This paper proposes a novel approach to improve accuracy in long-term
forecasts based on nonlinear models, and provides an empirical analysis of
long-run electricity demand forecast for Korea. Our approach is based on the
notions of catch up and convergence, which exploits the fact that various
economic characteristics of developing countries evolve following closely with
some time lags the historial patterns of corresponding characteristics of
developed countries. To forecast electricity demands in developing countries,
for instance, we may therefore utilize the historical patterns of the
determinants of electricity demands in developed countries. In particular, we
find that the time series of income elasticity of electricity demand in Korea
evolves in the same pattern as that of electricity demand in Japan. The
patterns are extracted using cointegrating regressions with time-varying
coefficients, and the actual gap and catch-up periods are determined to
maximize the performance of our forecasting model. Our procedure is shown
to perform substantially better than other models that are commonly used in
long-term forecasts. In particular the optimal forecasts are obtained under the
assumption that Korea’s catch-up period is shorter than the gap period
between Korea and Japan. Our empirical results are also consistent with those
done by the studies of the convergence of productivity.