Solar's fall from a space-age niche to the cheapest new electricity came from efficiency gains, manufacturing scale, policy and decades of accumulated learning.
By The Long View · Edited by Lachlan Brown
Published August 13, 2026 · How we edit
In 1956, usable solar photovoltaic capacity was so expensive that its first serious market was not a house or a power station, but a spacecraft. The often-cited historical estimate is $1,865 for one watt, adjusted for inflation and expressed in 2019 dollars. A modern module rated at 400 watts would represent $746,000 of capacity at that price.
By the end of 2024, solar modules sold in Europe for roughly 8 to 27 cents per watt, depending on type. In 2020, the International Energy Agency went further, calling solar photovoltaic power the cheapest electricity in history under the best combinations of sunshine and finance.
Those three numbers measure different things. The first is a historical capacity price, the second is a modern module price, and the third concerns the lifetime cost of electricity from a complete power plant. The transformation is real, but understanding it requires keeping the units straight.
Bell Laboratories unveiled a practical silicon photovoltaic cell in 1954. Its efficiency was about 6 percent, and commercial production was tiny. A historical account from the National Renewable Energy Laboratory says that a one-watt cell cost almost $300 in 1956, when building a conventional power plant cost about 50 cents per watt of capacity.
The $1,865 number used in the headline comes from an Our World in Data reconstruction. It explicitly reports the value in 2019 prices. “Today’s money” is therefore best read as inflation-adjusted money, not as a fresh conversion into 2026 dollars. Different historical products and inflation series produce different estimates, but none changes the economic conclusion: a terrestrial power producer could not justify early solar at anything close to those prices.
A watt is a unit of capacity, the maximum output under specified conditions. It is not a watt-hour of generated energy and not an electricity bill. At $1,865 per watt, even the capacity of a contemporary 400-watt module would have cost nearly three-quarters of a million dollars before any structure, wiring or installation.
Solar survived because spacecraft valued properties that ordinary electricity markets did not. A satellite needs a light, durable source of power where fuel deliveries and grid connections are impossible. Price per watt mattered less than operating for months or years in orbit.
Vanguard 1, launched in March 1958, carried six small solar panels that powered one of its radio transmitters. Its chemical battery lasted only weeks, while the solar-powered transmitter continued operating for years. That success turned satellites into an early customer for photovoltaic cells.
Production for space did not make solar cheap overnight. It gave manufacturers a reason to keep improving efficiency, reliability and fabrication. As prices fell, terrestrial niches followed: remote telecommunications, lighthouses, railroad crossings and vaccine refrigeration in places where extending a grid connection was unusually costly.
The modern figure in the title is conservative. The International Renewable Energy Agency’s 2024 cost review reports that modules sold in Europe ranged from about $0.08 per watt for the low-cost category to $0.27 for bifacial products in December 2024. Mainstream modules averaged roughly $0.11 per watt.
That is the price of the panel, not the installed system. A functioning solar array also needs an inverter, structural supports, cables, switches, labor, engineering, permitting and a connection to the grid. Residential projects carry sales and administrative costs. Utility projects need land, civil work and transmission access.
The US Department of Energy’s Q1 2024 solar cost benchmark divides installed costs into modules, inverters, structural and electrical equipment, fieldwork, office work and other developer expenses. The distinction explains why a homeowner can encounter a system quote of several dollars per watt while the underlying modules trade for cents per watt. As panels become cheaper, the rest of the stack accounts for a larger share of the bill.
There was no single invention that cut solar prices by three or four orders of magnitude. Cells became more efficient, so the same area produced more power. Silicon wafers became thinner. Factories grew larger, manufacturing yields improved, equipment became faster, supply chains thickened and modules lasted longer.
An MIT-led analysis covering 1980 to 2012 attributed the decline to several interacting technical and economic mechanisms. As an earlier Science Blog report on that work described, public policies that expanded markets played a large role, while efficiency improvements were the most important device-level factor. Research and development mattered especially early; economies of scale became increasingly important later.
The underlying 2018 Energy Policy paper by Goksin Kavlak, James McNerney and Jessika Trancik estimated that market-stimulating policies accounted for about 60 percent of the overall module-cost decline during its study period. That is not the same as saying subsidies mechanically paid for 60 percent of every price cut. The policies expanded demand, which supported manufacturing scale, learning and further technical work.
This feedback is often called a learning curve. Our World in Data estimates that, across more than four decades, each doubling of cumulative installed solar capacity was associated with a 20.2 percent decline in module price. It is an empirical pattern, not a physical law guaranteeing that the same rate will continue forever.
The International Energy Agency’s World Energy Outlook 2020 used unusually direct language: for projects with low-cost financing and high-quality solar resources, solar PV had become the cheapest source of electricity in history. The conditions in the first half of that sentence are essential to the claim in the second.
The IEA was discussing levelized cost of electricity, or LCOE. This estimates the average revenue per unit of electricity that a new plant would need to recover its construction, financing, operating and maintenance costs across its lifetime. A sunny site produces more energy from the same equipment. A lower interest rate reduces the cost of financing an asset whose expenses are concentrated at the start.
The statement did not mean that every rooftop system in every city was the cheapest power source ever built. It did not compare a solar module’s wholesale price with a household retail tariff. Nor did it say that building a new solar plant is always cheaper than continuing to operate an existing plant whose construction costs have already been paid.
Solar generates when sunlight is available, while electricity demand follows a different clock. A grid with a large solar share may need transmission, storage, flexible demand, dispatchable generation or some combination of all four. LCOE is useful for comparing generators, but it does not by itself capture every system-level cost or the changing value of electricity at different hours.
Module prices can also move upward as well as downward over short periods. Shipping constraints, tariffs, commodity prices, factory overcapacity and trade policy affect what buyers pay. IRENA attributed the especially low prices of 2024 partly to Chinese overproduction and unsold inventories, conditions that also placed financial pressure on manufacturers.
None of this cancels the historical achievement. It shows where the engineering and economic problem has moved. The photovoltaic module began as the impossibly expensive center of the system. Today it is often a commodity component, while permitting, labor, finance, grid access and balancing determine how fully cheap panels become cheap and dependable electricity.
That is the deeper meaning of the journey from $1,865 to cents per watt. Solar did not become competitive through one dramatic discovery. A protected space-age niche created a foothold, decades of production created learning, and deployment policy helped scale the market. By 2020, the best projects had moved from being merely competitive to setting the cost benchmark for new electricity.
Standards
Articles on Science Blog are edited and fact-checked before publication. We use AI tools in the newsroom. See our editorial standards and masthead.
A Science Blog column
The Long View steps back. A single finding rarely means much on its own, so this desk connects new research to what came before it, across physics, biology, medicine, and everything in between.
Clear science for curious people
A Brown Brothers Media publication. All rights reserved © 2026.
Clear science for curious people — free PDF, then our science email.
Also available via RSS.

source