| Summary: | 碩士 === 國立暨南國際大學 === 經濟學系 === 92 === The objective of this thesis is to apply a research method that integrates band spectrum regression (BSR, a frequency domain approach) and vector autoregression (VAR, a time domain approach) to analyze the long-term and short-term dynamic linkages between energy consumption and macroeconomic variables in Taiwan and the United States. Based on the vector autoregression model, we first filter out frequencies of variable series by spectral analysis, and then convert the filtered series into time domain representations for conducting linkage analysis among variables. We also use cross-spectral analysis to estimate the lead-lag structures of two variable series at different frequencies. Variables included in this study are: total energy consumption, energy price, M2 and real industrial production. Monthly data covering periods from 1980:01 to 2001:09 are used for empirical estimation and analysis.
The results of this study reveal that, in Taiwan, long-run change in money growth will affect energy consumption. However, the change in real industrial production (no matter it is short-, mid-, or long-term change) will not affect energy consumption, whereas the mid-term energy price change will affect energy consumption significantly. For the United States, the results indicate that the change in money growth will not affect U.S. energy consumption. Change in energy consumption also will not affect real industrial production as well. However, changes in energy price do have significant effect on energy consumption.
Due to the existence of differential endowment in natural resources, difference in industrial structures, and differential level of economic development between Taiwan and the United States, we did find different linkage relationship between energy consumption and macroeconomic for the two economies. In addition, the results of cross-spectral analysis show that Taiwan’s energy consumption leads all macroeconomic variables, while the results are reversed for the United States.
|
|---|
