Long-Run Projections of Environmental Impact in the OECD : The Importance of Uncertainty and Age Structure

This paper considers the influence of uncertainty with respect to GDP and population — both total and age structure — on long-run environmental impact projections. The paper takes as its starting point a simple, transparent, published model that is based on the stochastic version of the IPAT equation [Liddle, B. (201l) ‘Consumption-driven environmental impact and age-structure change in OECD countries: A cointegration-STIRPAT analysis'. Demographic Research, Vol. 24, pp. 749-770.]. That model is used to generate projections — for the OECD as a whole — of two environmental impacts for which population has a demonstrated influence — carbon emissions from transport and residential electricity consumption. The paper examines the sensitivity of those long-run (i.e., to 2050) projections to uncertainty regarding possible future GDP and population growth/change. The Energy Information Agency's reference, high, and low GDP cases are combined with the United Nations medium, high, and low population variants to create nine different “scenarios.” Two particularly interesting results are demonstrated: (i) if the medium path for either population or GDP is to be believed/trusted, then uncertainty with respect to the growth/change of the other factor is unimportant; and (ii) uncertainty with respect to population projections has a greater effect on the environmental impact projections than does uncertainty with respect to GDP projections — despite that the degree of uncertainty associated with population is smaller than the uncertainty associated with aggregate economic growth. That second, perhaps surprising, finding was driven by the environmental impact projection models' inclusion of population age structure's effect and the expectation of aging underlying the population assumptions/projections