Electrification and Cooking Fuel Choice in Rural India: Summer Research
Written by: Ridhima Gupta
Edited by: Alicia Oberholzer
We investigate the existence of a causal link between electrification and the adoption of LPG for cooking. Our hypothesis is that the increase in income generated by electrification (see Chakravorty et al., 2014; Lipscomb et al., 2013; Dinkelman, 2011) should push households to increase their use of LPG, a normal good, and decrease their use of fuelwood and other cheaper polluting fuels. However, identifying the impact of electrification on modern fuel use presents a number of empirical challenges. This is because electrification is not randomly assigned and therefore it may be correlated with other determinants of fuel use. We tackle this problem by using an instrumental variables approach. Following Bartik (1994), we construct our instrument by multiplying state level shifts in hydroelectric power availability (Allcott et al., 2016) by the initial level of electrification of each district, measured using the mean light intensity emanating from it.
We use data from three waves of the National Sample Survey (NSS), 2004, 2009 and 2011. This data allows us to construct a nationally representative survey of rural Indian households. We then analyse the intensive and the extensive margin of cooking fuel decisions. The extensive margin focuses on the adoption of a given cooking fuel. In this case, the outcome of interest is a binary variable indicating whether a household uses a particular cooking fuel or not. The extensive margin denotes the intensity of use of a given fuel, measured by calculating the share of expenditure on a given fuel type over total monthly expenditure by a household. We examine the choice of LPG and fuelwood.
We find that electrification leads to an increase in the probability of adoption of fuelwood and a decrease in the probability of adoption of LPG. These results may seem counter-intuitive in light of the energy ladder theory. Yet, once we split the sample according to median income, we find that these effects are statistically significant only for the poorest 50% of the sample. The same is true for the intensive margin. Two hypotheses may explain these results. First, it seems that for poor households, receiving an electricity connection and, therefore, facing a new expenditure creates an additional strain. In order to still be able to cook and also enjoy the benefits of electricity (e.g. reading or irrigation), these households revert to free cooking fuels, such as fuelwood. Another hypothesis is related to the quality of the power supply. As found in Chakravorty et al. (2014), a poor power supply has a much smaller impact on income, not allowing households’ income to increase sufficiently to push them to the next step of the energy ladder.
Our findings imply that, for the poor, access to electricity makes the switch to cleaner cooking fuels even harder. One of the takeaways of the paper for policy-makers is that the introduction of electricity should be accompanied by subsidies for the adoption of cleaner cooking fuels, to avoid a worsening of indoor air pollution for the poorest households.