Solar got cheap in factories, not in subsidy bills, and a Hormuz shock tests the difference

Last year the world added 600 terawatt-hours of solar photovoltaic generation, enough to power Canada for a year, and the first time a renewable source led global supply growth ^[5]. The International Energy Agency records it as the largest single-year increase for any electricity-generation technology ^[6], and it happened around a hard fuel shock: according to Scientific American's account, the U.S.-Iran conflict closed the Strait of Hormuz for months and spiked fossil-fuel prices, with solar already climbing before the closure ^[4].

That sequencing is the whole argument. A fuel price spike cannot propagate into a technology with no fuel line, which is the point Jimmy Carter made in 1979 when he said no one can embargo the sun or interrupt its delivery ^[2][5]. What has changed since is the capital cost, and it changed slowly, in factories.

Carter's White House array was 32 bulky panels installed during the oil crisis 47 years ago ^[1]. It cost roughly $160,500 in 2026 dollars and generated no electricity at all; it heated water ^[3]. A home system now runs about $25,000 and produces power ^[9]. In inflation-adjusted terms, the water heater cost about 6.4 times a modern home electricity system that does strictly more ^[1].

Efficiency followed the same grind. A state-of-the-art IBM research panel in 1977 reached 22 percent; the best home systems now hit that figure and standard systems run about 20 percent ^[10]. Roughly 49 years to move laboratory-grade conversion into a catalogue item ^[2]. Two device changes carried much of it: the passivated emitter and rear cell of the late 1980s, which added a reflective rear layer to capture more light and block damaging longer wavelengths ^[11], and since 2024 the crystalline-silicon tunnel oxide passivated contact cell, which has pulled ahead on both efficiency and high-volume manufacturability ^[12]. Walajabad S. Sampath of Colorado State University explains why efficiency shows up on invoices: higher efficiency means a smaller panel for the same output, so less land, less wiring, less racking ^[13]. The rest is process work, purer materials and more precise manufacturing at the scale of a highly integrated Chinese industry ^[14]. Even the energy accounting flipped. In the early 1990s the field assumed a panel could never repay the energy used to build it; a standard setup now does so in one to two years ^[8].

The subsidy-artifact theory is being tested in the United States right now, and it is losing. The Trump administration has cut billions of dollars from affordable solar projects and stopped issuing new permits through the Department of the Interior ^[15]. Deployment kept its momentum anyway, helped by tax incentives and investments that predate the administration, with the economics still attractive ^[16]. This May, solar outpaced coal's monthly share of U.S. generation for the first time ^[17].

Scale keeps the claim honest. Global solar generation now equals the European Union's entire electricity demand, Ember said in April ^[7], and solar's share of generation grew nearly 19-fold between 2014 and 2025, an 11-year run, while still trailing natural gas and coal ^[18][3]. "Today solar makes the cheapest electricity," says Sampath ^[19]. That is one researcher's characterisation, and this material says nothing about storage or firming costs, which is where the cheapest-kilowatt-hour claim actually gets argued.

Worth watching: whether solar holds the top of the IEA's supply-growth table once Hormuz-driven fossil prices unwind ^[4][6], whether the permit freeze at Interior shows up as a hole in U.S. additions two years out ^[15], and how quickly tunnel oxide cells displace the older PERC design in shipments ^[12].