Climate Change Taxes and Energy Efficiency in Japan

Abstract

In 2003 Japan proposed a Climate Change Tax to reduce its CO₂ emissions to the level required by the Kyoto Protocol. If implemented, the tax would be levied on fossil fuel use and the revenue distributed to encourage the purchase of energy efficient equipment. Analysis using the MIT Emissions Prediction and Policy Analysis (EPPA) model shows that this policy is unlikely to bring Japan into compliance with its Kyoto target unless the subsidy encourages improvement in energy intensity well beyond Japan’s recent historical experience. Similar demand-management programs in the US, where there has been extensive experience, have not been nearly as effective as they would need to be to achieve energy efficiency goals of the proposal. The Tax proposal also calls for limits on international emission trading. We find that this limit substantially affects costs of compliance. The welfare loss with full emissions trading is 1/6 that when Japan meets its target though domestic actions only, the carbon price is lower, and there is a smaller loss of energy-intensive exports. Japan can achieve substantial savings from emissions trading even under cases where, for example, the full amount of the Russian allowance is not available in international markets.

Appendix A. Trends in Japan’s Energy and Carbon Intensity

Carbon intensities for all major industrial countries have decline over the 1980–2000 period, however, Japan’s intensity shows little if any improvement in the 1990’s, as presented in Figure A1 (For these comparisons, GDP in Purchasing Power Parity (PPP) terms as computed by the World Bank was used). As a result of this slower improvement, France and Italy overtake Japan. A lower aggregate carbon intensity for a country does not necessarily translate to greater efficiency in individual processes, and thus, for example, this evidence on aggregate intensity does not necessarily contradict the data on energy efficiency of particular processes in material terms that are the basis of the Nippon Keidanren (2004a, b) claim, discussed in Section 2.

Figure A1.

Carbon intensity in selected developed countries Source: Institute of Energy Economics, Japan (2004).

Broad structural differences among countries, in addition to difference in technical efficiency in individual processes can lead to differences. For example, relative to Japan, France relies more heavily on nuclear electricity, while Italy imports electricity from other countries and thus any emissions related to this electricity production does not show up in Italy’s carbon account. The mix of fossil fuels, industrial structure, land area, climate, and availability of non-fossil resources such as hydro capacity can also have a large effect on international comparisons of both energy and carbon intensity.

The apparent slowdown in Japan’s improvement is more obvious when we extend the time series back to 1970 (Figure A2). We use ordinary least squares (OLS) to fit an exponential time trend for each decade. The estimated coefficient on the year (Y) is then an estimate of the average annual rate of change in energy intensity (EI) and carbon intensity (CI). These calculations show that energy intensity of GDP declined by an estimated 1.9% annually in the 1970’s. This was a period with rapidly rising energy prices brought about by the first oil crisis of the early 1970’s. Rising prices and concerns about fuel embargos produced strong incentives for energy efficiency and structural change. At the same time, Japan’s economy grew at 4% annually. In the 1980’s the average annual improvement in energy intensity fell to about 1.6% while the economy continued to grow by 4% per year. The second oil shock occurred late in the 1970’s and brought about another round of rising energy prices that likely affected energy use at least into the early 1980’s. By the middle of the decade world oil prices had fallen considerably from their peak. While energy intensity improvement slowed somewhat in the 1980’s, carbon intensity improved more rapidly in the 1980s (2.5%/yr.) than in the 1970s (2.2%/yr.). The direct reason for this divergence of energy and carbon intensity was the increased share of nuclear and natural gas in primary energy supply.

Figure A2.

Energy Intensity and Carbon Intensity in Japan. Source: authors’ calculation. Data source: Institute of Energy Economics, Japan (2004). Note: EI: energy intensity, CI: carbon intensity, Y: year. 95% confidence limits on energy intensity improvement rate 1970s: $-2.7\% < \times < -1.1$%, 1980s: $-2.1\% < \times < -1.2\%$, 1990s and after: $-0.1\% < \times < 0.3\%$.

Too much should not be made of the slight differences between the performance in the 1970’s and 1980’s. For one thing, the estimated time trends between decades are not significantly different from one another at a 95% confidence level. Beyond that, using a decade as the breakpoint is arbitrary. The overall conclusion is that at least in terms of the aggregate performance of energy and carbon intensity the 1970s and 1980s were, for Japan, very similar.

The 1990s offer a strong contrast from the 1970’s and 1980’s. Both energy and carbon intensity are nearly flat and not significantly different than zero. The point estimate of the average energy intensity actually increased very slightly, while the carbon intensity average decrease was less than 0.5% per year. A number of things changed in Japan in the 1990’s. For one, it was a period of slow GDP growth with a rate of 1.4% per year. Unlike the 1970’s and 1980’s, international energy prices were relatively stable and low. The slight improvement in carbon intensity can again be traced to greater use of nuclear and natural gas, as their share of primary energy supply rose from 9% to 12% and 10% to 14%, respectively (Institute of Energy Economics, Japan 2004). The actual data plotted in Figure A2 suggest the slowdown of both carbon and energy intensity may have had its roots as early as 1986. Removing the transition years in the late 1980’s from the earlier decade estimates would likely thus show an even greater difference between the rapid improvements through about 1985 and the negligible improvement since 1990.

What can Japan expect in terms of energy and carbon intensity improvement in a reference (i.e., no policy) case through the Kyoto commitment period? Will it return to the rapid improvements of the 1970’s and early 1980’s, or can it expect performance more like the 1990’s? The answer depends in large part on what is behind the change and there are several possible explanations. More detailed statistical analysis that attempted to sort out the effect of energy prices and other factors might help explain the reason for this change, but unfortunately many of the changes occurred together and this high correlation will make it difficult to separate the effect of each factor. Here we identify the leading hypotheses that could explain this change in performance in the Japanese economy and their very different implications for projections of Japan’s future and the difficulty of meeting the Kyoto target.

If the differences are mainly due to energy prices, then possibly the recently high oil and gas prices may spur improving energy and carbon efficiency. However, the fuel mix change that led to more rapid increases in carbon than energy intensity is unlikely to continue. Gas could penetrate further but unlike the previous periods when oil prices rise led the way and encouraged a switch to gas, in recent years gas prices have risen rapidly for independent reasons and so there may be less economic incentive to switch from oil to gas even with rising oil prices. As in most other industrial countries, nuclear expansion in Japan has slowed dramatically with few new plants expected to come on line in the near term. Thus, even if rising energy prices spur improvements in energy intensity, these may not be accompanied by an additional improvement in carbon intensity as occurred in the 1980’s. It is worth examining again the cross country comparison of carbon intensity in Figure A1 to note that Japan’s slowdown in the 1990’s is more pronounced than most other countries. This suggests that energy prices are not the sole explanation for it because energy price changes reflect to a large degree changes in international markets that were felt in all countries.

Another hypothesis is that rapid economic growth stimulates rapid energy efficiency improvement. The argument here is that with rapid growth comes a rapidly expanding capital stock, and perhaps rapid replacement of old capital. If the newer equipment and technologies are more energy efficient then improvement in energy efficiency will result. It is also likely that structural change is hastened by rapid economic growth, and if structural change is toward sectors that are less energy and emissions intensive—the service sectors are often offered as a candidate—then the aggregate energy intensity improvement will be faster when economic growth is faster. If these processes are at work then the low economic growth projections for Japan over the next decade would be inconsistent with rapid improvement in energy and carbon intensity. On the other hand, rising energy prices may also stimulate structural shifts and replacement of capital stock, and so there is likely some interaction between growth and price change as a source of intensity change.

Another possibility is that Japan has exhausted many of the immediate sources of improvement in energy intensity while other industrial countries have not yet done so. The case to be made here is that Japan has been at the leading edge of intensity change among the industrial countries. With highly efficient processes already widely adopted in Japan, the next set of energy efficient processes may not be yet proven or possess as clear an economic advantage as did the technologies contributing to the rapid decline in intensity in the 1970’s and early 1980’s. Another general trend among industrial countries in recent decades was a significant reduction in production of basic materials that are often very energy intensive, such as smelting of ores and energy intensive chemical production. In some cases this reduction in developed countries led to imports of these materials from developing countries. In other cases it reflected increased recycling of scrap or a structural change toward less demand for these materials. This source of structural change may have reached a limit in Japan. If this is a major reason for a slowdown in intensity improvement then it is difficult to project whether little or no improvement is only a pause until new sources of improvement are identified or the 1990’s experience is a new pattern that will persist for decades. In fact, one can not rule out the possibility that future development of the economy in Japan may lead instead to energy intensity increases, with the 1990’s being the turning point in the trend. These patterns could have important implications for Japan, and the world. What had seemed like a reliable trend of improving energy intensity may be ending, and as other industrial countries catch up to Japan they may face a similar situation. We return to these trends in the sections when we consider our modeling results and compare them to the MOE study results.