Using incentives to coordinate responses to a system ...

Ecosystem Services 32 (2018) 1–8

Contents lists available at ScienceDirect

Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser

Using incentives to coordinate responses to a system of payments for watershed services: The middle route of South–North Water Transfer Project, China Jichuan Shenga,b,c, Michael Webberc,

T

⁎

a

Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, 219 Ningliu Road, Nanjing 210044, China b School of Economics and Management, Nanjing University of Information Science & Technology, 219 Ningliu Road, Nanjing 210044, China c School of Geography, University of Melbourne, 221 Bouverie Street, Parkville, VIC 3010, Australia

A R T I C LE I N FO

A B S T R A C T

Keywords: South–North Water Transfer Project Middle route Payments for watershed services Extra incentives Danjiangkou reservoir

In any scheme of payments for watershed services, the incentive size can directly affect the watershed services which are provided by water-source areas. It is necessary to understand the payments for watershed services in the Middle Route of China’s South–North Water Transfer (SNWT) from the perspective of the relations among the central and local governments. The distributions of interests among governments need to be coordinated under the Chinese authoritarian system, which is characterised by a combination of political centralisation and economic decentralisation. This paper analyses the interactions between the central and local governments in water-source areas to determine the watershed service efforts of local governments, using Stackelberg game models. In particular, the potential effects of incentives on payments for watershed services on the middle route are analysed. Numerical simulation is adopted to examine watershed service strategies, both with and without central government coordination of incentives. The results demonstrate the following: First, by designing and coordinating extra incentives, the central government could achieve its maximum interests without causing losses to the local governments. Second, extra incentives could increase the watershed service efforts of some local governments, thereby efficiently improving the water source quality of the Middle Route. Third, local governments with better watershed service capabilities are likely to improve their watershed services under coordination, thereby obtaining extra incentives.

1. Introduction China’s South-North Water Transfer (SNWT) has become the most massive inter-basin water transfer megaproject in the world, costing approximately $20 billion and resettling more than 300,000 people (Ministry of Water Resources, 2002). The SNWT can provide 27.8 billion m3 fresh water from the Yangtze River Basin in the south to the drier plains in the north by the Eastern and Middle Routes, through 2900 km of canals (Ministry of Water Resources, 2002). The SNWT connects China’s four major basins, six provinces and hundreds of millions of water users and polluters (Barnett et al., 2015), affecting almost one-third of China’s landmass (Zhang et al., 2009). The Eastern Route (SNWT-ER) was completed in 2013, and the first phase of Middle Route (SNWT-MR) formally started Phase I operation in 2014, six years behind schedule. Given its task of bringing clean water onto the North China Plain,

⁎

the Middle Route has created a conflict over water quality and quantity between water-source areas and water-receiving areas (Yang and Zehnder, 2005). As the water-source areas of Middle Route, Danjiangkou Reservoir and its upstream regions include 43 counties in Henan, Hubei and Shaanxi provinces with a watershed area of 95,200 km2 (Fig. 1), which can provide watershed services of pollution prevention and soil and water conservation (State Council, 2012). The central government has taken a number of measures to ensure the water quality of Danjiangkou Reservoir. One of the most controversial measures for water quality protection is the decision to ban cage aquaculture and turmeric processing in the reservoir area. Sacrifice is certainly evident: turmeric planting and processing (the core industry in Yunxi County, Hubei Province) had to be abandoned to protect water quality, affecting the livelihood of farmers and workers. At the same time, more than 200,000 aquaculture cages had been removed from the reservoir area by the end of 2015 (Pohlner, 2015).

Corresponding author at: School of Geography, University of Melbourne, 221 Bouverie Street, Parkville, VIC 3010, Australia. E-mail address: [email protected] (M. Webber).

https://doi.org/10.1016/j.ecoser.2018.05.005 Received 30 July 2017; Received in revised form 26 January 2018; Accepted 14 May 2018 2212-0416/ © 2018 Elsevier B.V. All rights reserved.

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber

62 Fig. 1. Danjiangkou Reservoir and its upstream regions. Source: Author’s work.

construction of major water conservancy projects in Central and Western China. The fund is collected by the Ministry of Finance from the beneficiary provinces of the Three Gorges Project and the SNWT, and turned over to the central treasury (Ministry of Finance et al., 2009). Thus, the central government will pay the local governments in watersource areas for their watershed services from the National Fund through a fiscal transfer mechanism (Ministry of Finance et al., 2009). In a word, the current payments in Middle Route rely on central government decisions, rather than inter-provincial negotiations (Pohlner, 2016). The Chinese system is characterised by authoritarian politics that combine political centralisation and economic decentralisation (Xu, 2011). Economic decentralisation is regarded as a significant growth driver in China (Blanchard and Shleifer, 2001). Local governments launched fierce regional economic competition to obtain more fiscal revenue from the central government, thus promoting the rapid growth of their regional economies (Qian and Weingast, 1996). Meanwhile, the economic performance of local officials is also positively related to their political promotion probability (Enikolopov and Zhuravskaya, 2007). Local officials strive to develop their regional economies for political purposes, thereby intensifying competition among regions. The current PWS for the Middle Route only considers the game relations between central and local governments, but ignores competition among local governments. If the central government could improve the design of incentives to encourage local governments to provide more watershed service efforts, then water quality may be enhanced through the competition among local governments (Bao and Fang, 2011). However, current water management is characterised by a hierarchical

Considering the sacrifice in water-source areas, the central government proposed financial incentives for the water-source areas of the Middle Route through a system of payments for watershed services (PWS), to improve the water quality and mitigate the conflict between water-source areas and water-receiving areas (State Council, 2013). As an essential measure of global environmental management, PWS based on payments for ecosystem services (PES) can provide financial incentives to reduce water pollution in upstream regions (Bennett et al., 2014). PWS allows government agencies to purchase watershed services from landowners in upstream areas to reduce water pollution in downstream areas (Webb and Martin, 2016). China is the world’s largest investor in PWS schemes: it invested $11.5 billion in PWS projects in 2013, accounting for 94 percent of global PWS investments (Bennett and Carroll, 2014). China is concerned with ecological services and pollution control related to watersheds, encouraging cooperation between upstream and downstream governments (Zhang and Bennett, 2011). In Asia, since a poverty alleviation goal is typically included, governments rather than private sectors are generally involved in PWS (Huang et al., 2009). This is true of the Middle Route, with the added complication that the compensation for watershed services (such as water pollution control and soil and water conservation) in the Danjiangkou Reservoir and its upstream regions are incorporated into the overall investment budget of the SNWT project (State Council, 2006), which is mainly from the National Fund for Major Water Conservancy Construction (the National Fund). The National Fund is a government fund established to support the SNWT construction, to solve the followup issues of the Three Gorges Project, and to strengthen the

2

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber

Ri′ > 0, which suggests that the benefit increases as effort rises. Due to the law of diminishing marginal returns, the benefit function is also a concave function of effort, i.e. Ri″ < 0. A concave benefit function ensures satisfaction of the first-order conditions and the uniqueness of the optimal solution. Following Cvitanić and Zhang (2007), the benefit function of ith local government is assumed to be

authoritarian system with multiple horizontal and vertical lines of authority (Lieberthal and Oksenberg, 1988). Segmentation of governments often leads to a lack of sufficient coordination and cooperation, significantly increasing transaction costs (Cheng and Hu, 2012). Therefore, the critical issue for this PWS is how to design a suitable incentive contract to coordinate the distribution of interests and benefits between the central authority and various local governments to realise a Pareto improvement for society. Studies of incentive coordination are widespread in the supply chain literature; the goal of principals is to design an incentive plan that tempts managers to take action to maximise the expected return to business owners (Krishnan and Winter, 2010; Lee and Whang, 1999; Porteus and Whang, 1991; Schneeweiss and Zimmer, 2004). In our case, we are particularly interested in how incentive coordination induces local governments in water-source areas along the Middle Route of the SNWT to choose to deliver watershed services so as to maximise the revenue to the central government. The Middle Route uses new canals to transfer water from Danjiangkou Reservoir to North China. The Danjiangkou Reservoir and its upstream regions are the only sources of water pollution (Ministry of Environmental Protection, 2001). Therefore, there are clear producers and consumers of watershed services in Middle Route. This paper focuses mostly on the expensive and impactful Middle Route to model the potential effects of incentive coordination, and makes the following three contributions: (i) modelling the game behavior for watershed services among central and local governments in water-source areas through Stackelberg game models; (ii) illustrating the potential effects of incentive coordination on watershed service efforts and income of central and local governments through numerical simulations and related data about Danjiangkou Reservoir and its upstream regions, to seek a suitable incentive coordination contract for the PWS of Middle Route; and (iii) confirming that the central government can raise the environmental welfare of the whole society without causing local governments any loss, through designing and coordinating extra incentives. The paper is structured as follows. Section 2 establishes Stackelberg game models to describe incentive coordination for water service efforts theoretically. Section 3 presents the results of numerical simulations to compare the potential effects of incentive coordination by analysing scenarios with and without incentive coordination. Section 4 discusses the results, and Section 5 presents the conclusion.

1

Ri (ei ) = 2·αi·ei2

(1)

where αi represents the benefit coefficient for watershed services. When the ith local government makes a maximum effort for watershed services (ei = 1), the local government obtains a benefit of 2·αi. If the ith local government performs no watershed services (ei = 0), it receives no payments from the central government. However, watershed service efforts cause opportunity costs (Ci), which are assumed to be an increasing and convex function of effort, i.e. Ci′ > 0 and Ci″ > 0 (Gürtler and Gürtler, 2014). The opportunity costs and marginal costs both increase as effort rises. Following Cvitanić and Zhang (2007) and Cvitanic et al. (2008), the opportunity cost function of ith local government is assumed to be

Ci (ei ) =

1 ·β ·ei2 2 i

(2)

where βi represents the opportunity cost coefficient for watershed services. The costs paid by the ith local government to protect water quality in Danjiangkou Reservoir are 0.5·βi when it makes a maximum effort (ei = 1). If the ith local government performs no watershed services, it incurs no costs. The water-transfer income of the central government generated by the watershed services of the ith local government (Pi) is also an increasing and concave function of efforts (ei), i.e. Pi′ > 0 and Pi″ < 0. As efforts rise, the water-transfer income of the central government also increases, but at a decreasing rate. Thus, the water-transfer income function of the central government from the ith local government is assumed to be 1

Pi (ei ) = 2·θi ·ei2

(3)

where θi represents the water-transfer income coefficient generated by the watershed services of the ith local government. After deducting payments for watershed services, the central government could obtain 2 · θi from the ith local government when that government makes a maximum effort for watershed services (ei = 1). Under that condition, the water in Danjiangkou Reservoir meets the water quality requirements for the transfer. Since θi ≠ αi, the impacts of the efforts (ei) on the benefits of local governments and the water-transfer income of the central government are different. The central government could design extra incentives (I) to coordinate the efforts of local governments in the PWS to achieve the goal of maximising social environmental welfare. Such incentive coordination would encourage local governments in water-source areas to choose reasonable levels of effort for watershed services, in accord with their watershed service capabilities and the extra incentives. The fixed extra incentives, which are finite, encourage local governments to compete by improving their efforts. When the central government adopts extra incentives to coordinate local governments, there is a competitive relationship among local governments, and rational local governments would invest more effort on watershed services to compete for more benefits. Table 1 shows the specific notations. However, the statuses of central and local governments are different. The central government decides whether (and how much) to pay local governments for watershed services and whether to provide extra incentives. The local governments are followers, only choosing their efforts according to their watershed service capabilities and the payments from the central government. These governments are thus in a leader–follower game, of the kind analysed through a Stackelberg game model. In this model, the basic idea is that one player (leader) has the

2. Incentive coordination for watershed service efforts 2.1. Stakeholders of payments for watershed services in Middle Route Since the payments for watershed services on the Middle Route mainly rely on fiscal transfers from the central government, the local governments in water-receiving areas do not directly pay the local governments in water-source areas. The payments to the National Fund from water-receiving areas are not directly related to watershed services from water-source areas. Therefore, there are only two stakeholders involved in the PWS of Middle Route: local governments in water-source areas and central government. Local governments in water-source areas can receive payments from the central government for providing watershed services, and pay out the corresponding costs for delivering watershed services. The central government receives clean water due to watershed services provided by local governments in the water-source areas, and can obtain the water-transfer income through the water sales to the water-receiving areas. We first assume a scenario without incentive coordination (Scenario 1), which is the current PWS of Middle Route. Under this scenario, the benefit from watershed services of the ith local government in watersource areas (Ri) depends on its watershed service efforts (ei) (0 ≤ ei ≤ 1). To simplify the analysis, several assumptions about the benefit function of local governments are proposed. The benefit of the ith local government (Ri) is an increasing function of efforts (ei), i.e. 3

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber

Table 1 The notation in the models.

πi1 =

Notation

Description

Notes

ei αi

Watershed service efforts of ith local government Benefit coefficient for watershed services of ith local government Opportunity cost coefficient of ith local government Water transfer income coefficient of central government generated by the watershed services of the ith local government Extra incentive from central government

0 < ei ≤ 1 αi > 0

βi θi

I

n

βi > 0 θi > 0

I>0

2.3. Watershed services under the scenario with incentive coordination (Scenario 2) When there is incentive coordination (Scenario 2), the central government is assumed to provide extra incentives (I) to coordinate and improve the watershed service efforts of local governments. Thus, the local governments not only obtain the original payments for watershed services, but also benefit from extra incentives. Following Sheng et al. (2017), we assume that local governments need to consider both their benefits and opportunity costs in choosing their optimal effort, which can be identified as their watershed service capabilities (αi/βi). Thus, local governments will select their effort (ei) to obtain extra incentives based on their watershed service capabilities. The profit function of the ith local government under Scenario 2 is 1

(10)

where I represents extra incentives from the central government. Once the central government decides the optimal extra incentives to maximise its income (I), the amount is fixed. However, the extra incentives available to each local government are not the same. Therefore, local governments must select their effort to compete for corresponding extra incentives based on their watershed service capabilities. At the same time, the water-transfer income function of the central government under Scenario 2 is

(4)

1

(5)

The transactions are maintained by the mutual interest of the stakeholders in the PWS of the Middle Route. Since a stakeholder cannot make deals that would never be profitable, there is always a stable and equilibrium game relation among the stakeholders in a long-term transaction system (Corley and Kwain, 2014). Under the scenario without incentive coordination (Scenario 1), each local government will choose its efforts to maximise revenue, eventually forming a Nash equilibrium among local governments. Since each local government is relatively independent, its optimal strategy would be not affected by any other. When πi1 (ei∗) ≥ πi1 (ei) for any ei∗ ≠ ei, the effort (ei∗) is the optimal effort to maximise the revenue of ith local government under Nash equilibrium. Backward induction can be adopted to obtain the subgame perfect Nash equilibrium. Letting ∂πi1/∂ei = 0, the Nash equilibrium solutions of n local governments can be obtained under Scenario 1 as follows:

n

1

πs2 = 2· ∑ θi ·ei2 −I i=1

(11)

The original intention of incentive coordination is to enhance the efforts of local governments, to achieve a Pareto improvement for both sides. When the central government adopts extra incentives to coordinate the interests of local governments in water-source areas, the incomes of local governments and the central government under Scenario 2 should not be less than those under Scenario 1:

πi2 ⩾ πi1, i = 1,2,…,nπs2 ⩾ πs1

(12)

Apart from the uncertainty of the parameters, it is assumed that decisions by stakeholders are based on complete information (Konrad and Leininger, 2007). As in Scenario 1, the central government, as the leader of the Stackelberg game model, first determines the extra incentive to maximise its income. After understanding the extent of the extra incentives, the local governments as followers choose reasonable efforts according to their watershed service capabilities. Based on these assumptions, the relations between central and local governments in Scenario 2 are regarded as a master–slave Stackelberg game with complete information. The first order conditions of profit maximisation for the ith local government are obtained through letting ∂πi2/∂ei = 0:

(6)

The following can be obtained by solving Eq. (6): 2 3

α ei = ⎜⎛ i ⎟⎞ , i = 1,2,…,n ⎝ βi ⎠

αi ·ei βi ·I n α ∑ βi ·ei i i=1

s. t . 0 ⩽ ei ⩽ 1, i = 1,2,…,n

The water-transfer income function of the central government under Scenario 1 is

−1 ∂πi1 = αi·ei 2 −βi ·ei = 0, i = 1,2,…,n ∂ei

1

maxπi2 = 2·αi·ei2 − 2 ·βi ·ei2 +

1

i=1

(9)

respectively.

maxπi1 = 2·αi·ei2 − 2 ·βi ·ei2

n

−1 3

i=1

Through a fiscal transfer mechanism, the central government pays local governments in water-source areas in accord with their watershed service efforts. Since the central government does not provide extra incentives to local governments in the current PWS of Middle Route, the profit function of ith local government under Scenario 1 is

πs1 = 2· ∑ θi ·ei2

1

πs1 = 2· ∑ θi ·αi3 ·βi

2.2. Watershed services under the scenario without incentive coordination (Scenario 1)

s. t . 0 ⩽ ei ⩽ 1, i = 1,2,…,n

(8)

and

right to take the first step due to, for example, its historical status, size, reputation, innovation and information (Von Stackelberg, 1934). Then, other players (followers) determine their corresponding activities by observing the leader's strategy. The Stackelberg game model is mainly used to describe imperfect competition in markets, but its format fits the interaction of central and local governments in the PWS for the Middle Route.

1

3 34 − 13 ·αi ·βi , i = 1,2,…,n 2

(7)

At this moment, the efforts to maximise revenue of local governments (ei∗) are determined by their benefit coefficients and opportunity cost coefficients for watershed services. Thus, the incomes of the ith local government and the central government are 4

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber −1 ∂πi2 = αi·ei 2 −βi ·ei ∂ei αi ⎛ ·I ·⎜ βi

n

∑

⎝ i=1

+

n

n

maxπs2 = 2· ∑ α αi ⎞ ·e − i ·e βi i βi i ⎟

⎠

⎠

2 αi ⎞ · e i ⎟ βi

n

⎛ ⎜∑ ⎝ i=1

1

n

+ αi·I ·ei2 · ∑

⎠

j=1

αj βj

βi ·ei ⎜∑ ⎝ i=1

Finally, analytical or simulation methods could be used to solve the model.

·ej

2

n 1 2 ·⎛

=0

αi ⎞ ·e βi i ⎟

3. Numerical simulation

⎠ 3.1. Simulation design

(13)

i = 1,2,…,n

The numerical simulation is employed to compare the efforts in the two scenarios to examine the potential effects of incentive coordination on the income of central and local governments in the water-source area. Since Danjiangkou Reservoir and its upstream regions involve three provinces, i = 1, 2, 3 is used to represent Hubei, Henan and Shaanxi, respectively. According to State Council (2012), the central government will allocate 11 731 million yuan to pay for watershed services in Danjiangkou Reservoir and its upstream regions during 2011–2015, which is mainly used to compensate for water pollution control and water and soil conservation. The water-source areas of Hubei, Henan and Shaanxi can obtain 3585 million yuan, 2437 million yuan and 5709 million yuan, respectively. Consequently, annual average payments for the water-source areas of Hubei, Henan and Shaanxi are 717 million yuan per year, 487.4 million yuan per year and 1141.8 million yuan per year, respectively, which are set to the benefit coefficients (αi). Local governments’ efforts to improve watershed service lead to reductions in the gross outputs of the water-source areas. Since microeconomic data about opportunity costs is difficult to obtain (such as the benefits related to turmeric and cage aquaculture), the reduced outputs are regarded as the opportunity costs of the watershed services. According to State Council (2006), the annual average gross outputs of the water-source areas of Hubei, Henan and Shaanxi are 27,950 million yuan, 6150 million yuan, 10,300 million yuan, respectively, which are set to the opportunity cost coefficients (βi) in the simulation. According to the master plan, 13 billion m3 per year of fresh water is to be transferred from Danjiangkou Reservoir through the Middle Route by 2050 (Ministry of Water Resources, 2002). The unit water-transfer income calculated in 2014 was 0.13 yuan per m3 (National Development and Reform Commission, 2014). Thus the central government could obtain an annual water-transfer income of 1690 million yuan from the water transfer through Middle Route. Since wastewater discharged into rivers is the primary source of water pollution in Danjiangkou Reservoir (Zhu et al., 2008), the water in Danjiangkou Reservoir would maintain high quality if local governments make 100 percent effort and successfully eliminate wastewater discharge into rivers. Therefore, if the water-transfer income is assumed to originate from the watershed services of local governments in Danjiangkou Reservoir and its upstream regions, the water-transfer income coefficients from the efforts of each province can be calculated according to the amount of provincial wastewater (discharged into rivers) that they eliminate. According to State Council (2012), the amounts of wastewater discharged into rivers in the water-source areas of Hubei, Henan and Shaanxi are 181.55 million tons, 59.77 million tons and 172.12 million tons in 2010, respectively (43.91%, 14.46% and 41.63% of the total). According to the total water-transfer income and the proportion of provincial wastewater discharged into rivers, the watertransfer income coefficients of the central government generated by the watershed services in each province (θi) are 742.11 million yuan, 244.32 million yuan and 704.57 million yuan, respectively. The numerical specification is shown in Table 2.

Thus, 2

2

n

n n 3 1 αj α ⎞ α ⎞ ⎛ ⎛ ·ej = 0 αi·βi ·⎜∑ i ·ei ⎟ −βi2·ei2 ⎜∑ i ·ei ⎟ + αi·I ·ei2 · ∑ β β j = 1 βj ⎝ i=1 i ⎠ ⎝ i=1 i ⎠

i,j = 1,2,...,n,i ≠ j

(14)

Transposing the terms of Eq. (14), then n

I· ∑ j=1

2

n

2

−1 α ⎞ β ⎛ ·ej = ⎜∑ i ·ei ⎟ ·⎛⎜ i ·ei−βi ·ei 2 ⎞⎟ βj β α ⎠ ⎝ i=1 i ⎠ ⎝ i

αj

i,j = 1,2,...,n,i ≠ j

(15)

According to Eq. (15), n equations could be obtained when i is 1, 2, …, n, respectively. n

I· ∑ j=1 n

I· ∑ j=1 n

I· ∑ j=1

2 − 12 αi ⎞ ⎛ β12 ⎞ · e i ⎟ · α1 ·e1−β1·e1 βi

n

⎛ ·e = ⎜∑ βj j ⎝ i=1

αj

αj βj

⎠ ⎝ 2

n

αi ⎞ ·e βi i ⎟

⎛ ·ej = ⎜∑ ⎝ i=1

⋯

2 ⎛ β2

· α ·e2−β2 2 ⎠ ⎝

⎠

−1 ·e 2 ⎞ 2

⎠

2

n

−1 α ⎞ β2 ⎛ ·e = ⎜∑ βi ·ei ⎟ ·⎛ αn ·en−βn·en 2 ⎞ βj j n i ⎠ ⎠ ⎝ ⎝ i=1 j = 1,2,...,n,i ≠ j

αj

(16)

Accumulating these n equations, then n

I ·(n−1)· ∑ i=1

n n β2 −1 ⎤ αi α ⎡ ·ei = ( ∑ i ·ei )2 ·⎢∑ ⎜⎛ i ·ei−βi ·ei 2 ⎟⎞ ⎥ βi βi α i=1 ⎠⎦ ⎣ i=1 ⎝ i

(17)

Simplifying Eq. (17) to obtain the extra incentives to maximise income (I) n

⎛ ⎜∑ i=1 I= ⎝

αi ⎞⎡ ·e · βi i ⎟ ⎢

(19)

⎠

2 3 αi ⎞ 2 2 · e i ⎟ −βi ·ei βi

n

πs2 ⩾ πs1 0 ⩽ ei ⩽ 1

αi ⎞ ·e βi i ⎟

⎛ ⎜∑ ⎝ i=1 ⎛ αi·βi ·⎜∑ ⎝ i=1

i=1

s. t . πi2 ⩾ πi1,i = 1,2,…,n

2

n

n

−1 ⎤ β2 ⎛ α ⎞⎡ ∑ βi ·ei⎟·⎢ ∑ ⎛⎜ αi ·ei − βi·ei 2 ⎞⎟ ⎥ 1 ⎜ i i ⎢i = 1 ⎝ ⎠⎥ ⎦ ⎠⎣ θi ·ei2 − ⎝i = 1 n−1

n

∑

2 1 ⎛ βi ·ei−βi ·ei− 2 ⎞ ⎤ αi ⎝ ⎠⎥ ⎦

⎠ ⎣ i=1 n−1

(18)

According to Eq. (18), the income-maximising extra incentives are related to the efforts of local governments and their benefit coefficients and opportunity cost coefficients for watershed services. Meanwhile, the extra incentives to maximise income are also related to the number of local governments in the water-source areas. The nonlinear bi-level programming model can be transformed into a nonlinear single-level programming model by substituting Eq. (12) and Eq. (18) into Eq. (7), i.e.

5

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber

Table 2 Numerical specification. Notation

Description

Unit

α1 α2 α3 β1 β2 β3 θ1 θ2 θ3

Hubei’s benefit coefficient Henan’s benefit coefficient Shaanxi’s benefit coefficient Hubei’s opportunity cost coefficient Henan’s opportunity cost coefficient Shaanxi’s opportunity cost coefficient Water transfer income coefficient generated by watershed services in Hubei Water transfer income coefficient generated by watershed services in Henan Water transfer income coefficient generated by watershed services in Shaanxi

Million Million Million Million Million Million Million Million Million

Stakeholders

Optimal water service effort

Optimal profits (million yuan)

Scenario 1

Hubei Henan Shaanxi Central government

0.0870 0.1844 0.2308

317.1925 314.0366 822.7456 1,323.64

Scenario 2

Hubei Henan Shaanxi Central government

0.0790 0.2478 0.2935

317.6603 314.0366 822.7456 1,374.10

Proportion (%)

717.00 487.40 1,141.80 27,950.00 6,150.00 10,300.00 742.11 244.32 703.57

30.56 20.77 48.67 62.95 13.85 23.20 43.91 14.46 41.63

Therefore, incentive coordination can not only ensure that local governments’ interests are not damaged, but can also stimulate local governments to improve watershed service efforts to maximise society’s environmental welfare. Under incentive coordination, the central government realises increased water-transfer income while the interests of local governments are not damaged.

Table 3 Simulation results. Scenario

yuan yuan yuan yuan yuan yuan yuan yuan yuan

Value

4. Discussion The Stackelberg game models for watershed services developed in this paper describe the interaction among the central and local governments in the water-source areas of the Middle Route. The authoritarian regime with its combination of political centralisation and economic decentralisation is a typical Chinese institution (Xu, 2011). Under this regime, local governments, which are similar to multiple divisions of an enterprise, have both relative economic autonomy and approximately homogeneous tasks. Local governments engage in a fierce regional competition to maximise their interests to obtain more financial payments for watershed services from the central government (Jin et al., 2005). Thus, to understand the PWS in the Middle Route, it is necessary to examine the roles of institutions, government and policy from the perspective of the relations among central and local governments. Especially, the coordination of interest distribution among central government and local governments is particularly important in China’s water management. Some laws and policies related to water cannot be implemented in China due to the lack of sufficient coordination among central and local governments (Liu and Yang, 2012). China’s water management system arising from the SNWT tends to produce more central authority, and the central government's financial objectives can directly determine the balance sheets of local governments (Pohlner, 2016). Therefore, the central government can make use of its dominant position and competition between local governments to coordinate the incentives for local governments’ watershed services. The central government can design and control extra incentives to maximise income, to effectively induce the local government to improve watershed service efforts with the lowest incentive costs. A Pareto improvement can be achieved that increases society’s environmental welfare while not damaging local governments’ interests. The water quality of Middle Route depends mostly on the watershed service efforts of local governments, which also shows the importance of agents in ensuring the effectiveness of the PWS. Local governments in water-source areas could limit the overall utility of the Middle Route by delaying watershed services, to gain more concessions from the central government. For example, cage aquaculture and turmeric processing, as important industries in Danjiangkou Reservoir area, produce a significant amount of wastewater. During 2006–2014, the total nitrogen content increased from 0.62 mg/L to 1.55 mg/L, and the overall phosphorus content increased from 0.005 mg/L to of 0.02 mg/L in Danjiangkou Reservoir (Pohlner, 2015). Given this, it is necessary for the central government to use persuasive resources, such as information, cooperation and compensation, to closely coordinate with the provincial governments and achieve strategic objectives in the field of

3.2. Results Given the specific parameter values, the nonlinear programming models can be solved using MATLAB. Thus, not only the optimal efforts (to maximise revenue) and the corresponding income of local governments, but also the maximum water-transfer income of central government under the two scenarios, can be obtained by solved Eqs. (8), (9) and. (19). The results are shown in Table 3. The optimal extra incentive to maximise the central government’s income (I) is 48.6281 million yuan. According to Table 3, the optimal effort to maximise revenue in the water-source areas of Shaanxi is highest (e31 = 0.2308) under current PWS, while that in the watersource areas of Hubei is lowest (e11 = 0.0870). Since the watershed service capability in the water-source areas of Hubei is the weakest (α1/ β1 = 0.0257), Hubei chooses to reduce its efforts (e12 = 0.0790) according to its capability after incentive coordination. At the same time, the efforts to maximise revenue in the water-source areas of Henan and Shaanxi are increased by 34.33% and 27.17%, respectively. Therefore, incentive coordination can enhance the majority of local governments' watershed service efforts. The total effort of the three local governments (weighted by their share of wastewater discharge into Danjiangkou reservoir) rise by nearly 20%, from 0.161 to 0.193. Table 3 also indicates that the income of the three provinces does not decrease after incentive coordination, while there is a slight increase in Hubei’s income (a rise of 0.15%). Therefore, incentive coordination from the central government does not damage the interests of local governments in the water-source areas. The local governments with a higher ratio of benefit coefficients to opportunity cost coefficients have better watershed service capabilities, and due to competition, they tend to increase efforts to obtain more extra incentives. The water-transfer income of central government increases by 3.82% after deducting the extra incentive of 48.6281 million yuan. Incentive coordination improves the efforts of most local governments to protect water quality, thereby enhancing the environmental welfare of society. Furthermore, the increased water-transfer income not only offsets the costs of extra incentive, but also provides additional surplus (an increase of 50.46 million yuan) through incentive coordination. 6

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber

environment (Moore, 2014). The extra incentives from the central government are fixed; however, the extra incentives available to each local government are not equal. To achieve maximum interests, competition among local governments is necessary. Since the income does not fall after incentive coordination, the local governments with high watershed services capabilities are likely to improve watershed service efforts. Local governments are willing to compete for extra incentives from the central government by choosing the effort that maximizes their revenue. However, excessive competition among local governments may also have negative impacts, mainly caused by the entrepreneurial behavior of local governments in China’s water management system, such as local governments’ tension between profit and motive, uncertainty about roles and responsibilities, and obstacles to national water resource coordination and management (Barnett et al., 2015). Therefore, the central government needs to coordinate the local government competition effectively, and to provide sufficient positive incentives to local governments, to ensure that local governments have a high degree of adaptability (Xia and Pahl-Wostl, 2012). These measures could contribute to maximising society’s environmental welfare through effective competition under the existing water management system. Incentive coordination for local governments in the water-source areas would also help the central government to achieve strategic objectives in the environmental field. Due to competition among local governments, extra incentives could encourage the efforts of local governments. The higher the efforts of local governments, the higher the water-transfer income of the central government. Thus, the central government also has incentives to implement incentive coordination. With a combination of administrative fiscal transfers and market-oriented incentives, the SNWT reflects typical Chinese governmentality, incorporating the market and interventionist reasoning and techniques (Rogers et al., 2016). Therefore, the PWS of the Middle Route is only a weak version of ecological modernisation, and one more focused on technological solutions rather than the general adoption of ecological modernisation by Western countries (Moore, 2014). The prominent feature of China’s PWS is that the central government plays an important role. Despite the existence of defects and deviations from the original intention of the policy design, the PWS of the Middle Route reflects policy innovation for China’s reality (Wang et al., 2016).

Through the contemporary Chinese political and economic analysis of water governance, the analysis augments the literature on PWS of inter-basin water transfer megaprojects in the world. The latter are mostly generated by carefully observing the fully market-oriented PWS that operate in Western democratic political systems (such as Bennett et al. (2014), Potter and Wolf (2014), Huber-Stearns et al. (2015) and Webb and Martin (2016)). The present study develops an understanding of the incentive scheme of PWS in the non-democratic system. This paper makes the following three contributions: (i) theoretically analyzing the potential effects of incentive coordination on watershed service efforts of local governments by modelling the game behavior for watershed services among central and local governments in watersource areas; (ii) simulating the effects of incentive coordination on the stakeholders’ interest distribution by using the data about Danjiangkou Reservoir and its upstream regions, which could help policymakers to seek a suitable incentive coordination contract for the PWS of Middle Route; (iii) confirming that the central government can maximise its own interests through designing and coordinating extra incentives, which can also ensure that local government interests do not suffer any loss. Through incentive coordination, most local governments (Henan and Shaanxi) would improve their watershed service efforts, thereby effectively improving water quality along the Middle Route. Local governments with better watershed service capabilities are more likely to improve watershed services, and thus obtain extra incentives. The PWS of Middle Route illustrates more than anything else the fundamental importance of coordination among governments in China’s authoritarian environmental politics.

5. Conclusion

Bao, C., Fang, C.-L., 2011. Water resources flows related to urbanization in China: challenges and perspectives for water management and urban development. Water Resour. Manage. 26, 531–552. Barnett, J., Rogers, S., Webber, M., Finlayson, B., Wang, M., 2015. Sustainability: transfer project cannot meet China’s water needs. Nature 527, 295–297. Bennett, D.E., Gosnell, H., Lurie, S., Duncan, S., 2014. Utility engagement with payments for watershed services in the United States. Ecosyst. Serv. 8, 56–64. Bennett, G., Carroll, N., 2014. Gaining Depth: State of Watershed Payments 2014. Forest Trends, Washington, DC. Blanchard, O., Shleifer, A., 2001. Federalism with and without political centralization: China versus Russia. IMF Econ. Rev. 48, 171–179. Cheng, H., Hu, Y., 2012. Improving China’s water resources management for better adaptation to climate change. Clim. Change 112, 253–282. Corley, H.W., Kwain, P., 2014. A cooperative dual to the nash equilibrium for two-person prescriptive games. J. Appl. Math. 2014, 4. Cvitanic, J., Wan, X., Zhang, J., 2008. Principal-agent problems with exit options. J. Theor. Econ. 8 1935-1704. Cvitanić, J., Zhang, J., 2007. Optimal compensation with adverse selection and dynamic actions. Math. Financ. Econ. 1, 21–55. Enikolopov, R., Zhuravskaya, E., 2007. Decentralization and political institutions. J. Public Econ. 91, 2261–2290. Gürtler, M., Gürtler, O., 2014. The interaction of explicit and implicit contracts: a signaling approach. J. Econ. Behav. Organ. 108, 135–146. Huang, M., Upadhyaya, S.K., Jindal, R., Kerr, J., 2009. Payments for watershed services in Asia: a review of current initiatives. J. Sustainable Forest. 28, 551–575. Huber-Stearns, H.R., Goldstein, J.H., Cheng, A.S., Toombs, T.P., 2015. Institutional analysis of payments for watershed services in the western United States. Ecosyst. Serv. 16, 83–93. Jin, H., Qian, Y., Weingast, B.R., 2005. Regional decentralization and fiscal incentives: federalism, Chinese style. J. Public Econ. 89, 1719–1742. Konrad, K.A., Leininger, W., 2007. The generalized Stackelberg equilibrium of the all-pay auction with complete information. Rev. Econ. Design 11, 165–174. Krishnan, H., Winter, R.A., 2010. Inventory dynamics and supply chain coordination.

Acknowledgements The authors are grateful to the financial support provided by the National Natural Science Foundation of China (Nos. 71774088, 71303123), the Social Science Foundation in Jiangsu Province (17GLD014), the Six Talent Peaks Project in Jiangsu Province (2017JNHB-058). This paper is also funded by the Qing Lan Project. The authors are also grateful to the Australian Research Council for funding in support of this research (DP170104138). References

Due to the complexity and profound spatial effects of the PWS in the SNWT, it is necessary to think from the perspective of the relations among central and local governments, and to understand that coordination among government interests is the key to the successful implementation of PWS under the Chinese non-democratic and nonWestern system. Local governments in water-source areas compete to gain more payments from the central government. However, the existing water management system with multiple horizontal and vertical lines of authority may easily lead to insufficient coordination and cooperation (Lieberthal and Oksenberg, 1988). Thus, the central government needs to explore suitable incentive coordination contracts, which not only could stimulate local governments to improve watershed service efforts through competition, but also could coordinate the interest distribution among governments. Such incentive coordination can often produce better results than the standard system of PWS, as it can improve the overall efforts for watershed services and the water-transfer income of the central government. The Stackelberg game models for watershed services developed in this paper describe the interaction among the central and local governments in the PWS of Middle Route. The numerical simulations compare the potential effects of incentive coordination on the revenue-maximising watershed service efforts of local governments in Danjiangkou Reservoir and its upstream regions, thus illustrating one way in which to stimulate watershed service efforts. 7

Ecosystem Services 32 (2018) 1–8

J. Sheng, M. Webber

the conduct of water: China's South-North Water Transfer Project. Trans. Inst. Br. Geogr. 41, 429–441. Schneeweiss, C., Zimmer, K., 2004. Hierarchical coordination mechanisms within the supply chain. Eur. J. Oper. Res. 153, 687–703. Sheng, J., Zhang, S., Li, Y., 2017. Heterogeneous governance capabilities, reference emission levels and emissions from deforestation and degradation: A signaling model approach. Land Use Policy 64, 124–132. State Council, 2006. Reply of State Council on Plan of Water Pollution Control and Water and Soil Conservation in Danjiangkou Reservoir Area and Upstream Region (国务院 关于丹江口库区及上游水污染防治和水土保持规划的批复) (Official communication 国函[2006]10号). State Council, Beijing. State Council, 2012. Reply of State Council on 12th Five-year Plan of Water Pollution Control and Water and Soil Conservation in Danjiangkou Reservoir Area and Upstream Region (国务院关于丹江口库区及上游水污染防治和水土保持“十二五”规划 的批复) (Official communication 国函[2012]50号). State Council, Beijing. State Council, 2013. Report on the Construction of Ecological Compensation Scheme (国 务院关于生态补偿机制建设工作情况的报告). State Council, Beijing. Von Stackelberg, H., 1934. Marktform und gleichgewicht. J. Springer. Wang, H., Dong, Z., Xu, Y., Ge, C., 2016. Eco-compensation for watershed services in China. Water Int. 41, 271–289. Webb, A.A., Martin, P.V., 2016. Potential of a payments for ecosystem services scheme to improve the quality of water entering the Sydney catchments. Water Policy 18, 91–110. Xia, C., Pahl-Wostl, C., 2012. The process of innovation during transition to a water saving society in China. Water Policy 14, 447–469. Xu, C., 2011. The fundamental institutions of China’s reforms and development. J. Econ. Lit. 49, 1076–1151. Yang, H., Zehnder, A.J., 2005. The south-north water transfer project in China: an analysis of water demand uncertainty and environmental objectives in decision making. Water Int. 30, 339–349. Zhang, Q., Bennett, M.T., 2011. Eco-compensation for Watershed Services in the People’s Republic of China. Asian Development Bank, Manila, Philippines. Zhang, Q., Xu, Z., Shen, Z., Li, S., Wang, S., 2009. The Han River watershed management initiative for the South-to-North water transfer project (Middle Route) of China. Environ. Monit. Assess. 148, 369–377. Zhu, Y.P., Zhang, H.P., Chen, L., Zhao, J.F., 2008. Influence of the south-north water diversion project and the mitigation projects on the water quality of Han River. Sci. Total Environ. 406, 57–68.

Manage. Sci. 56, 141–147. Lee, H., Whang, S., 1999. Decentralized multi-echelon supply chains: Incentives and information. Manage. Sci. 45, 633–640. Lieberthal, K., Oksenberg, M., 1988. Policy Making in China: Leaders, Structures, and Processes. Princeton University Press, Princeton, NJ, USA. Liu, J., Yang, W., 2012. Water sustainability for China and beyond. Science 337, 649–650. Ministry of Environmental Protection, 2001. Pollution Control Plan for Eastern Route of South-North Water Transfer Project (南水北调东线工程治污规划). Ministry of Environmental Protection, Beijing. Ministry of Finance, National Development and Reform Commission, Ministry of Water Resources, 2009. Notice from Ministry of Finance, National Development and Reform Commission, Ministry of Water Resources on the Release of Interim Administrative Measures for Collection and Use of National Major Water Conservancy Construction Fund (财政部、国家发展改革委、水利部关于国家重大水利工程建设基金征收使用管 理暂行办法的通知) (Official communication 财综[2009]90号). Ministry of Finance, Beijing. Ministry of Water Resources, 2002. South-North Water Transfer Project Masterplan (Summary) (南水北调工程总体规划(简要本)). Ministry of Water Resources, Beijing. Moore, S.M., 2014. Modernisation, authoritarianism, and the environment: the politics of China’s South-North Water Transfer Project. Environmental Politics 23, 947–964. National Development Reform Commission, 2014. Notice from National Development and Reform Commission on Water Price Policy in the Initial Operation Period of the First Phase Project of Middle Route of South-North Water Transfer (关于南水北调中 线一期主体工程运行初期供水价格政策的通知) (Official communication 发改价格 [2014]2959号). National Development and Reform Commission, Beijing. Pohlner, H., 2015. Can South-North water transfer project and industry co-exist?, China Dialogue. https://www.chinadialogue.net/article/show/single/en/8236-Can-SouthNorth-water-transfer-project-and-industry-co-exist. Pohlner, H., 2016. Institutional change and the political economy of water megaprojects: China’s south-north water transfer. Global Environ. Change 38, 205–216. Porteus, E.L., Whang, S., 1991. On manufacturing/marketing incentives. Manage. Sci. 37, 1166–1181. Potter, C.A., Wolf, S.A., 2014. Payments for ecosystem services in relation to US and UK agri-environmental policy: disruptive neoliberal innovation or hybrid policy adaptation? Agric. Hum. Values 31, 397–408. Qian, Y., Weingast, B.R., 1996. China’s transition to markets: market-preserving federalism, Chinese style. J. Policy Reform 1, 149–185. Rogers, S., Barnett, J., Webber, M., Finlayson, B., Wang, M., 2016. Governmentality and

8