Associations between reported socio-economic hardships and key household energy-use outcomes (kerosene used for lighting, primary cooking fuel is polluting and any polluting cooking fuel used) from regressions carried out using the fixest package in R that contains (1) a dummy variable for reporting to have experienced any socio-economic hardship in that round and (2) fixed effects for household and round, with standard errors clustered at the household level. The coefficient for the socio-economic hardship dummy variable was exponentiated to estimate the odds ratio and 95% confidence interval plotted in the panel. The sample sizes for these regressions were n =2,868, n =2,761 and n =2,499, respectively.
