Comparison to partial-factor CO 2 intensity

Table 17 shows that the partial-factor CO2 intensities, of all economies except China and Singapore are steady with small changes. Hong Kong, Peru, and the Philippines are the three most efficient economies, and Australia is the worst. China improved its CO2 intensity, but Chile’s CO2 intensity was increasing slightly at the same time. By applying Wilcoxon signed-rank test to compare the total-factor CATR to CO2 intensity, the result (W+ = 87, W- = 83, Z = -2.129, p-value = 0.03) is significant at the 5% level and shows that the total-factor CATR has significantly different rank patterns with the partial-factor CO2 intensity. In addition, the relation

between CO2 intensity and per capita GDP does not have a significant pattern as with the inverted U-shape relation between per capita CATR and per capita GDP.

This shows a significant substitution effect of other inputs such as labor and capital on the energy input to produce the GDP. The CO2 abatement efficiency could be over-estimated or under-estimated if CO2 emission is taken as a single factor to measure. This study hence applies a factor framework, with which the total-factor IRTR is established.

Table 17 CO2 intensity for each APEC economies (1991-2000)

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Australia 0.21 0.21 0.21 0.21 0.21 0.21 0.20 0.20 0.20 0.20 Note: The unit is toe/US$ billion purchasing power parity, at 1995 international prices.

4.6 Relation among ESTR and industrial structure indicators

We use panel data regression models to find out the relation between industrial structure and ESTR. Fifteen economies are selected, except South Korea and Singapore, since there is a limitation of data. The data of value-added percents of GDP by industry and service sectors for every economy are taken from World

Development Indicators (World Bank, 2005). The Hausman test does not reject the random-effects model at the 5% level (CHISQ = 0.70, p-value = .71). In the random-effects model’s estimates shown in Table 18, ESTR has a positive relation with value-added percent of GDP by industry sector and a negative relation with that of the service industry - that is, ESTR increases with industrialization and then decreases with the rising service industries. According to this relation, a newly industrializing economy will have lower total-factor energy efficiency than agriculture-dominant and service-dominant economies. The industrial structure of an economy is hence a crucial factor for energy efficiency and thus the energy-saving ratio. An industry-dominant economy can improve its energy efficiency and save energy more efficiently and effectively via shifting the economy structure toward services.

Due to data limitation, we can only find the retail prices of oil in 1997 for 17 economies from APERC (2000). However, there is neither a significant relation found between ESTR and the retail price of oil nor one between per capita EST and the retail price of oil. This may be because energy prices alone cannot determine the total energy efficiency and energy saving of an economy. The structure of energy mixes, energy efficiency, taxation, and relative prices for all energy resources includes the factors influencing energy use and the energy saving of an economy.

Table 18 Relation among ESTR and industrial structure indicators for APEC economies (random-effects panel data model estimation)

Variable Coefficient t-statistic p-value Value-added percentage of

GDP by the industry sector 0.60 2.328 0.020**

Value-added percentage of

GDP by the service sector -0.73 -3.991 <0.001***

Constant 0.41 2.291 0.022**

R² 0.432

Note: **represents significance at the 0.05 level; ***represents significance at the 0.01 level.

4.7 Relation among CATR and industrial structure indicators

The same dataset as 4.2.1 is used with panel data regression models to find out the relation between industrial structure and CATR. The Hausman test does not reject the random-effects model at the 5% level (CHISQ = 0.97, p-value = .61). In the random-effects model’s estimates shown in Table 19, CATR has not a significant relation with value-added percent of GDP by industry sector, but has a significant negative relation with that of the service industry – that is, CATR does not change with industrialization but then decreases with the rising service industries.

According to this relation, a service-dominant economy has higher total-factor CO2

abatement efficiency. On the other hand, the industrial structure of an economy may not be a crucial factor for CO2 abatement and thus the CO2 abatement ratio.

APEC economies have followed the industrialization pattern of developed countries shifting one after another from labor-intensive to capital-intensive industries, and then recently, even to technology-intensive industries. While technology transfer itself does not necessarily cause such shifts, they are at least promoted thereby. However, we should note that technology transfer could have a double-edged effect on emissions. On the one hand, it can lead to enlarged industrial capacity, resulting in increased pollutant emissions, but on the other hand, abatement technology can be also made available to the recipient economy.

Whether the net effect on emissions is positive or negative depends in part on the characteristics of the technology and on levels of public and government awareness (Iwami, 2004).

The former negative effect is related to the degree of industrialization achieved by an economy, whereas the latter positive effect is not easily measured. Because the correlation is either positive or negative, it implies that factors other than the negative effects of technology transfer are, in fact, at work. An “advantage of latecomer” would imply that economies industrializing later would complete the process in a shorter time and/or with better performances. Factors related to this

issue include not only technology transfer and initiatives on the part of government and private institutions (for example, banks), but also learning from the experiences of advanced economies. The existence of the “advantage of latecomer” can explain the no significant relation between CATR and value-added percent of GDP by industry sector.

Table 19 Relation among CATR and industrial structure indicators for APEC economies (random-effects panel data model estimation)

Variable Coefficient t-statistic p-value Value-added percentage of

GDP by the industry sector 0.38 1.431 0.152

Value-added percentage of

GDP by the service sector -1.09 -5.879 <0.001***

Constant 0.68 3.745 <0.001***

R² 0.521

Note: **represents significance at the 0.05 level; ***represents significance at the 0.01 level.