On November 10, 2026, Bayes Business School presented research that explains the spread of silicon solar cells through the relationship between mass-production equipment and large-scale demand. The original paper, a peer-reviewed study published online on June 23, analyzes the U.S. solar power industry from 1956 to 2020. It compares silicon-based cells with thin-film cells, a category that covers several materials and manufacturing methods. The authors argue that standardized manufacturing equipment supported mass production and helped win the trust of customers and financial institutions that demand long-term reliability. Rather than focusing on conversion efficiency, the study looks at the conditions under which a technology takes hold in the market. It also distinguishes between companies exiting the business and manufacturing capacity disappearing.
Tracing 109 solar cell makers and 149 equipment makers in U.S. industry history
The study examined 109 solar cell manufacturers and 149 manufacturers of production equipment worldwide. It also analyzed detailed business histories of 13 major equipment makers and 10 large U.S. electric utilities. These last two groups are selected samples of leading companies, not the total number of firms in the industry. The 149 equipment makers also include companies outside the United States.
The research was carried out by Mara Guerra of Bayes Business School, City St George's, University of London, and Rajshree Agarwal of the University of Maryland. The paper appears in Organization Science. The authors cross-checked quantitative company data against corporate histories, newspaper articles, and industry reports to examine the relationships between cell makers and equipment suppliers, which entry and exit figures alone cannot capture. However, the sources cover different periods, so continuous data from 1956 onward are not available for every item.
The analysis emphasized an interaction in which equipment makers' activities support cell makers' production, and expanded production in turn spurs more activity among equipment makers. The authors checked multiple sources and the sequence of events to explain how the gap in adoption between silicon and thin-film arose. This is an observational study based on industry history, however. It did not experimentally isolate the causal effect of each factor, nor did it produce numerical forecasts of solar cell performance.
Standardized production equipment helped silicon solar cells spread
The paper explains that policies such as renewable energy procurement mandates and tax credits expanded the market for grid-connected solar power. U.S. policy, however, does not specify which type of solar cell to use, and both silicon and thin-film makers had the chance to enter the growing market. In other words, the paper does not argue that policy favored silicon alone and thereby created the gap in adoption.
Even when demand grows, orders are unlikely to expand unless manufacturers can supply what customers want reliably. Large customers such as electric utilities value long-term reliability, and a manufacturer's ability to stand behind product warranties also bears on financial institutions' lending decisions. For those financing power generation equipment, what matters is not only near-term performance but also whether a business can operate stably over many years. The paper analyzes demand-side requirements and the activities of equipment suppliers as closely linked.
Silicon-based solar cells had the advantage of drawing on manufacturing technologies accumulated in neighboring industries such as semiconductors, electronics, and optics. According to the authors, equipment makers applied this existing technology and knowledge to standardize manufacturing equipment for solar cells. Standardized equipment is easier to supply to multiple cell makers. Manufacturers can also share common equipment and production know-how, which makes it easier to scale up production than if each company developed its own manufacturing technology.
Standardized production equipment also helps limit variation in product quality. If stable production capacity supports long-term warranties and earns the trust of customers and financial institutions, it becomes easier to capture large-scale demand. Growing demand prompts new capital investment and gives equipment makers opportunities to expand their own businesses. The authors explain that this created a virtuous cycle in which improved supply capacity and expanding demand reinforced each other.
Thin-film, by contrast, is not a single material or manufacturing method. It includes different technologies, such as CdTe, which uses cadmium telluride, and CIGS, which uses copper indium gallium selenide. Because each method requires its own equipment adjustments and company-specific partnerships, the paper argues, manufacturing know-how developed at one company was hard to transfer to another. As a result, scaling up production proved difficult, and obstacles arose in meeting requirements for long-term reliability and financing, leading to a vicious cycle that held back growth. This does not prove, however, that all thin-film products are physically less reliable. The point is that, even with the same market opportunity, the two approaches differed greatly in the equipment supply systems that supported mass production.
Differences in exit rates and the inheritance of manufacturing capacity
Fifty-nine companies entered commercialization of thin-film, more than the 54 that entered silicon. The slower spread of thin-film therefore cannot be explained by a smaller number of entrants alone. The breakdown by type for 1956–2020, from Table 2 of the paper by Guerra and Agarwal, is shown below. The exit rate is the value reported in the paper, using the number of entrants of each type as the denominator. Four companies attempted commercialization of both types, so the combined 113 entrants across the two types is consistent with the 109 companies surveyed.
| Category in the paper's tabulation | Silicon | Thin-film |
|---|---|---|
| Number of entrants | 54 | 59 |
| Companies that exited (share of entrants) | 30 of 54 (55.56%) | 44 of 59 (74.58%) |
| Exited during R&D stage | 10 | 27 |
| Exited during manufacturing stage | 20 | 17 |
Of the 44 thin-film exits, 27 companies left the business at the R&D stage. For silicon, 10 of the 30 exits occurred at the R&D stage. Exits at the manufacturing stage numbered 20 for silicon and 17 for thin-film. The overall exit rate alone obscures how far each company had advanced its business. Exit in the paper also includes acquisitions, dissolutions, and withdrawals from the solar cell business, so it does not all mean bankruptcy.
In silicon, the paper explains, 10 of the 20 companies that exited at the manufacturing stage were acquired by other silicon makers, and their manufacturing capacity was carried over within the industry. Even when a company leaves a business, the manufacturing capacity it built up is not necessarily lost. If acquisitions carry that capacity forward, it may continue to support production across the industry.
For some companies that exited at the manufacturing stage, however, sufficient records on what happened afterward do not exist. According to the notes to Table 2, reliable detailed records could be obtained for 16 companies of each type, which limits precise comparison of how often manufacturing capacity was inherited. What this tabulation shows is that the stage at which firms exited differed between the two types, and that for silicon there are confirmed cases of manufacturing capacity being passed on through acquisitions. It does not quantify how much the virtuous cycle of equipment supply and demand boosted adoption.
Conditions that will shape thin-film adoption
These acquisition cases show that it matters not only how many companies remain, but also how manufacturing equipment and know-how are passed along within an industry. In evaluating the commercialization of a new technology, it is important not just whether individual companies can develop superior products but also whether the manufacturing method can be rolled out to multiple companies. Whether a production system can be built to meet customers' demands for long-term warranties, and whether it can capture large-scale demand, are also conditions that shape adoption.
Some thin-film technologies may hold advantages in areas such as material usage and flexibility. The authors write in the paper that, given latent demand in markets where thin-film has a major technical advantage, the possibility remains that adoption will take off in the future. This is an outlook premised on latent demand in specific markets, however, and does not guarantee adoption of thin-film as a whole.
The study covers the U.S. industry only through 2020 and analyzes thin-film, which includes different technologies, as a single category. The results therefore cannot be used to directly judge the merits of individual products today or the success of newer technologies such as perovskite solar cells. The authors also say that applying an analysis of a single industry to other industries requires additional research. Independent reanalyses or replication studies in other industries could not be found within the scope of what was checked.
To judge the prospects for thin-film adoption going forward, it will be necessary to check, company by company and market by market, whether production equipment can be rolled out to multiple companies, whether the manufacturing capacity that underpins long-term warranties is maintained, and whether it connects to large-scale demand. If these conditions are met, there is a possibility that thin-film's material-level technical strengths could be turned into products that can be supplied reliably.