Review of existing empirical studies
The development of theoretical research is inseparable from the continuous verification of empirical research. Table 1.1 lists some important empirical documents. Through combining and summarizing these documents, the following conclusions can be drawn.
In terms of research objects, foreign countries focus on micro-scale research, and domestic focus on macro-level analysis
The research objects of most foreign literatures are mainly based on the production activities of micro-enterprises. In the selection of micro-enterprises, it is important to consider industrial production-oriented enterprises that are heavily invested or dependent on certain polluting raw materials in the production process, such as thermal power stations, which use fossil energy to burn and emit a lot of pollution. Gases, such as SO,, not only cause greater pollution to the atmosphere, but also affect human health. Therefore, in the existing literature, research on environmentally sensitive productivity studies using SO, emitted from power stations as undesired outputs is dominant
Table 1.1 Summary of evidence-based literature on studies about environmentally sensitive productivity
|
Authors |
Evidence-based methods |
Data |
Variables |
Main results |
|||||
|
Model |
Functions |
Assessment |
Input |
Desirable Output |
Undesirable Output |
Productivity |
Shadow price |
||
|
Gallop and Roberts (1985) |
Cost function |
- |
Minimum cost method |
56 power plants in the US from 1973 to 1979 |
Labor/capital/ low-sulfur fuels/high-sulfur fuels |
Electricity generated |
SO, |
- |
SO,: 0.195 (USS/pounds, 1979 price) |
|
Fare et al. (1993) |
DF(O) |
Translog |
Parameter method |
30 papermaking factories in the US in 1976 |
Paper pulp/ energy/ capital/ labor |
Paper |
BOD/TSS/PART/ so. |
Efficiency: 0.9182 |
BOD: 1,043.4, TSS: 0, PART: 25,270, SO,: 3,696 (USS/ton, 1976 price) |
|
Coggins and Swinton (1996) |
DF(O) |
Translog |
Parameter method |
14 heating and power plants in Wisconsin in 1990-1992 |
Sulfide/energy/ labor/ capital |
Electricity generated |
SO, |
Efficiency: 0.946 |
SO,: 305(1990), 251.6(1991), 322.9(1992), Average 292.7 (USS/ton, 1992 price) |
|
Boyd et al. (1996) |
DDF |
- |
DEA |
Coal-fueled power plants in the US Yaisawarng and Klein 1994 data |
Fuel/labor/ capital (fixed input)/sulfur (undesirable output) |
Net electricity |
SO, |
Average efficiency: 0.933 |
SO,: 1,703 (USS/ton, 1973 price) |
|
Chung et al. (1997) |
DDF |
— |
DEA |
30 Swedish papermaking factories in 1986-1990 |
Labor/wood fiber/energy/ capital |
Pulp |
BOD/COD/TSS |
M index:0.997 (improved efficiency:0.977; technological progress: 1.02) |
— |
|
ML index: 1.039 (improved efficiency: 0.955; technological progress: 1.088) |
|||||||||
|
Kolstad and Turnovsky (1998) |
- |
Quadric form |
— |
51 coalgenerated power plants in eastern America in 1970-1979 |
Sulfur/ash/ capital/ thermal energy |
Electricity generated |
SO, |
- |
SO,: 0.071; Ash: 0.121 (USS/pounds, price in 1976) |
|
Swinton (1998) |
DF(O) |
Translog |
Parameter method |
Coal-fueled power plants in 1990-1998 Florida |
Energy/labor/ capital/ sulfur |
Electricity generated |
SO, |
Efficiency: 0.978 |
SO,: 157.10 (USS/ton, 1996 price) |
|
Murty and Kumar (2003) |
DF(O) |
Translog |
Parameter method SFA |
Industries with water pollution issues in India (samples from 60 companies) |
Capital/labor/ energy/ materials |
Turnover |
BOD/COD/TSS |
Efficiency: 0.899 |
BOD: 0.246, COD: 0.0775 (million Rupee/ ton, 1994/95 price) |
|
Hailu and Veeman (2000) |
DF(I) |
Translog |
Parameter method |
Aggregated data of Canadian papermaking sector for 36 years between 1959 and 1994 |
Energy/wood residue/ wood pulp/ other raw materials/ production labor/ managerial labor/ capital |
Wood pulp/ newsprint/ paper board/ other types of paper |
BOD/TSS |
TE: 0.996, M Index: 0.878, ML Index: 1.044 |
BOD: 123, TSS: 286 (USS/million ton, 1986 price) |
(continued)
|
Authors |
Evidence-based methods |
Data |
Variables |
Main results |
|||||
|
Model |
Functions |
Assessment |
Input |
Desirable Output |
Undesirable Output |
Productivity |
Shadow price |
||
|
Reig-Martinez et al. (2001) |
DF(O) |
Translog |
Parameter method |
18 pottery factories in Spain |
Raw materials/ capital/ labor |
Ceramic pavement |
Cement/waste oil |
Average efficiency: 0.927 |
Cement: 336.6 (euro/ton), waste oil: 125.5 (euro/kg) |
|
Lee et al. (2002) |
DDF |
- |
DEA |
43 Korean power plants of 1990-1995 |
Installed capacity/ fuels/labor |
Power generation |
SO,/NO,/TSP |
- |
SO,: 3,107, NO,: 17,393, TSP: 51,093 (USS/ton) |
|
Salnykov and Zelenyuk (2005) |
DDF |
Translog |
Parameter method |
50 countries |
Labor/arable land/energy/ capital/ |
GNP |
COJSOJNO, |
Efficiency: 0.8433 |
CO,: 331.89, SO,: 59,997.95 NO,: 154,583.63 (USS/ton) |
|
Atkinson and Dorfman (2005) |
DF(I) |
Translog |
Parameter method |
43 US private for-profit power plants in 1980, 1985, 1990, 1995 |
Energy/labor/ capital |
Power generation |
SO, |
Classical efficiency: 0.564277 LIBS efficiency: 0.553187 |
SO,: 1,871.7; 1990 SO,: 556.8: 1995 SO,: 486.7 (USS/ton) |
|
Lee(2005) |
DF(I) |
Translog |
Parameter method |
51 US thermal power generating units in 1977-1986 |
Capital/heat/ sulfide/ash |
Power generation |
Sulfur/ash |
TE: 0.945 |
SO,: 167.4, Àsh: 127.7 (USS/pounds, 1976 price) |
|
Fare et al. (2005) |
DDF |
Quadric form |
Deterministic parameters |
209 US thermal power plants in 1993/1997 |
Labor/ installed capacity/ fuel |
Power generation |
SO, |
|
|
|
SFA |
1993: 0.798, 1997:0.804 |
|
|||||||
|
Kumar (2006) |
DDF |
DEA |
41 countries in 1971-1992 |
Labor /capital/ energy |
GDP |
CO, |
M index: 0.9998 (Improved efficiency: 1.0019; technological progress:0.9981) ML index: 1.0002 (Improved efficiency: 0.9997; Technological progress: 1.0006) |
||
|
Fare et al. (2007) |
DDF |
- |
DEA |
92 thermal power plants in the US in 1995 |
Capital/labor/ fuel heat (coal, oil and gas) |
Power generation |
SO, NO, |
- |
- |
|
Ke et al. (2008) |
DF(O) |
Translog |
Parameters method |
30 provinces in China from 1996 to 2002 |
Capital/labor |
GDP |
SO, |
East: 0.831, Middle: 0.706, West: 0.682 |
East: 0.516, Middle: 0.508, West: 0.529 (hundred million yuan/ ton, 1996 price) |
|
Van Ha et al. (2008) |
DF(O) |
Translog |
Parameter/ measurement assessment |
63 papermaking workshops in Vietnam in 2003 |
Capital/labor/ energy/ waste paper/other materials/ social capital |
Paper |
BOD/COD/TSS |
Efficiency: 0.72 |
BOD: 575.2, COD: 1,429.7, TSS: 3,354.8 (USS/ton, 2003 price) |
(continued)
Table 1.1 (Cont. )
|
Au Ihors |
Evidence-based methods |
Data |
Variables |
Main results |
|||||
|
Model |
Functions |
Assessment |
Input |
Desirable Output |
Undesirable Output |
Productivity |
Shadow price |
||
|
Ghorbani and Motallebi (2009) |
DF(O) |
Translog |
Parameters method |
85 dairy farms in Iran in 2006 |
Farm area/ energy/ labor/feed |
milk |
CH4/CO/N,O |
- |
CH4: 0.61. CO,: 0.058, N,6: 0.59 (price ratio against milk) |
|
Hu et al. (2008) |
DDF |
DEA |
30 provinces in China from 1999 to 2005 |
Capital/labor |
GDP |
CO./COD/SO./ waste water/ solid waste |
Highest in the east and lowest in the west (specific value depends on the type of undesirable output) |
||
|
Tu(2008) |
DDF |
DEA |
Industrial enterprises above a designated size in 30 provinces from 1998 to 2005 |
Capital/ energy/ labor |
Industrial added value |
SO, |
East: relatively harmonious relationship between industry and environment Middle and west: imbalances between environmental protection and industrial growth |
||
|
Wang et al. (2008) |
DDF |
- |
DEA |
1980-2004 APEC 17 countries and regions |
Capital/labor |
GDP |
CO, |
ML: 1.0056 (Technological progress:0.76%) |
- |
(continued)
|
Tu (2009) |
DDF |
DEA |
Industrial enterprises above a designated size in 30 provinces from 1998 to 2005 |
Capital/ energy/ labor |
Industrial added value |
SO2 |
SO: 2.09 (RMB, hundred million yuan/ ten thousand tons, 1998 price unchanged) |
|
|
Wu (2009) |
DDF |
DEA |
1998-2007 China's 31 provinces industrial sector |
Capital/labor |
Industrial added value |
COD/ SO, |
National average ML: 1.085 (contribution of technological progress 95.29%) |
- |
|
Yue & Liu (2009) |
Inversed _ output Reciprocal method DDF |
DEA |
36 industrial sectors in China between 2001 and 2006 |
Capital/labor |
Industrial added value |
SO, |
Efficiency value:inverse algorithm: 0.55; reciprocal approach: 0.49; directional distance function: 0.68 |
|
|
Yang & Shao (2009) |
DDF |
DEA |
30 provincial industrial sectors from 1998 to 2007 |
Capital/labor |
Industrial added value |
SO, |
East: 0.886, Middle: 0.703, West: 0.686 |
- |
|
Zhou & Gu (2009) |
DDF |
DEA |
Industry data of large and mediumsized industrial enterprises in Shanghai from 1997 to 2004 |
Capital/labor/ energy |
Total industry |
SO, |
2006 technological efficiency index: 0.6437 (heavy industry); 0.7396 (light industry) |
|
Authors |
Evidence-based methods |
Data |
Variables |
Main results |
|||||
|
Model |
Functions |
Assessment |
Input |
Desirable Output |
Undesirable Output |
Productivity |
Shadow price |
||
|
Chen et al. (2010) |
DDF |
DEA |
Industrial enterprises above a designated size in 11 provinces in the east of China from 2000 to 2007 |
Capital/labor |
Output value |
SO, |
ML: 0.902(2007) |
||
|
Wang et al. (2010) |
DDF |
- |
DEA |
30 provinces in China from 1998 to 2007 |
Capital/labor/ energy |
Industrial added value |
COD/ SO, |
National average:0.712 (VRS); 0.657 (CRS) |
- |
|
Dong et al. (2010) |
- |
- |
DEA |
|
Capital/labor/ sown area/ energy |
Actual regional GDP |
Reciprocal of environmental pollution index |
ML: 1.008 |
- |
|
Wu et al. (2010) |
DDF |
- |
DEA |
East/Central/ Western region, 2000-2007 |
Capital/labor/ human capital |
GDP |
COD/ SO2 |
- |
- |
Note: price shown is the price of the year unless there is special explanation.
DF(O): Output distance function; DF(I): input distance function; DDF: directional distance function; DEA: Data Envelope Analysis method; SFA: stochastic frontier approach; M index: Malmquist productivity index; ML index: Malmquist-Luenberger productivity index; VRS: variable scale; CRS: invariant scale; BOD: biochemical oxygen demand; COD: chemical oxygen demand; TSS: total suspended solids; PART: particles; SO, sulfur oxides:
(Gallop & Roberts, 1985; Coggins & Swinton, 1996; Lee et al., 2002; Fare et al., 2005, 2007); in addition, some industrial producers whose emissions are easily metered, such as BOD/COD emissions from paper mills, water pollutants (Fare et al., 1993; Chung et al., 1997; Hailu & Veernan, 2000; Van Ha et al., 2008), waste oils discharged from ceramic plants (Reig-Martinez et al., 2001) are also used in the environment.
As countries continue to pay attention to greenhouse gas (GHG) emissions issues, some scholars have begun to shift their perspectives to macro-level research. They usually measure and compare the environmental sensitivity productivity of major GHG emissions such as CO2 and NO,, which are equivalent in economic development level (such as OECD, transition economies) or geographically similar countries (such as APEC, ECJ) (Salnykov & Zelenyuk, 2005; Kumar, 2006; Wang Bing et al., 2008).
Limited by the lack of data at the domestic enterprise level, especially the pollutant data is difficult to obtain. Most of the research on China is based on the macro-level, mainly to measure the environmental sensitivity productivity and the marginal abatement cost of pollutants in different provinces or industries. For example, Ke et al. (2008) used the output distance function and the super-logarithmic function form to measure the environmental sensitivity productivity of 30 provinces in mainland China from 1996 to 2002, and estimated the shadow of SO, pollutants. Hu Angang et al.
- (2008) used the directional distance function earlier, using CO,, COD, SO,, total wastewater discharge and total solid waste discharge as indicators of undesired output, measuring 30 provinces in mainland China environmentally sensitive productivity during 1999-2005. At the industry level, there are mainly Tu Zhengge (2008, 2009), Wu Jun (2009), Yue Shujing and Liu Fuhua
- (2009), Yang Jun and Shao Hanhua (2009), Zhou Jian and Gu Liuliu (2009), Chen Ru et al. (2010). Based on the SO, data of China's provincial industrial sector, the scholars used the directional distance function to measure the environmental sensitivity productivity of the industrial sectors in various provinces; at the regional level, Wang Bing et al. (2010), Dong Feng et al.
- (2010), Wu Jun (2010) adopts provincial input-output data. In addition to capital and labor, the input includes energy consumption, human capital and other factors. The output end includes two Eleventh Five-Year Plan of COD and SO,. The pollutants required to be forced to reduce emissions use the directional distance function to calculate the environmentally sensitive productivity at the provincial level. In addition, in the study of Tu Zhenge (2008), Wang Bing et al. (2010) scholars, the factors affecting environmentally sensitive productivity were further analyzed and the environmental Kuznets curve and pollution paradise hypothesis have been empirically tested. These studies have important implications for understanding the differences in environmentally sensitive productivity between industries and regions, but they are limited by data factors, and their microscopic mechanisms are often not revealed.
