In 1977, a watt of solar power cost $76. Today, that same watt sells for less than 10 cents. If gasoline prices had fallen by the same proportion, filling your tank would cost less than a piece of candy [1].
And this isn’t an exaggeration — it’s reality. It has already happened, and, like so many things in our society, we get swept up in the shocking, negative headlines of the evening news and miss the important ones.
I happen to think we’re living through years of extraordinary change that go largely unnoticed simply because we’re the ones living inside them — but humanity is absorbing the biggest technological shifts in its entire history (computing, the internet, smartphones, genetics, neuroscience, artificial intelligence, energy…) compressed into the last five years. These advances are reshaping how we relate to one another and the balances of power, in ways that are hard to overstate.
The curve behind solar’s growth
Known as Swanson’s Law (after Richard Swanson, founder of SunPower), it describes the process by which every time cumulative solar panel production doubles, the price drops by roughly 20% [2]. It’s the force of mass manufacturing, iterative engineering, competition between companies, and economies of scale feeding back into each other. The more panels you produce, the cheaper you make them; the cheaper they are, the more people buy them; the more people buy them, the more you produce. Repeat.
Gregory Nemet, a professor at the University of Wisconsin-Madison, devoted an entire book to unpacking this process, and showed that the fall in photovoltaic costs isn’t explained by a single technical breakthrough, but by the accumulation of hundreds of small improvements in manufacturing, installation, financing, and public policy — especially in Germany, China, and the United States between 2000 and 2015 [3]. There was no “eureka moment.” There were ten thousand Tuesdays of engineers tuning a silicon diffusion furnace to be 0.3% more efficient.
According to IRENA, the levelized cost of solar electricity generation fell from $0.417/kWh in 2010 to $0.043/kWh in 2024 [4]. A 90% drop in fourteen years. And here’s a figure that should knock anyone out of their chair: 91% of the renewable projects commissioned in 2024 were cheaper than the least expensive available fossil fuel alternative [4]. Not 51%, not “most.” Nine out of ten. Over that same period, solar went from being more expensive than almost everything to being cheaper than coal and gas in most markets on the planet — with no subsidies required.
The records that broke every forecast
Here’s the part that still blows my mind. In 2024, 452 gigawatts of new solar photovoltaic capacity were connected to the grid worldwide, according to the official country-by-country statistics compiled by IRENA [5]. To put that in perspective: that’s more power than Spain’s entire installed electrical capacity, times four, added in a single year. And in 2025 that figure rose to about 511 GW — 75% of the 692 GW of total renewables installed that year [5]. Growth, year after year, without letting up.
China alone connected 315 GW of new solar capacity in 2025 — more than the entire world installed in 2022 [6]. India wasn’t far behind: it installed a record 45 GW in fiscal year 2025-2026, nearly double the year before [7]. Brazil, Germany, Saudi Arabia, and the United States all broke their own national records.
In 2026, even though China slowed its pace somewhat in the first quarter (41 GW in Q1 versus 60 GW in Q1 2025, according to China’s NEA) [8], the global picture remains one of massive expansion. The International Energy Agency projects that between 2026 and 2030, more than 400 net GW of renewable capacity will be added every year.
It’s worth remembering that solar isn’t growing because it’s “green.” It’s growing because it’s the cheapest option, full stop. In electricity auctions in Saudi Arabia, India, or Chile, solar bids beat gas and coal on price alone, with no green premium involved [4]. The engine behind this revolution isn’t environmental conscience.
A key piece of solar’s future success is batteries — the electricity that goes unused during the day, the surplus, has to be stored so it isn’t wasted. In recent years, batteries have been undergoing a major revolution driven by the rise of electric mobility, and the latest cost analyses increasingly fold batteries into the calculation.
The transition doesn’t run on silicon alone
Onshore wind has followed a respectable learning curve of its own: its LCOE fell 70% between 2010 and 2023, settling around $0.033/kWh [4]. Offshore wind, more expensive but with steadier wind hours, is expanding rapidly in the North Sea, the U.S. East Coast, South Korea, and China.
But what’s really interesting is what’s coming next. Perovskite solar cells — a cheap, easy-to-manufacture crystalline material — have gone from 3.8% lab efficiency in 2009 to over 34% in tandem configuration (perovskite layered on silicon). LONGi Green Energy set a world record certified by NREL of 34.85% in April 2025 [9]. The promise: cheaper, flexible, semi-transparent panels that can be manufactured at low temperatures. The open problem is long-term durability; perovskites degrade with humidity and heat, and the accelerated 25-year trials aren’t complete yet. It’s a field where lab efficiency is outrunning industrialization, so it’s worth reading the headlines with a bit of caution.
Other lines gaining traction: agrivoltaics (elevated panels that allow crops to grow underneath, successfully tested in French vineyards and Japanese rice paddies; market estimates put the sector at around $5 billion in 2026, though the figure varies by consultancy) [10], floating solar on reservoirs (which also reduces water evaporation, relevant in India and Southeast Asia), and green hydrogen as a seasonal storage vector and industrial fuel — though it still needs to improve on efficiency and solve storage and logistics problems.
What might come next
It’s possible that if progress in solid-state battery research keeps up its current pace, seasonal storage could stop being the transition’s bottleneck by around 2030-2032. This is supported by the fact that prototypes from Toyota, QuantumScape, and CATL are already showing energy densities 40-50% higher than current cells under lab conditions.
Another possibility is that the combination of ultra-cheap solar and high-efficiency electrolyzers could make green hydrogen competitive with grey hydrogen (produced from natural gas) before 2030, in regions with more than 2,000 hours of sunlight a year — essentially the entire equatorial belt, North Africa, the Middle East, Australia, and southern Europe. Right now, in 2026, green hydrogen costs between $4.50 and $6/kg in Europe, versus $1.50-$2/kg for grey hydrogen [11]. If electrolysis costs fall the way solar costs did, decarbonizing heavy industry (steel, cement, fertilizers) stops being a technical problem and becomes purely a matter of investment.
What’s coming
For once, a massive technological transition is happening not because anyone is being forced into it, but because the numbers work. That doesn’t mean the problems have disappeared: lithium and cobalt mining, recycling panels at the end of their life, geopolitical dependence on Chinese manufacturing, integrating all this into electrical grids designed for thermal power plants. All of that is real, and it takes work.
On recycling, there’s good news: companies like ROSI are investing more than €20 million in recycling plants across Europe, including a facility in Teruel, Spain, capable of processing 10,000 tonnes of modules a year [12]. The EU’s WEEE directive requires recovery rates of 85% and reuse rates of 80%, and Europe’s solar recycling market — still small, at around $180 million in 2026 — is growing at a double-digit annual pace, driven by the wave of panels that will start reaching end-of-life from 2030 onward [13].
The question is no longer “will solar energy work?” It works. The question is how fast we let it work. And that, my friend, is no longer a question of physics. It’s a question of political and social will — and part of today’s technological progress may hinge on it.
REFERENCES
[1] Nemet, G. F. (2019). How Solar Energy Became Cheap: A Model for Low-Carbon Innovation. Routledge. — Theoretical framework. Compiles historical cost data going back to 1977 and analyzes the learning mechanism. The $76/W figure comes from space-program records and early commercial cells from the 1970s.
[2] Swanson, R. M. (2005). “A Vision for Crystalline Silicon Photovoltaics.” Progress in Photovoltaics: Research and Applications, 13(5), 443-453. — Reliable. The foundational paper behind what’s now called Swanson’s Law. The ~20% learning rate has been repeatedly confirmed by later market data.
[3] Nemet, G. F. (2019). Work cited in [1]. Chapters 3-5, analysis of policy in Germany, Japan, China, and the U.S. — Theoretical framework.
[4] IRENA (2025). Renewable Power Generation Costs in 2024. International Renewable Energy Agency, Abu Dhabi. Published July 2025. — Reliable. Annual report based on auction and PPA data from more than 150 countries. Methodology: LCOE calculated using capital costs, operations and maintenance, capacity factor, and standardized lifespan. The $0.417/kWh 2010 baseline and the 90% drop to $0.043/kWh by 2024 come from IRENA’s consolidated historical series. The figure on 91% of renewable projects being cheaper than the fossil alternative comes from the same report.
[5] IRENA (2026). Renewable Energy Statistics 2025 and Renewable Capacity Statistics 2026. International Renewable Energy Agency, Abu Dhabi. — Reliable. Grid-connected capacity data confirmed by national statistics. Methodology: IRENA compiles official government figures (China’s NEA, India’s MNRE, Germany’s BNetzA, the U.S. EIA, etc.) and consolidates them. Figures: ~452 GW of new solar in 2024; ~511 GW in 2025 (75% of the 692 GW of total renewables). IRENA is used as the single source for the entire annual-installation series, avoiding a mix with Ember/BNEF market estimates, which use different methodologies (module shipments, pipeline estimates).
[6] China’s NEA (2026). Official data from the National Energy Administration, published January 2026. Confirmed by PV Magazine (Jan. 28, 2026): “China adds 315 GW of solar in 2025.” — Reliable. Official Chinese government data.
[7] JMK Research / PV Magazine India (2026). “India installs record 45 GW solar capacity in FY2026.” Data from India’s MNRE (Ministry of New and Renewable Energy). — Reliable. Official Indian government data, verified by multiple sources.
[8] China’s NEA (2026). Q1 2026 data published by the National Energy Administration. Confirmed by PV Tech: “China’s new PV installations plunge 51% YoY in January-April.” — Reliable. Official quarterly data.
[9] LONGi Green Energy / NREL (2025). Perovskite-silicon tandem cell efficiency record: 34.85%, certified by NREL in April 2025. — Reliable. Independent certification by the U.S. National Renewable Energy Laboratory. Methodology: measured under standard AM1.5G conditions, defined area, verified by an external lab.
[10] Coherent Market Insights (2026). Agrivoltaics Market Size, Share & Opportunities, 2026-2033. — With reservations. Market estimate from a private consultancy based on active-project and pipeline data. The exact figure varies between consultancies (Coherent, Mordor Intelligence, Grand View Research); the order of magnitude (~$5 billion) is consistent, but it’s not an empirical or official figure. Included as an indication of scale, not a fixed number.
[11] IEA (2026). Global Hydrogen Review 2026: Cost Acceptability. Paris. — Recent. Sixth edition of the annual report. Data on green hydrogen costs ($4.50-6/kg in Europe) and grey hydrogen ($1.50-2/kg). Published in 2026, with a transparent IEA methodology.
[12] ROSI Solar (2026). “ROSI secures over €20 million to scale solar panel recycling across Europe.” Confirmed by PV Tech and Mercom Capital. — Reliable. Official company announcement, verified by multiple industry sources.
[13] Mordor Intelligence (2026). Europe Solar Panel Recycling Market Growth Report 2031. — With reservations. Market analysis from a private consultancy. The ~$180 million 2026 figure is consistent with the sector’s current size, but the 2031-2034 projections (double-digit growth) depend on assumptions about the pace of panel retirement and regulatory developments. Included as an indication of trend, not a fixed forecast.
Note on [5]: IRENA was chosen as the single source for the entire annual-installation series (2024 and 2025) because its methodology — grid-connected capacity confirmed by official national statistics — is the most conservative and verifiable. Other sources such as Ember, BloombergNEF, or SolarPower Europe use different methodologies (module-shipment estimates, capacity under construction, pipeline projections) that produce higher figures (~593 GW for 2024). Mixing both methodologies in the same paragraph would create a false impression of slowdown. With IRENA, the series is internally consistent: 452 GW (2024) → 511 GW (2025), roughly 13% growth.