The Downside to Made in the USA

Keith Bradsher has a largely excellent article in Wednesday’s Times that’s focused on the (possibly illegal) advantages that the Chinese government is providing its clean energy firms. In reporting the piece, though, he falls prey to a deceptive story about the success of Chinese solar at the expense of U.S. industry. Since it’s a story that regularly shows up in one form or another, it’s important to understand why there’s less than meets the eye. (I’ll have more to say on the broader trade issues in several upcoming articles and posts.)

Here’s the basic outline: Chinese solar businesses are booming. In this particular story, Hunan Sunzone Optoelectronics, which exports 95% of its panels, is held up as an example. U.S. businesses, meanwhile, are being hurt. Here the example is Evergreen Solar, which, we’re told, “plans to move the final manufacturing steps for its solar panels from Devens, Mass., to China next summer, eliminating 300 American jobs”. It’s a simple story: China wins, the United States loses.

But let’s take a more careful look at what’s going on. There are four basic stages in solar module manufacturing: silicon purification, ingot and wafer manufacturing, cell production, and module assembly. Evergreen Solar, according to its website, derives its competitive advantage through a proprietary low-cost technology for making wafers. Hunan Sunzone Optoelectronics, meanwhile, advertises its focus as being on cell production and module assembly. The two types of firms are not entirely, or even mostly, in competition.

This should not be particularly surprising. In what is quickly become one of my favorite obscure academic papers, Arnaud de la Tour and his colleagues at MINES ParisTech took a careful look (PDF) earlier this year at the structure of the Chinese solar industry. They found that China (circa 2008) was strong in the later stages of the solar value chain (27% of the cell and module market), but that it lagged far behind in the earlier stages (2.5% of the ultrapure silicon market and less than 5% of the ingot and wafer market). Those two later stages accounted for 60% of the cost of a module but only 18% of its profit. That’s because they’re less technologically sophisticated than the earlier stages, which accounted for only 40% of the cost but a whopping 82% of the profit. Those higher-value-added steps, in turn, support higher-wage jobs.

Seen from this vantage, the Sunzone/Evergreen story is decidedly less depressing. U.S. firms are unable to hold on to cell and module manufacturing (“the final manufacturing steps”) but still have an edge in wafers and silicon, where there is far more profit to be made. Indeed by lowering the cost of turning Evergreen Solar’s wafers into finished solar products, companies like Hunan Sunzone Optoelectronics help grow the market for the things that Evergreen Solar makes. Evergreen may have lost 300 cell and module manufacturing jobs to China, but it’s quite possible – indeed even likely – that it’s gaining (or retaining) high-wage jobs elsewhere in its value chain because of the same low-cost Chinese developments.

Indeed if the United States were to insist that all parts of the solar value chain stay in the United States, the result might not be more jobs – it might be less. Unable to reduce the cost of cell and module manufacturing, the cost of solar might stay too high, reducing the overall solar market, and with it jobs in wafer and silicon production too.

This dynamic is no different from what happens in many other sectors. China assembles computers that used to be made in the United States. Does anyone think that this means America is losing from the computer and IT revolutions? Of course not: the United States is making its contributions primarily in areas that yield far greater profits, while cheap Chinese computer assembly is enlarging the market for everything computer and IT-related. The same is true in solar: it’s quite possible for the United States and China both to win, with China lowering the cost of relatively low-tech parts of the value chain, in turn growing the market for the higher-tech parts that are still handled by the United States.

Of course, this state of affairs isn’t guaranteed to last forever. The MINES ParisTech study looks at 2008 data; things have probably tilted in China’s direction since then, and they may continue to do so, particularly given unfair trade policies from China. But the lesson remains: just because cleantech products don’t get stamped “Made in the USA” in the finishing stages doesn’t mean that the United States can’t win from the clean energy race.