Water, food and markets : household-level impact of irrigation water policies and institutions in northern China

Water is increasingly becoming a limiting factor for sustainable economic growth and development, particularly in developing countries. Besides technical innovations, water institution reforms may contribute to improving water allocation decisions. Appropriately designed water institutions can motivate water users to conserve and use water efficiently for irrigation and other uses. In northern China, growing demands on agricultural water due to relatively low water availability and increasing grain production are putting more and more pressure on improving water resource management. The Ministry of Water Resources of the P.R. China has initiated a number of pilot projects to gain experience with the development of water-saving irrigation systems. These pilot projects focus on the construction of engineering systems as well as institutional innovations in water resource management. Analysing the household-level effects of the implemented measures is hence of great importance for further policy development. The project ‘Building a Water-saving Society in Zhangye City’, initiated early 2002 in Zhangye City in northwest China, is the first pilot project of this kind in China. It provides a unique opportunity to examine the economic effects of changes in water policies and institutions. Minle County, the research area for this study, is located within Zhangye City. A large potato processing company was established in Minle County in 2008. After the factory started its activities, the local government intervened in the allocation of irrigation water within the region by assigning more water to a specific variety of potatoes (i.e. Atlantic potatoes) that the factory needs for processing. This further makes Minle County an interesting case for analysing the link between output market development and institutional change in irrigation water management. The general objective of this study is to empirically investigate the household-level impacts of policies and institutional changes in irrigation water use. From this general objective, the following four specific objectives are defined and analysed in separate chapters. 1) To examine the impact of the institutional setup of Water Users Associations (WUAs) on productivity of irrigation water use by the WUA member households, based on a user-based resource governance framework. 2) To analyse the effects of a policy affecting the availability of water for different crops on farmers’ acreage allocation among crops. 3) To evaluate the internal valuation (i.e. marginal value) of irrigation water, before and after the introduction of the water policy as explained above. 4) To investigate the effects of output market development on irrigation water trading. The information used for the empirical analyses mainly comes from two surveys that were carried out in Minle County in May 2008 and May 2010. These surveys cover information for the years 2007 and 2009, that is before and after the potato processing factory became operational. A stratified sampling approach was used for selecting the households and WUAs to be interviewed in the surveys. Additional interviews were held by the author in August 2010 with the Water Management Bureaus (WMBs) that are responsible for water allocation within the seven irrigation areas in Minle County. Chapter 2 investigates the underlying causes of differences in WUA performance by analysing the impact of WUA characteristics on the productivity of irrigation water use. Total crop production value and household income obtained from crop production, both expressed per m³ of water, are used as dependent variables in the empirical analysis. The explanatory variables in the analysis are derived from an established user-based resource governance framework, that specifies the conditions under which user groups are expected to sustainably govern common-pool resources. These conditions are grouped into resource characteristics, group characteristics, relationships between resources and user groups, and the external environment (markets, technology). Applying a random intercept model, the estimation results show that group characteristics, particularly group size and number of water users groups, and the existing pressure on available water resources are important WUA characteristics explaining water productivity. Chapter 3 analyses the impact of the local government intervention in irrigation water allocation on farmers’ crop planting decisions. A system of unconditional crop acreage demand functions depending on prices of variable inputs, levels of quasi-fixed inputs and prices of outputs is estimated. Two hypotheses are tested: Firstly, the government intervention results in an increase in land allocated to Atlantic potatoes and a decrease in land allocated to other crops; Secondly, among the alternative crops (i.e. other crops than Atlantic potatoes), the water policy is expected to cause a relatively small response for grain crops, because grains are mainly used for domestic consumption. The empirical results do not support the first hypothesis. The increased water allocation to Atlantic potatoes does not significantly affect the land allocated to this crop, because its planting decisions are mainly taken by village leaders instead of households. Instead, the intervention results in a shift from planting potatoes towards grains with relatively low water requirements.The second hypothesis is partly supported by the empirical results. The estimated impact of the government intervention is found to be stronger for local potato varieties than for grains, but the impact on the area planted with cash crops does not differ significantly from zero. Output prices seem to play a more important role in cash crop planting decisions than the water allocation intervention. Chapter 4 examines the economic valuation (i.e. marginal value) of irrigation water, before and after the local government intervention in water allocation. To accomplish this, a system of translog production functions is estimated. Two hypotheses are tested: Firstly, the valuation of irrigation water is expected to be equal across different crops before the start of the new water policy. And secondly, valuation of irrigation water is expected to be lower for Atlantic potatoes as compared to the alternative crops after the water policy change. The empirical results do not support the first hypothesis. The valuation of irrigation water used on grain crops is very low, and is even below the actual water prices charged to farm households. This is probably due to self-consumption of grain by households, and to government subsidies for grain farmers that are based on the planted area with grains. The second hypothesis is supported by the empirical results, except for grains. The valuation of irrigation water used on Atlantic potatoes is lower than the value of water used on other (non-grain) crops. Moreover, the returns for irrigation water used on other crops are higher in the year after the water allocation intervention than in the year before the intervention. Chapter 5 aims to provide insights into the impact of output market development on the trading of water use rights by farm households. Theresults of the two farm household surveys indicate that water markets have emerged at a small scale in response to the development of the potato market in Minle County. Observed water trade in the second survey, that was held after the establishment of the potato processing factory, consists mainly of the exchange of water without payment between relatives or neighbours, and seems to be meant to improve the timing of water applications to crops with different seasonal water requirements. Those who have started trading water rights tend to have more land with water use rights than other potato farmers. High transaction costs and information asymmetry between the government and water users, however, severely constrain the trading of water use rights in the region. Chapter 6 summarizes and integrates the main findings, discusses the policy implications and the limitations of the research, and presents some suggestions for further research.