Chinese Studies
2012. Vol.1, No.3, 18-22
Published Online November 2012 in SciRes (
Copyright © 2012 SciRes.
Impact of Soft-Input on the Cost of Resources and Environment
in Economic Growth
——Empirical Research Based on Chinese Provincial Panel Data
Chunling Pan, Jie Lü*
The College of Economics and Management, Shenyang Agricultural University, Shenyang, China
Email:, *
Received May 31st, 2012; revised July 28th, 2012; accepted August 26th, 2012
China’s rapid economic growth has proceeded at considerable cost of resources and environment. To find
out how to reduce the cost while maintaining economic growth, we took panel data of 28 provinces from
2000 to 2009 as our samples and conducted comparative analysis on the said cost among 28 provinces.
From the perspective of soft-input, we subsequently examined soft-input’s influence on the cost of re-
sources and environment of in economic growth (CREIEG). Results show that: government spending on
scientific research and environmental protections has a significant role in reducing energy consumption
and wastewater discharge, but it has limited impact on gas emissions and solid waste emissions reduction.
Moreover, financial development also has a positive role in reducing energy consumption. Finally, we
propose a number of initiatives to reduce the cost of resources and environment of economic growth
based on the analysis results.
Keywords: Soft-Input; Economic Growth; Cost of Resources and Environment
China has sustained a rapid economic growth. Her per capita
GNI, after exceeding $1000 for the first time in 2003, reached
$1500 in 2004, and rose to $1740 in 2005 as a result of further
improvement. Her world ranking rose from No. 138 in 2001 to
No. 128 in 2005 (National Bureau of Statistics, 2010). However,
China’s remarkable economic growth has proceeded at consid-
erable cost of both resources and environment. Compared with
developed countries, China has relatively low resource-reco-
very rate. Likewise, her comprehensive utilization rate is also
not high. Many usable and reusable resources have been wasted.
China is now the world’s largest energy-consuming country,
and also a country with the largest CO, SO2, CO2 emissions as
well as other environmentally hazardous substances. The CO2
emissions in China in 2007 took up 22.3% of the total global
emissions (293 million tons), which for the first time exceed the
emissions in the United States (International Energy Agency,
2007). The question how to reduce environmental damage to
the minimum while maintaining economic growth has become
a focus of governmental as well as academic attention in Eco-
nomists; however, mostly focus their attention on tangible fac-
tors, such as capital, raw materials, and labor. These factors
invariably have a material form, which is often known as hard-
input. But economic growth is something larger than a combi-
nation of physical and chemical processes. Human participation
will give rise to factors in non-material form. These elements in
non-specific material form are known as soft input among eco-
nomists. They include system, technology, and so on. Com-
pared with research on hard-inputs, there has been little re-
search on soft input and its effect on the cost of resources and
environment in economic growth (CREIEG henceforward)
because of its unmeasured characteristic. In fact, soft-input has
at least the same extent of impact on CREIEG as hard-input
In summary, the impact of soft input on CREIEG is an im-
portant economic issue in both theory and practice. Taking the
fast-growing Chinese economy as a case study we use the pro-
vincial panel data from 2000-2009 to examine the difference in
CREIEG among the provinces under study, and to analyze the
impact of soft-input on CREIEG.
The rest of the paper is organized as follows: Section 2 is re-
search design, including a brief introduction of variables, selec-
tion of the regression model, and statistical characteristics of
variables. Section 3 presents the empirical results. Section 4 is a
section of conclusions and suggestions.
Empirical Specifications
Variable Definiti ons
Dependent Variable
We mainly investigate CREIEG’s dimensions of energy
efficiency and pollution emission level.
Energy Efficiency Indicators
Energy consumption per unit GDP (Represented by enegdp)
is the ratio of the total energy consumption (tons of standard
coal) to Domestic (regional) GDP (million). Energy consump-
tion per unit industry GDP (Represented by eneind) is the ratio
of the total energy consumption (tons of standard coal) to do-
mestic (regional) industrial GDP (million). And power con-
sumption per unit GDP (Represented by eneele) is the ratio of
the total electricity consumption (1000 kWh) to domestic (re-
gional) GDP (million).
*Corresponding author.
C. L. PAN, J. LÜ
Pollution Emission Indicators
Sewage emissions per unit GDP (Represented by watgdp) is
the ratio of Sewage emissions (tons) to Domestic (regional)
GDP (million). Gas emissions per unit GDP (Represented by
gasgdp) is the ratio of gas emissions (standard cubic meters) to
domestic (regional) production GDP (million). And solid waste
emissions per unit GDP (Represented by solgdp) is the ratio of
solid waste emissions (tons) to domestic (regional) GDP (mil-
Explanatory Variables
Soft-input is the sum of non-material elements in addition to
the material elements of the production process in certain
economies. As defined in (Li, 1995), soft-input divides itself
into the following three categories: comprehensive policy in-
vestment, comprehensive science and technology investment,
and the workers’ enthusiasm investment. Comprehensive policy
investments include institutional policy and economic man-
agement. Since it is difficult to quantify workers’ enthusiasm,
this paper mainly examines the impact of comprehensive policy
investment and comprehensive science-and-technology invest-
ment on the environmental cost. Taking the literatures of (Jiang,
Hou, & Liu, 2004; Chao, 2009; Cheng, 2010) and other studies
as references, we mainly choose the following variables shown
in Table 1.
Regression Specification
To test the effect of soft input on CREIEG, we use the fol-
lowing regression specification:
01 234
567 8
gov open save edu
scinonenv cap
itititit it
itititit it
yaa aaa
aaa a
 
 
In the specific regression analysis, we employ Stata statisti-
cal software, using fixed effects methods to estimate the above
Description of Variables
All data in this paper came from “China Statistical Year-
book” of Chinese National Bureau of Statistics over the years
and “China Compendium of statistics”, results of variable de-
scrip- tion are shown in Table 2.
Table 1.
Variable description.
Variable Definition
gov Government spending in addition to education divided by
open Total imports and exports divided by GDP
save Financial institutions’ deposits at the end of year divided
by GDP
edu Government spending on education divided by GDP
sci Government spending on scientific research divided by
env Government spending on environment protection divided
by GDP
cap Amount of share capital per Unit employees
Table 2.
Parameters variables’ summary statistics from 2000 to 2009.
Variable Obs Mean Std. Dev. Min Max
enegdp 150 1.53 0.77 0.61 4.14
eneind 150 2.81 1.55 0.81 9.03
eneele 150 1.52 0.95 0.68 5.71
watgdp 310 38.97 28.20 3.98 159.90
gasgdp 279 4.35 2.79 0.49 15.54
solgdp 310 2.45 1.93 0.21 10.89
gov 310 15.91 11.59 5.52 92.69
open 310 33.78 44.99 3.68 184.29
save 279 147.40 65.62 73.19 530.87
edu 310 2.76 1.50 1.30 13.83
sci 310 0.16 0.17 0.03 1.43
nonnat 93 58.05 18.31 19.81 89.16
env 93 0.70 0.53 0.09 2.68
cap 303 1.62 1.35 0.24 10.24
Source: National Bureau of Statistics, “China Statistical Yearbook” (2001-2010),
China Statistics Press.
Empirical Results
Abbreviations and Acronyms Comparative Analysis
of the Cost of Resources and Environment
Provincial and Regional Energy Efficiency Comparison
As shown in Table 3, energy consumption per unit of GDP
in eastern provinces is much lower than in central and western
provinces. For instance, energy consumption per unit GDP is
very high in Shanxi, Qinghai, Inner Mongolia, and Guizhou.
But the energy consumption per unit of industrial GDP and
electricity consumption per unit GDP show similar geographi-
cal characteristics. Therefore, from the perspective of energy
consumption, CREIEG in eastern provinces is significantly
lower than in central and western provinces. So, when they
attempt to change the mode of extensive economic growth,
central and western provinces face much more challenge than
their eastern counterparts.
Provincial and Regional Pollution Emission Comparison
From Figure 1, we can see that pollution levels in central
and western provinces are much higher than in eastern prov-
inces. Such western provinces as Guangxi, Shanxi, and Chong-
qing have a relatively high pollution level. From the perspective
of environmental pollution, there is a big gap between central
and western provinces on the one hand and eastern provinces
on the other in terms of quality of economic growth.
Therefore, economic growths in central and western cities
are mainly relying on extensive development model due to their
lower level of technology and management. While economic
growth is inevitably accompanied by high energy consumption
and thus high pollution, conditons in various eastern areas i
Copyright © 2012 SciRes. 19
C. L. PAN, J. LÜ
Copyright © 2012 SciRes.
Table 3.
Trends of energy efficiency.
Enegdp Eneind Eneele
2005-2007 2008-2009 2005-2007 2008-2009 2005-2007 2008-2009
Heilongjiang 1.41 1.25 2.22 1.64 0.96 0.83
Jilin 1.59 1.33 2.81 1.80 1.00 0.85
Liaoning 1.77 1.53 2.89 2.34 1.36 1.17
Beijing 0.76 0.63 1.34 0.97 0.79 0.70
Fujian 0.91 0.83 1.38 1.17 1.15 1.07
Guangdong 0.77 0.70 1.03 0.84 1.17 1.04
Hainan 0.91 0.86 3.17 2.61 0.96 0.95
Hubei 1.90 1.68 4.16 3.16 1.53 1.47
Jiangsu 0.89 0.78 1.55 1.19 1.21 1.11
Shandong 1.23 1.09 2.02 1.62 1.06 0.99
Shanghai 0.86 0.76 1.13 0.96 0.96 0.85
Tianjin 1.07 0.89 1.33 0.98 1.03 0.85
Zhejiang 0.86 0.76 1.41 1.15 1.24 1.19
Gansu 2.19 1.94 4.63 3.79 2.52 2.47
Guangxi 1.19 1.08 2.89 2.29 1.26 1.27
Guizhou 3.17 2.61 5.16 4.32 2.59 2.39
Inner Mongolia 2.40 2.08 5.31 3.87 1.91 1.79
Ningxia 4.06 3.57 8.61 6.82 5.41 4.90
Qinghai 3.09 2.81 3.52 3.09 3.99 3.96
Shaanxi 1.42 1.23 2.45 1.69 1.37 1.17
Sichuan 1.49 1.36 2.99 2.36 1.26 1.12
Xinjiang 2.08 1.95 2.90 3.05 1.23 1.37
Yunnan 1.69 1.53 3.37 2.79 1.66 1.62
Chongqing 1.37 1.22 2.60 1.98 1.15 0.99
Anhui 1.17 1.05 2.88 2.22 1.10 1.10
Henan 1.34 1.19 3.75 2.89 1.28 1.24
Hubei 1.46 1.27 3.28 2.51 1.19 1.06
Hunan 1.35 1.21 2.71 1.78 1.06 0.94
Jiangxi 1.02 0.90 2.71 1.81 0.98 0.93
Shanxi 2.86 2.46 5.96 4.72 2.39 2.11
Source: National Bureau of Statistics, “China Statistical Yearbook” (2001-2010), China Statistics Press.
have become more mature thanks to the reform and opening-up
policy. So, those areas can devote more energy to transferring
the economic growth mode, saving energy, and protecting en-
vironment, which explains why eastern provinces are taking
lead in this regard.
Impact of Soft-Input to the Cost of Resources and
Effect of Soft-Input on Energy Consumption
The effect of soft-input on energy consumption can be seen
in Table 4.
From Column 1, we can see that save, edu, sci, env, and cap
can significantly reduce energy consumption per unit GDP. But
in Column 3, we find no significant impact of industrial edu on
energy consumption of per unit industry GDP, and in Column 5,
there are no significant effects of env, and edu on power con-
sumption per unit GDP. These results indicate that government
spending on scientific research and environmental protection
have significant role in the reduction of energy consumption.
Financial development, likewise, also has a positive role in
reducing energy consumption.
C. L. PAN, J. LÜ
Fujia n
Hain an
shan gh ai
Inner Mongolia
Qin ghai
Sich ua n
Sin k ia n g
Chongqi ng
Hen a n
Hun a n
Sh an x i
0 20406080100120
gagdp solgdpwat gdp
Figure 1.
Provincial average index of industrial pollution from 2000
to 2009. Source: National Bureau of Statistics, “China Sta-
tistical Yearbook” (2001-2010), China Statistics Press.
Soft-Input’ Effect on Pollution Emission
Results of soft-input’ effect on pollution emission can be
seen in Table 5.
From Column 1, we can see that save, sci, non, and env can
significantly reduce sewage emissions per unit GDP, but they
have no significant effect on gas emissions and solid waste
emissions, which can be seen in Columns 3 and 5. This shows
that government spending on scientific research and environ-
mental protections has distinct role in the reduction of sewage
discharge, but its impact on the reduction of gas emissions and
solid waste emissions is rather limited.
The results of empirical analysis have led us to the following
conclusions: 1) Compared with eastern cities, central and west-
ern cities have higher energy consumption, higher pollution,
and hence greater CREIEG; 2) As of the effect of soft-input on
CREIEG, governmental soft-input of education investment,
scientific-research investment, and environmental investment
can significantly reduce the energy consumption per unit GDP,
so can the soft-input of financial-development level and share
capital per unit employees. Moreover, government expenditures
on scientific research and environmental protection can de-
crease Sewage discharges.
The following measures can be taken to reduce CREIEG:
since government is currently the main soft-input body, we
should make it possible for more participants, especially com-
panies, to join the soft-input projects. We should continue to
deepen the reform of financial system, improve capital alloca-
tion efficiency, strengthen cooperation between economically
developed regions and less developed regions, apply differen-
tial financial institutions and more preferential policies for
economically underdeveloped areas. Finally, since educational
investment and science-and-technology investment, through the
Table 4.
Effect of soft-input on energy consumption Soft-input’ effect on pollu-
tion emission.
Enegdp Eneind Eneele
(1) (2) (3) (4) (5) (6)
b t b t b t
gov 0.01 0.98 0.02 0.59 0.00 0.01
open 0.00 0.91 0.00 0.72 0.00 0.55
save 0.00*** 3.12 0.01** 2.15 0.01** 2.55
edu 0.09** 2.26 0.06 0.36 0.00 0.03
sci 0.94*** 2.872.30* 1.76 1.78*** 2.96
non 0.00 0.54 0.02 1.17 0.00 0.15
env 0.08** 2.010.63* 1.84 0.05 0.31
cap 0.09*** 3.00 0.30** 2.30 0.15** 2.70
Cons 2.38*** 12.925.95*** 7.48 2.20*** 6.52
Obs 60 60 60
R2 0.78 0.30 0.66
Note: ***, **, * indicate significant at 1%, 5% and 10% respectively.
Table 5.
Effect of soft-input on pollution emission.
watgdp gasgdp solgdp
(1) (2) (3) (4) (5) (6)
B t b t b t
gov 0.30 0.93 0.01 0.09 0.01 0.24
open0.04 0.54 0.01 0.17 0.01 1.11
save 0.13** 2.31 0.03 1.15 0.00 0.44
edu 2.11 1.02 0.71 0.78 0.10 0.43
sci 27.14*1.7310.04 1.44 0.63 0.36
non 0.61*** 3.32 0.06 0.75 0.00 0.07
env 4.28** 2.04 2.34 1.29 0.37 0.81
cap 1.93 1.260.46 0.68 0.32*1.87
Cons 71.93*** 7.84 11.92*** 2.94 3.83*** 3.71
Obs 62 62 62
R2 0.74 0.33 0.35
Note: ***, **, * indicate significant at 1%, 5% and 10% respectively.
technological innovations they bring about, play a significant
role in decreasing CREIEG, it is extremely important to create
an institutional environment that is favorable for technological
National Bureau of Statistics (2010). China statistical yearbook. Bei-
jing: China Statistics Press.
International Energy Agency (2007). Key world energy statistics. URL
Copyright © 2012 SciRes. 21
C. L. PAN, J. LÜ
Copyright © 2012 SciRes.
(last checked 12 February 2007).
Li, G. Z. (1995). Several theoretical issues on soft-input research. Re-
search and Developmen t, 5, 11-14.
Jiang, R. X., Hou, X. L., & Liu, Y. (2004). Evaluating China’s provin-
cial sustainability by integrated indicators. Journal of the Graduate
School of the Chinese Academy of Sciences, 21, 63-70.
Chao, X. J. (2009). The quality of economic growth: A theoretical ex-
planation and empirical analysis in China. Ph.D. Thesis, Xi’an:
University of Northwest.
Cheng, W. H. (2010). Soft-input and region quality of economic gro wth.
Ph.D. Thesis, Lanzhou: University of Lanzhou.