Natural gas access and household fuelwood harvesting: Two-wave difference-in-differences evidence

Household energy transitions are widely documented, yet isolating the causal effect of modern-fuel access on fuelwood harvesting remains challenging because much of the evidence relies on cross-sectional comparisons or aggregated data. Using two waves of household survey data (pre- and post-expansion) and a comparison group of villages that never received gas, we estimate two-period difference-in-differences models in first differences with generalized estimating equations and village-clustered robust inference. Relative to never-gas villages, gas access is associated with roughly a one-third reduction in total fuelwood harvesting and about a one-half reduction in firewood harvesting. By contrast, estimated effects on charcoal are small and statistically indistinguishable from zero, consistent with fuel stacking and the distinct livelihood and market role of charcoal. Heterogeneity analyses indicate significant firewood reductions across baseline income groups, with the largest declines among middle-income households, and larger reductions among households primarily engaged in agriculture/livestock than among others. Reported reasons for continued biomass use emphasize affordability pressures and task-specific preferences—especially bread baking—alongside limited salience of health risks. Overall, infrastructure expansion can meaningfully reduce routine firewood extraction, while charcoal-related forest pressures are less responsive and may require complementary approaches beyond access alone.