The behavioral change after 1973 was largely due to the oil shocks of 1973 and 1979, which rapidly altered energy prices, changed perceptions of the future price and availability of fossil fuels, and brought about policy changes. Energy users made three effective kinds of responses (U.S. Department of Energy, 1989; Schipper et al., 1990). First, they changed the way they operated energy-using equipment, curtailing heat and travel, and improving management, such as by tighter maintenance of furnaces. Such changes accounted for 10-20 percent of national energy savings achieved in 1986 (compared with the pre-1973 trend; estimates from U.S. Department of Energy, 1989) but are easily reversed when energy prices drop or incomes rise, as they did in the 1980s.
Second, energy users adopted more energy-efficient technology to provide the same service with less energy use, either by retrofitting existing equipment (e.g., insulating buildings, installing reflecting windows) or by replacing existing equipment with more energy-efficient models. These improvements were responsible for 50-60 percent of total energy savings by 1986.
Third, the mix of products and services in the economy changed. Demand fell sharply in energy-intensive industries, such as primary metals, relative to less energy-intensive industries; small cars got an increased share of the automobile market; and commercial airlines improved the match between aircraft size and demand on passenger routes. Together, such shifts accounted for about 20-30 percent of the energy savings achieved in 1986.
Higher real energy prices are generally considered the most important single explanation for these responses (International Energy Agency, 1987; U.S. Department of Energy, 1989). However, price is not the whole story. Although the two energy shocks of the period had very similar price trajectories, the effects on the economic productivity of energy differed markedly after the first two years (see Figure 4-2). For the first two years of each shock, real energy prices increased about 40 percent and energy productivity increased about 5 percent. But over the longer-term, the second shock had much more effect than the first. A five-year price increase of about 45 percent in 1973-1978 increased energy productivity 7 percent; a similar increase in 1978-1983 increased energy productivity 18 percent. Moreover, the trend continued through several years of falling real energy prices.
Why the different reactions to the two energy shocks? One explanation is perceptions: it took the second shock to get energy