Energy, the economy and the future: Energy and the economy

1.1 Every economic activity is an energy conversion

Mainstream growth theory explains output with labour, capital and technology, and treats energy as one input among many, worth only what is spent on it. In rich economies spending on energy is usually a small share of GDP, often well under a tenth, so by that logic energy looks marginal. Smil argues that this badly understates its role. Without a steady supply of fuels and electricity, labour and capital produce almost nothing: no farm machinery, no fertiliser, no steel, no freight, no data centres. Energy is less like one ingredient in a recipe and more like the heat that cooks the whole meal.

Across countries, the link between energy and income is strong. Plot energy use per person against GDP per person for the world's nations and the two rise together; Smil's Global Ecology reported a worldwide correlation above 90%. At the extremes the gap is stark: the poorest countries use a few tens of GJ of commercial energy per person a year at most, the richest well over a hundred.

Among countries at similar incomes, however, the link is loose. Japan and the large European economies use roughly half as much energy per person as the United States or Canada, at broadly comparable living standards. Climate, the size of the country, how densely people live, the weight of heavy industry, fuel prices and plain habit all matter.

Smil's long historical view adds a caution. Energy sets the limits of what a society can do, but it does not decide what that society does within those limits. Institutions, ideas, trade and politics shape whether abundant energy produces broad prosperity or waste. In Grand Transitions he links five great changes, in population, food, energy, economies and the environment, and shows that each depended on the others. None of the economic transformations of the past two centuries could have happened without fossil fuels, but fossil fuels alone did not make them happen.

• Smil, Energy at the Crossroads: Global Perspectives and Uncertainties (2003) · via Smil Mastery Course, M13 'Energy and economic growth'
• Smil, Global Ecology: Environmental Change and Social Flexibility (1993) · via Smil Mastery Course, M13 'Energy and economic growth'
• Smil, Grand Transitions: How the Modern World Was Made (2021) · via Smil Mastery Course, M13 'Energy and economic growth'
• Smil, Energy and Civilization: A History (2017) · via Smil Mastery Course, M13 'Energy and economic growth'

1.2 Energy intensity falls as economies mature, but comparisons need care

Energy intensity is primary energy used per unit of GDP, for example megajoules per dollar. A falling intensity means an economy is producing more output from each unit of energy.

The typical path has a hump. In early industrialisation intensity rises, because a country is building railways, steel mills, cement plants, power stations and cities, all of which use a lot of energy per dollar of output. Later it falls, as the economy shifts towards services, machines become more efficient and the basic infrastructure is finished. China is a remarkable case. Smil showed that its energy intensity began to fall even before the economic reforms of 1979 and then dropped steeply through the 1980s and 1990s, unusually early for a country still industrialising so fast.

Before comparing countries, Smil insists on taking the number apart. Converting GDP at market exchange rates makes poor countries look extremely energy-intensive; converting at purchasing-power parity, which reflects local prices, shrinks that apparent gap considerably. Whether traditional biomass is counted changes the energy figure for countries such as India or Nigeria. Cold climates need more heating and large countries more transport. An economy built on aluminium smelting or oil refining will look intensive even if each of its plants is efficient.

Two traps follow. First, falling intensity is not falling use. World energy intensity has declined for decades while total energy use kept rising, because GDP grew faster still. Second, rich countries flatter their figures by importing steel, cement, chemicals and manufactured goods. The energy used to make them, and the emissions, are counted where the factories are, often in China or India.

Intensity remains a useful policy target. At the COP₂₈ climate conference in 2023, governments agreed to aim to double the global rate of energy intensity improvement to about 4% a year by 2030. The IEA's Energy Efficiency 2025 report shows progress well short of that: about 1% in 2024 and, on preliminary estimates, 1.8% in 2025.

• Smil, Energy at the Crossroads: Global Perspectives and Uncertainties (2003) · via Smil Mastery Course, M13 'Energy intensity'
• Smil, China's Past, China's Future: Energy, Food, Environment (2004) · via Smil Mastery Course, M13 'Energy intensity'
• Smil, Global Ecology: Environmental Change and Social Flexibility (1993) · via Smil Mastery Course, M13 'Energy intensity'
• IEA, Energy Efficiency 2025 (2025) · improvement of 1% in 2024 and 1.8% (preliminary) in 2025, against the COP28 doubling goal, https://www.iea.org/reports/energy-efficiency-2025

1.3 Energy prices fell for a century, but real costs include what prices leave out

For most of the twentieth century the real, inflation-adjusted prices of the main energy services fell. Electricity became cheaper, and much cheaper still relative to wages. Light became astonishingly cheap compared with candles or kerosene lamps. Oil, despite the shocks of the 1970s, was cheap for most of the century. Falling prices drove consumption, and cheap energy is one of the hidden foundations of modern productivity.

Market prices, however, leave out many costs. Burning coal and oil causes air pollution that shortens millions of lives a year; acid deposition damaged lakes and forests; CO₂ changes the climate; oil spills foul coasts; and some countries spend heavily on military power to keep oil flowing. Economists call these external costs. Smil reviews many attempts to put a money value on them and stresses their wide uncertainty: estimates can differ tenfold or more depending on assumptions about health effects, discount rates and future damage. Uncertainty is not a reason to treat them as zero, which is exactly what an unadjusted market price does.

Subsidies distort prices further. Many governments, India's among them in the past, have held fuel and electricity prices below cost for consumers, and others subsidise producers. New renewables have also received large support through feed-in tariffs, tax credits and guaranteed contracts. Comparing technologies on subsidised prices can mislead in either direction.

The newest price story is the collapse in the cost of clean technologies. The IEA estimates that since 2010 the cost of solar power has fallen by about 90%, wind by about 70% and batteries by about 90%. In many markets, including India, new solar is now the cheapest source of bulk electricity. But the cost of a kWh at the plant gate is not the cost of a reliable supply at the moment it is needed. Integration costs, for storage, grids and backup, rise as variable sources grow, and they are where the next arguments about price will be fought.

• Smil, Energy at the Crossroads: Global Perspectives and Uncertainties (2003) · via Smil Mastery Course, M13 'Prices, real costs and energy in war'
• Smil, Energy in Nature and Society: General Energetics of Complex Systems (2008) · via Smil Mastery Course, M13 'Prices, real costs and energy in war'
• Carbon Brief, 'IEA: Fossil-fuel use will peak before 2030 unless stated policies are abandoned' (November 2025) · summary of IEA World Energy Outlook 2025 scenario results, https://www.carbonbrief.org/iea-fossil-fuel-use-will-peak-before-2030-unless-stated-policies-are-abandoned/
• IEA, World Energy Outlook 2025 (2025) · Overview and key findings, https://www.iea.org/reports/world-energy-outlook-2025/overview-and-key-findings

See it in the table

Check your understanding

Why does energy look unimportant in standard growth accounting?

A poor country's energy intensity looks very high when GDP is converted at market exchange rates. What usually happens if purchasing-power parity is used instead?

According to the IEA, roughly how much did the cost of solar power fall between 2010 and 2025?