Conventional economic models were based on the premise that consumers and companies engaged in rational decision making. In such models, there was no such thing as profitable energy-saving behavior. Low carbon was considered to be another form of restriction that led to an increase in costs, and was therefore a negative influence on the economy.
However, in terms of service life, energy-saving light bulbs are of definite benefit to household finances, meaning that low carbon produces positive economic results. However, the conversion to energy-saving bulbs has not yet reached 100%. This indicates that consumers and companies are not totally rational (bounded rationality).
This year, we looked at time discount rates and by adopting a hyperbolic discounting function (based on the results of a survey on time discount rates) instead of the exponential discounting used in economic analysis, and allowing paying in installments rather than lump-sum, we were able to develop a model for promoting the use of low carbon devices (specifically, residential solar power generation systems) without lowering utility. We also conducted trial calculations.
By considering the actual decision-making mechanism of consumers, we were able to propose effective measures that could not be achieved within a 'rational' framework. We will conduct further studies on the difference in discount functions resulting from differences in cost and household income and link this with our general equilibrium model.
Using data on the number of newly registered private vehicles per annum provided by the Automobile Inspection and Registration Information Association and data compiled from the websites of used car dealers for all models of cars sold in Japan, we developed a multinomial logit model to show the relationships between sales and price, fuel efficiency and other characteristics of a particular vehicle model. Using a utility function of five explanatory variables: previous annual sales, new vehicle effect coding (code 1 for new vehicles, 0 for others), interior volume, annual working expenditures and subsidies, we are able to estimate the choice probability or share of each model.
Using our multinomial logit model, we calculated how changing subsidy amounts would affect the market share of each model. The results led us to estimate that the introduction of subsidies for purchasing eco-friendly vehicles resulted in an annual 70,000 ton reduction in CO2 emissions. This is the equivalent of roughly 1.8% of predicted CO2 emissions from the total number of cars sold that year. However, since the cost of CO2 reductions per vehicle, derived from dividing reduction costs by reduction amounts, comes to around 150,000 yen per ton, this is a very costly way to mitigate global warming. The policy of providing subsidies and tax breaks for purchasing eco-friendly vehicles should therefore be regarded as an economic rather than an environmental measure.
In order to propose a system with the potential to achieve greater CO2 emission reductions than the eco-car subsidy system actually in effect, we created a hypothetical automobile tax system with rates in proportion to fuel efficiency, and calculated the effect it had on reducing CO2 emissions. Unlike the current automobile tax system, in which tax rates are determined by car type and displacement, our hypothetical system calculates tax rates according to fuel efficiency. Under this system, hybrid cars would basically be tax free, while the tax rate of cars with the worst fuel efficiency on the market would be triple that of the current tax system or roughly 200,000 yen a year. Introducing such a system could potentially reduce CO2 emissions by 280,000 tons a year, or four times the amount achieved by the eco car subsidy system. This demonstrates the importance of designing appropriate systems to achieve maximum CO2 emission reductions. We therefore intend to analyze preferences per region in order to develop more efficient systems.