Decomposition of China’s CO₂ emissions from agriculture utilizing an improved Kaya identity

Abstract

In recent decades, China’s agriculture has been experiencing flourishing growth accompanied by rising pesticide consumption, fertilizer consumption, energy consumption, etc. and increasing CO₂ emissions. Analyzing the driving forces of agricultural CO₂ emissions is key requirements for low-carbon agricultural policy formulation and decomposition analysis is widely used for this purpose. This study estimates the agricultural CO₂ emissions in China from 1994 to 2011 and applies the Logarithmic Mean Divisia Index (LMDI) as the decomposition technique. Change in agricultural CO₂ emissions is decomposed from 1994 to 2011 and includes a measure of the effect of agricultural subsidy. Results illustrate that economic development acts to increase CO₂ emissions significantly. Agricultural subsidy acts to reduce CO₂ emissions effectively and has increased in recent years. Policy is needed to significantly optimize agricultural subsidy structure and change agricultural development pathway, if China’s low-carbon agriculture target is to be achieved. This requires not only decreasing pesticide consumption, fertilizer consumption, energy consumption, etc. but also transformation of China’s agricultural development path for optimal outcomes.

Introduction

With the rapidly rising of global average temperature, global warming has become a focus of many countries. Many experts have contributed the root reasons of global warming to rapid global economy growth, large amounts of energy consumption, and increasing CO₂ emissions. According to the Environmental Kuznets Curve (EKC) hypothesis, further economic growth will overcome the environmental degradation. However, there is no critique suggesting that the EKC concept can be valid (Kaika and Zervas 2013a, b). China has become the world’s biggest emitter of carbon dioxide with rapid economic growth and increasing energy consumption, and agriculture is a large contributor to CO₂ emissions. Agricultural CO₂ emissions are the CO₂ emissions in the agricultural production process (Johnson 2007). As a big agricultural country in the world, China’s agricultural CO₂ emissions increased sharply from 46.33 million tons in 1994 to 85.75 million tons in 2011. In Europe, CO₂ emissions from agriculture are 10 % of the total annual CO₂ emissions (Robaina-Alves and Moutinho 2014). In the UK, the production processing and retailing of food were estimated to produce approximately 20 % of the national CO₂ emissions (Franks and Hadingham 2012). In our study, the CO₂ emissions estimated from agricultural production process to produce approximately 15 % of national CO₂ emissions and the ratio of agricultural CO₂ emissions to national CO₂ emissions continued to increase in recent years. In order to actively respond to climate change and to develop low-carbon economy, China has an ambitious target to reduce CO₂ emission intensity by 40–45 % by 2020, from 2005 levels. It is widely accepted that agriculture plays an important role in achievement of the goals of decreasing national CO₂ emissions (Luc et al. 2012; Michal et al. 2013; Francesco et al. 2013). Therefore, the success or failure of decreasing agricultural CO₂ emissions will have a significant effect on achieve these ambitious targets. Agricultural CO₂ emissions in China has become a focus of policy makers, and analyses of driving forces of agricultural CO₂ emissions constitute a vital part of low-carbon agriculture.

Logarithmic Mean Divisia Index (LMDI) theory was first proposed by Ang and Liu (2001) and has been widely used in decomposing changes in CO₂ emissions due to its perfect decomposition, consistency in aggregation, path independency, and an ability to handle zero values (Ang 2005). LMDI model is applied to study energy consumption and CO₂ emissions in the world (Peters and Hertwich 2008; Pani and Mukhopadhyay 2010). Greening et al. (2001) and Greening (2004) conducts the secondary decomposition of CO₂ emissions from sectors. Based on completed decomposition technique, many scholars study CO₂ emissions in different countries, such as the UK (Kwon 2005), South Korea (Oh et al. 2010), Thailand (Bhattacharyya and Ussanarassamee 2004), Brazil (Luciano and Shinji 2011), India (Paul and Bhattacharya 2004), Greece (Hatzigeorgiou et al. 2011), Turkey (Tunc et al. 2009), and Ireland (Tadhg 2013). Completed decomposition technique is also applied to research energy-related CO₂ emissions in China (Sun et al. 2010; Xu et al. 2012; Li and Ou 2013; Zhang et al. 2013).

Recent studies have analyzed CO₂ emissions and related problems in the national level and industry sector. However, few studies analyze the driving forces of agricultural CO₂ emissions. In this paper, we estimate China’s agricultural CO₂ emissions based on six kinds of main carbon sources in agricultural production process and analyze the trend of China’s agricultural CO₂ emissions from 1994 to 2011. An extended Kaya identity and LMDI method were applied to identify, quantify, and explain major driving forces acting on change China’s agricultural CO₂ emissions. It is the first study to measure the impact of agricultural subsidy excitation and agricultural subsidy intensity on agricultural CO₂ emissions. The paper is organized as follows. “Decomposition methodology” details the decomposition methodology adopted. “Data” describes the data used. “Results and discussion” estimates and describes China’s agricultural CO₂ emissions and discusses the results from decomposition analysis. “Conclusions and policy recommendations” provided the major conclusions and policy implications.

Decomposition methodology

Kaya identity was first proposed by Professor Yoichi Kaya (1990) on Intergovernmental Panel on Climate Change (IPCC) in 1990. The “drivers” or “effects” leading to observed change in CO₂ are population effect, energy intensity effect, carbon emissions intensity effect, and economic development effect in the Kaya identity. The Kaya identity is expressed as follows.

$$ \begin{array}{c}\hfill C=\frac{C}{E}\times \frac{E}{\mathrm{GDP}}\times \frac{\mathrm{GDP}}{P}\times P\hfill \\ {}\hfill \mathrm{CE}=\frac{C}{E},\mathrm{EI}=\frac{E}{\mathrm{GDP}}, G=\frac{\mathrm{GDP}}{P}\hfill \end{array} $$

(1)

where C, E, GDI, and P denote CO₂ emissions, energy consumption, gross domestic product, and population, respectively. CE, EI, G, and P denote CO₂ emission coefficient effect, energy intensity effect, the GDP per capita (affluence) effect, and population effect, respectively.

In this paper, agricultural CO₂ emissions is decomposed to predefined factors of CO₂ emission intensity effect, agricultural subsidy excitation effect, agricultural subsidy intensity effect, economic development (affluence) effect, and population effect based on real situation of agricultural production process in China as the following extended Kaya identity.

$$ C=\frac{C}{\mathrm{YA}}\times \frac{\mathrm{YA}}{\mathrm{IA}}\times \frac{\mathrm{IA}}{Y}\times \frac{Y}{P}\times P=\mathrm{CI}\times \mathrm{OE}\times F\times G\times P $$

(2)

The following variables are applied in Eq. (2).

C :

Total agricultural CO₂ emissions

YA:

Agricultural subsidy excitation

IA:

Agricultural subsidy intensity

Y :

GDP

P :

Population

$\mathrm{CI}=\frac{C}{\mathrm{YA}}$ :

The CO₂ emission intensity in agricultural sector

$\mathrm{OE}=\frac{\mathrm{YA}}{\mathrm{IA}}$ :

The agricultural output per agricultural subsidies

$F=\frac{\mathrm{IA}}{Y}$ :

The ratio of agricultural subsidies to GDP

$G=\frac{Y}{P}$ :

The GDP per capita or affluence

P :

Population

The decomposition of an observed change in C associated with these factors are referred to as, carbon emission intensity effect (ΔC _cei), agricultural subsidy excitation effect (ΔC _ase), agricultural subsidy intensity effect (ΔC _asi), affluence effect (ΔC _ecm), and population effect (ΔC _pop). The index of annual change in total agricultural CO₂ emissions (C_tot) can be expressed in an addition form as follows.

$$ \varDelta {C}_{\mathrm{tot}}=\varDelta {C}_T-\varDelta {C}_0=\varDelta {C}_{\mathrm{cei}}+\varDelta {C}_{\mathrm{ase}}+\varDelta {C}_{\mathrm{asi}}+\varDelta {C}_{\mathrm{ecm}}+\varDelta {C}_{\mathrm{pop}} $$

(3)

LMDI analysis is then applied as the method to decompose Eq. (3). From Ang and Liu (2001), the following LMDI II formulae apply to each of the effects.

$$ \varDelta {C}_{\mathrm{cei}}= L\left({W}^T,{W}^0\right)\times \ln \left(\frac{{\mathrm{CI}}^T}{{\mathrm{CI}}^0}\right) $$

(4)

$$ \varDelta {C}_{\mathrm{ase}}= L\left({W}^T,{W}^0\right)\times \ln \left(\frac{{\mathrm{OE}}^T}{{\mathrm{OE}}^0}\right) $$

(5)

$$ \varDelta {C}_{\mathrm{asi}}= L\left({W}^T,{W}^0\right)\times \ln \left(\frac{F^T}{F^0}\right) $$

(6)

$$ \varDelta {C}_{\mathrm{ecm}}= L\left({W}^T,{W}^0\right)\times \ln \left(\frac{G^T}{G^0}\right) $$

(7)

$$ \varDelta {C}_{\mathrm{pop}}= L\left({W}^T,{W}^0\right)\times \ln \left(\frac{P^T}{P^0}\right) $$

(8)

Where W = CI × OE × F × G × P and $L\left({W}^T,{W}^0\right)=\raisebox{1ex}{$\left({W}^T-{W}^0\right)$}\!\left/ \!\raisebox{-1ex}{$\left( \ln {W}^T- \ln {W}^0\right)$}\right.$.

Data

This study collected annual data on the pesticide consumption, fertilizer consumption, agricultural diesel consumption, plastic sheeting consumption, tilling area, and actual irrigated area for 1994–2011 (presented in Appendix Table 3), from China Rural Statistical Yearbook (1995–2012). Agricultural subsidies, real GDP, and population from 1994 to 2011 were collected from Statistical Yearbook of the People’s Republic of China (2012). The agricultural subsidies and real GDP were measured in China Yuan at 1994 prices.

Typically, agricultural CO₂ emissions comes from six areas: fertilizer consumption, pesticides consumption, plastic sheeting consumption, fossil fuels consumption for agricultural machinery (mainly agricultural diesel consumption), agricultural tilling, and electrical power consumption in agriculture irrigated. This study estimates agricultural CO₂ emissions in China from 1994 to 2011 as follows:

$$ C={\displaystyle \sum_{i=1}^6{C}_i}={\displaystyle \sum_{i=1}^6{S}_i\times {\rho}_i} $$

(9)

Where C, C _i , S _i , and ρ _i denote total amount of agricultural CO₂ emissions, CO₂ emissions from source i, the amount of carbon source i consumption, and carbon emission coefficients for carbon source i of CO₂ emissions, respectively. The CO₂ emission coefficients of different sources of CO₂ emissions are given in Table 1.

Table 1 The CO₂ emission coefficients of different resources

Results and discussion

Agricultural CO₂ emission estimation

According to Eq. (9), Table 1, and Appendix Table 3, the trend of China’s agricultural CO₂ emissions is presented in Fig. 1.

Fig. 1

Agricultural CO₂ emissions of six carbon sources in China from 1994 to 2011

Figure 1 shows the changes in China’s agricultural CO₂ emissions and agricultural CO₂ emission mix based on six sources of CO₂ emissions from 1994 to 2011. In recent years, China’s agricultural CO₂ emissions continue to increase sharply in correlation to the levels of fertilizer consumption, pesticide consumption, plastic sheeting consumption, and agricultural diesel consumption. From 46.33 million tons in 1994 to 85.75 million tons in 2011, agricultural CO₂ emissions increased by approximately 85 %. Fertilizer consumption is the biggest contributor to China’s agricultural CO₂ emissions followed by agricultural diesel consumption, plastic sheeting consumption, and pesticide consumption. From 1994 to 2011, China’s fertilizer consumption increased sharply from 29.72 to 51.09 million tons, and the ratio of CO₂ emissions contributed by fertilizer consumption to total agricultural CO₂ emissions is approximately 60 %. With the accelerating of agricultural modernization, CO₂ emissions caused by agricultural diesel consumption, plastic sheeting consumption, and pesticide consumption continue to increase rapidly in recent years, and the annual average growth rates are 6.3, 8.8, and 4.6 %, respectively. According to the growth rate of CO₂ emissions, the changes of China’s agricultural CO₂ emissions could be divided into four stages: 1994–1997, 1998–2003, 2004–2007, and 2008–2011.

The annual growth rate of China’s agricultural CO₂ emissions surpasses 5 % in the 1994–1997 period. One reason for the rapid CO₂ emission growth is that a large amount of fertilizer was consumed in agricultural production process to improve the output per unit area. From 31.52 million tons in 1994 to 39.81 million tons in 1997, China’s fertilizer consumption increased by approximately 20 % in only 4 years. The other reason for the rapidly increasing CO₂ emissions is the rapid process of agricultural modernization accompanied by rising agricultural diesel consumption.

In the 1998–2003 period, China’s agricultural CO₂ emissions continued to increase slowly. The burden of farmers overweighs in China from 1998 to 2003. Many farmers gave up agricultural production and went to urban for work which had a negative impact on agricultural production sources demands and slowed down the increasing of agricultural CO₂ emissions. China’s grain production was facing enormous challenges due to the decreasing output of grain.

During 2004–2007, China’s agricultural CO₂ emission growth was accelerated. In 2004, China’s government promulgated the policies of “two tax breaks, three subsidies” for agriculture in the central committee of the Chinese communist party which greatly improved the farmers’ enthusiasm for agricultural production process. Two tax breaks policy is deductions and exemptions of agricultural tax and agricultural special tax except tobacco tax, and three subsidies policy is the direct subsidy, seed variety subsidy, and purchase sublimations for large farm machinery to farmers who produce gains.

Faced with the serious environmental pollution and climate changes, Chinese government realized the importance of resources conservation and environmental protection and advocated low-carbon agriculture and green agriculture since 2008. China began to promote the use of biogas, solar, and other renewable energy consumption in rural areas, and decreased fertilizer consumption, pesticide consumption, and plastic sheeting consumption to control China’s agricultural CO₂ emissions.

Results

The complete time-series decomposition results for China’s agricultural CO₂ emissions are presented in Appendix Table 4 based on the extended Kaya identity established in “Decomposition methodology” and China’s agricultural CO₂ emissions data estimated in “Results and discussion.” The accumulated effects by period are available in Table 2 and Fig. 2.

Table 2 Decomposition of China’s agricultural CO₂ emissions from 1994 to 2011

Fig. 2

Accumulated decomposition of China’s agricultural CO₂ emissions 1994–2011

The accumulated effects over the entire period (Fig. 2) illustrate that the dominant positive effect is economic development effect, as is often identified in other studies (Tadhg et al. 2012; Chen et al. 2013; Xu et al. 2012; Xiaoli et al. 2012). The population effect as a scale effect also had a positive but relatively small effect on emissions. Agricultural subsidy intensity effect, as a significant positive in general, had a fluctuant effect on emissions. The agricultural subsidy excitation effect was the most significant negative effect on emissions, while CO₂ emission intensity had a small negative effect also. The total negative effects are heavily outweighed by the total positive effects, which resulted in the increase in agricultural CO₂ emissions. These decomposition results of China’s agricultural CO₂ emissions from 1994 to 2011 are summarized as follows:

• (1) The scale effects (economic development effect and population effect) are the dominant drivers in China’s agricultural CO₂ emissions. Economic development effect has a great effect on emissions, and population effect has a weak effect on emissions.

• (2) The intensity effect (CO₂ emission intensity effect) reducing the intensity of CO₂ emissions per unit agricultural output is a weak factor in limiting increasing in emissions.

• (3) Agricultural subsidy effects (agricultural subsidy intensity effect and agricultural subsidy exCitation effect) have different effects on emissions. Agricultural subsidy intensity effect is the most significant factor in driving growth in China’s agricultural CO₂ emissions. The agricultural subsidy excitation effect contributes to decreasing emissions effectively. The impact of agricultural subsidy intensity effect to emission changes with the variation of agricultural subsidy intensity. Between 1994 and 1998, in order to increase grain production, China increased agricultural subsidies and formulated policies to reduce the burden of farmers and ensure grain purchase. A lot of fertilizer, pesticide, plastic sheeting, etc. were adopted in agricultural production process accompanied by rising CO₂ emissions. Agricultural subsidy intensity effect had a positive effect on agricultural CO₂ emissions from 1994 to 1998. During 1999 and 2003, China’s government decreased agricultural subsidy intensity and formulated policies to accelerate the reform of agricultural industry and widen the ways to increasing farmers’ income. In the 1999–2003 period, the positive effect of agricultural subsidy intensity to CO₂ emissions became weak and CO₂ emissions continued to increase at a slow growth rate. China’s government improved agricultural subsidy intensity to promote the development of green agriculture and circular agriculture since 2004. Agricultural subsidy intensity effect played a significant factor in driving growth in CO₂ emissions.

• (4) The total positive effects heavily exceeded the total negative effects, which resulted in the rapidly increasing agricultural CO₂ emissions.

Conclusions and policy recommendations

Understanding the driving forces of agricultural CO₂ emissions is essential for policy-making process. In this paper, it is attempted to decompose the driving forces of agricultural CO₂ emissions. The decomposition analysis reveals the following important conclusions. First, the positive effects for agricultural CO₂ emissions are economic development effect, population effect, and agricultural subsidy intensity effect. Second, the negative effects for agricultural CO₂ emissions are agricultural subsidy excitation effect and CO₂ emissions intensity. Third, agricultural subsidy excitation effect had a significant negative effect on CO₂ emissions. The total positive effects heavily outweigh the total positive effects, which resulted in agricultural CO₂ emissions increasing. Therefore, low-carbon agriculture should be paid more efforts to reduce China’s agricultural CO₂ emissions, and some policy suggestions are presented as follows.

• (1) Improving efficiency of agricultural sources and decreasing fertilizer consumption and pesticide consumption

  In traditional agricultural production process, agricultural output was heavily relied on the fertilizer consumption and pesticide consumption which led to a rapid increasing in agricultural CO₂ emissions. China should increase investment in agricultural science and technology innovation, which could promote the popularization of economizing farming, planting, fertilizing, spraying irrigation, and other low-carbon farming techniques. New scientific and technological means should be adopted widely in rural area to improving efficiency of agricultural sources and inhibit agricultural CO₂ emission growth.

• (2) Increasing agricultural subsidies

  Agricultural subsidies play an important role in decreasing agricultural CO₂ emissions and guide the development way of agriculture. China should increase agricultural subsidies and accelerate agricultural infrastructure construction, which can provide basic guarantee for development of low-carbon agriculture. Increasing agricultural subsidy intensity could encourage farmers to join in the development of low-carbon agriculture and circular agriculture and achieve the goal of agricultural sustainable development.

Appendix

Table 3 Annual time series of six carbon sources in China’s agriculture from 1994 to 2011

Table 4 Annual time-series decomposition results from 1994 to 2011. Unit: million tons