Water rights reform and water-saving irrigation: evidence from China

A natural resource that is vital to human existence and economic growth is water. As economies grow, water scarcity eventually becomes a global issue. The primary cause of the water shortage is that human usage of water resources exceeds the supply of freshwater (Savenije 2000). This also reflects the contradiction between human use of water resources and the natural availability (Rijsberman 2006). The increase in food demand caused by global population growth will continue to increase the demand for water resources, resulting in more serious water scarcity. Water resource is also an important natural resource for food production (Mueller et al. 2012), and utilizing water resources wisely is one of the key elements in promoting the production of food (D'Odorico et al. 2018). In order to increase the yield of agriculture, agricultural production needs to shift from rain-fed agriculture to irrigated agriculture, which will also increase the demand for freshwater (Rosa et al. 2018). However, the popularization of irrigation may increase the demand for water resources, which may lead to an increase in water stress (Elliott et al. 2014). Facing water scarcity due to constraints in supplying enough fresh water, efficient water resource management is required to handle water issues (Mohseni et al. 2022). To deal with the contradiction between water stress and food production, adopting water-saving irrigation (WSI) technology has become one of the important measures to avoid the increase of water stress and promote the increase of food production.

Current research concentrates on a variety of influencing factors for the adoption of WSI. Factors influencing the adoption of WSI technologies include institutional factors (Liu et al. 2019; Hu et al. 2022), adoption of climate change (Mi et al. 2021; Surendran et al. 2021; Chen et al. 2022), household characteristics (Namara et al. 2007; Kumar 2012; Chandran & Surendran 2016). Because WSI devices are expensive, in order to promote the adoption of efficient water-saving technologies, the government will promote the adoption of WSI technologies through various subsidy policies (Liu et al. 2019; Hu et al. 2022). Due to the change and instability of rainfall caused by climate change, farmers will adopt high-efficiency technologies for irrigation to ensure production in agriculture (Mi et al. 2021; Surendran et al. 2021; Chen et al. 2022). For characteristics of households, Namara et al. (2007) and Kumar (2012) indicate that older irrigators with enough experience may adopt high-efficiency irrigation, while younger irrigators with new knowledge may adopt WSI for agricultural production (Chandran & Surendran 2016).

The current water resource management mode is gradually changing to marketization, especially the water rights reform (WRR), such as the groundwater market in India (Tamuli et al. 2022). As a part of the hydrological cycle, water resources are fluid and constantly updated, which limits the applicability of traditional legal methods to natural resources. Therefore, traditional water rights institutional arrangements are associated with land ownership institutional arrangements. Due to water scarcity and the escalating disparity between the supply and demand of water resources, the government gradually formed a reform of the separation of rights arrangements and land ownership, also known as modern WRR. The reform is an essential institutional arrangement to realize the efficient use of limited water resources under the condition of the market economy. This institutional arrangement can promote water rights to be transacted based on market principles so that scarce water resources will flow to high-efficiency uses, thereby achieving efficient use of water resources (Brown 2006). China is also facing the problem of water scarcity and the transformation of water resource management. The reform of China's water rights system mainly began to mature in 2014. First of all, the government allocates the amount of water that can be used in agriculture. This allocation amount is mainly the remaining water after deducting reasonable living, non-agricultural production, ecological environment water requirement and reserved water for natural flows from the distributable water. The savings of water resources that can be realized through water-saving technologies can be traded to obtain benefits. Thus, the WRR may be beneficial to adopt high-efficiency irrigation.

This study takes China's WRR as an example, with China's panel data from 2004 to 2019, and adopts a difference-in-differences (DID) model to empirically analyze the impact of WRR on the land area of WSI. The DID strategy is an econometric analysis method for studying pilot policy effects. Also, the WRR is a typical pilot policy, because only some areas have carried out this reform. So, it is suitable to use this method to estimate the impacts of the WRR on WSI. Also, using this method to estimate the impact of WRR on WSI areas can lead to more credible research conclusions. In addition, the impact on WSI areas is of great significance for ensuring agricultural production and promoting food security. Since the research in this article is based on the analysis of province data, it is difficult to clarify the microscopic mechanism of farmers' adoption of WSI technologies, which is worth exploring in future research.

Since WRR is a pilot policy in China, the policy can be regarded as a quasi-natural experiment. The DID is a policy evaluation method used to estimate the impact of pilot policies, so this article mainly uses the empirical strategy of DID for estimation. The adoption of the method requires parallel trend testing, so this article will further use the event study method to test it to ensure the credibility of the estimation results.

Table 2 shows the estimation results of the baseline specification. In addition to the fixed effects, there are no control variables in column 1 at all, columns 2–4 include a natural water-related variable (precipitation), a land-related variable (land potential), and a geography variable (slope) respectively, and column 5 is a specification with full variables mentioned above. Column 1 shows that the WRR can increase the land area for WSI at a significance level of 1%. Specifically, the WRR can expand the land area for high-efficiency irrigation by an average of 13.69%. Columns 2–4 are all consistent with the previously mentioned results in column 1, and the relationship is all significant at the 1% level. According to the estimation results including all the control variables, compared with other provinces, the land area for WSI can be increased by an average of 13.63% in the provinces with WRR, which is a relatively large increase. Therefore, the WRR can promote the expansion of land area for WSI, and the increase is relatively large.

The impact of WRR on WSI areas has been analyzed previously. In order to ensure the robustness of the estimation results, re-estimation will be carried out considering alternative outcome measures, confounding policies, and subsample.

The benchmark model analyzes the influence of the WRR on the absolute quantity of land area for WSI, which is further estimated by replacing alternative variables to ensure robustness. In the following, we will further consider the size of the irrigated land area and the cultivated land area, and use the relative size of the land area for WSI to the irrigated area and the cultivated area.

A potential problem with the above estimates is that the impact of the WRR on the land area for WSI may come from other policies occurring at the same time. In China, the policy that is close to the WRR and may also affect the land area for WSI is the Agricultural Water-Saving Outline proposed in 2012. This policy is implemented to save agricultural irrigation water, mainly through subsidies to promote the land area for WSI. In order to avoid the estimation bias caused by the policy, according to Xu & Yang (2022), we try to avoid the impact of the contemporaneous policy by controlling the interaction between suitable arable land area and whether the policy occurs.

The previous section has analyzed the impact of WRR on the overall WSI area, but it is not clear which types of water-saving technologies have increased. Based on this, it is further discussed below which types of WSI technologies have been adopted. Since WSI technology plays an important role in agricultural production, the impact of WRR on agricultural production is further discussed below. Since this article is based on the analysis of provincial panel data, it is difficult to conduct further analysis of the internal mechanism of individual farmers adopting WSI technology, which needs to be further examined by other micro-level studies.

The results are consistent with the previous research by Mu et al. (2022), Zhang et al. (2021) and Fang & Zhang (2020). Mu et al. (2022) indicate that the WRR can effectively improve WSI technology. Fang & Zhang (2020) suggest that farmers tend to actively promote irrigation technology and reduce agricultural water use after the WRR. Zhang et al. (2021) also indicate that the main channels of water-saving effects from the policy are the promotion of agricultural technology innovation and the water rights transfer to high water-consuming industries. Based on this, what kind of WSI area will increase the most after the WRR deserves further discussion. Based on this, the impact of the WRR on WSI technologies will be further analyzed below. The Table 8 shows the results estimating the effect of the WRR on various kinds of WSI technologies.

As a comparison group, column 1 shows that the WRR can increase the land area for WSI by 13.63% on average, and it is significant at the level of 1%. The dependent variable in column 2 is the logarithmic form of the spray and drip irrigation area. Column 2 shows that the WRR can increase the spray and drip irrigation area by 43.11% on average, and it is also significant at the level of 1%. The dependent variable in column 3 is the logarithmic form of the micro-irrigation area. Column 3 shows that the WRR can increase the micro-irrigation area by 67.30% on average, and it is also significant at the level of 1%. The dependent variable in column 4 is the logarithmic form of the low-pressure tube irrigation area. Column 4 shows that the WRR can increase the low-pressure tube irrigation area by an average of 13.72%, but the estimated result is not significant. The dependent variable in column 5 is the logarithmic form of other WSI areas, including other non-mainstream WSI technologies. Column 5 shows that the WRR can reduce other WSI area by 2.98% on average, but the estimated result is not significant. To sum up, the WRR has mainly promoted the increase of spray and drip irrigation areas, and micro-irrigation land area, which are high-efficiency technologies for irrigation.

Gao et al. (2014) showed that WSI is an important technology to promote agricultural production. Rosa et al. (2018) indicate that a shift from rain-fed agriculture to irrigated agriculture will increase the yield of agriculture. The reform of the water rights system can promote the popularization of WSI technology, so whether it is beneficial to agricultural production deserves further discussion. The WRR can further guarantee agricultural production after promoting the expansion of the land area for WSI. However, whether the WRR can significantly promote agricultural production needs further testing with empirical data. Table 9 shows the results of estimating the effect of the WRR on agricultural production. Note that all the dependent variables of the model are in logarithmic form.

Column 1 indicates that the WRR can increase water use in agriculture by about 0.39%, but the estimated coefficient is not significant. Column 2 clarifies the impact of the WRR on the agricultural water use productivity (agricultural value added per unit of water used). The WRR can increase water productivity in agriculture at a significant level of 5%. Specifically, compared with other provinces, the agricultural water productivity of provinces with WRR can increase by an average of 5.66%. Although the impact of the WRR on agricultural water use is not significant, the WRR can promote the added value that can be realized per unit of water resources used in agricultural production. Column 3 estimates the impact of the WRR on agricultural value added. Column 3 shows that the WRR can significantly increase the added value of agriculture, specifically, the WRR can increase the added value of agriculture by an average of 6.05%. The promotion effect of the WRR on the added value of agriculture and the estimated results in columns 1–2 can be mutually confirmed. Column 4 clarifies the impact of the WRR on the sown area of crops. Column 4 shows that the WRR can increase the sown area of crops at a significant level of 1%, and its increasing effect can reach 7.47%. This shows that the possible reason why the WRR promotes the increase of agricultural added value is due to the expansion of crop sown area. Columns 5–6 clarify the impact of the WRR on the sown area of grain crops and the sown area of cash crops, respectively. Columns 5–6 show that the WRR not only promotes the expansion of the sown area of grain crops at the 1% significant level, but also significantly increases the sown area of cash crops. Specifically, the WRR increases the sown area of grain crops by 9.57% and the sown area of cash crops by 14.10%. This shows that the WRR has increased the sown area of cash crops more than the sown area of food crops, which indicates that the WRR is more beneficial to cash crops than grain crops.

This study takes China's WRR as an example, with China's panel data from 2004 to 2019, and adopts a DID model to empirically analyze the impact of WRR on the land area for WSI. The estimated results indicate that the WRR can increase the land area for WSI at a significant level of 1%. Specifically, compared with other provinces, the land area for WSI could increase by an average of 13.63% in the provinces with WRR, which is a relatively large increase. The event study indicates that participation in WRR does not significantly affect the land area for WSI before the policy occurs, and its impact starts to become significant in the fourth year after the policy occurs. The robust results could be obtained when alternative outcome measures are used, a contemporaneous policy is considered, and subsample regression is applied. Furthermore, the WRR could promote the expansion of spray and drip irrigation areas and micro-irrigation land areas, which are high-efficiency irrigation technologies. Specifically, the WRR could increase the spray and drip irrigation area by 43.11% on average and the WRR could increase the micro-irrigation area by 67.30% on average. In addition, the WRR also could improve agricultural production by increasing agricultural water productivity and planting area (including the sown area of grain crops and cash crops), but it has not reduced agricultural water extraction. Therefore, the WRR could increase agricultural production without increasing agricultural water extraction.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.