Grouping factors and intergroup heterogeneity in cooking fuel choice: evidence from post-earthquake Nepal

Which socioeconomic factor as a basis for grouping population yields the highest intergroup, but the lowest intragroup, heterogeneity in cooking-fuel choice? In this paper, using post-earthquake data on 747,137 households from Nepal, we construct a Euclidean dissimilarity matrix that exhibits the link between the households’ cooking-fuel choice and their socioeconomic group identities. We then employ PERMANOVA, a distance-based multivariate semiparametric method, and find that ethnicity as a grouping factor leads to about 39.1% of intergroup variance in cooking fuel choice, followed by income (26.3%), education (12.6%), and location (4.1%). We also find two distinct clusters of ethnic groups exhibiting similar fuel-choice behaviors. These findings underscore the importance of ethnic-group specific policies in promoting clean cooking in post-earthquake Nepal.<br>