Farmers’ adoption of water-saving irrigation technology alleviates water scarcity in metropolis suburbs: A case study of Beijing, China

Agricultural Water Management 212 (2019) 349–357

Contents lists available at ScienceDirect

Agricultural Water Management journal homepage: www.elsevier.com/locate/agwat

Farmers’ adoption of water-saving irrigation technology alleviates water scarcity in metropolis suburbs: A case study of Beijing, China Biao Zhanga,b, Zetian Fua, Jieqiong Wanga, Lingxian Zhanga,c,

T

⁎

a

China Agricultural University, Beijing, 100083, China Fuyang Normal University, Fuyang, 236037, China c Beijing Laboratory of Food Quality and Safety, Beijing, 100083, China b

A R T I C LE I N FO

A B S T R A C T

Keywords: Water-saving irrigation technology Farmers Adoption Water scarcity Metropolis suburbs China

Water scarcity has threatened the food security and been a critical concern in China. Promoting modern agricultural irrigation technologies has been identified as an important measure against water scarcity. The overall goal of this study was to analyze the adoption of water-saving irrigation technology by farmers and to identify the major influencing factors of this decision for metropolis suburbs. Based on a field survey of Beijing of China, the results showed that 53.1% of farmers adopted water-saving irrigation technologies to cope with water scarcity, most of which adopted engineering water-saving technologies. The number of adopted water-saving irrigation technologies followed a strong negative correlation with the share of adopters. Econometric analysis revealed that education, farm size, on-farm demonstration, cooperative, training, groundwater, access to information, water use associations, drought-prone area, neighboring farmers, and policy subsidies significantly improved the adaption to water scarcity. Age, production specialization, and cost posed a negative effect on famers’ adoption of water-saving irrigation technologies. These results and implications provide an understanding of farmers’ sustainable irrigation practices and offer an insight to influencing factors to frame improved strategies and policies that enable to cope with water scarcity of metropolis suburbs.

1. Introduction Water scarcity and drought have been of critical concern, and pose severe threats to both food security and economy throughout many parts of the world (Alam, 2015; Chen et al., 2014). About 40% of the global population lives in regions where water resources are over-allocated due to scarcity and competition (Wheeler et al., 2015). Water shortage in China is very severe, particularly in the north and northwest of the country (Zhang et al., 2013). With water scarcity becoming an increasing constraint for food production in northern China, the pressure to procure sufficient agricultural water is growing, and has been the most important factor to threaten China's food security (Chen et al., 2014; Yang et al., 2003). In light of increasingly severe water scarcity conditions, appropriate measures were taken to mitigate its impacts for China. Despite an improvement of water use efficiency in the agricultural sector from about 40% in 2006 to nearly 50% in 2013, it is very low compared to the 70% to 90% reached by most developed countries (Deng et al., 2006; Zhu et al., 2013). In addition, future climate change will further require irrigators to improve irrigation water-use efficiency, thus enabling a

⁎

decreased reliance on water as input (Wallace, 2000; Wheeler et al., 2015). More widespread use of water-saving irrigation is one central way to promote water conservation in irrigated areas (Nikouei et al., 2012; Peck et al., 2004). In the water-saving irrigation extension system, how to apply watersaving technologies is key for the production practice to achieve better economic, social, and environmental benefits. This has aroused concern on how to promote the adoption of water-saving irrigation by producers. For instance, Bjornlund et al. (2009) indicated that the major drivers for the adoption of improved irrigation technologies and management practices were to ensure the security of the water supply during droughts, to increase quantity and quality of crops, and to decrease the costs involved. Alam (2015) reported that farmers with more experience of farming, better schooling, more secure tenure rights, better access to electricity and institutional facilities, and an awareness of climatic effects are more likely to embrace adaptation strategies to avoid water scarcity. Cremades et al. (2015) emphasized that governmental subsidies and extension service policies have played an important role in promoting the adoption of modern irrigation technology.

Corresponding author at: China Agricultural University, Beijing, 100083, China. E-mail addresses: [email protected] (B. Zhang), [email protected], [email protected] (L. Zhang).

https://doi.org/10.1016/j.agwat.2018.09.021 Received 11 January 2018; Accepted 10 September 2018 0378-3774/ © 2018 Elsevier B.V. All rights reserved.

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

use efficiency are part of water-saving irrigation. Fig. 1 shows that agricultural irrigation water, from water sources to final crop use, has undergone roughly three stages. During the first state, the water was drawn from the water source and moved to the field through canals. During this process, there will be leakage, seepage, and further means of water loss. During the second state, the water is drawn from the field to the crop root system to facilitate water absorption by the crops. This is also the traditional irrigation concept, including traditional irrigation (e.g., flood irrigation and furrow irrigation) and modern irrigation (e.g., drip irrigation and micro sprinkler irrigation). The final state, accompanies the whole process of crop growth, where water from the roots of the crop is transported throughout the entire plant, which is also the ultimate use of water resources for crop growth. The core of watersaving irrigation lies in minimizing water loss during these three processes while improving the efficiency of water resources utilization. Various definitions for the adaption options against water scarcity have been classified. To make these easy to understand for farmers, Chen et al. (2014) classified the adaptation options against drought as types of either engineering (e.g., investment or maintenance options) or non-engineering (e.g., technological, regulatory, or market options). Alam (2015) defined adaptation as the increased use of surface water, increased use of ground water, crop diversification, and calendar adjustment, all of which are changed land use strategies to cope with drought. Pereira et al. (2002) summarized adaptation methods to drought, including supplemental irrigation, deficit irrigation, improved irrigation methods and performance, distribution uniformity, and various soil and water conservation practices. Since we collected information on whether farmers have actually adopted WSIT, our analysis investigated adaptation practices of adopting WSIT rather than of potential adaptation options and other adaptation measures. Based on theoretical analysis and personal experience during a field survey, we classified the WSIT as types of engineering water-saving technology (EWST), agricultural water-saving technology (AWST), biological and chemical water-saving technology (BCWST), and managerial water-saving technology (MWST) (Fig. 2). EWST is mainly accomplished through the construction of a variety of water-saving irrigation projects that reduce water consumption, including the delivery of water, the distribution of water, the irrigation of water, and other processes of water leakage and evaporation to improve the utilization of irrigation water (Chen et al., 2014; Wang et al., 2002). AWST is mainly accomplished through specific measures that control the field of water and optimize the production structure while improving water-use efficiency (Jiménez and Chávez, 2003; Zou et al., 2013). BCWST is a realtime and reasonable irrigation system that promotes crop growth with biological and chemical technology. It also involves the use of biotechnology for breeding and cultivating of crop varieties that are drought-resistant and water saving (Wang et al., 2007). MWST can achieve a reasonable allocation of irrigation water resources and an optimal scheduling of irrigation measures to reach a specific target, which is deriving the maximum benefit from limited water resources (Gebrehiwot and van der Veen, 2013; Naranjo-Gil, 2017; Reilly et al., 2003).

However, the majority of adoption studies to date focused on rural areas and major grain producing areas, while studies specifically focusing on agriculture in metropolis suburbs are rare (Abdulai et al., 2011; Alam, 2015; Cai and Rosegrant, 2004; Mostafa and Fujimoto, 2014; Wang et al., 2014, 2015; Zhang et al., 2013). Agriculture is an essential component of the sustainable development of metropolis. It plays an important role to ensure steady farmers’ income and thus promote rural economic development. The important function of agriculture is to serve the metropolis, generating as stable supply of safe agricultural products, improving the ecological environment, as well as enhancing leisure tourism and cultural education for citizens. With the enormous population and limited area of metropolis, increasing demand for water from urban and industrial sectors exert increased pressure on agricultural water (Nikouei et al., 2012; Zhang et al., 2013). However, farmers in metropolis suburbs are different from those of other areas, benefiting from multi channels of revenue sources, multi functions of agricultural production, higher production requirements, and increased pressure on environmental resources. This leads to differences in farmers’ attitudes and adoption behaviors of water-saving irrigation technology (WSIT) in metropolis suburbs and rural areas. Studies are required that investigate metropolis suburbs in light of water scarcity to increase our understanding of farmers’ adoption behaviors and influencing factors to ultimately promote a sustainable development of agriculture. Beijing, located on the northern North China plain of China, is a metropolis with more than 21.7 million people (in 2015), making it one of the world’s most water-scarce cities (NBSC, 2016). In this area, water resources per capita (124.0 m3) are far below the national average (of 2039.2 m3), which is only 1/4 of the world’s average (Cheng et al., 2015; NBSC, 2016). Therefore, the objective of this study was to provide an understanding of farmers’ sustainable irrigation practices used to cope with water-stress in water scarcity environments of Beijing, China. Furthermore, identifying major factors that influence farmers’ adoption of WSIT in the study area is a further important research aim. The results obtained in the present study provide insights that could help goverments to understand barriers and drivers of farmers’ adoption of WSIT. The remainder of this paper is arranged as follows: the next section provides a brief overview and classification of WSIT. Section 3 considers theoretical arguments, while also discussing the conceptual framework and hypothesis. Section 4 presents the empirical model and data. Section 5 discusses empirical results. The final section provides conclusions and policy implications. 2. Water-saving irrigation technology For crop irrigation, ideal water efficiency refers to reducing losses caused by evaporation, runoff, or subsurface drainage while increasing production (Yang, 2012). Water-saving irrigation is required to reduce agricultural water waste while timely and fully meeting the requirements of crop water. Using WSIT not only saves water and increases production, it also improves the nutrition of agricultural products ensuring food safety by improving the environment (Abdulai et al., 2011; Cremades et al., 2015; Jalota et al., 2009; Wang et al., 2002). Water-saving includes the entire irrigation process. All measures, techniques, and methods for reducing water loss and improving water

3. Conceptual framework and hypothesis According to the theory of innovation diffusion (Rogers, 2003), the

Fig. 1. The process of crop irrigation. 350

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

Fig. 2. Classification and illustrations of WSIT.

et al., 2012; Wang et al., 2015). In this study, we developed a conceptual framework, suggesting that farmers’ adoption of WSIT is a result of common influence of householder characteristics, family characteristics, farm characteristics, production conditions, perceptions of technology, and environmental factors. The conceptual framework is shown in Fig. 3.

3.1. Householder characteristics Some characteristics of the surveyed households tend to influence the adoption of irrigation technologies (Abdulai et al., 2011). The probability for the adoption of irrigation technologies declines with increasing age (Abdulai and Huffman, 2005; Wang et al., 2015). The education level of farmers played an important role in the adoption of modern irrigation technology (Alam, 2015; Cremades et al., 2015; Fernandez-Cornejo et al., 2005; Saha and Schwart, 1994). Better-educated farmers are often in a better position to understand and appreciate the advantages of adopting irrigation technologies. This finding is consistent with the notion that education is important in helping farmers to reach decisions about adopting new innovations and technologies (Abdulai and Huffman, 2005; Abdulai et al., 2011). Farming experience affects farmers’ decision-making behavior (Hisali et al., 2011; Seekao and Pharino, 2016). Alam (2015) suggested that the greater the experiences in agricultural farming, the more likely farmers would be to adopt alternative adaptation strategies. Based on this, we

Fig. 3. The conceptual framework.

innovation-decision process is an information-seeking and informationprocessing activity, with five steps in the farmers’ adoption process, including knowledge of an adoption, persuasion to form an attitude, decision, implementation, and confirmation (either reinforcement or rejection). Based on the theory of planned behavior, farmers’ adoption behaviors have been shown to be related to multiple factors (Ajzen, 1991). Previous studies reported that an economic incentive, social capital, natural environment, technical characteristics, family and farm characteristics, and other factors significantly affected farmers' adoptions of agricultural technologies (Abdulai et al., 2011; Alam, 2015; Bjornlund et al., 2009; Chen et al., 2014; Cremades et al., 2015; Nikouei 351

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

technologies (Feder et al., 1985). Greater distance between the farm and the nearest market indicates poor access, which will constrain adoption (Mariano et al., 2012). The two main sources of agricultural water (ground and surface water) are regulated separately (Yoo et al., 2013). The water source has a significant influence on the adoption of WSIT for farmers (Green et al., 1996). Previous studies confirmed that farmers exclusively using groundwater are more likely to adopt modern irrigation technology (Alam, 2015; Caswell and Zilberman, 1985; Cremades et al., 2015; Yang et al., 2003). Therefore, we hypothesized that:

hypothesized that: H1. Householder characteristics are related to farmers’ adoption of WSIT 3.2. Family characteristics The family is the basic unit for farmers to engage in agricultural production. Household size and production specialization are the main family characteristics. As pointed out by previous studies, farmers’ adoption decisions are influenced by household size (Abdulai et al., 2011; Bekele and Drake, 2003; Mariano et al., 2012). In contrast, Chen et al. (2014) and Wang et al. (2015) found that the likelihood for an adoption behavior of farmers is not significantly related to household size. Thus, we assume that a larger household size will lead to a better chance WSIT adoption. Moreover, Adoption of agricultural innovations is also linked to off-farm income through managerial time (FernandezCornejo et al., 2005). Deressa et al. (2009) indicated that both the farm income and nonfarm income have a significant relationship with the adoption behavior. Furthermore, total income was found to be related to the adoption decision of agriculture (Goswami et al., 2012). To address this concern, the production specialization was used as a household income index that refers to the proportion of farm income to the total income. In line with this suggestion, we hypothesized that:

H4. Production conditions will have a significant effect on farmers’ adoption of WSIT

3.5. Perceptions of technology Information is the foundation of any adoption. Farmers generally adopt agricultural technologies only if they fully understand the provided technical information (Negatu and Parikh, 1999). In particular, when the respective technologies involve different practices, information is critical for their adoption (Abdulai et al., 2011). Clear associations exist between the access to technical information and the level of adoption of agricultural technology (Wang et al., 2015). Moreover, an adoption of modern irrigation technology occurs when incremental net benefits of the modern irrigation technology exceed its cost (Dridi and Khanna, 2005). Farmers’ subjective perceptions of the costs of adoption are key factors to adoption (Bandiera and Rasul, 2006). Consequently, the following hypothesis has been derived:

H2. Family characteristics will affect farmers’ adoption of WSIT 3.3. Farm characteristics In general, farm size is based on the total cultivable acres in operation, both owned and rented (Mariano et al., 2012). Abdulai et al. (2011) noted that farm size, which normally indicates wealth, was frequently used in adoption models to capture the impact of wealth on the decision process. Studies on adoption of agricultural technologies indicate that farm size exerts both negative and positive effects on adoption, indicating that the effect of farm size on technology adoption is inconclusive (Bradshaw et al., 2004; Deressa et al., 2009). The studies confirm that farm size influences farmers’ adoption of novel agricultural technologies (Alam, 2015; Gebrehiwot and van der Veen, 2013; Goswami et al., 2012). On-farm easy, informal demonstrations are the most effective extension conditions, which allow the farmers to enhance their own autodidactic approach (Badrage et al., 2009). Previous studies show on-farm demonstration to be an effective way to transfer technology by improving farmers’ adoption of agricultural technologies (Barao, 1992). Participating in a cooperative helps farmers to enter social networks for the flow of information, which tends to enhance the adoption of new agricultural technologies (Bandiera and Rasul, 2006). Abdulai et al. (2011) and Mariano et al. (2012) suggested that being a member of a farmer's cooperatives increases the probability to adopt new agricultural technologies. Based on these publications, we hypothesized that:

H5. Perceptions of technology will have a significant effect on farmers’ adoption of WSIT

3.6. Environmental factors In general, water use associations (WUAs) are voluntary, non-governmental, nonprofit, and have been established and are managed by a group of irrigators. WUAs have become the dominant form of irrigation in several areas (Wang et al., 2014). Zhang et al. (2013) examined which WUAs play a significant role in promoting water productivity among households that belong to a WUA in northern China. Previous studies found that adoption of irrigation technology is positively and significantly related to the natural and geographical environments (Cremades et al., 2015; Schuck et al., 2005). When the resource is scarce, the technology needed to save the resource is more likely to be adopted (Ruttan and Hayami, 1984). Correspondingly, farmers are more willing to adopt WSIT in water-scarcity regions (Cai and Rosegrant, 2004; Caswell and Zilberman, 1986; Schuck et al., 2005; Shrestha and Gopalakrishnan, 1993). Moreover, between neighboring farmers, it is important to model extension and diffusion of agricultural technology (Jørs et al., 2016). Islam et al. (2013) reported that neighboring farmers’ adoption influenced farmers’ adoption of agricultural technology. Finally, policy subsidies cut the cost of technology and provide important incentives for farmers to adopt WSIT (Bjornlund et al., 2009; Blanke et al., 2007; Cremades et al., 2015; Dinar et al., 1992; Nikouei et al., 2012; Wheeler et al., 2015). More importantly, when subsidies are available for farmers, they are more likely to adopt irrigation technologies (Cremades et al., 2015). Therefore, we assumed that:

H3. Farm characteristics will have a significant effect on farmers’ adoption of WSIT 3.4. Production conditions Training seems to be an effective tool to improve knowledge (Cunha et al., 2014). Attendance of agricultural technology training encourages adoption since exposure to information reduces subjective uncertainty about the technology (Mariano et al., 2012). Based on this, we assumed the attendance by farmers at training sessions had a positive impact on the adoption of WSIT. Market was identified as an important determinant of adoption since markets also served as a platform for farmers to exchange information (Gebrehiwot and van der Veen, 2013). Market access is an important factor to affect adoption of agricultural

H6. Environmental factors will have a significant effect on farmers’ adoption of WSIT

352

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

Table 1 Definition and descriptive statistics of variables of the adoption model. Variable Householder characteristics Age Education Farming experience Family characteristics Household size Production specialization Farm characteristics Farm size On-farm demonstration Cooperative Production conditions Training Distance to nearest market Groundwater Perceptions of technology Access to information Cost Environmental factors WUAs Drought-prone area Neighboring farmers Policy subsidies

Definition

Mean

S.D.

Years Years of formal education by the farmer Years

49.84 9.68 14.71

7.08 3.22 9.41

Number of family members Percentage of farm income to total income

3.96 0.52

1.23 0.41

Total area planted in hectares 1 if the farmer participated in on-farm demonstration; 0 otherwise 1 if the farmer is a member of a farmer’s cooperative; 0 otherwise

0.32 0.34 0.49

0.69 0.48 0.51

1 if the farmer attended agricultural technology training; 0 otherwise Distance of the farm to the nearest market in kilometers 1 if the farmer use groundwater as main source of agricultural water; 0 otherwise

0.88 5.49 0.88

0.95 3.92 0.55

1 if the farmer obtains information resources easily and accurately; 0 otherwise 4-point scale was used to represent farmer’s perception of input cost for adoption of WSIT ranging from low (1) to high (4)；

0.37 3.36

0.48 0.84

1 1 1 1

0.18 0.65 0.65 0.29

0.38 0.69 0.48 0.46

if if if if

the farmer is a member of WUAs; 0 otherwise the farmer’s field is prone to drought; 0 otherwise some neighboring farmers adopt WSIT; 0 otherwise the farmer access to policy subsidies of WSIT; 0 otherwise

Δpi ∂p | all other x consant = i Δx i ∂x i

4. Empirical model and data sources

(2)

4.1. Empirical model Most of the technology choices that farmers consider for their decision-making are of a ‘use or do not use’ nature (Mariano et al., 2012). In this study, responding farmers were asked whether they used WSIT or not. The advantage of the binary logit choice model was used to analyze the choice between two categories and the model is widely used in the adoption research (Abdulai et al., 2011; Chen et al., 2014; Cremades et al., 2015; Goswami et al., 2012; Mariano et al., 2012). Therefore, each farmer’s choice is represented by a dummy variable:

4.2. Data and definition of variables To meet the objectives of this research and study the behavior of farmers regarding their adoption of WSIT, as well as the factors that influenced this behavior, we designed a structured questionnaire with several aspects. First, we collected basic characteristics about farmers, their family, and their farms (e.g., age, education, income, and farm size). Second, data about the adoption of WSIT was to be collected as the focus of our data collection. Furthermore, we also recorded other information, such as training, access to information, cost, and policy subsidies. The data used in this study were collected from one field survey via face-to-face interview in Beijing during 2015. It was conducted in the top seven crop-producing districts of Beijing, including Daxing, Changping, Fangshan, Huairou, Shunyi, Miyun, and Tongzhou. After approval by the Beijing Municipal Bureau of Agriculture, each district elected one outstanding agricultural technology extension worker as investigator who was then trained on the questionnaire content and specific considerations from us. Then, stratified random sampling was used in field surveys. First, villages were chosen from each district via a system sampling technique based on spatial location, and their geographical position was dispersed. Then, we chose several farmers in each selected village via simple random sampling method. A total of 490 valid samples originated from 121 villages of seven districts. The data were widely distributed; therefore, we considered the data to be representative of the overall population of the entire study area. In the remainder of this paper, both the descriptive statistical analysis and the econometric analysis refer to this sample of 490 farmers. The definitions and descriptive statistics of explanatory variables used for the binary model are provided in Table 1. On average, farmers were 49.84 years of age, had 9.68 years of formal schooling, 14.71 years of farm experience, and approximately four household members. For production specialization, most farmers had other jobs and farm income formed only part of their total family income. The mean farm size was 0.32 ha. The proportion of farmers that belonged to demonstration farmers was 34.4% and 48.6% farmers had joined cooperatives. The rate of farmer attendance to training was 88.0%, and the mean

1 if the farmer adopts WSIT yi = ⎧ ⎨ ⎩ 0 if the farmer does not adopt WSIT To better quantify the influences of different factors on farmer decisions to adopt WSIT, based on our framework, we constructed a binary model as follows:

ln(

pi )=α+ 1 − pi

n

∑ βn xni n=1

(1)

where the dependent variable pi represents the probability of technology adoption, α represents the intercept parameter, β represents the vector of regression coefficients and x = (x1,…, xn) denotes a vector of n independent variables (e.g., householder, family, and farm characteristics). When the dependent variable is also the categorical variable, an ordinary least squares (OLS) estimation method is not effective due to the nonlinearity of the relationship. Therefore, a maximum likelihood estimation (MLE) method is required for the logit regression model (Mzoughi, 2011; Pampel, 2000). Importantly, the estimated parameters are not directly tied to the actual magnitude of change but only provide the direction of the effect of independent variables in the logit regression model (Deressa et al., 2009; Zhang et al., 2017). We determined the marginal effects to analyze the level of contribution of the factors that influence farmers’ adoption of WSIT (Alam, 2015; Wang et al., 2015). The marginal effect was defined as the effect of a unit change in x (while all other factors remained constant) on the probability that a farmer chose to adopt WSIT (Mariano et al., 2012). It can be shown that: 353

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

Table 2 Farmers’ adoption of WSIT against water scarcity.

Total samples WSIT EWST AWST BCWST MWST

Table 3 Farmers’ adoption of specific WSIT.

Number

Share of total samples (%)

Share of total adopters (%)

490 260 206 136 81 96

100 53.1 42.0 27.8 16.5 19.6

— 100.0 79.2 52.3 31.2 36.9

EWST

distance between farm and the nearest market was 5.49 km. However, perceptions of technology were poor, such as access to information was difficult and connected to high input cost. Only 17.8% of farmers have been members of WUAs, and 29.1% had access to policy subsidies. Most fields are prone to drought, but had better adoption environment from neighboring farmers.

AWST

5. Results and discussion

BCWST

5.1. Farmers’ adoption of water-saving irrigation technologies MWST

The field surveys showed that half of the farmers took water-saving action in response to drought. For example, of the 260 farmers who faced drought during the study period, 53.1% adopted WSIT (Table 2). Of those farmers who were able to apply WSIT in response to water scarcity, further analyses indicate that 79.2% adopted EWST. This is understandable, as the water-saving effect of EWST is remarkable, the facilities and equipment can be reused for a long period of time, and maintenance is simple. In addition, 52.3% of these farmers adopted AWST to make full use of the provided advantages, including simple operation and lower cost. Operating the equipment needed for MWST and BCWST is more difficult and requires a higher level of professionalism, which is difficult for many farmers to master. This leads to the relatively low adoption of both technologies. For example, the share of farmers adopting BCWST and MWST accounted for 31.2% and 36.9%, respectively. Further analysis revealed that six technologies were adopted at more than 25% in the survey of 23 specific types of WSIT (Table 3). These are drip irrigation technology (41.9%), field coverage technology (39.6%), field cultivation technology (36.9%), reasonable irrigation system (28.5%), on-film irrigation technology (28.1%), and drought resistant varieties (26.5%). The highest rate of adoption of a specific technology in each of the four types of WSIT was: drip irrigation technology (EWST), field coverage technology (AWST), drought resistant and water saving varieties (BCWST), and reasonable irrigation system (MWST). It is worth noting that the adoption of 10 types of WSIT remained below 10%, and three types remained below 5%. Examples of some of these minimally-adopted technologies include chemical reagent water saving, using impervious membranes and concrete anti-seepage technology. According to our understanding (based on our survey) of farmers' adoption of WSIT, adoption of a technology occurs when the technology has reached a high degree of maturity, is easy to operate, causes low cost, and offers obvious benefits. In contrast, when the technology is high cost, complicated to operate, requires a high degree of professional knowledge, and does not offer an immediate benefit, farmers are not likely to adopt them. However, the water-saving effect is a result of the interaction of a number of WSIT. The results show that a strong negative correlation existed between the number of WSIT adopted and the share of adopters (Fig. 4). The share of farmers who adopted two technologies in combination was maximal, accounting for 20.8%, and 20.4% of adopters who had only used single WSIT. When the farmers used a small amount of WSIT, the effect was better. With the increasing number of adopted technologies, the benefits of WSIT gradually decreased while the cost of

Specific WSIT

Number

Share of total adopters (%)

Infiltration irrigation Subsurface drip irrigation Surge irrigation On-film irrigation Sprinkler irrigation Drip irrigation Micro sprinkler irrigation Concrete anti -seepage Channel lining stone Impervious membrane High-pressure water transmission pipeline Low-pressure water delivery pipeline Field coverage technology Field cultivation technology Irrigating-sowing Water-fertilizer coupling technique Chemical reagent water saving Drought-resistant varieties Insufficient irrigation Soil moisture monitoring Determination of crop water requirement Reasonable irrigation system Measuring irrigation

19 41 17 73 29 109 46 7 22 10 41

7.3 15.8 6.5 28.1 11.2 41.9 17.7 2.7 8.5 3.8 15.8

52

20.0

103 96 23 44 3 69 24 14 28

39.6 36.9 8.8 16.9 1.2 26.5 9.2 5.4 10.8

74 25

28.5 9.6

Fig. 4. Distribution of total WSIT adopted by farmers.

investment increased. This led to a reduction in overall efficiency, which caused farmers to abandon the increase of other WSIT. 5.2. Estimation results of the econometric model We used STATA 14.0 software to estimate the logit model and marginal effects. The results are shown in Table 4. The chi-squared test statistic was significant at the 1% level, implying joint significance of the WSIT adoption variables (Mariano et al., 2012). The pseudo R2 values of the logit model were 0.363, which is a reasonably high value for a multivariate analysis based on cross-sectional data (Wang et al., 2015). Importantly, 14 of the 18 selected variables were significant at a 10% level of significance in the logit model. Therefore, the logit model developed in this study was effective and explained the farmers’ adoption of WSIT. 5.3. Householder characteristics As expected, older farmers had a lower probability of WSIT adoption. For instance, a unit increase in age of the householder resulted in a 1.1% decline in the probability of adoption of WSIT. Education is one of the key determinants for adopting WSIT. Results of Table 4 indicate that education appeared to have a positive effect on the adoption of WSIT. 354

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

The positive effect of farm size for the adoption is consistent with the findings of this study. This is probably because larger farms contribute to highlight a number of shortcomings of traditional irrigation, such as time-consumption and labor intensity, and waste of water resources, which drives farmers to pay more attention to save resources and reduce costs via WSIT adoption. The on-farm demonstration variable shows a significantly positive sign in the adoption equation, indicating that farmers who participated in on-farm demonstrations were more likely to adopt WSIT. More importantly, the marginal effects revealed that if farmers had participated in on-farm demonstration, their likelihood of adoption of WSIT increased by 31.5%. Similarly, results in Table 4 indicate a positive and significant correlation between being a member of cooperatives and the likelihood of adoption of WSIT to cope with water scarcity. Marginal effects indicated that if a farmer participated in a cooperative, the likelihood of adoption of WSIT increased by 31.1%. These results imply that on-farm demonstration and cooperative tremendously promotes the adoption and diffusion of WSIT.

Table 4 Estimated coefficients and marginal effects of the logit model. Variable

Constant Householder characteristics Age Education Farming experience Family characteristics Household size Production specialization Farm characteristics Farm size On-farm demonstration Cooperative Production conditions Training Distance to nearest market Groundwater Perceptions of technology Access to information Cost Environmental factors WUAs Drought-prone area Neighboring farmers Policy subsidies Log likelihood Number of obs LR chi2 Prob > chi2 Pseudo R2

Coefficient estimates

Marginal effects

Coef.

Std error

Coef.

Std error

−10.931***

3.994

—

—

−0.085* 0.003** −0.039

0.045 0.096 0.036

−0.011* 0.004** −0.005

0.006 0.013 0.005

−0.071 −0.104**

0.218 0.902

−0.009 −0.014**

0.029 0.118

0.154** 2.407*** 0.238***

0.082 0.831 0.577

0.020** 0.315*** 0.311***

0.010 0.096 0.075

3.545** 0.041 0.928***

1.238 0.087 0.925

0.264** 0.005 0.121***

0.144 0.011 0.119

0.606* −0.675***

0.559 0.360

0.079* −0.088***

0.072 0.045

1.839*** 0.168* 1.434** 1.491*** −52.553 490 59.92 0.000 0.363

0.938 0.241 0.620 0.709

0.240*** 0.022* 0.188** 0.195***

0.116 0.031 0.075 0.087

5.6. Production conditions As predicted by hypothesis H4, some production conditions correlate with farmers’ decisions to adopt WSIT. For example, results show that attendance at training sessions had a significant positive influence on farmers’ WSIT adoption probability. Marginal effects indicated that if farmers received training on agricultural technology, their probability of adoption of WSIT increased by 26.4%. Moreover, the estimated results also showed that the groundwater variable were significantly positive at the 1% level on farmers’ adoption of WSIT. Marginal effects indicated that compared to other farmers, the likelihood of WSIT adoption by farmers with groundwater for irrigation increased by 12.1%.

Note: *, ** and *** are significant at 10%, 5%, 1% significant level, respectively.

5.7. Perceptions of technology

According to the calculated results on the marginal effects, a one-unit (year) increase of schooling led to a 0.4% increase in the probability of adoption while the effect on the remaining options was negligible. As suggested by Mariano et al. (2012), farmers who are better educated have greater ability to process information and search for technologies that are suitable for their production constraints compared to farmers who are less educated. These results reinforced the findings of other studies showing that farmer characteristics were significantly related to the choice of adoption (Alam, 2015; Mzoughi, 2011; Torkamani and Shajari, 2007; Wang et al., 2015).

The perceptions of technology significantly influenced the farmers’ adoption of WSIT. The coefficient for the access to information was significant and positive, suggesting that farmers who have easily access to information of technology were more likely to adopt WSIT. This result confirms the importance of information for the farmers’ adoption of agricultural technologies (Mariano et al., 2012). Marginal effects suggested that the access to information increased the likelihood of WSIT adoption by 7.9%. In contrast, the results of this analysis reconfirmed that increasing cost significantly decreased the likelihood of WSIT adoption. The probability of WSIT adoption was 8.8% lower when the cost increased by one unit. Based on farmers’ small-scale production, high cost improved the barrier of adoption with respect to capital input. Mzoughi (2011) suggested that farmers are more likely to adopt technology, which reduces production costs. This result implies that it is important to reduce the cost of a farmer’s initial investment for WSIT adoption to effectively promote a technology’s popularization and application.

5.4. Family characteristics The coefficient of the production specialization variable was negative and statistically significant for farmer’s adoption of WSIT. Marginal effects indicated that each additional unit in production specialization reduced the probability of WSIT adoption by 1.4%. These results indicate that farmers with higher production specialization were less likely to have the ability to withstand risks arising from the adoption of WSIT. This finding suggests that reducing risk is an important pathway for promoting the adoption of WSIT by farmers (Barham et al., 2015). For adoption of WSIT, increasing household size did not significantly reduce the probability, although the coefficient was negative.

5.8. Environmental factors As predicted by hypothesis H6, the results show that environmental factors significantly influenced farmers’ WSIT adoption. First, WUAs posed a significant positive influence on farmers’ adoption of WSIT. For example, a farmer who is a member of WUAs is 24.0% more likely to adopt WSIT. Second, results indicate that farmers who were exposed to drought were encouraged to use WSIT. Technology adoption in such adverse areas is higher by 2.2%. Ongoing water scarcity in combination with future anticipated drought conditions, increased the likelihood of farmers who use WSIT to appreciate water resources and improve irrigation efficiency. In addition, the variable of neighboring farmers had a significantly positive impact at the 1% level regarding WSIT adoption.

5.5. Farm characteristics As for control variables, results show that farm size had a positive and significant impact on the adoption of WSIT. Marginal effects indicated that a unit (ha) of increase in farm size increased the probability of farmers’ WSIT adoption by 2%. As noted by Filho et al. (1999) and Wang et al. (2015), farmers with larger farms are more likely to adopt new agricultural technologies compared to those with smaller farms. 355

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

Overall, the aims of this study have been achieved. However, it is also worth noting that our study only focused on farmers’ adoption of WSIT and factors influencing whether they adopted WSIT against water scarcity. Factors influencing farmers’ adoption of different types and number of WSIT are also important and provide interesting research topics for future studies. Finally, we also believe that the results and implications of this study are important for the extension of WSIT under conditions of water scarcity and drought of other metropolis across the world.

The results for marginal effects indicated that if neighboring farmers adopted WSIT, the probability of farmer’s adoption increased by 18.8%. Finally, we also noted the importance of policy subsidies to farmers’ decisions to adopt WSIT. Based on estimation results, the coefficient of policy subsidies is significantly positive in the model, implying that there is a positive relationship between policy subsidies and the adoption of WSIT. In other words, if farmers had access to policy subsidies, the possibility that they will adopt WSIT increases by 19.5%. Based on these findings, it can be concluded that improving the environments farmers face can help to promote WSIT adoption.

Acknowledgement 6. Conclusions and policy implications This work was supported by the Beijing Social Science Foundation (16YJA007) and Beijing Leafy Vegetables Innovation Team of Modern Agro-industry Technology Research System (BAIC07-2018).

Based on field surveys conducted in Beijing (China), this study examined farmers’ practices when facing water scarcity and identified the major factors that affect farmers’ decisions on whether or not to adopt WSIT to alleviate water scarcity. The results show that, when facing water scarcity, about half of the farmers (53.1%) in Beijing adopted WSIT. For those farmers, the results of the analysis indicate that 79.2%, 52.3%, 31.2%, and 36.9% adopted EWST, AWST, MWST, and BCWST, respectively. It is also worth noting that the results show a strong negative correlation between the number of WSIT adopted and the share of adopters. Further analysis from the utilized econometric model indicates that farmers’ adoption decisions on WSIT are a result of common influence of householder characteristics, family characteristics, farm characteristics, production conditions, perceptions of technology, and environmental factors. In detail, education, farm size, participation in on-farm demonstration, cooperative, training, groundwater, access to information, WUAs, drought-prone area, neighboring farmers, and policy subsidies significantly improved farmers’ adaption to water scarcity, while age, production specialization, and cost posed a negative effect on famers’ adoption of WSIT. The results of this study have several potential policy implications. First, there is significant room for the government to promote WSIT extension against water scarcity. Based on farmers’ production demand, differential promotion measures will be conducted to help farmers adopt WSIT for the sustainable development of agriculture in metropolis suburbs. Second, improving performance, increasing the stability, and reducing the cost of WSIT is a way to help farmers reduce adoption barriers. For example, farmers were more likely to adopt WSIT if the technology had a high degree of maturity and was easy to operate. According to field surveys and interviews, farmers encountered many WSIT problems, such as quickly aging equipment and ease of breakdown, having a short service life, being ineffectual, and being complex to operate. Third, the results of this analysis reconfirm that increasing cost significantly decreases the likelihood of WSIT adoption. Results indicate that 54.4% of farmers thought the costs of WSIT were high. This finding suggests that reducing cost is important to drive farmers to adopt WSIT. Fourth, the results indicated a positive relationship between policy subsidies and the adoption of WSIT. Therefore, in future, local governments may need to continue to expand their policy of supporting farmers to adopt WSIT against water scarcity, such as subsidies, extension policies, and technical services. Lastly, the study also reveals that improving farmers’ social capital and environments for the adoption of WSIT can play an important role to help farmers against water scarcity. For example, policies that aimed at promoting the adoption of WSIT need to emphasize the critical role of providing timely and accurate information on improved techniques and water scarcity. Then, the government should encourage and guide farmers to take part in WUAs, on-farm demonstrations, and cooperatives through incentives and demonstration measures. Moreover, local governments should also offer education and training programs to help farmers improve their understanding of the necessity to improve water-use efficiency and to increase their capacity to use innovative practices or technologies.

References Abdulai, A., Huffman, W.E., 2005. The diffusion of new agricultural technologies: the case of crossbred-cow technology in Tanzania. Am. J. Agric. Econ. 87, 645–659. Abdulai, A., Owusu, V., Bakang, J.E.A., 2011. Adoption of safer irrigation technologies and cropping patterns: evidence from Southern Ghana. Ecol. Econ. 70, 1415–1423. Ajzen, I., 1991. The theory of planned behavior. Organ. Behav. Hum. Decis. Processes 50, 179–211. Alam, K., 2015. Farmers’ adaptation to water scarcity in drought-prone environments: a case study of Rajshahi District, Bangladesh. Agric. Water Manag. 148, 196–206. Badrage, A., Dinpanah, G., Aboeye, F., Sinaki, M., 2009. An investigation of effective factors involved in perception of wheat farmers regarding on-farm demonstration in Isfahan Township, Iran. Am.-Eur. J. Agric. Environ. Sci. 6 (5), 616–624. Bandiera, O., Rasul, I., 2006. Social networks and technology adoption in northern Mozambique. Econ. J. 116, 869–902. Barao, S.M., 1992. Behavioral aspects of technology adoption: the role of on-farm demonstration. J. Ext. 30, 13–15. Barham, B.L., Chavas, J.P., Fitz, D., Ríos-Salas, V., Schechter, L., 2015. Risk, learning, and technology adoption. Agricultural Economics. 46, 11–24. Bekele, W., Drake, L., 2003. Soil and water conservation decision behavior of subsistence farmers in the Eastern highlands of Ethiopia: a case study of the Hunde-Lafto area. Ecol Eng. 46, 437–451. Bjornlund, H., Nicol, L., Klein, K.K., 2009. The adoption of improved irrigation technology and management practices: a study of two irrigation districts in Alberta, Canada. Agric. Water Manag. 96, 121–131. Blanke, A., Rozelle, S., Lohmar, B., Wang, J., Huang, J., 2007. Water saving technology and saving water in China. Agric. Water Manag. 87, 139–150. Bradshaw, B., Dolan, H., Smit, B., 2004. Farm-level adaptation to climatic variability and change: crop diversification in the Canadian Prairies. Clim. Change 67, 119–141. Cai, X., Rosegrant, M.W., 2004. Irrigation technology choices under hydrologic uncertainty: a case study from Maipo River Basin, Chile. Water Resour. Res. 40, 1149–1155. Caswell, M., Zilberman, D., 1985. The choices of irrigation technologies in California. Am. J. Agric. Econ. 67, 224–234. Caswell, M.F., Zilberman, D., 1986. The effects of well depth and land quality on the choice of irrigation technology. Am. J. Agric. Econ. 68, 798–811. Chen, H., Wang, J., Huang, J., 2014. Policy support, social capital, and farmers’ adaptation to drought in China. Global Environ. Change 24, 193–202. Cheng, K., Fu, Q., Li, T., Jiang, Q., Liu, W., 2015. Regional food security risk assessment under the coordinated development of water resources. Nat. Hazards 78, 603–619. Cremades, R., Wang, J., Morris, J., 2015. Policies, economic incentives and the adoption of modern irrigation technology in China. Earth Syst. Dyn. 6, 399–410. Cunha, D.T., Stedefeldt, E., de Rosso, V.V., 2014. The role of theoretical food safety training on Brazilian food handlers’ knowledge, attitude and practice. Food Control 43, 167–174. Deng, X.P., Shan, L., Zhang, H., Turner, N.C., 2006. Improving agricultural water use efficiency in arid and semiarid areas of China. Agric. Water Manag. 80, 23–40. Deressa, T.T., Hassan, R.M., Ringler, C., Alemu, T., Yesuf, M., 2009. Determinants of farmers’ choice of adaptation methods to climate change in the Nile Basin of Ethiopia. Global Environ. Change 19, 248–255. Dinar, A., Campbell, M.B., Zilberman, D., 1992. Adoption of improved irrigation and drainage reduction technologies under limiting environmental conditions. Environ. Resour. Econ. 2, 373–398. Dridi, C., Khanna, M., 2005. Irrigation technology adoption and gains from water trading under asymmetric information. Am. J. Agric. Econ. 87, 289–301. Feder, G., Just, R.E., Zilberman, D., 1985. Adoption of agricultural innovations in developing countries: a survey. Econ. Dev. Cult. Change 33, 255–298. Fernandez-Cornejo, J., Hendricks, C., Mishra, A., 2005. Technology adoption and off-farm household income: the case of herbicide-tolerant soybeans. J. Agric. Appl. Econ. 37, 549–563. Filho, H.M.D.S., Young, T., Burton, M.P., 1999. Factors influencing the adoption of sustainable agricultural technologies: evidence from the State of Espı́rito Santo. Braz. Technol. Forecast. Soc. Change 60, 97–112. Gebrehiwot, T., van der Veen, A., 2013. Farm level adaptation to climate change: the case

356

Agricultural Water Management 212 (2019) 349–357

B. Zhang et al.

Ruttan, V.W., Hayami, Y., 1984. Toward a theory of induced institutional innovation. J. Dev. Stud. 20, 203–223. Saha, A., Schwart, R., 1994. Adoption of emerging technologies under output uncertainty. Am. J. Agric. Econ. 76, 836–846. Schuck, E.C., Frasier, W.M., Webb, R.S., Ellingson, L.J., Umberger, W.J., 2005. Adoption of more technically efficient irrigation systems as a drought response. Int. J. Water Resour. Dev. 21, 651–662. Seekao, C., Pharino, C., 2016. Key factors affecting the flood vulnerability and adaptation of the shrimp farming sector in Thailand. Int. J. Disaster Risk Reduct. 17, 161–172. Shrestha, R.B., Gopalakrishnan, C., 1993. Adoption and diffusion of drip irrigation technology: an econometric analysis. Econ. Dev. Cult. Change 41, 407–418. Torkamani, J., Shajari, S., 2007. Adoption of new irrigation technology under production risk. Water Resour. Manag. 22, 229–237. Wallace, J.S., 2000. Increasing agricultural water use efficiency to meet future food production. Agric. Ecosyst. Environ. 82, 105–119. Wang, H., Liu, C., Zhang, L., 2002. Water-saving agriculture in China: an overview. Adv. Agron. 75, 135–171. Wang, Y., Chen, X., Xiang, C.-B., 2007. Stomatal density and bio-water saving. J. Integr. Plant Biol. 49, 1435–1444. Wang, J., Huang, J., Zhang, L., Huang, Q., 2014. Do incentives still matter for the reform of irrigation management in the Yellow River Basin in China? J. Hydrol. 517, 584–594. Wang, J., Klein, K.K., Bjornlund, H., Zhang, L., Zhang, W., 2015. Adoption of improved irrigation scheduling methods in Alberta: an empirical analysis. Canadian Water Resources J. 40, 47–61. Wheeler, S.A., Zuo, A., Loch, A., 2015. Watering the farm: comparing organic and conventional irrigation water use in the Murray–Darling Basin, Australia. Ecol Eng. 112, 78–85. Yang, C., 2012. Technologies to improve water management for rice cultivation to cope with climate change. Crop Environ. Bioinform. 9, 193–207. Yang, H., Zhang, X., Zehnder, A.J.B., 2003. Water scarcity, pricing mechanism and institutional reform in northern China irrigated agriculture. Agric. Water Manag. 61, 143–161. Yoo, J., Simonit, S., Connors, J.P., Maliszewski, P.J., Kinzig, A.P., Perrings, C., 2013. The value of agricultural water rights in agricultural properties in the path of development. Ecol Eng. 91, 57–68. Zhang, L., Heerink, N., Dries, L., Shi, X., 2013. Water users associations and irrigation water productivity in northern China. Ecol. Econ. 95, 128–136. Zhang, B., Fu, Z., Wang, J., Tang, X., Zhao, Y., Zhang, L., 2017. Effect of householder characteristics, production, sales and safety awareness on farmers’ choice of vegetable marketing channels in Beijing, China. Br. Food J. 19, 1216–1231. Zhu, X., Li, Y., Li, M., Pan, Y., Shi, P., 2013. Agricultural irrigation in China. J. Soil Water Conserv. 68, 147–154. Zou, X., Li, Ye., Cremades, R., Gao, Q., Wan, Y., Qin, X., 2013. Cost-effectiveness analysis of water-saving irrigation technologies based on climate change response: a case study of China. Agric. Water Manag. 129, 9–20.

of farmer’s in the Ethiopian highlands. Environ. Manage. 52, 29–44. Goswami, K., Choudhury, H.K., Saikia, J., 2012. Factors influencing farmers’ adoption of slash and burn agriculture in North East India. For. Policy Econ. 15, 146–151. Green, G., Sunding, D., Zilberman, D., Parker, D., 1996. Explaining irrigation technology choices: a microparameter approach. Am. J. Agric. Econ. 78, 1064–1072. Hisali, E., Birungi, P., Buyinza, F., 2011. Adaptation to climate change in Uganda: evidence from micro level data. Global Environ. Change 21, 1245–1261. Islam, Q.M.S., Miah, M.A.M., Rahman, M.S., Hossain, M.S., 2013. Adoption of bari mung varieties and its constraints to higher production in southern region of Bangladesh. Bangladesh J. Agric. Res. 38, 85–96. Jalota, S.K., Singh, K.B., Chahal, G.B.S., Gupta, R.K., Chakraborty, S., Sood, A., Ray, S.S., Panigrahy, S., 2009. Integrated effect of transplanting date, cultivar and irrigation on yield, water saving and water productivity of rice (Oryza sativa L.) in Indian Punjab: field and simulation study. Agric. Water Manag. 96, 1096–1104. Jiménez, B., Chávez, A., 2003. Low cost technology for reliable use of Mexico city’s wastewater for agricultural irrigation. Technology 9, 95–108. Jørs, E., Konradsen, F., Huici, O., Morant, R.C., Volk, J., Lander, F., 2016. Impact of training bolivian farmers on integrated pest management and diffusion of knowledge to neighboring farmers. J. Agromed. 21, 200–208. Mariano, M.J., Villano, R., Fleming, E., 2012. Factors influencing farmers’ adoption of modern rice technologies and good management practices in the Philippines. Agric Syst. 110, 41–53. Mostafa, H., Fujimoto, N., 2014. Water saving scenarios for effective irrigation management in Egyptian rice cultivation. Ecol Eng. 70, 11–15. Mzoughi, N., 2011. Farmers adoption of integrated crop protection and organic farming: do moral and social concerns matter? Ecol Eng. 70, 1536–1545. Naranjo-Gil, D., 2017. Accomplishing water strategy policies in hospitals: the role of management information systems and managerial styles. Water 9, 107. NBSC, 2016. China Statistical Yearbook. China Statistics Press, Beijing. Negatu, W., Parikh, A., 1999. The impact of perception and other factors on the adoption of agricultural technology in the Moret and Jiru Woreda (district) of Ethiopia. Agric. Econ. Agric. Econ. 21, 205–216. Nikouei, A., Zibaei, M., Ward, F.A., 2012. Incentives to adopt irrigation water saving measures for wetlands preservation: an integrated basin scale analysis. J. Hydrol. 464–465, 216–232. Pampel, F.C., 2000. Logistic Regression : A Primer. Sage University Papers. Peck, D.E., McLeod, D.M., Hewlett, J.P., Lovvorn, J.R., 2004. Irrigation-dependent wetlands versus instream flow enhancement: economics of water transfers from agriculture to wildlife uses. Environ. Manag. 34, 842–855. Pereira, L.S., Oweis, T., Zairi, A., 2002. Irrigation management under water scarcity. Agric. Water Manag. 57, 175–206. Reilly, J., Tubiello, F., McCarl, B., Abler, D., Darwin, R., Fuglie, K., Hollinger, S., Izaurralde, C., Jagtap, S., Jones, J., Mearns, L., Ojima, D., Paul, E., Paustian, K., Riha, S., Rosenberg, N., Rosenzweig, C., 2003. U.S. Agriculture and climate change: new results. Clim. Change 57, 43–67. Rogers, E., 2003. Diffusion of Innovations. Free Press, New York.

357