Carbon dioxide marginal abatement cost research

According to the method of deriving the MAC of carbon dioxide, the current research can be divided into three categories.

Based on expert carbon dioxide abatement costs

The basic idea is to use the most advanced available technical solutions as the reference line to conduct technical evaluations of various emission reduction measures in different countries and different industries. After summing up, calculate the emission reduction potential and abatement costs. Then, according to the order of their cost from low to high, they form a carbon dioxide MAC curve. This type of thinking is mainly based on engineering schemes for evaluation and summing up, so it is a kind of “bottom-up” research ideas. The most typical case is McKinsey’s global carbon dioxide MAC curve (Mckinsey Company, 2009). Different emission reduction measures, such as nuclear power generation technology and waste water recycling technology, are ranked from low to high in terms of their emission reduction costs (CO, equivalent per ton). For measures with negative emission reduction costs, they are generally considered as priority measures or “regretless choices.” In addition to different research institutions (such as the World Bank) and scholars using carbon dioxide marginal abatement cost curve for Poland, Mexico, Ireland, etc., the country’s carbon dioxide emission reduction potential and abatement costs are evaluated and analyzed (Johnson et al., 2009; Motherway & Walker, 2009; Poswiata & Bogdan, 2009).

Although the expert-based MAC curve is easy to understand and provides policy-makers with a rich set of tools and their respective priorities, the theoretical community is very controversial and believes that it has many shortcomings (Kesicki & Strachan, 2011). For example, there are differences in the boundaries and connotations of cost and benefit definitions that cause them to ignore other potential costs and benefits (Ekins et al., 2011); they do not take into account the interaction between emission reduction measures; rebound effect (Greening et al., 2000); does not evaluate the institutional barriers and transaction costs associated with implementing emission reduction measures, resulting in “negative” abatement costs (Brechet & Jouvet, 2009), in addition, the MAC curve is mostly based on static technical characteristics and does not take into account the intertemporal dynamics and inertia characteristics of different abatement measures (Adrien & Stephane, 2011).

Carbon dioxide abatement costs based on economic-energy models

Such methods generally first construct a partial equilibrium or general equilibrium model, and then change the constraints; such as increasing the emission reductions to obtain the corresponding shadow price, you can get the MAC information at different emission reduction levels (Kesicki & Strachan, 2011). According to the model setting, it can be further divided into two types: one is to use a bottom-up energy system model, such as MARKAL, POLES model, etc. (Criqui et al., 1999; Gao Pengfei et al., 2004). Paying more attention to the energy sector, use non-aggregated data, and achieve optimal technology set through linear programming and set certain constraints. Most energy system models are used to analyze the situation of one country, and some can be used for international emissions trading analysis. The model uses a top-down computable general equilibrium analysis, such as EPPA, GEM-E3, and GREEN models (Ellerman & Decaux, 1998), using aggregated data from all sectors, and is subjected to simulation economic systems. The new equilibrium state after external disturbances (such as carbon taxes) derives the marginal abatement costs.

The MAC based on the economic-energy model can show the emission reduction potential of different sectors, but it is limited by the characteristics of the derived model itself, and there are some inherent defects. When using energy system models to derive marginal abatement costs, energy demand is exogenous and limited to the energy sector itself, ignoring linkages with other economic sectors; when using CGE models to derive marginal abatement costs, the impact of energy policy on other sectors and international trade can be captured, but CGE cannot accurately provide its adjustment path when calculating the new equilibrium after disturbance, and therefore may

Marginal abatement cost 75 underestimate marginal abatement costs (Springer, 2003). In addition, the MACs estimated by different economic-energy models vary widely, mainly due to such economic-energy models themselves, such as setting a higher Aminton trade elasticity coefficient, or assuming alternative elasticity between elements. Higher will make the carbon dioxide marginal abatement cost lower, and the division of regions and sectors will lead to an overestimation of the carbon dioxide MAC (Fischer & Morgenstern, 2006); therefore, the assumptions imposed on the economic-energy model and the setting of the parameters will influence the final derivation of the carbon dioxide MAC distribution (Marklund & Samakovlis, 2007).

Taking different models at home and abroad as examples of China’s marginal carbon abatement cost in 2010 (see Table 4.1), it can be seen that due to differences in parameter assumptions, model structure settings, and data sources of different models, the conclusions of model evaluation are often inconsistent (Gao Pengfei et al., 2004); from the current research progress of this type of model, it is not enough to provide reliable and sufficient information for decision-makers, and the theoretical model still needs to be improved.

Carbon dioxide abatement cost curve based on micro supply side

This type of model is based primarily on the micro level, defining the set of production possibilities by setting detailed production techniques and

Table 4.1 Results of China’s 2010 marginal carbon abatement costs

Note: ^aThe original text is expressed in RMB, and is converted at the current exchange rate for comparison purposes.

economic constraints. The derived carbon dioxide MACs can be interpreted as: carbon dioxide emissions under given market, technical conditions, and opportunity costs (De Cara & Jayet, 2011). Most of these models use production functions to quantitatively characterize the relationship between carbon dioxide MACs and emission reductions. Typical is the linear carbon dioxide MAC function defined by Nordhaus: this function can be used for research at the national level, MC is the marginal cost, r is the emission reduction rate, and the unknown parameters a and p are observed through the engineering data, such as cost-to-fit estimates (Nordhaus, 1991). Although the model can describe the trend of marginal abatement costs, it is difficult to obtain real data for countries.

This type of model has recently emerged as a new branch and development. With the expansion of production theory and the intersection with environmental economics, researchers have incorporated pollutants including carbon dioxide into production models based on the production theory framework. By constructing environmental production techniques to estimate the shadow price of carbon dioxide (Fare et al., 1993), due to the lesser theoretical assumptions of applied and realistic observations, such models have been used in a large number of carbon dioxide shadow prices at different levels. For example, Rezek and Campbell (2007) used the generalized maximum entropy to estimate the MACs of atmospheric pollutants such as carbon dioxide and sulfur dioxide in thermal power plants in the USA, and the feasibility of constructing an emissions trading market for different pollutants were discussed (Marklund & Samakovlis, 2007). The directional distance function is used to estimate the carbon dioxide abatement costs of EU member states. On this basis, the fairness and efficiency of EU carbon emission reduction target allocation are discussed in Park and Lim (2009) and are based on transcendental logarithmic form. The distance function estimates the carbon dioxide MACs of thermal power plants in Korea and discusses the costs of different abatement options; Choi et al. (2012) use nonradial redundancy-based data envelopment analysis for China’s interprovincial carbon dioxide MACs. Domestic scholars have also begun to use this idea to evaluate industrial MACs. For example, Chen Shiyi (2010c, 2011) evaluated the marginal abatement cost of carbon dioxide in different sectors of China’s industry, and initially discussed the issue of environmental tax; Tu Zhengge (2012) also examined the carbon dioxide abatement costs of China’s eight major industrial sectors and discussed the choice of emission reduction strategies.

In summary, the above three research methods and perspectives have their scope and shortcomings. The expert-based marginal abatement cost curve is simple and easy to read, but its bottom-up analysis based on static individuals makes it difficult to dynamically evaluate the combined effects of abatement measures; the MAC results from the economic-energy model estimates. It is relatively stable, but the model construction is complex, and it is sensitive

Marginal abatement cost 77 to assumptions and parameters. The conclusion is lack of consistency. The MAC based on the production supply side is simple and intuitive, but the current research is still in a discrete “point” shape.

This chapter will focus on the third method, which uses the pollutant price model derived from the production function to estimate the marginal cost of CO, reduction.