Analysis of Japan's New Energy Strategy 3/4: Demand Forecast

The previous two posts of the analysis on Japan' new energy plan - the Innovative Strategy for Energy and the Environment - focused on the supply side of the planned energy system. Now the focus shifts to the demand side, especially on the energy demand forecast used in the plan. It is imperative to thoroughly analyze the demand forecast because it determines how much supply, ranging from fossil fuel to renewable energy, will be needed in the long run, and ultimately affects what strategies will be needed to secure resources, maintain economic competitiveness, and protect the environment. In this post, I will argue that the energy demand forecast in the plan overestimates future energy demand by assuming unrealistic economic growth and neglecting demographic, industrial, and lifestyle changes that Japan has been experiencing.

Economic Outlook

The changes in energy demand have traditionally been driven by economic growth. For this reason, the first order of business in forecasting energy demand is to predict the future economic growth. It is obviously not an easy task to forecast economic growth up to 2030 - the plan's target year. The government originally prepared two different scenarios for economic growth: (1) Safe Scenario and (2) Growth Scenario. The below table summarizes the assumed average annual growth rates for real GDP in Japan. They seem fairly low even in Growth Scenario, but many critiques argue that they are still too high for a nation with declining population, particularly a rapid decrease in working age population. In response to the criticisms, a new scenario called Low Growth Scenario was added to reflect the actual growth rate per working age population in the past decade.

Real GDP Annual Growth Rate Assumptions (Source: National Policy Unit)

	2011-2020	2021-2030	Note
Growth Scenario	1.8%	1.2%	Assumption in Rebirth of Japan: A Comprehensive Strategy
Safe Scenario	1.1%	0.8%	Assumption in Fiscal Management Strategy
Low Growth Scenario	0.2%	0.4%	Continuation of the average growth rate per working age population in the past decade

These growth rates are one of the most critical determinants of future energy demand. The figure below suggests that the energy demand in 2030 can be substantially lower under Low Growth Scenario. The difference between Growth Scenario and Low Growth Scenario in 2030 is about 20% of the energy demand in Growth Scenario in 2030.

Figure: Energy Demand Forecast (Source: National Policy Unit)

This indicates that the amount of power plants needed to meet the energy demand should be significantly different under the three scenarios. For example, if the electricity demand is substantially lower than today as in Low Growth Scenario, it is possible to shut down most coal-fired power plants to curb greenhouse gas emissions, or alternatively, shut down most nuclear power plants as about half of the public desires nowadays without compromising reliability of electric grid. The fuel mix for electricity generation in this way can be altered under different growth assumptions. The plan however only focuses a single fixed fuel mix for the three growth scenarios, which further undermines the plan's cohesiveness and feasibility in my view.

Industrial Sector

The discussion above shows the lack of interaction between supply and demand scenarios. The next problem to point out is the energy demand itself. There is a variety of indicators used to predict energy demand for each sector, and in this blog post I focus on parameters used to forecast energy use in industrial and transportation sector. As the figure below shows, most drivers of industrial activities shows either upward or steady trend (click the figure to enlarge).

Figure: Industrial Output Trend and Forecast (Source: National Policy Unit)

I would argue that this assumption is unrealistic for two reasons. One reason is that the level of materialistic consumption has been declining due to changing demographics and lifestyle. Take for example, the plan assumes the level of paper production to be steady at a current level until 2030. On the contrary, the population, particularly working age population is on the decline, and the society as a whole is depending more and more on paperless transaction. While the plan assumes the level of paper production will remain steady, paper producers expects their businesses to shrink in coming decades; Oji Paper, a major paper producer in Japan, just announced massive payoffs and factory closings despite increasing profits, primarily to cope with expected decline in paper demand. Another shrinking energy-intensive industry is cement; the most massive infrastructures such as dam, highway, and subway are already built out in Japan, and the aging society along with declining population will obviously curb the appetite for new housings and office buildings. There will of course be demand for maintenance and replacement, but it won't be as much as new development.

The other reason is the structural changes that the Japanese economy is facing today. The level of industrial output has declined steadily through the 2000's for most products as assembly lines continued to move to China and Southeast Asia. In fact, once enjoying enormous trade surplus, Japan is now in red in trade balance due not only to increasing import of fossil fuel but also to declining exports. Japan managed to maintain positive balance of payments through returns on investment abroad, and Japan's economy itself is becoming more service-oriented. Unless this trend is somehow suddenly reversed, for good or bad, industrial activities are likely to continue to decline. For these reasons, while it is understandable for politicians and officials to have a motivation to draw upward trend, the energy demand in industrial sector is likely to be significantly lower than the forecasts due to lower-than-expected production level.

Transportation Sector

The demand for transportation fuels, mostly gasoline and diesel for automobiles, are determined by the distance traveled (VKT: vehicle kilometers traveled) and fuel efficiency. The fuel efficiency is expected to increase significantly thanks to continuous improvement in internal combustion engine and large deployment of hybrid and electric vehicles. There is a variety of predictions and opinions on the level of alternative vehicle penetration, but I would say the plan's assumptions on fuel efficiency are neutral and realistic.

Figure: Total Vehicle Travel Demand Trend and Forecast (Source: National Policy Unit)

There is a problem on the assumed VKT, or distance traveled, however. The above figure shows that vehicle kilometers traveled by passenger vehicles are assumed to decline continuously til 2030. When looking at this figure on per capita basis, it is assumed to be more in 2030 than in 2020. This is an odd assumption in my view, because the amount of driving has been on decline on per capita basis in aging societies (elderlies usually drives less), and it is difficult to come up with an explanation for the sudden change in the trend in 2020. Furthermore, Generation Y'ers - also known as Millennial Generation - tends to drive less and less as they prefer to live close to urban centers and rely on public transportation to get around. For whatever reasons, this is causing automobile travels to decline in most industrial nations. These demographic and lifestyle changes could significantly reduce driving on per capita basis, and the assumed VKT in the plan is likely to overestimate the energy demand from automobile use.

Figure: Vehicle Travel Demand Trend and Forecast (Source: National Policy Unit)

The freight side of vehicle kilometers traveled is more difficult to predict. Yet, I would argue that the demographic change and more service-oriented economy mentioned earlier is likely to curb the demand for traditional bulky goods such as paper and construction materials. The challenging part is to incorporate these social changes into demand forecast, and I don't have a clear solution for it. My point here is that the overall energy demand should be much less than the plan's forecast.

The implication of these forecasts on a variety of topics such as greenhouse gas emissions and fuel mix for electricity generation will be discussed in the next post, which is scheduled to be published in December.