 billion in investment due to the higher prices of solar modules resulting from these tariffs, these figures are misleading. In fact, Wood McKenzie counted 19 gigawatts of new solar capacity in 2020, a 43 percent increase from 2019.
The SEIA posture reflects the fact that it is more a creature of companies that install solar (which want low prices) than those committed to domestic manufacturing (which resist unfair Chinese-subsidized competition). Short term, there is a tension between the goal of installing more solar at the lowest possible price and the goal of expanding domestic production. But long term, if the U.S. can rebuild supply chains and domestic technological leadership, prices will keep falling.
Putting It All Together
One tool a president can use is the Buy American Act of 1933, which imposes domestic-production rules on federal procurement and federal grants to states and counties for procurement. Another is the Buy America Act (the two are often confused). The latter is a set of rules that apply to purchases of iron, steel, and manufactured goods used in Department of Transportation–funded infrastructure projects.
The Buy America rules for DOT spending haven’t been committed under our free-trade agreements and don’t allow waivers for trading partners that are signatories of free-trade agreements with the U.S., while the Buy American Act does. Currently, the WTO Agreement on Government Procurement (GPA) gives the other 47 countries in the agreement the same Buy America status as domestic producers. President Biden’s Buy America executive order calls for renegotiating those rules, as did Katherine Tai in her confirmation hearing to become the United States trade representative.
In practice, the U.S. can unilaterally exclude certain goods from GPA in trade agreements with a simple declaration that we are taking this category of goods out of commitments. We should do that with the solar supply chain. As one trade restriction supporter noted, “We can’t let trade policy be set in Geneva.”
Infrastructure build-out can promote domestic solar production. The Coalition for a Prosperous America proposes public investment in electric charging infrastructure for the national highway system using domestic solar and battery storage equipment. The proposal points out that the Federal-Aid Highway Program, with average annual spending of $40 billion between 2016 and 2020, is exempt from the WTO government procurement agreement. The U.S. also needs a major public investment in battery storage.
Congress should also reinstate the Department of Energy 48C Advanced Manufacturing Tax Credits offered under the American Recovery and Reinvestment Act to offset subsidized producers in China. This credit provided a 30 percent investment tax credit to clean-energy manufacturers and required that they pay prevailing wages. The value of the credits could be adjusted to target producers that locate in places that will be hard-hit as the fossil fuel industry declines and in environmental-justice communities. These credits should be extended over decades to create a sustained investment environment that produces innovation.
These disparate initiatives on research, installation, domestic production, trade, and incentives for utilities to shift to solar cry out for a coherent and coordinated national strategy. President Biden’s commitment to a green transition, a large-scale infrastructure investment, and a turning away from past, America-last trade policies, creates the opportunity. Now we need to put it all together.
