A critique of Saunders' 'historical evidence for energy efficiency rebound in 30 us sectors'

A recent article in Technological Forecasting & Social Change presents a calculation of historical rebound effects in thirty sectors of the United States economy over the period 1960-2005 (Saunders 2013). Here, we show that the empirical data set used to generate those findings-a prominent input-output data set developed by Jorgenson (2007)-is not appropriate for the use to which Saunders puts it. Saunders' model requires annual data on the price and quantity of energy consumed by each sector; however, the Jorgenson data are inferred from national prices, not prices observed at the sector level and disaggregated by geographic region. Furthermore, Jorgenson reports average prices, rather than marginal prices; yet the rebound effect is caused by changes in marginal price of energy services. We compare the differences between national prices and sector-specific prices across geographic regions in the United States, demonstrating that Saunders' use of national average energy prices is inappropriate for investigating the rebound effect.