A key issue in water resources management in China is the adoption of a total volume control framework of water supply at a regional level corresponding to socioeconomic development. This requires more efficient water demand management in order to achieve a balance between water supply and demand. The purpose of this paper is to conduct a thorough analysis of the water use structure of production sectors in order to identify the major impact factors and thus provide insights for demand management policy design. Taking Jiangsu Province as a case study, we compile a series of extended input-output (I-O) tables at a constant price and establish an I-O structural decomposition model. The major factors leading to change in water use by production sectors (primary, secondary and tertiary) in the Jiangsu Province during five time periods from 1997–2010 are categorized into structure effect, water use efficiency effect, and demand effect for each of three levels: (1) aggregated sectors, (2) sub-aggregated sectors, and (3) individual sectors. Within the study period, the demand effect consistently leads to an increase in water utilization and the increase effect becomes weaker over time, while the other two factors consistently lead to a decrease in water utilization and their dampening effect becomes stronger.

Introduction

Water shortages and water quality degradation have seriously restricted sustainable and economic development in China (China Sustainable Development Strategy Study Group, 2007; Cai, 2008). China has focused on engineering development for reliable water supply for several decades. However, in many regions of the country water supply development is limited by water availability, as well as capital, technological and environmental pressures. Recently, water resources management in China has focused on water demand management. In 2011, the State Council of China issued an order to control the total amount of water use, wastewater discharge and water quality, and set targets for water use efficiency (CPC, 2011) at the various levels of administration, from province to district to county. The decree also requests water uses at the river basin scale to follow the regulations. It is expected that these controls will create economic incentives, technology development and changes in water use behavior (Wang, 2004; McKay, 2005; Gao & Dang, 2008).

Within China, 86.1% of water withdrawals are used for production (including the primary, secondary and tertiary sectors), 12.1% for domestic use, and 1.8% for the environment (Ministry of Water Resources of China, 2013). As people's standard of living improves, the domestic and environmental water use quantity will increase. A key question is how to continue regional economic and social development under growing water demands and diminishing available water supplies. The purpose of this paper is to analyze the driving factors of water uses in the production sectors including agricultural, industrial and municipal (referred to as water use in short below), understand the mechanisms of water use change, and provide policy implications for the water demand of production sectors in regions of China with rapid industrialization and urbanization.

In this paper we adopt a factor decomposition model (FDM) to identify the various factors of water use at the regional scale. Numerous studies used structural decomposition analysis to analyze the changes in energy use, greenhouse gas emission (Gould & Kulshreshtha, 1986; Rose & Chen, 1991; Weber, 2009; Su & Ang, 2012), and air pollutants and solid wastes (Haan, 2001). However, to our knowledge, there have not been any studies that apply structure decomposition analysis to analyze changes of water use based on a series of input-output (I-O) tables in a region. As an application, the method is applied to Jiangsu Province in southeastern China. The series of tables of input-output at a constant price were compiled and used to develop a water use change decomposition model. The driving factors of water use change of production sectors in Jiangsu Province were classified into structure effect, water use efficiency effect, and water demand effect. The effects of these three types of factors on the evolution of the water use structure was analyzed at three levels: (1) aggregated sectors (i.e., the entire production system), (2) sub-aggregated sectors, and (3) individual sectors. Fuzzy clustering was used to cluster the three types of influence factors into different levels (e.g., strong, medium, and weak) for each of the sectors. Because decomposition residuals exist in most structure decomposition models (Ang & Zhang, 2000; Boyd & Roop, 2004), this study uses the path based method (PBM; Hoekstra & Bergh, 2002), which represents a general expression of the I-O structure decomposition. This avoids the impact of different expressions on the results. Identifying and quantifying the socioeconomic and technological factors leading to water use change provides policy implications regarding production structure, technology adaptation and water use efficiency improvement.

Table 1.

Sub-aggregated sectors.

Sub-aggregated sectors Sectors 
Primary Agriculture 
Secondary Coal mining and processing, other mining, oil and gas, food, textile, forest, paper, chemical, building materials, metallurgical, machinery and equipment, electronic instrument, other manufacturing, power, production and supply of tap water, construction 
Tertiary Transportation and telecommunication, accommodation and restaurants, wholesale and retail trade, other services 
Sub-aggregated sectors Sectors 
Primary Agriculture 
Secondary Coal mining and processing, other mining, oil and gas, food, textile, forest, paper, chemical, building materials, metallurgical, machinery and equipment, electronic instrument, other manufacturing, power, production and supply of tap water, construction 
Tertiary Transportation and telecommunication, accommodation and restaurants, wholesale and retail trade, other services 

Method

FDMs are based on a variety of decomposition methods, which provide effective tools to explain fundamental determinants of variable change and diagnose the mechanisms of resource usage (Hoekstra & Bergh, 2003). FDMs adopt a systems approach to examine the relationship between the change of a dependent variable and the change of a number of independent variables. The dependent variable change is decomposed into a number of components, each of which is related to a subset of independent variables by the measurement of their contribution to the change in the dependent variable. Two models are commonly used in resources economics: index decomposition analysis (IDA) and structural decomposition analysis (SDA). Both methods have been used to assess the driving forces of economic growth and technology changes over a variety of socioeconomic and environmental indicators (Li, 2004). Index decomposition analysis uses the aggregated data of sectors; for example, it analyzes water use change induced by the total output, without considering the intermediate production. Thus, IDA can only reflect direct effects. SDA uses input-output tables (i.e., consumption coefficient matrices) and thus can be used to conduct a detailed analysis on not only the direct effect of each of the factors but also the indirect effects. SDA, unlike other methods, accounts for the connections among the sectors based on the input-output relationships.

IDA has been used to analyze the change in water availability in China. For example, Sun & Wang (2009) established an IDA model of total water use in Liaoning Province to measure the driving forces of water utilization change, including economic growth, economic structure, and water use efficiency. Zhang et al. (2011) applied the logarithmic mean division index (LMDI) method to analyzing the influencing factors of water resource utilization change in Dalian City, including the water use quota, industrial structure, economic growth and population.

Using I-O analysis as the basic method, we adopt SDA to explore the water decomposition factors. Assuming W is the total water consumption, (I-A)−1 is the Leontief inverse matrix L, Y is the final demand vector, M is the inflow vector (including interregional inflows and import), and Q is the unit direct consumption of output water (also named direct water quota) column vector, we use: 
formula
1
Equation (1) indicates that the total amount of water use depends on the unit water consumption, intermediate product demand and final demand. Using zero as the base period and t as the target period, the water use change can be expressed in increments as follows: 
formula
2
The decomposition of Equation (2) can have various forms. This paper introduces a path-based method to build a general decomposition equation, thereby differentiating it from other decomposition models in other studies (Miller & Shao, 1994; Harrison et al., 2000). First, suppose the value of z and xi vary between the initial time (0) time (t), thus: 
formula
3
The expression of differential change of Z from time 0 to 1 is: 
formula
4
Variable xi is a time function, differential change of z can be written as: 
formula
5
The total change of Z is the sum of the change of every differential: 
formula
6
Each factor variable's impact on z can be written as: 
formula
7
Assume that the time path of each factor variable xi is determined by the parameter . In order to describe the influence path variable time parameter, a monotone function type (8) is introduced. 
formula
8
Put (8) in (7), then, the influence of each factor can be written as: 
formula
9
If there are only initial time and end time data, it is assumed that each time path parameter is equal (that is θ1 = θ2 = …= θn): 
formula
10
Then the factors affecting the water use change expression are as follows: 
formula
11
 
formula
12
 
formula
13
where Equations (11), (12) and (13) are used to calculate water use changes driven by water use efficiency, structure and demand effects, respectively. , and refer to the water use change driven by water use efficiency, structure and demand effects.

Water use change of production sectors is caused by economic structure change, water utilization intensity change, as well as final demand change. The expression represents the effect of a change in water utilization intensity on water use (referred to as water use efficiency effect). represents the change of water use due to the change of economic structure and production technology resulting in intermediate demand adjustments (referred to as structure effect). represents the effect of final demand changes on water use (referred to as demand effect). , and represent the change of water use due to the changes of two of the above three effects in each case. represents the change of water use due to the changes of the above three effects.

In the same way, if is represented as , is the vector order diagonal matrix of the production sectors. In this way, we can obtain the decomposition type of water use change for every production sector.

Indicators and data sources

Indicator selection

China's statistics sectors conduct a national, large-scale input-output survey every five years for the preparation of basic input-output tables, which are released in years ending in digits two and seven (e.g., 1997, 2002, and 2007). Simple input-output tables are prepared from small-scale surveys and basic coefficient table adjustments in years ending in digits zero and five (e.g., 2000, 2005, and 2010). From these different input-output tables, data were collected for Jiangsu Province from 1997 to 2010 (1997, 2000, 2002, 2005, 2007 and 2010). Five time periods were formed from the data (1997–2000, 2000–2002, 2002–2005, 2005–2007 and 2007–2010), which were used to decompose the factors leading to change in water consumption in Jiangsu Province. The unit of water use volume is billion cubic meters (BCM, 109 m3); water consumption by unit output is ‘m3/104 yuan’.

Data sources and data processing

Data were collected from the ‘Jiangsu Statistical Yearbook,’ ‘Chinese Statistical Yearbook,’ and ‘Jiangsu Water Resources Bulletin’ for the years ranging from 1997 to 2011.

Jiangsu Province's input-output tables reflect the current water prices and do not report the physical water quantity transferred and consumed among the sectors. Therefore, we created and compiled a series of input-output tables related to water use.

We did this by first collecting each economic sector's ex-factory price index (Liu & Peng, 2010) and using the previously mentioned price index to process the corresponding data of I-O tables for Jiangsu Province. We compiled the constant price I-O tables using 2005 as the baseline year. Then, on the basis of the I-O tables, we added the water use volume into the ‘input block’ of the I-O table as the IV quadrant, which can reflect the water usage of economic sectors. The extended I-O table structure can be found in Appendix Table A1 (available with the online version of this paper). Along the way, the extended tables in 1997, 2000, 2002, 2005, 2007 and 2010 years were compiled. The inconsistency of the item classification within the I-O tables from 1997 to 2010 and the limited water use data availability for every economic sector led us to merge the multitude of economic sector categories within the I-O tables into 21 sectors. The internal relations between each sector, the ‘classification of economic sectors,’ and the experience of experts in related fields guided the formation of the new sector categories. The 21 sectors are listed in Table 1 and they are aggregated into three categories: primary, secondary and tertiary.

Results and analysis

The driving factors of water use change are assessed at three sector levels, i.e., (1) aggregated sectors, (2) sub-aggregated sectors (i.e. the primary, secondary and tertiary sectors), and (3) the individual sectors.

Water use change decomposition characteristics for the aggregated sectors

The decomposition method was applied for five periods, respectively, to determine the effects of production structure, water use efficiency, and demand on water usage (see Figure 1).
Fig. 1.

Factor decomposition of total water use change of production sectors in Jiangsu Province.

Fig. 1.

Factor decomposition of total water use change of production sectors in Jiangsu Province.

Water use from 1997 to 2000 grew by 8.26 BCM. The structure effect is responsible for 1.58 BCM of the increased water consumption. An improvement in water use efficiency caused a reduction in water consumption of 2.64 BCM. The demand effect, however, increased water consumption by 9.32 BCM. Water use from 2000 to 2002 had a growth of 3.61 BCM. The demand effect caused an increase of 11.78 BCM in water consumption. This was partially offset by the structure effect reducing water consumption by 2.92 BCM. Water consumption was further reduced by 5.26 BCM by water use efficiency improvement. Water use grew by 3.43 BCM from 2002 to 2005. The demand effect increased water consumption by 22.64 BCM, but this was partially offset by the structure effect (reducing by 2.47 BCM) and water use efficiency effect (reducing by 16.56 BCM). Water use from 2005 to 2007 experienced a growth of 2.57 BCM. The water consumptions were reduced by 0.24 BCM by the structure effect and by 15.00 BCM by the water utilization efficiency effect, meanwhile water consumption was increased by the demand effect by 17.82 BCM. Finally, water use from 2007 to 2010 experienced a net growth of 0.51 BCM. Structure effect and water use efficiency effect reduced water consumption by 3.15 BCM and 21.03 BCM, respectively, while the demand grew by 24.70 BCM.

Decomposition results show that the demand effect is the primary driving force of water use increase amongst the production sectors in Jiangsu Province, since the year 2002; however, the impact on water consumption from the demand effect has gradually decreased. The structure effect and the water use efficiency effect are the two key factors in reducing the region's water use, and the gross contribution of the two effects to water use saving is increasing over the years. However, the degree to which the structure effect decreases water use (i.e., the marginal contribution) appears to be shrinking, especially in more recent years, indicating that marginal gains from implementing economic structure change are diminishing and future water savings should be limited from this aspect. Meanwhile the water use efficiency effect is playing a substantial increasing role, showing continual improvements in water savings from 1997 to 2010. Improvements in water use efficiency between 2002 and 2005 within the agriculture sector, the largest water user, reduced overall water use in Jiangsu Province by 2.39 BCM.

Water use change decomposition characteristics in sub-aggregated sectors

The effect of structure, water use efficiency, and demand are quantified and analyzed for the primary, secondary and tertiary sectors separately, as shown in Figures 25.
Fig. 2.

Effect of water use efficiency, structure, and demand on water use change from 1997 to 2010 in Jiangsu Province.

Fig. 2.

Effect of water use efficiency, structure, and demand on water use change from 1997 to 2010 in Jiangsu Province.

Fig. 3.

Water use change from 1997–2010 within the primary sectors driven by the effect of water use efficiency, structure, and water demand for each time period.

Fig. 3.

Water use change from 1997–2010 within the primary sectors driven by the effect of water use efficiency, structure, and water demand for each time period.

Fig. 4.

Water use change from 1997–2010 within the secondary sectors driven by the effect of water use efficiency, structure, and water demand for each time period.

Fig. 4.

Water use change from 1997–2010 within the secondary sectors driven by the effect of water use efficiency, structure, and water demand for each time period.

Fig. 5.

Water use change from 1997–2010 within the tertiary sectors driven by the effect of water use efficiency, structure, and water demand for each time period.

Fig. 5.

Water use change from 1997–2010 within the tertiary sectors driven by the effect of water use efficiency, structure, and water demand for each time period.

From 1997 to 2010, the water utilization change of the primary sectors was 11.31 BCM. The contribution rates attributed to demand, structure, and water use efficiency are 80.3%, 15.3%, and 4.3%, respectively. The water use change of the secondary sectors was 6.95 BCM, the increase in demand for water led to a 45.39 BCM increase in water consumption, which was partially offset by the improvement in structure and water use efficiency that led to reductions in water use of 34.27 BCM and 4.17 BCM, respectively. The water utilization change of the tertiary sectors was 0.31 BCM, the increase for water use by demand was 3.02 BCM, and the reductions for water use by structure and water use efficiency were 0.58 BCM and 2.13 BCM, respectively (Figure 2).

The effect of increasing water demand on overall water use within the primary sectors is declining. From 1997 to 2000, water use increased by 6.62 BCM, while from 2007 to 2010 water use increased by 1.71 BCM only. Growth slowed in the primary sectors because the overall importance of the primary sectors in gross domestic production (GDP) shrunk during the period of 1997–2010, thereby leading to a steady decline in water demand increase over the five sub-periods. Water use within agriculture (the main sector of the primary) is easily affected by weather patterns and uncertainty. Thus, the structure effect and water use efficiency effect on water use change demonstrate some variation (Figure 3).

The secondary sectors had only a slight increase in water use from 10.10 BCM during the years 1997–2000 to 10.44 BCM during 2007–2010. Water demand had a steady, increasing effect because of the increasing added value of the secondary sectors and their increasing share of GDP over years. The effect of water use efficiency steadily decreased water use, resulting from water-saving technology application. Structure effect made water use steadily decrease, indicating the continuous improvement of production technology and the more rational economic structure (Figure 4).

Water use in the tertiary sector had an upward trend, rising from 0.07 BCM over 1997–2000 to 0.28 BCM over 2007–2010. The impact of increasing water demands on overall water use is evident across all five study periods. From 1997 to 2000, the effect of increasing water demand was relatively weak because the development of the tertiary sectors was slow; however, as the relative importance of the tertiary sectors gradually grew, so did the water demands of those sectors. The production structure had little effect on water use in the tertiary sectors. Water use efficiency significantly decreased water use with small fluctuations over years. Although the tertiary sectors account for a small fraction of total water use in the region (2.35% in 2010), the growth of those sectors along the overall economic growth can lead to greatly increased water use, and water use efficiency improvement is needed to offset the net increase in water use.

Water use change decomposition characteristics of individual sectors

Examining the various driving forces for individual sectors provides some insights for particular sectors. Factors decomposition of water use of individual sectors from 1997 to 2010 results are shown in Table 2. Over the five time periods during 1997–2010, the effect of structure reduces water use across most production sectors, especially coal mining and processing, other mining industry, other manufacturing, power, and construction. Compared to the structure effect, improvements in water use efficiency led to even larger water use reductions, indicating that water use efficiency is the primary factor offsetting the overall increase in water use in the Province. Increasing water demands are the main factor leading to water use increases in the various sectors (except for ‘other mining industry’ and ‘other manufacturing’). The increase in water use due to expanding production demands exceeded the reduction due to economic structure change and water use efficiency improvements. This explains why the overall water use increases in the region.

Table 2.

Factors decomposition of water use of individual sectors from 1997 to 2010 (unit: 109 m3).

  1997–2010
 
1997–2000
 
2000–2002
 
Sector number Water efficiency effect Structure effect Demand effect Water use efficiency effect Structure effect Demand effect Water efficiency effect Structure effect Demand effect 
−0.81 −2.86 14.98 −0.02 −2.06 8.71 −1.47 0.19 4.06 
−0.02 0.01 0.00 −0.01 0.00 0.01 −0.03 0.00 0.00 
−0.47 −0.16 0.51 −0.06 −0.01 0.10 −0.06 −0.01 0.00 
−0.01 0.02 −0.02 −0.01 0.00 0.03 0.27 0.02 −0.22 
−0.79 −0.20 0.64 −0.03 −0.01 0.14 −0.46 0.08 −0.06 
−2.16 −0.29 1.85 0.29 0.08 0.22 −1.35 0.07 0.24 
−7.22 −7.61 14.75 −0.04 −0.60 0.68 −0.15 −0.16 0.24 
−0.84 −0.41 1.20 −0.01 −0.07 0.17 −0.33 −0.06 0.09 
−1.80 −0.16 3.22 −0.10 −0.12 0.52 −0.80 0.02 0.39 
10 −0.60 −0.38 0.85 −0.01 0.10 0.00 −0.03 −0.23 0.15 
11 −2.91 −0.53 3.19 −0.04 −0.15 0.50 −0.66 −0.20 0.59 
12 −8.09 −4.75 12.86 −0.04 −0.27 0.69 −0.83 −0.39 0.58 
13 −5.85 −3.20 9.19 −0.02 −0.15 0.26 −0.20 −0.02 0.40 
14 −0.03 0.02 −0.02 0.01 0.00 0.01 0.00 0.01 −0.03 
15 −12.79 0.03 20.08 −1.42 9.25 −8.01 −0.28 −12.42 15.25 
16 −0.36 −0.01 0.18 0.01 −0.02 0.09 0.82 0.33 −0.24 
17 −0.23 0.17 0.10 −0.09 0.02 0.03 0.00 0.01 0.01 
18 −0.25 −0.04 0.31 −0.04 −0.01 0.03 0.00 0.01 −0.01 
19 −0.35 −0.21 0.54 −0.09 −0.02 0.06 0.17 −0.04 0.06 
20 −0.01 −0.03 0.21 −0.03 0.00 0.02 −0.01 −0.01 0.04 
21 −1.48 −0.25 1.86 −0.25 0.13 −0.10 0.20 −0.22 0.31 
  2002–2005 2005–2007 2007–2010 
Sector number Water efficiency effect Structure effect Demand effect Water use efficiency effect Structure effect Demand effect Water efficiency effect Structure effect Demand effect 
1.32 −1.45 −2.07 −1.02 −0.81 2.31 0.38 1.53 1.71 
0.00 0.00 0.00 0.02 0.00 0.00 −0.02 0.00 0.00 
0.11 0.03 0.07 −0.27 −0.05 0.08 −0.06 −0.03 0.04 
−0.02 0.02 −0.02 −0.08 0.02 0.01 −0.04 0.00 0.01 
−0.03 −0.06 0.06 0.09 0.03 0.01 −0.15 −0.02 0.07 
−0.15 −0.12 0.27 0.09 0.06 0.13 −0.52 −0.13 0.22 
−0.03 −0.09 0.12 −0.06 0.06 −0.01 −0.08 −0.04 0.09 
−0.02 −0.02 0.04 0.29 0.01 0.07 −0.32 −0.02 0.13 
−0.03 0.10 0.28 −0.30 −0.11 0.62 0.34 −0.02 0.46 
10 −0.08 −0.20 0.22 −0.06 0.07 0.02 −0.21 −0.03 0.14 
11 0.00 −0.02 0.54 −0.73 0.14 0.39 −0.76 −0.08 0.23 
12 −0.35 −0.54 1.18 −0.85 0.04 0.39 −0.14 −0.13 0.68 
13 −0.50 −0.28 1.06 −0.46 −0.18 0.45 −0.49 −0.03 0.29 
14 0.00 0.00 −0.03 −0.03 −0.03 0.08 −0.01 0.03 −0.04 
15 −1.42 4.06 3.03 −0.80 −3.72 6.46 −8.89 −1.42 7.67 
16 −0.04 0.02 −0.93 −0.18 0.02 0.02 −0.55 −0.15 0.62 
17 0.00 0.04 0.00 0.00 0.00 0.02 −0.10 0.08 0.02 
18 −0.02 −0.01 0.05 0.02 0.00 0.03 −0.08 −0.01 0.07 
19 −0.30 −0.09 0.16 0.01 0.00 0.01 0.00 0.00 0.05 
20 0.01 0.02 −0.01 0.11 0.00 0.03 −0.01 −0.04 0.08 
21 −0.60 −0.01 0.31 −0.08 −0.03 0.27 −0.16 −0.04 0.38 
  1997–2010
 
1997–2000
 
2000–2002
 
Sector number Water efficiency effect Structure effect Demand effect Water use efficiency effect Structure effect Demand effect Water efficiency effect Structure effect Demand effect 
−0.81 −2.86 14.98 −0.02 −2.06 8.71 −1.47 0.19 4.06 
−0.02 0.01 0.00 −0.01 0.00 0.01 −0.03 0.00 0.00 
−0.47 −0.16 0.51 −0.06 −0.01 0.10 −0.06 −0.01 0.00 
−0.01 0.02 −0.02 −0.01 0.00 0.03 0.27 0.02 −0.22 
−0.79 −0.20 0.64 −0.03 −0.01 0.14 −0.46 0.08 −0.06 
−2.16 −0.29 1.85 0.29 0.08 0.22 −1.35 0.07 0.24 
−7.22 −7.61 14.75 −0.04 −0.60 0.68 −0.15 −0.16 0.24 
−0.84 −0.41 1.20 −0.01 −0.07 0.17 −0.33 −0.06 0.09 
−1.80 −0.16 3.22 −0.10 −0.12 0.52 −0.80 0.02 0.39 
10 −0.60 −0.38 0.85 −0.01 0.10 0.00 −0.03 −0.23 0.15 
11 −2.91 −0.53 3.19 −0.04 −0.15 0.50 −0.66 −0.20 0.59 
12 −8.09 −4.75 12.86 −0.04 −0.27 0.69 −0.83 −0.39 0.58 
13 −5.85 −3.20 9.19 −0.02 −0.15 0.26 −0.20 −0.02 0.40 
14 −0.03 0.02 −0.02 0.01 0.00 0.01 0.00 0.01 −0.03 
15 −12.79 0.03 20.08 −1.42 9.25 −8.01 −0.28 −12.42 15.25 
16 −0.36 −0.01 0.18 0.01 −0.02 0.09 0.82 0.33 −0.24 
17 −0.23 0.17 0.10 −0.09 0.02 0.03 0.00 0.01 0.01 
18 −0.25 −0.04 0.31 −0.04 −0.01 0.03 0.00 0.01 −0.01 
19 −0.35 −0.21 0.54 −0.09 −0.02 0.06 0.17 −0.04 0.06 
20 −0.01 −0.03 0.21 −0.03 0.00 0.02 −0.01 −0.01 0.04 
21 −1.48 −0.25 1.86 −0.25 0.13 −0.10 0.20 −0.22 0.31 
  2002–2005 2005–2007 2007–2010 
Sector number Water efficiency effect Structure effect Demand effect Water use efficiency effect Structure effect Demand effect Water efficiency effect Structure effect Demand effect 
1.32 −1.45 −2.07 −1.02 −0.81 2.31 0.38 1.53 1.71 
0.00 0.00 0.00 0.02 0.00 0.00 −0.02 0.00 0.00 
0.11 0.03 0.07 −0.27 −0.05 0.08 −0.06 −0.03 0.04 
−0.02 0.02 −0.02 −0.08 0.02 0.01 −0.04 0.00 0.01 
−0.03 −0.06 0.06 0.09 0.03 0.01 −0.15 −0.02 0.07 
−0.15 −0.12 0.27 0.09 0.06 0.13 −0.52 −0.13 0.22 
−0.03 −0.09 0.12 −0.06 0.06 −0.01 −0.08 −0.04 0.09 
−0.02 −0.02 0.04 0.29 0.01 0.07 −0.32 −0.02 0.13 
−0.03 0.10 0.28 −0.30 −0.11 0.62 0.34 −0.02 0.46 
10 −0.08 −0.20 0.22 −0.06 0.07 0.02 −0.21 −0.03 0.14 
11 0.00 −0.02 0.54 −0.73 0.14 0.39 −0.76 −0.08 0.23 
12 −0.35 −0.54 1.18 −0.85 0.04 0.39 −0.14 −0.13 0.68 
13 −0.50 −0.28 1.06 −0.46 −0.18 0.45 −0.49 −0.03 0.29 
14 0.00 0.00 −0.03 −0.03 −0.03 0.08 −0.01 0.03 −0.04 
15 −1.42 4.06 3.03 −0.80 −3.72 6.46 −8.89 −1.42 7.67 
16 −0.04 0.02 −0.93 −0.18 0.02 0.02 −0.55 −0.15 0.62 
17 0.00 0.04 0.00 0.00 0.00 0.02 −0.10 0.08 0.02 
18 −0.02 −0.01 0.05 0.02 0.00 0.03 −0.08 −0.01 0.07 
19 −0.30 −0.09 0.16 0.01 0.00 0.01 0.00 0.00 0.05 
20 0.01 0.02 −0.01 0.11 0.00 0.03 −0.01 −0.04 0.08 
21 −0.60 −0.01 0.31 −0.08 −0.03 0.27 −0.16 −0.04 0.38 

To identify which of the three driving factors (water use efficiency, production structure and demand) have strong, medium or weak effects on those individual sectors in terms of their water use during 1997–2010, we used the fuzzy C means clustering method (Fang, 1982; Yang et al., 2006) to cluster the 21 production sectors to the categories of ‘strong,’ ‘medium’ and ‘weak’ as shown in Table 3.

Table 3.

Driving cluster of three effects on production sectors.

Effect Drive intensity Sectors 
Structure Strong (1) 
Medium (2)–(14), (16)–(21) 
Weak (15) 
Water use efficiency Strong (1), (6), (11), (12), (15) 
Medium (4), (9), (13), (17), (21) 
Weak (2), (3), (5), (7), (8), (10), (14), (16), (18)–(20) 
Demand Strong (15) 
Medium (1), (7), (9), (11)–(13), (21) 
Weak (2)–(6), (8), (10), (14), (16)–(20) 
Effect Drive intensity Sectors 
Structure Strong (1) 
Medium (2)–(14), (16)–(21) 
Weak (15) 
Water use efficiency Strong (1), (6), (11), (12), (15) 
Medium (4), (9), (13), (17), (21) 
Weak (2), (3), (5), (7), (8), (10), (14), (16), (18)–(20) 
Demand Strong (15) 
Medium (1), (7), (9), (11)–(13), (21) 
Weak (2)–(6), (8), (10), (14), (16)–(20) 

Notes: (1) agriculture, (2) coal mining and processing, (3) other mining, (4) oil and gas, (5) food, (6) textile, (7) forest, (8) paper, (9) chemical, (10) building materials, (11) metallurgical, (12) machinery and equipment, (13) electronic instrument, (14) other manufacturing, (15) power, (16) production and supply of tap water, (17) construction, (18) transportation and telecommunication, (19) accommodation and restaurants, (20) wholesale and retail trade, and (21) other services.

The agriculture sector is identified as being strongly effected by structure change. Recent improvements in agricultural science and technology have achieved a higher input-output ratio, thereby reducing the demand for agricultural water. Meanwhile the power industry is weakly effected by structure change. The other 19 sectors belong to the category of which structure change has medium effect on the water use of those sectors.

Water use efficiency is found to have a strong effect on water saving in the agriculture, textile, metallurgical, machinery and equipment, and power industry sectors. Oil and gas, chemical industry, electronic instrument, construction, and other services are classified as sectors with medium effect from the improvement of water saving technology. The rest of the sectors are identified as ones that are weakly effected by water use efficiency. These sectors may have larger potential for saving water through water use efficiency improvement than other sectors. Power is the only sector identified that was strongly effected by water demand. The rapid economic growth in China prompted a corresponding growth in the energy industry, which led to a significant increase in water demand from the sector. The water demands of agriculture, forest, chemical, metallurgical, machinery, electronic instruments, and other services industries have a medium effect on the increased water use of those sectors. The demands of the rest of the sectors have a relatively weak impact on overall water use.

Conclusions

This paper analyzed the factors that most influence the change in water use in production sectors of Jiangsu Province, China. The analysis is based on input and output (I-O) sequence tables, which consider water use at a constant price. The structure decomposition method of I-O analysis has been used.

The effects of demand, structure and technology, and water use efficiency vary considerably over time and by sector. From 1997 to 2010, the average annual water use grew by 1.41 BCM. The change in water demand was responsible for an increase in water consumption by 6.64 BCM, while an improvement in water use efficiency and economic structure change decreased water consumption by 4.65 BCM and 0.55 BCM, respectively (refer to Appendix Figure A1, available with the online version of this paper). Thus the economic development in Jiangsu Province has increased water use more than all water savings realized from technological advances.

From 1997 to 2010, the average annual water use growth of the primary, secondary and tertiary sectors were 0.87 BCM, 0.54 BCM, and 0.02 BCM, accounting for 60.91%, 37.41%, and 1.68% of overall water use growth, respectively. The water use changes caused by the demand, water use efficiency and structure effects are, respectively, 4.34, 15.33 and 80.33% for the primary sectors, 40.88, 4.98 and 54.14% for the secondary sectors, and 37.20, 10.07 and 52.73% for the tertiary sectors (see details in Figure A2 of the Appendix, available with the online version of this paper). The increase in demand is the main factor driving water use growth for all the sectors, while water use efficiency has a larger role in reducing water use than economic structure change.

Furthermore, the effect of water use efficiency within the primary sectors is relatively small. Agriculture is the major sector in this category. Rice and corn are the major crops in the Province. Irrigation water use efficiency is low due to out-of-date irrigation systems which have not been maintained reasonably. Thus investment in irrigation systems will have large potential to reduce the agricultural water requirement. Meanwhile, for the secondary and tertiary sectors, economic structure regulation and production technology improvement have cut off water use growth during the past decade.

At the province level, structure change is a medium driver of water saving for all sectors except for agriculture and power. Nearly half of the 21 sectors have experienced medium or large water savings through the improvement of water use efficiency. Demand effect is a strong driving force for the power industry only.

The findings on water use effects provide considerable policy implications for water demand management in the case study region. Given the identified trend of decreasing demand effect (increasing water use) and increasing effect of production structure adjustment and water use efficiency improvement (reducing water use), the water demand management measures are effective and should be continued in Jiangsu Province. However, the effects of the three driving factors vary considerably by sector (and over time). How to reach the goal of sustainable water use will still challenge future water resources management research and practices. For example, the growing water demands required for economic growth are offset by economic structure change, technology and water use efficiency improvement, especially for the primary sectors and some secondary sectors (e.g., power industry). In addition, due to the input-output tables and water use data limitations, we used the same path to describe the changes of the three factor effects driving water use change. Measurement of parameters of different factor effects driving water use change, based on additional economic and water use data, is suggested as an area for further research.

Acknowledgements

Support for this research was provided by a grant from the National Natural Science Foundation of China (No. 51109055, 51279223, 51479119), Public-services Foundation of Ministry Water Resources of China (No. 201301003, 201201022), Social Science Foundation from Jiangsu Province (No. 11GLA001), and The National Soft Science Research Program (2014GXS4B047).

References

References
Boyd
G. A.
Roop
J. M.
(
2004
).
A note on the fisher ideal index decomposition for structural change in energy intensity
.
Energy
25
(
1
),
87
101
.
Central Committee, State Council (CPC)
(
2011
).
CPC Central Committee and State Council on accelerating the development of water conservancy reform decision
.
China Water Diary
2
,
1
.
China Sustainable Development Strategy Study Group
(
2007
).
China sustainable development strategy report 2007 – Water: Governance and Innovation
.
Science Press
,
Beijing
.
Fang
K. T.
(
1982
).
Cluster Analysis
.
Geology Press
,
Beijing
.
Gao
E. K.
Dang
L. W.
(
2008
).
Empirical Summary of Water Right Institution Construction
.
China Water Press
,
Beijing
.
Haan
M. D.
(
2001
).
A structural decomposition analysis of pollution in the Netherlands
.
Economic Systems Research
13
(
2
),
181
196
.
Harrison
W. J.
Horridge
J. M.
Pearson
K. R.
(
2000
).
Decomposing simulation results with respect to exogenous shocks
.
Computational Economics
15
(
3
),
227
249
.
Hoekstra
R.
Bergh
J.
(
2002
).
Structural decomposition analysis of physical flows in the economy
.
Environmental and Resource Economics
23
(
3
),
357
378
.
Hoekstra
R.
Bergh
J.
(
2003
).
Comparing structural decomposition analysis and index
.
Energy Economics
25
,
39
64
.
Li
J. H.
(
2004
).
Economic Structure Analysis and Application
.
China Science Academy
,
Beijing
.
Liu
Q. Y.
Peng
Z. L.
(
2010
).
1992–2005 Input-Output Series Tables with Comparable Prices and Analysis in China
.
China Statistic Press
,
Beijing
.
McKay
J.
(
2005
).
Water institutional reforms in Australia
.
Water Policy
7
(
1
),
35
52
.
Miller
R. E.
Shao
G.
(
1994
).
Structural change in the US multiregional economy
.
Structural Change and Economic Dynamics
5
(
1
),
41
72
.
Ministry of Water Resources of China
(
2013
).
China Water Resources Bulletin
.
China Water Power Press
,
Beijing
.
Sun
C. Z.
Wang
Y.
(
2009
).
Driving force measurement of water utilization change in Liaoning Province and analysis of their spatial temporal difference based on factor decomposition model
.
Arid Land Geography
32
(
6
),
850
858
.
Wang
Z.
(
2004
).
Policy Simulation of Environment and Economics Safety in China
.
Science Press
,
Beijing
.
Yang
H.
Zhong
B.
Liu
Q. Y.
(
2006
).
Application of Mathematical Statistics
.
Tsinghua University Press
,
Beijing
.
Zhang
Q.
Wang
B. D.
Cao
M. L.
(
2011
).
Analysis of water resource utilization change based on factor decomposition model
.
Journal of Natural Resources
26
(
7
),
1209
1216
.

Supplementary data