Time-varying human capital and energy consumption for a panel of OECD countries in the long-run

AbstractWe examine how human capital influences energy consumption in a panel of OECD nations in the long run. We make important contributions to understanding how education affects energy consumption. First, much of the existing research on how human capital affects energy consumption, employs time series or panel data which typically span a few decades. We utilise a newly assembled long-run panel spanning 150 years, disaggregated by primary, secondary, and tertiary levels, for a core set of OECD countries. This long panel enables us to examine how the human capital–energy relationship evolved during the Industrial Revolution, multiple energy transitions, and major global shocks. Second, most prior studies rely on parametric models that assume constant relationships over time. Such approaches yield average effects but fail to capture how the education–energy nexus shifts in response to changes in policies, technologies, and macroeconomic conditions. To overcome this limitation, we employ both a parametric and a semi-parametric estimator, which generates time-varying elasticities. Our parametric results highlight the heterogeneous effects of education, with primary and secondary schooling associated with higher energy consumption and tertiary education linked to lower energy consumption. Semi-parametric findings show that primary and secondary education contributed strongly to energy-intensive growth in the early stages of development, but their influence diminished as economies shifted toward services, while tertiary education became increasingly connected with lower energy use in later decades.