MPIDR Working Paper

Demographic dynamics and per capita environmental impact: using panel regressions and household decompositions to examine population and transport

Liddle, B.

MPIDR Working Paper WP-2003-029, 21 Seiten (August 2003).
Rostock, Max-Planck-Institut für demografische Forschung

Revised August 2003; Also published in: Population and Environment 26(2004)1: 23-39.

Schlagworte: OECD countries, energy consumption, environmental policy, household size, transport


Demographic variables have tended to be ignored in many environment-development analyses. This paper examines how population changes (in aging, households, and urbanization/density) can help explain changes/differences in personal transport using both macro- and micro- level data. First, panel regressions are performed with IEA-OECD road sector energy use data (spanning 1960-2000) on spatial population measures, average household size, and age structure data. Then US household data is used to determine the extent compositional changes in the nature of households can explain changes in per capita driving. An Environmental Kuznets Curve for per capita road energy use was rejected—the coefficients on the GDP squared terms were insignificant and the implied turning points were well outside the sample range; instead, the relationship between wealth and road energy was found to be monotonic (log-linear). The ideas that more densely populated countries have less personal transport demands, the young drive more, and smaller households mean higher per capita driving were confirmed. The basic result from the household decompositions was that changes in demand were more important than compositional changes, however, during some periods the compositional change component was considerable. A few policy implications can be drawn from these analyses. First, the look at micro data implies that there is much potential for policy to affect transport behavior since the compositional component of change—more difficult for policy to alter—is smaller than the behavioral or demand component. However, the look at the macro data implies that spatial factors, like population density and urbanization—which also can be difficult to alter—are significant in influencing personal transport demand.