Introduction

Water is an essential resource for human existence; however, water use, especially in developing countries, has exceeded sustainable levels; a growing scarcity of fresh water relative to human demands is now evident in many parts of the world (Postel 2000; Guo et al. 2016). Alarm bells are already ringing as climate change, economic and population growth, more and more unpredictable weather, and water patterns have brought serious challenge to the global water resource security. Much of the world is already feeling the strain on water resources, with the United Nations predicting that by 2025, 1.8 billion people around the world be suffering from the water shortages.

China’s economy is developing rapidly, but it still has not got rid of the development model of wasting resources and sacrificing the environment for economic growth. The average annual growth rate of water use has hit 0.63% (the calculation interval is 1979–2017 due to missing data). Currently, the situation of water resources in China is very serious. China’s per capita water resource is 28% of the world’s average. Moreover, China lacks more than 50 billion m3 of water resources each year, and more than 400 of the more than 600 cities are “not enough to drink.” Water shortage, serious water pollution, and water ecological deterioration have become increasingly serious problems, which have become the major bottleneck restricting the sustainable development of social economy. Jiangsu is located in the Southeast Coastal area of China. Figure 1 shows its geographical location in China. During the period from 2000 to 2016, Jiangsu’s annual regional GDP growth was 11.73% based on the constant price of 2000; at the same time, the total water use increased by 13.18 billion m3 from 44.56 billion m3 in 2000 to 57.74 billion m3 in 2016, with an average annual growth rate of 1.63%; the proportion of China’s total water use increased from 8.10 to 9.56%, indicating that the economic growth of Jiangsu is still highly dependent on water resources. The coupled increasing relationship between GDP and water use clearly indicates that such a development model is not sustainable and may result in many issues, such as wastewater pollution and water ecological environment deteriorates.

Fig. 1
figure 1

The geographical location of Jiangsu in China

Full size image

Decomposition analysis is a method of decomposing an aggregate indicator into several predefined and easily understood factors. Specifically, there are mainly two kinds of decomposition analysis methods. They are structural decomposition analysis (SDA) and index decomposition analysis (IDA). Compared with SDA, IDA has been widely applied in many research areas. IDA is easier to apply for its simplicity and flexibility (Ang 2004; Ang and Zhang 2000). In this sense, we will apply IDA to decompose the driving factors of changes on water use in Jiangsu China. Recently, IDA is widely used in water resource research. The study on the use of IDA to decompose the change of water resource consumption indicators in China and its regions is shown in Table 1. From existing research, the research areas mainly include China’s overall, provinces, cities, and basins; decomposition indicators are mainly agricultural water use indicator and industrial water use indicator; and the decomposition methods are all LMDI method which has been proved to be the best decomposition method in IDA. Crop irrigation efficiency, planting structure, and planting scale are common drivers of change in agricultural water use indicator. Other factors are also considered; Zhao and Chen (2014) found that the diet structure has a weak promoting effect on the growth of China’s agricultural water footprint. According to Zou et al. (2018), climate change is only second to planting scale in promoting irrigation water consumption in Heihe River basin. Zhang et al. (2015) studied the influence of water resource endowment and water resource development level on the water efficiency of Urumqi. The change of industrial water indicator is decomposed into economic output, industrial structure, and industrial water intensity.

Table 1 Representative literature on the IDA method
Full size table

Summarizing the previous literature (see Table 1), we find that although several researchers have used the LMDI method to investigate the driving forces of water use or water intensity from China and its regions, some omissions still exist. More specifically, the authors pay attention to industrial water use without considering the driving effect of change from the two dimensions of production water use and domestic water use. As an important province in the Yangtze River Delta region, the control of its water use will play an important role in the construction of water-saving society in China. This paper, for the first time, takes Jiangsu as the research object and decomposes the driving effect of its production water use and domestic water use changes.

Methodology and data

Decomposition model

This paper adopts LMDI method to decompose the driving factors of water use change.

Total water use, TW, can be formulated as follows:

$$ \mathrm{TW}=\mathrm{PW}+\mathrm{DW}=\sum \limits_i{\mathrm{PW}}_i+\mathrm{DW}=\sum \limits_i\frac{{\mathrm{PW}}_i}{G_i}\times \frac{{\mathrm{G}}_{\mathrm{i}}}{G}\times G+\frac{\mathrm{DW}}{P}\times P $$
(1)

Here, PW is the production water use, DW is the domestic water use, PWi is the production water use of the ith industry, Gi is the added value of the ith industry, and G satisfy the equation\( G=\sum \limits_i{G}_i\left(i=1,2\right) \); this paper covers only agriculture and industry, and P is the population scale.

Equation (1) can be reconstructed as:

$$ \mathrm{TW}=\sum \limits_i{\mathrm{PI}}_i\times {S}_i\times G+\mathrm{DI}\times P $$
(2)

Here, PIi = PWi/Gi is the ith industry water use intensity, Si = Gi/G is the industrial structure, and DI = DW/I is the domestic water use intensity.

We assume the year goes from t − 1 to t, the change of total water use ΔTWt − 1, t can be decomposed into production effect and domestic effect, and the production effect consists of production intensity effect, industrial structure effect, and economic development effect; domestic effect consists domestic intensity effect and population effect and is given as:

$$ \varDelta {TW}^{t-1,t}=\left(\varDelta {W_{PI}}^{t-1,t}+\varDelta {W_S}^{t-1,t}+\varDelta {W_G}^{t-1,t}\right)+\left(\varDelta {W_{\mathrm{DI}}}^{t-1,t}+\varDelta {W_P}^{t-1,t}\right) $$
(3)

here

$$ \varDelta {W_{\mathrm{PI}}}^{t-1,t}=\sum \limits_i\frac{{\mathrm{PW}}_i^t-{\mathrm{PW}}_i^{t-1}}{1{\mathrm{nPW}}_i^t-1{\mathrm{nPW}}_i^{t-1}}1\mathrm{n}\left(\frac{{\mathrm{PI}}_i^t}{{\mathrm{PI}}_i^{t-1}}\right) $$
(4)
$$ \varDelta {W_S}^{t-1,t}=\sum \limits_i\frac{{\mathrm{PW}}_{\mathrm{i}}^{\mathrm{t}}-{\mathrm{PW}}_i^{t-1}}{1{\mathrm{nPW}}_{\mathrm{i}}^{\mathrm{t}}-1{\mathrm{nPW}}_i^{t-1}}1\mathrm{n}\left(\frac{S_i^t}{S_i^{t-1}}\right) $$
(5)
$$ \varDelta {W_G}^{t-1,t}=\sum \limits_i\frac{{\mathrm{PW}}_i^t-{\mathrm{PW}}_i^{t-1}}{1{\mathrm{nPW}}_{\mathrm{i}}^{\mathrm{t}}-1{\mathrm{nPW}}_i^{t-1}}1\mathrm{n}\left(\frac{G^t}{G^{t-1}}\right) $$
(6)
$$ \varDelta {W_{\mathrm{DI}}}^{t-1,t}=\sum \limits_i\frac{{\mathrm{DW}}^t-{\mathrm{DW}}^{t-1}}{1{\mathrm{nDW}}^t-1{\mathrm{nDW}}^{t-1}}1\mathrm{n}\left(\frac{{\mathrm{DI}}^t}{{\mathrm{DI}}^{t-1}}\right) $$
(7)
$$ \varDelta {W_P}^{t-1,t}=\sum \limits_i\frac{{\mathrm{DW}}^t-{\mathrm{DW}}_i^{t-1}}{1{\mathrm{nDW}}^t-1{\mathrm{nDW}}_i^{t-1}}1\mathrm{n}\left(\frac{P^t}{P^{t-1}}\right) $$
(8)

Here, ΔWPIt − 1, t, ΔWSt − 1, t, ΔWGt − 1, t, ΔWDIt − 1, t, and ΔWPt − 1, t are defined as production intensity effect, industrial structure effect, economic development effect, domestic intensity effect, and population scale effect, which reflects the impact of change of production water intensity, industrial structure, economic growth, per capita domestic water use, and population scale on total water use, respectively.

Since Eq. (1) does not reflect the contribution of per capita domestic water use difference between urban and rural area to the change of domestic water use, Eq. (1) can be reconstructed as shown in Eq. (9).

$$ \mathrm{DW}=\sum \limits_i{\mathrm{DW}}_i=\sum \limits_i\frac{\mathrm{DW}}{P_i}\times \frac{P_i}{P}\times P=\sum \limits_i{\mathrm{DI}}_i\times {\mathrm{SP}}_i\times P $$
(9)

Here, DWi(i = 1, 2) indicates the urban resident domestic water use and rural resident domestic water use, respectively. Pi(i = 1, 2) indicates the population scale of urban and rural, respectively. DIi = DWi/Pi(i = 1, 2) indicates the per capita domestic water use intensity of urban and rural resident. SPi = Pi/P(i = 1, 2) indicates the proportion of urban and rural population in the total population, respectively.

Assuming that period change from t − 1 to t, the change of domestic water use can be decomposed into domestic intensity effect, urbanization effect, and population scale effect, as shown in Eq. (10).

$$ \varDelta {\mathrm{DW}}^{t-1,t}=\varDelta {W_{\mathrm{DI}}}^{t-1,t}+\varDelta {W_{\mathrm{SP}}}^{t-1,t}+\varDelta {W_P}^{t-1,t} $$
(10)

Here,

$$ \varDelta {W_{\mathrm{DI}}}^{t-1,t}=\sum \limits_i\frac{{\mathrm{DW}}_i^t-{\mathrm{DW}}_i^{t-1}}{1{\mathrm{nDW}}_i^t-1{\mathrm{nDW}}_i^{t-1}}1\mathrm{n}\left(\frac{{\mathrm{DI}}_i^t}{{\mathrm{DI}}_i^{t-1}}\right) $$
(11)
$$ \varDelta {W_{SP}}^{t-1,t}=\sum \limits_i\frac{{\mathrm{DW}}_i^t-{\mathrm{DW}}_i^{t-1}}{1{\mathrm{nDW}}_i^t-1{\mathrm{nDW}}_i^{t-1}}1\mathrm{n}\left(\frac{{\mathrm{SP}}_i^t}{{\mathrm{SP}}_i^{t-1}}\right) $$
(12)
$$ \varDelta {W_P}^{t-1,t}=\sum \limits_i\frac{{\mathrm{DW}}_i^t-{\mathrm{DW}}_i^{t-1}}{1{\mathrm{nDW}}_i^t-1{\mathrm{nDW}}_i^{t-1}}1\mathrm{n}\left(\frac{P^t}{P^{t-1}}\right) $$
(13)

\( \varDelta {W_{DI}}^{t-1,t} \) shows the domestic intensity effect, \( \varDelta {W_{SP}}^{t-1,t} \) shows the urbanization effect, and ΔWPt − 1, t shows the population scale effect, which reflects the impact of per capita domestic water use, urbanization level, and population scale on domestic water use.

Figure 2 shows the process of using LMDI to decompose the driving effects of total water use and domestic water use changes.

Fig. 2
figure 2

The driving effect decomposition process of water use change

Full size image

Data and variable

The research period in the current study starts in 2000 and ends in 2016. The data are all collected from China Statistical Yearbook, China Water Bulletin. The description about the indicators is shown as follows.

  1. (1)

    Water use index. China Water Bulletin divides China’s overall water use into domestic water use, production water use, and ecological and environmental water use. Domestic water use covers urban household water use and rural household water use, and production water use covers primary industry water use, secondary industry water use, and tertiary industry water use. However, China Water Bulletin does not publish three types of water use data for each province and divides total water use into agricultural water use, industrial water use, domestic water use, and ecological and environmental water use. To achieve the research goal of this paper, it is necessary to adjust the pro-categorized data of water use. We merge agricultural water use and industrial water use into production water use which do not cover tertiary water use, because the tertiary water use cannot be calculated separately from the domestic water use. Table 2 shows the composition of water use in China and Jiangsu.

Table 2 Composition of total water use in China
Full size table
  1. (2)

    Output index. The agricultural added value and industrial added value of Jiangsu province are all adjusted based on the constant prices in 2000 to exclude the effect of price factors.

Results

Descriptive statistical analysis

Figure 3 shows the change trend of total water use, production water use, and domestic water use from 2000 to 2016 in Jiangsu province. The total water use increased from 445.6 × 108 m3 in 2000 to 575.4 × 108 m3 in 2016, an increase of 129.8 × 108 m3, with an average annual growth rate of 1.61%. Within this period, there was a considerable decline in 2003. The result was mainly attributed by insufficient water demand arising from insufficient water supply and the development of water-saving society. The production water use increased by 115.47 × 108 m3 from 403.83 × 108 m3 in 2000 to 519.30 × 108 m3 in 2016, with an average annual growth rate of 1.58%, stably accounting for more than 90% of total water use. The change trend of production water use was very similar to that of total water use for the correlation coefficient which reached 0.9984. Also, 2003 witnesses a significant decline in production water use. Domestic water use increased by 14.33 × 108 m3 from 41.77 × 108 m3 in 2000 to 56.10 × 108 m3 in 2016, with an average annual growth rate of 1.86%, accounting for less than 10% of total water use. It can be seen that the change of total water use mainly comes from the change of production water use.

Fig. 3
figure 3

The trend of total water use, production water use, and domestic water use from 2000 to 2016 in Jiangsu

Full size image

Driving effect decomposition of total water use

During the period of 2000–2016, the total water use in Jiangsu decreased significantly in 2003 and increased significantly in 2014. Therefore, the overall time period 2000–2016 is divided into 2000–2003, 2003–2014, and 2014–2016 three sub-time periods. According to Eqs. (3)–(8), the driving effect decomposition results of the total water use change are calculated, as shown in Table 3. Among the five driving factors examined, the production intensity effect ΔWPI and industrial structure effect ΔWS have always been promoted reduction in total water use, and the economic development effect ΔWG and the population scale effect ΔWP have been maintained to promote increase, while the domestic intensity effect ΔWDI has both effect on the promoting reduction and increase. Apparently, it can be seen that the improvement of production water use efficiency and the upgrading of industrial structure have effectively promoted the decline in total water use, especially the former, while economic growth and population scale increase have contributed to the increase in total water use, especially economic growth; except for the 2000–2003 period, in the two periods of 2000–2014 and 2014–2016, the domestic water use intensity has increased, which has promoted the increase in the total water use, reflecting the increase in the living standards of the people and the increase in the demand for domestic water.

Table 3 Driving effect decomposition results of the total water use change in Jiangsu (108 m3)
Full size table

Among the effects of promoting the reduction, the production intensity effect ΔWPI is the dominant factor that promotes the decline in total water use. In the three sub-periods, the total water use decreased by 98.53 × 108 m3, 227.05 × 108 m3, and 62.59 × 108 m3, respectively, that is, the average decreased by 32.84 × 108 m3, 20.64 × 108 m3, and 31.30 × 108 m3. The reason for the increase of reduction effect in 2014–2016 may be that at the end of the 12th 5-year plan period, the assessment requirements of the most stringent water resource management system need to be completed, so the water-saving efforts are intensified. The industrial structure effect ΔWS on the decline in total water use is weaker than the production intensity effect ΔWPI, resulting in a decrease in total water use of 58.5 × 108 m3, 199.24 × 108 m3, and 25.99 × 108 m3 respectively. The transfer of agriculture to non-agricultural industries will effectively promote the decline in total water use. The main reason is that agricultural water use efficiency is far lower than industry. Among the effects of promoting the increase, the economic development effect ΔWG is the dominant factor in promoting the increase in total water use, resulting in an increase in total water use of 132.1 × 108 m3, 583.13 × 108 m3, and 72.14 × 108 m3, respectively. During the period 2003–2014, the economic development effect ΔWG on increase is greater than the production intensity effect ΔWPI and the industrial structure effect ΔWS on reduction, resulting in an increase in production water use of 156.84 × 108 m3, while in 2000–2003 and 2014–2016, the production intensity effect ΔWPI and the industrial structure effect ΔWS on reduction have completely offset the increase from the economic development effect, leading to a decline in production water use. The domestic intensity effect ΔWDI is a secondary factor driving the increase in total water use (except 2000–2003); in 2003–2014 and 2014–2016, the total water use increased by 9.93 × 108 m3 and 3.00 × 108 m3, respectively, resulting from the domestic intensity effect ΔWDI. The population scale effect ΔWP on the increase in total water use is relatively weak, resulting in an increase in total water use of 0.72 × 108 m3, 3.00 × 108 m3, and 0.27 × 108 m3, respectively. In the period of 2000–2003 and 2003–2016, the trend of production water use is similar to the inverted U-shaped curve, with the highest points in 2002 and 2014, respectively.

The year 2000 is selected as the base period to investigate the change trend of driving factors of total water use from 2000 to 2016. The value of cumulative effect of each factor is calculated by accumulating the contribution of each factor to the total water use evolution year by year, as shown in Table 4. During the period 2000–2016, the total water use increased by 129.8 × 108 m3, of which, in 2000–2003, it decreased by 26.8 × 108 m3 and reached a maximum of 478.74 × 108 m3 in 2002. The trend is similar to the inverted U-shaped curve. In 2004–2016, the cumulative increase was 63.68 × 108 m3 (129.80–66.12), with an average annual growth rate of 5.78%. The total water use reached a maximum of 588.57 × 108 m3 in 2014. The trend is also similar to the inverted U-shaped curve. Compared with 2015, it has increased by 2.90 × 108m3 (129.80–126.90) in 2016.

Table 4 Cumulative contribution of the driving effect of water use change in Jiangsu from 2000 to 2016 (108 m3)
Full size table

From the impact of production effect and domestic effect on total water use, production effect is the dominant factor for the increase of total water use, resulting in an increase in water use from 20.1 × 108 m3 in 2001 to 115.47 × 108 m3 in 2016. In 2000–2003, the cumulative production effect dropped by 24.93 × 108 m3, but increased by 48.15 × 108 m3 (115.47–67.32) in 2004–2016, with an average annual growth rate of 4.60%, reaching the highest point in 2014, which also caused the highest point of total water use. The domestic effect is a secondary factor in the increase in total water use, resulting in an increase in water use from 0.68 × 108 m3 in 2001 to 14.33 × 108 m3 in 2016. Among them, the cumulative domestic effect in 2000–2004 has dropped by 1.20 × 108 m3, but increased by 13.05 × 108 m3 (14.33–1.28) in 2005–2016, with an average annual growth rate of 24.56%. It can be seen that the total water use control in Jiangsu should be mainly from the perspective of production water, including improving the farmland irrigation water use coefficient and industrial water reuse rate. Although the domestic effect on the total water use change is weak, with the improvement of people’s living standard, the demand for water resources will gradually increase, and the savings in domestic water use will also play an important role in controlling the total amount of water use.

From the perspective of all driving effects, the economic development effect ΔWG is the dominant factor to promote the increase in total water use, leading to total water use increase from 38.23 × 108 m3 in 2001 to 805.14 × 108 m3 in 2016, with an average annual growth rate of 22.53%. As one of the important provinces in Yangtze River Delta, the rapid economic growth of Jiangsu has increased the huge demand for water in agriculture and industry. The population scale effect ΔWP on the promotion of total water use is relatively weak, resulting in an increase in total water use from 0.18 × 108 m3 in 2001 to 4.26 × 108 m3 in 2016, with an average annual growth rate of 23.48%. With the full liberalization of the second child policy, the population scale will continue to increase to a certain extent, and the role of promoting the total water use will continue for a long time. The domestic intensity effect ΔWDI was negative in 2003, 2004, and 2005, but it had a cumulative increase of 7.62 × 108 m3 in 2006–2016. The effect of promoting the total water use is stronger than the population scale effect, indicating that the urban and rural household water efficiency needs to be further improved. The production intensity effect ΔWPI is the primary factor to promote the decline in total water use, resulting in a total decrease of 396.49 × 108 m3, with an average annual decline rate of 37.20%. The industrial structure effect ΔWS is a secondary factor that promotes the decline in total water use, resulting in a total decrease of 293.17 × 108 m3, with an average annual decline rate of 22.09%. The promoting reduction of production intensity ΔWPI exceeded the effect of industrial structure ΔWS in 2009, and the gap gradually widened. It can be seen that the factors that contribute to the reduction of total water use are production intensity effect ΔWPI and industrial structure effect ΔWS, while the factors that contribute to the increase in total water use are economic development effect ΔWG, domestic intensity effect ΔWPI, and population scale effect ΔWP.

There are differences between agriculture and industry in production intensity effect ΔWPI, industrial structure effect ΔWS, and economic development effect ΔWG, as shown in Table 4. The efficiency of industrial and agricultural water use has generally improved, which has effectively promoted the decline in total water use. Figure 4 shows the trend of industrial and agricultural water intensity. Among them, the industrial sector is the main source and the total water use decreased by 263.46 × 108 m3 in 2000–2016, accounting for 66.45% (293.46/396.49) of the production intensity effect ΔWPI, while the agricultural sector’s contribution to the decline in total water use was weaker than that of the industrial sector, resulting in a total reduction of water use of 133.03 × 108 m3, which is only half of the industrial sector, indicating that the agricultural sector has large water saving potential and space. The effect of industrial structure adjustment on the reduction in total water use mainly comes from the decline of the proportion of agriculture. Figure 5 shows the trend of the proportion of industrial and agricultural added value, the proportion of agriculture decreased from 21.41% in 2000 to 6.27% in 2016, the cumulative decline 15.14 percentage points, resulting in a total decline of water use of 326.74 × 108 m3, accounting for 114.45% of the industrial structure effect ΔWS, while the proportion of industry increased from 78.59% in 2000 to 93.73% in 2016, resulting in a cumulative increase in total water use by 33.57 × 108 m3. Figure 6 shows the trend of agricultural and industrial added value, agricultural and industrial economic growth have promoted the increase in total water use, the value added of agriculture increased from 1048.34 × 108 RMB in 2000 to 1790.36 × 108 RMB in 2016, with an average annual growth rate of only 3.40%, resulting in a cumulative increase in total water use of 469.15 × 108m3 and accounting for 58.27% (469.15/805.14) of the economic development effect ΔWG, while the industrial added value increased from 3848.52 × 108 RMB in 2000 to 26762.81 × 108RMB in 2016, with an average annual growth rate of 12.89%, resulting in a cumulative increase in total water use of 335.99 × 108m3, accounting for 41.73% (335.99/805.14) of the economic development effect ΔWG. The promoting effect of agriculture economic growth on total water use is stronger than that of industry, mainly because the agricultural water intensity is significantly higher than that of industry.

Fig. 4
figure 4

The trend of agricultural and industrial water intensity from 2000 to 2016 in Jiangsu

Full size image
Fig. 5
figure 5

The trend of proportion of agricultural and industrial added value from 2000 to 2016 in Jiangsu

Full size image
Fig. 6
figure 6

The trend of agricultural and industrial added value from 2000 to 2016 in Jiangsu

Full size image

Driving effect decomposition of domestic water use

Table 5 shows the driving effect decomposition results of domestic water use from 2014 to 2016 in Jiangsu.

Table 5 Driving effect decomposition of domestic water use from 2014 to 2016 in Jiangsu (108 m3)
Full size table

Period of 2014–2016

During 2014–2016, the domestic water use increased by 1.75 × 108 m3, among which the domestic intensity effect ΔWDI, urbanization effect \( \varDelta {W}_{\mathrm{SP}} \), and the population scale effect ΔWP were 1.26 × 108 m3, 0.31 × 108 m3, and 0.18 × 108 m3, accounting for 72.00%, 17.71%, and 10.29% of the total respectively. It can be seen that the three effects are the factors contributing to the increase in domestic water use. Domestic intensity effect ΔWDI is the dominant factor to promote the increase of domestic water use, namely the increase in per capita domestic water use. The per capita domestic water use of urban residents increased from 137 to 143 L/day, an increase of 7 L/day, leading to domestic water use increase by 1.16 × 108 m3, and rural residents per capita domestic water use increased from 97 to 98 L/day, an increase of 1 L/day, leading to domestic water use increase by 0.10 × 108 m3. It can be seen that the increase in the per capita water use of urban residents is the main reason for the increase of the domestic intensity effect ΔWDI. At present, the per capita domestic water use in urban and rural areas are very close to the upper limit of the residents living water ration in Jiangsu Province industry, service industry, and domestic water quota (revised in 2014). The urban and rural residents’ domestic water quota are 120–150 L/day and 80–100 L/day. Therefore, relevant departments need to formulate policies and measures to save water use in Jiangsu province.

The urbanization effect \( \varDelta {W}_{\mathrm{SP}} \) is a secondary factor that promotes the increase in domestic water use. The urbanization rate has increased from 65 to 68%, resulting in an increase of 1.02 × 108 m3 in domestic water use. On the contrary, the proportion of rural population has dropped from 35 to 32%, resulting in a decrease of 0.71 × 108 m3 in domestic water use. The increase of domestic water use caused by the transfer of population from rural to urban areas can be considered from the following aspects: the per capita domestic water use of urban residents is greater than that of rural areas, and the increase of domestic water use caused by the migration of rural population to urban areas is exactly the promoting effect of the domestic intensity effect. The urban residents’ domestic water metering facilities are more complete. However, there is a phenomenon of self-dug wells in rural areas. This part of water use has not been effectively measured yet, which indirectly reduces the water use of rural residents.

The population scale effect ΔWP has the weakest effect on the increase in domestic water use. The urban population increased from 51.91 million to 54.17 million with an increase of 2.26 million, resulting in an increase of 0.13 × 108 m3 in domestic water use. And a decline in the rural population from 27.69 million to 25.82 million people, with a decrease of 1.87 million, results in a drop in domestic water use of 0.05 × 108 m3. The expansion of population scale directly increases the demand for domestic water. With the liberalization of the two-child policy, the domestic water use will continue to increase even other conditions unchanged.

Periods 2014–2015 and 2015–2016

In the two periods of 2014–2015 and 2015–2016, promoting increase in domestic intensity effect ΔWDI on the domestic water use is stronger than the urbanization effect \( \varDelta {W}_{\mathrm{SP}} \) and population scale effect ΔWP. Compared with 2014–2015, the domestic intensity effect ΔWDI has declined in 2015–2016. The main reason is that the per capita water use of rural residents has not changed, which is 98 L/day, leading to a zero domestic intensity effect in rural areas, while the per capita water use of urban residents has still increased by 3 L/day, resulting in an increase of 0.02 (0.59–0.57) × 108 m3 in the domestic intensity effect in urban areas. The urbanization effect \( \varDelta {W}_{\mathrm{SP}} \) decreased from 0.16 × 108 m3 to 0.15 × 108 m3, because the urbanization rate increased by only 1 percentage point. And the urbanization effect \( \varDelta {W}_{\mathrm{SP}} \) is 0.04 (0.49–0.53) × 108 m3 less in the urban. The population scale effect ΔWP increased from 0.07 billion m3 to 0.11 billion m3. Both urban population and rural population increased, so the population scale effect increased by 0.03 (0.08–0.05) × 108 m3 and 0.01 (0.03–0.02) × 108 m3 respectively in urban and rural areas.

Conclusions and policy implications

In this paper, LMDI method is adopted to analyze the driving effect of water use change in Jiangsu province, and the following main findings are obtained:

  1. (1)

    The increase of total water use mainly comes from the increase of production water use. The increase in domestic water use is far weaker than the production water use.

  2. (2)

    The production intensity effect ΔWPI and the industrial structure effect ΔWS are the two dominating factors that induce the decline of total water use. Economic development effect ΔWG is the most important factor to promote the increase of total water use, and domestic intensity effect ΔWDI is the secondary factor except 2000–2003; population scale effect ΔWP on the total water use is relatively weak.

  3. (3)

    The improvement of water use efficiency in the agricultural sector is less effective than that of industrial sector in promoting the decline of total water use. The role of industrial structure adjustment in reducing the total water use is mainly due to the decline in the proportion of agriculture. The promotion of agricultural economic growth to total water use is stronger than that of industry.

  4. (4)

    The domestic intensity effect ΔWDI is the primary factor to promote the increase of domestic water use. Among them, the increase of per capita water use of urban residents is the main reason for the increase of domestic water use due to the domestic intensity effect; the urbanization effect \( \varDelta {W}_{\mathrm{SP}} \) is a secondary factor to promote the increase of domestic water use. The population scale effect ΔWP has the weakest effect on the increase of domestic water use.

Based on the above research conclusions, the following policy recommendations are proposed:

  1. (1)

    The water use control in Jiangsu should focus on industrial and agricultural production. However, with the improvement of people’s living standards, the demand for water use will gradually increase, and attention should be paid to water saving for urban and rural residents. It is recommended to implement water conservation management for residential water use and industrial and agricultural production water consumption. For residential water use, we should use water quota management. For industrial and agricultural production water use, water management should be carried out in combination with planning and quota. The management measures for refining the water consumption of residents should be incorporated into the Jiangsu Water Use Regulations and revised to enhance the mandatory and effective water-saving management.

  2. (2)

    Jiangsu should further improve the efficiency of industrial and agricultural water use, especially in the agricultural sector; promote the high standard farmland water conservancy construction; increase the proportion of high-efficiency water-saving irrigation project area in the province’s cultivated land area; and establish a water price system combining basic water price and metered water price. At the same time, strive to achieve the province’s average irrigation water utilization coefficient of more than 0.60 water saving target by 2020. We will set water quotas for the industry, strictly enforce the approval system for new water use, and improve the water pricing mechanism such as classified pricing, differential pricing, and stepped pricing. Use the price mechanism to promote and guide the whole society to save water. We should transfer from agriculture with high water use intensity to industries with low water use intensity.

  3. (3)

    The water use reduction strategy at the expense of slowing economic growth does not meet the sustainable development demands of developing countries with development as the top priority. Therefore, the water use control in Jiangsu province should be implemented according to the water resources conditions and economic and social development requirements of the region. It should center on some methods which can improve water use efficiency, such as the promotion of agricultural water-saving irrigation technology, industrial and agricultural production layout, transformation of agricultural, and industrial production methods.

  4. (4)

    Though urbanization development and population scale increase have a weak promoting effect on water use, they should not be ignored. It is necessary to raise people’s awareness of water conservation in their daily lives. The water quota for urban and rural residents is very close to the upper limit, so residents are encouraged to use water-saving appliances. The rural areas need to strengthen the construction of water metering facilities and promote one household with one meter to avoid false reports.