One of the most interesting and important papers I’ve recently read is “How Pro-Poor Growth Affects the Demand for Energy” (Gertler, Shelef, Wolfram, Fuchs). The paper begins with the notion that as household incomes increase, the demand for discrete assets that consume high amounts of electricity, such as fridges, grows rapidly. Consequently, economic development may result in a massive, discontinuous increase in residential electricity consumption. As large segments of the population leave extreme poverty, they start to purchase assets that lock in a high level of electricity consumption.
The evidence for the claim comes from impact evaluation data on a large cash transfer intervention in Mexico. The authors show that, by increasing household incomes, the impact evaluation caused a spike in the demand for refrigerators over time. Consistent with this result, cross-sectional data across households shows that electricity consumption went up.
The authors then conduct a simple prediction exercise for electricity consumption in developing countries between 1980 and 2006 using national data from 37 developing countries. The main finding is that higher income levels are strongly associated with electricity consumption in countries where economic growth is associated with low levels of income inequality.
Why is this such a big deal? Concerns about income inequality have prompted many emerging countries from China to India to invest in pro-poor policies. Our current demand forecast models do not account for these policies and may, therefore, underestimate the growth of electricity demand in the future. Since most demand growth is bound to come from developing countries, today’s forecasting models may understate the difficulty of providing billions of people with affordable energy while combating climate change. Further research on this topic should be at the very top of the research agenda in energy economics. For once, I kind of regret not switching from political science to economics. This is the kind of research that could and should have a huge impact on policy.