The rapid increase in global energy demand is likely to strain existing energy infrastructure, environmental resources, and the welfare of low-income households. Between 2010 and 2035, most of the increase in global energy consumption is expected to come from developing countries, as increasing incomes enable households to obtain appliances for the first time.1 In response, policymakers have designed programs to increase energy efficiency and address related environmental and economic factors.
Energy efficiency programs may lessen the demand on the energy supply infrastructure and help consumers lower their energy expenditures through reduced consumption. Few studies have been able to empirically measure the effectiveness of energy efficiency programs due to lack of high-quality data.
One such energy efficiency program, “Cash for Coolers” (C4C), was established in 2009 in Mexico. C4C subsidizes the replacement of refrigerators and air conditioners that are at least 10 years old with newer, more energy-efficient models. These appliances are key contributors to carbon emissions in the developing country context as households are more likely to purchase refrigerators and air conditioners before automobiles. As of 2009, 1.5 million households replaced their old appliances with more efficient models through C4C.
Independent analyses concluded that an appliance replacement program would be extremely cost effective in Mexico because of the energy savings from appliance efficiency.2 However, because energy-efficient goods are less costly to operate, households may actually use them more. The overall increase in consumption resulting from a price decrease is known as a “rebound effect,” and may be of particular importance to policymakers in developing countries.
This study empirically examines the effect of the C4C program by utilizing energy consumption micro data from Mexico’s exclusive supplier of electricity, the Mexican Federal Electricity Commission (CFE). Analysis was done using a panel dataset of bimonthly electricity consumption and expenditures from May 2009 to April 2011 obtained from billing records. Researchers were able to match 86 percent of C4C participants with their billing records, resulting in a sample of 957,080 households.
Because C4C was non-randomly implemented, the authors utilized a Difference-In-Differences approach, with controls for household location and month, to estimate the true effect of the program. This method compares electricity consumption before and after appliance replacement, using households in the same county that did not receive a replacement as a comparison group. The panel data allowed the researchers to control for the variation in electricity consumption over time that was not attributable to C4C and to provide an unbiased estimate of the program’s effect. The large-scale analysis using nationwide utility records gives the results a high degree of external validity, indicating they are nationally representative.
Results and Policy Implications
While the program lowered energy consumption from refrigeration by 132 kilowatt-hours per year, the program resulted in an increase in energy usage from replaced air conditioners of 80 kilowatt-hours per year. The researchers estimate that the total reduction in electricity consumption from the program was approximately 100 gigawatt hours annually, or 0.2 percent of the total residential electricity sales in Mexico in 2009. C4C also proved to be extremely cost-inefficient. The cost per ton of carbon dioxide emissions abated exceeded $400. By comparison, the 2012 price for an emissions permit from the European Union Emissions Trading System was about $10 per ton, a fraction of the cost of C4C.
According to the authors, the increase in electricity consumption from air conditioners is from the “rebound effect.” The decreased cost of operating the air conditioner resulted in higher use by households. In fact, usage was much higher during the hot summer months, which suggests households had been substituting less expensive and less energy-intensive cooling methods for air conditioning (e.g. fans, opening the windows, etc.). Overall reductions in energy consumption were modest, likely because energy-efficiency savings from the new appliances were offset by the energy intensity of larger appliances (which typically have more features). The results underscore the importance of empirical evidence to inform program scale-up, and the need to account for changes in utilization when considering promotion of more efficient technologies.
1 Wolfram Catherine, Orie Shelef and Paul Gertler (2012) “How Will Energy Demand Develop in the Developing World?” Journal of Economic Perspectives, 26(1), 119-138.
2 Johnson, Todd M., Claudio Alatorre, Zayra Romo, and Feng Lui (2009) “Low-Carbon Development for Mexico,” World Bank, Conference Edition; McKinsey and Company, “Low Carbon Growth – A Potential Path for Mexico,” presentation by Francisco Barnes Jr. at Carbon Markets Workshop, August 20, 2009.
Photo credit: Air Conditioner Row by ToddMorris via Flickr.