A stable and accessible power supply is pivotal for economic growth and development. We study the impact of nationwide residential electrification in Cambodia on sectoral employment. Leveraging annual village-level administrative data from 2013 to 2021, we estimate the employment effects of electrification using a novel difference-in-differences estimator designed to accommodate staggered adoption and treatment-effect heterogeneity across groups and over time. In our preferred specification, electrification reduced agricultural employment by around 2.4 percentage points in the long run but raised service employment by only 1.5 percentage points; the agriculture-to-services transition was also more pronounced among working-age women. The magnitude of these estimates is substantially smaller than those reported in earlier studies, consistent with a pattern of premature deindustrialisation in which electrification induces only limited structural transformation. The modest inter-sectoral shift in labour appears to operate through a mix of channels, including agricultural productivity, labour demand, investment in human capital, and migration, although the contribution of each channel, considered in isolation, is minor. Our findings highlight that agrarian workers who detach from the labour market may not necessarily re-enter it when complementary markets and demand are absent. We further underscore the importance of comparing model estimates across alternative constructions of the treatment variable, as the novel estimator we employ permits continuous, discrete, or binary treatments.