Do water Saving technologies save water: Empirical Evidence From North China

Qiuqiong Huang, Jinxia Wang & Yumin Li

Presented by

Nabittun Nahar

Washington State University

• Water is the most limiting factor in agricultural production

• More efficient irrigation is needed to lessen the pressure of water shortages on agricultural production

• China’s agriculture is threatened due to huge gap betweenwater demanded and available water

• China govt didn’t start technology until 1990

• Since 2000, water saving irrigation projects have been developed at a rapid pace.

• Both areas irrigated under water saving irrigation projects and their shares in China's total irrigated area have been increasing at a steady rate

Nabittun Nahar

Introduction

• In 2009, the State Council of China issued the document “National Water-saving Irrigation Planning,”

• By 2020 the area irrigated under water saving irrigation projects should reach 80% of the nation's effective irrigate area.

• The government planned to invest a total amount of about 150 billion yuan in water saving irrigation projects during the twelfth “Five-year” plan period (2011–2015).

• In January 2011, the government's top policy priorities was to set the goal of achieving an irrigation efficiency of 0.55 by 2020

• Evaluation of the effectiveness of water saving technologies is needed.

Govt. effort for Water Saving Technology

Evaluation of the WST

• Previous lit show that higher irrigation efficiency reduces the effective price of applied irrigation water, and the savings provide additional irrigation capacity.

• Not all water saving technologies can achieve their expected levels of water saving after adoption.

• Control for the factors such as farmers’ skills in implementing a technology and the production environment (e.g., soil characteristics).

• Therefore, the effects of more efficient irrigation technologies remain an empirical question

• Despite the importance of the issue few studies are available to examine the question with observational data (in contrast to simulation models).

• Most existing studies only used qualitative analysis (Zhang and Cai, 2001, Zhang et al., 2002, Deng et al., 2006).

• The few studies with quantitative analysis used data either from experiment fields or from only one or two pilot project sites (Wang and Li, 2005; Wang, 2007; Ding et al., 2009; Zhao and Wang, 2010).

• Studies without controlling for any confounding factors (e.g., Hu et al., 2002; Liu and Kang, 2006).

• Very few studies with data that cover large geographic areas to general broad policy implications.

Earlier Work

• Can water saving technologies save water?

• Does the productivity increase due to the WST?

• Is there any impact of WST on the sown area or irrigated sown area or crop mix?

• Objective: to provide estimates of the impacts of different types of water saving technologies on water use.

• The term water-saving is used in a loose sense in that “these technologies reduce the amount of applied water but do not necessarily generate real water saving (reduction in consumptive water use or evapotranspiration, Blanke et al., 2007)”.

Question(s)

• The authors have conducted a panel household survey in two of the most water scarce areas in north China: Yellow River Basin (YRB) and Hai River Basin (HRB).

• The sample areas of the survey cover Henan and Ningxia provinces in the YRB and Hebei province in the HRB.

• 77 sample villages were randomly selected from the three sample provinces.

• The first round of the survey in 2001 (collected information on years 1995 and 2001). The second round on year 2004 and the third round on year 2007.

• In the sample villages, the survey successfully tracked 118 households and 142 plots over the three panels of survey.

Data

Survey areaSource: Strokal, Maryna & Kroeze, C. & Wang, Mengru & Bai, Zhaohai & Ma, Lin. (2016). The MARINA model (Model to Assess River Inputs of Nutrients to seAs): Model description and results for China. The Science of the total environment. 562. 869-888. 10.1016/j.scitotenv.2016.04.071.

Data

• The survey has collected detailed information on a

owide range of water saving technologies used in the village,

o the characteristics of water resources and

o socio-economic characteristics of the villages and Farmers’

o input uses (e.g., water and fertilizer) in agricultural production at the plot level.

• Climate conditions: data from China Meteorological Administration

• To control heterogeneity : Use of HH level or plot level fixed effects (alpha) to control for any unobserved heterogeneity

• To address the potential endogeneity : IV used

Three types of Technologies:

Traditional

• Border irrigation, furrow irrigation and field leveling

• Not capital intensive

• Used by individual households

HH based

• Surface level plastic irrigation pipe. Drought resistant varieties, plastic sheeting and retaining stub/low till

• Not capital intensive

• Individual households

Community based

• Underground pipe systems, lined canals, sprinkler systems and drip irrigation systems

• Capital intensive

• Used by groups of households instead of individual households

Descriptive Statistics

Water use by Village (%)

Share of sown area using WST

Comparison of water use and crop allocation by WST

Econometric Analysis

𝑙𝑛 𝑊𝑖𝑘𝑚𝑡 = 𝛼𝑖𝑘𝑚 + 𝑇𝑖𝑘𝑚𝑡𝛽 + 𝑋𝑖𝑘𝑚𝑡𝛿 + 𝐶𝑡𝛾 + 𝜖𝑖𝑘𝑚𝑡 (1)

• Wikmt -amount of water per unit of land the kth household

applied on the ith plot to irrigate crop m in period t.

• Tikmt -the key variables of interest

• Xikmt -factors that may influence water use

• Ct -county-level annual total precipitation and average

temperatures

• Soil type and distance from farmers house are time invariant

Econometric Analysis

Plot level fixed effect

𝑙𝑛 𝑊𝑖𝑘𝑚𝑡−1 = 𝜶𝒊𝒌𝒎 + 𝑇𝑖𝑘𝑚𝑡−1𝛽 + 𝑋𝑖𝑘𝑚𝑡−1𝛿 + 𝐶𝑡−1𝛾 + 𝜖𝑖𝑘𝑚𝑡−1(1a)

𝑙𝑛 𝑊𝑖𝑘𝑚𝑡−1 − 𝑙𝑛 𝑊𝑖𝑘𝑚𝑡 = 𝑇𝑖𝑘𝑚𝑡𝛽 − 𝑇𝑖𝑘𝑚𝑡−1𝛽 +(𝑋𝑖𝑘𝑚𝑡 − 𝑋𝑖𝑘𝑚𝑡−1 )𝛿 + (𝐶𝑡−𝐶𝑡−1)𝛾 + 𝜖𝑖𝑘𝑚𝑡 − 𝜖𝑖𝑘𝑚𝑡−1 (1b)

Water productivity

𝑙𝑛 𝑄𝑖𝑘𝑚𝑡 = 𝛼′𝑖𝑘𝑚 + 𝑇𝑖𝑘𝑚𝑡𝛽′ + 𝑍𝑖𝑘𝑚𝑡𝛿′ + 𝐶𝑡𝛾′ + 𝜈𝑖𝑘𝑚𝑡 (2)

• Qikmt is the productivity of water in producing crop m on plot i of household k in period t. the unit is kg/m3

• The only difference is that the vector Zikmt contains one more variable than Xikmt

Issue with the model

• Attritioni) There is attrition of households which take off-farm work.• Solution: Included in 𝑋𝑖𝑘𝑚𝑡

ii) For households that stayed in the sample, only about half of their plots can be matched. • Reason: exchange of plots with other farmers in the

village, rent out plots, given to sons or daughters, confiscated for highway construction

• Solution: No selection problem found after conducting a test of attrition suggested by Verbeek and Nijman (1992)

• Sown area: 𝑆𝑘𝑚𝑡 = 𝛼′′𝑘𝑚 + 𝑇𝑘𝑡𝛽′′ + 𝑯𝒌𝒕𝜹′′ + 𝐶𝑡𝛾′′ + 𝜔𝑘𝑚𝑡

(3)

• Irrigated sown area 𝐼𝑆𝑘𝑚𝑡 = 𝛼′′𝑘𝑚 + 𝑇𝑘𝑡𝛽′′ + 𝐻𝑘𝑡𝛿′′ +

𝐶𝑡𝛾′′ + ꞃ𝑘𝑚𝑡 (4)

• Total irrigated sown area of HH K

𝑌𝑘𝑡 = 𝛼′′𝑘 + 𝑇𝑘𝑡𝛽′′ + 𝐻𝑘𝑡𝛿′′ + 𝐶𝑡𝛾′′ + 𝛹𝑘𝑡 (5)

• Hkt HH characteristics that may influence households’ responses at the extensive margins.

• Policy efforts by government to encourage the use of water saving technology as IV for water saving technology variables

• Good IV

Econometric Analysis

ResultsImpact of WST on water use

Calculation ( % of table 4)

Result

• Precipitation: The estimated coefficient on annual average precipitation is negative and statistically significant in corn.

• This coefficient is not statistically significant in wheat production, the growing season of wheat is outside the rainy season.

• There is some evidence that larger plots use less water per unit of land (negative and statistically significant for corn plots)

• As expected, in villages with more scarce water resource, water use is significantly lower.

Results

Results (Impact on crop mix)

Results (Impact on irrigated area)

Results (Impact on water productivity)

Calculation (% change in table 8)

Conclusion

• Report the extent of water saving technologies usage and evaluates the impacts of using various types of technologies on water use and crop allocations in the YRB and HRB.

• Three groups of technologies are used.

• Using water saving technologies can reduce crop water use and improve the productivity of water.

• No statistically significant impacts on crop mix, irrigated sown area or the share of a crop's sown area that is irrigated.

• The positive effects are generated mainly using household-based or community-based technologies.

Conclusion

• Future research on field level irrigation efficiencies associated with various water saving technologies would provide better estimates.

• Larger data sets should be collected to allow analysis of individual technologies.

• Results of this study is not final to conclude on the impact of water saving technologies.

• In the changing landscape of rural China, it is important to update the analysis to reflect any changes in policies or in the production environment.

Future Research

Thanks! Questions?