In 1980 electricity from solar panels was a luxury for satellites and remote lighthouses. Today new solar and wind farms are among the cheapest ways to make electricity in most of the world. Few technologies have fallen in price so far, so fast, and the reason says much about how technology changes.

Early sunlight and wind

People have used wind for sails and mills for thousands of years. In the 1860s and 1870s the French teacher Augustin Mouchot built solar engines that focused sunlight to make steam, but cheap coal made them uneconomic. In 1839 the French physicist Edmond Becquerel found that certain materials produce an electric current when light falls on them, the photovoltaic effect. In 1954 researchers at Bell Laboratories made the first practical silicon solar cell, which turned about six percent of sunlight into electricity.

A solar cell has no moving parts: light knocks electrons loose inside a material such as silicon, and they flow as a current. Early cells were fantastically expensive. Satellites became the first big market because in space reliability and low weight mattered far more than price; the American satellite Vanguard 1 carried solar cells in 1958.

Wind turbines followed a similar path from experiment to industry. Charles Brush lit his Cleveland mansion with a giant wind dynamo in 1888, and in the 1890s Poul la Cour ran wind electricity experiments in Denmark. The modern industry grew in Denmark after the oil shock of 1973, and turbines have since grown much taller, because stronger, steadier winds blow higher up and longer blades sweep more air. The first offshore wind farm opened off Denmark in 1991.

Learning curves

In 1936 the American engineer Theodore Wright studied airplane factories and found that each doubling of the total number of planes built cut the cost of making one by a fairly steady percentage. Workers learned, factories improved and suppliers grew. This pattern, now called a learning curve or Wright's law, fits solar panels strikingly well. Module prices have fallen by roughly a fifth with each doubling of total production, and by more than ninety nine percent since the mid 1970s.

Why did solar follow the curve so closely? A solar farm is built from millions of identical small panels, and a wind farm from dozens of identical turbines. Each unit is made in a factory, tested and improved, again and again. Small, standard units made in huge numbers tend to get cheaper faster than large projects built one at a time, such as dams or nuclear plants, where each site is different and lessons are learned slowly.

Policy and the factory floor

Learning needs buyers. In 2000 Germany passed a law that guaranteed fixed prices for twenty years for renewable electricity fed into the grid, a scheme called a feed-in tariff. Japan, California and others ran their own programs. China then built factories on an enormous scale with state support, and it now makes more than four fifths of the world's solar panels at most stages of production. Economists still argue over whether these subsidies were the cheapest route, and over the risks of depending on one country's factories.

Measuring cost fairly

No power plant runs flat out all the time. A plant's capacity factor is the share of its maximum possible output that it actually produces over a year. Solar farms often reach ten to twenty five percent; onshore wind roughly twenty five to forty percent; and nuclear plants in the United States above ninety percent. Analysts often compare sources by levelized cost, the total lifetime cost of a plant divided by all the electricity it will make. On that measure new solar and onshore wind now beat new coal and gas plants in much of the world. But levelized cost ignores when the power arrives and the extra costs of backup, storage and grid connections that variable sources need. Supporters reply that such system costs are real but have so far been modest at moderate shares of wind and solar.

Fair criticism

Solar and wind farms use a lot of land, and neighbors often object to the view. Turbines kill birds and bats, though far fewer birds than cats or glass windows, and poorly sited farms can harm rare species. Turbine blades are hard to recycle. Hydropower, the oldest large renewable source, has its own costs: China's Three Gorges Dam displaced over a million people. Even so, hydropower supplied more of the world's electricity in 2024 than any other low-carbon source.

Worked example: Germany's expensive gift

Under its 2000 law, Germany paid solar owners very high guaranteed prices. Households paid through a surcharge on power bills that reached billions of euros a year. Solar spread across German rooftops, and German factories boomed. Then, around 2010, Chinese manufacturers expanded so quickly that prices crashed, and most German panel makers went out of business.

Apply the ideas. Germany's demand pushed the world down the learning curve, and later buyers everywhere got much cheaper panels. Seen one way, German consumers overpaid for early, costly panels. Seen another, they bought the world a cheap technology. The case shows that a learning curve is a shared good: the country that pays for the early steps does not keep all the benefits.