We develop a quantitative macroeconomic theory of input-saving technical change to analyze how markets economize on scarce natural resources, with an application to fossil fuel. We find that aggregate US data call for a very low short-run substitution elasticity between energy and the capital/labor inputs. Our estimates imply that energy-saving technical change took off when the oil shocks hit in the 1970s. This response implies significant substitutability with the other inputs in the long run: even under ever-rising energy prices, long-run consumption growth is still possible, along with a modest factor share of energy.