**Meet the editors**

Dr. Teruaki Nanseki is a professor of Kyushu University, Japan. His research interests cover management, informatics and operational research in agriculture. Management of risk, information and talent is the major research topic in his ongoing national research projects. He received B.S. in Agronomy and M.S. in Farm Management from Okayama University in 1980 and 1982

respectively. He was a visiting fellow at Cornell University from 1980 to 1981 and received Ph.D. in Agricultural Economics from Kyoto University in 1990. He has published more than 300 publications in scholarly journals, books, conference proceedings and academic reports. He has received seven academic prizes from the academic societies.

Dr. Min Song is a professor of Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, P. R. China. He mainly engages in researches on resources and environment economics, intellectual property rights and public policy in agriculture. He received B.S. and M.S. in Agricultural Economics from Southwest Agricultural University of China

in 1985 and 1993 respectively. He was conferred Ph.D. in Agricultural & Resource Economics from Kyushu University in 2002. He has published more than 100 publications in scholarly journals and books, and received many prizes for the supervision of research projects, and offering policy suggestions etc.

### Contents




Index 165

## Preface

Is the 21st century going to be Asia's century, especially of China? The world is greatly concerned by this question, which is difficult to answer. According to Sommer (2006) [1], "*A power shift from the West to the East is in the offing. A new world order is taking shape. The 21st century is still young. The Americans are hoping for a second American century. But perhaps – given the rise of Asia, especially China – they will have to settle for an American half century. Only one thing is certain: the second awakening of China will bring about a power shift in the world.*" This prediction of China's rise is the subject of this book.

When we ponder the answer to the question, we need to consider important facts stated in Sommer (2006) [1]: "*Asia was the birthplace of the three Abrahamic religions. Epoch-making inventions and discoveries had reached the West from there… but this was only one side of the coin. Simultaneously, Asia was seen as static, inert, immobile… the Asia-Pacific region has become a model of modern development dynamics. Japan started first.*"

Is the environment an important issue in Asia? Kyushu University established the Research Institute for East Asia Environments (RIEAE) in 2009. The president of the universitystated the reasons for establishing the institute in Arikawa (2009) [2]: *"Currently, the entire world, especially the East Asian region, is facing serious environmental damage. This damage involves various issues including air pollution, river and marine pollution, waste problems, and food contamination—issues too numerous to innumerate in full—and which is far beyond an individual researcher's capability to address. Integrating relevant members of the university as an organization through the cooperation of relevant research institutes as well as the support of private companies, RIEAE is conducting a range of activities with an eye on making a social contribution at the national level to practically solve these environmental issues, which have become complicated."* 

What is the contribution of this book? The main idea of this book is *Kanshokufuji*— ⎔㣗 in Japanese and ⧟伏нҼ in Chinese. This new concept proposed in Nanseki (2008) [3] pertains to a sound food system in a sound agro-environment. It implies that food supply and demand are non-separable in terms of location due to the importance of retaining the suitable condition of the environment and biodiversity in the region. This concept is necessary and useful for coping with environmental issues and related food safety issues.

#### XII Preface

Food science, considered as a branch of agricultural science in its broad meaning, has made valuable contributions to food safety. However, in the real agro-food chain, agricultural products may be contaminated by environmental pollution, to which agricultural production may contribute. This dilemma implies that contemporary science is still missing the point of an agro-environment for crops and livestock in that agriculture production including agrochemical utilization, food processing and marketing, food consumption, and residue processing need to be viewed as a whole system. In this sense, the concept of *kanshokufuji* is also useful and needed in environmental science to prevent environmental damage as well as enhance food safety.

From this view point, several integrated surveys in both rural and urban areas of China were conducted by the food risk research group at the RIEAE to reveal the current status of the environment, food, and agriculture in this country. The results of the surveys are introduced in this book, along with implications and recommendations. This book presents the beginning of our research, and we are happy to receive comments from our readers. Finally, the authors gratefully acknowledge the financial support of the RIEAE, Kyushu University.

> **Teruaki Nanseki**  Faculty of Agriculture, Kyushu University, Japan

**Min Song** Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, P. R. China

#### **References**


#### **Chapter 1**

### **Introduction**

Min Song and Teruaki Nanseki

#### **1.1 Problem statement**

The Industrial Revolution which began in Britain in the mid-1700s, and spread to the rest of the world around the mid-1800s, not only brought the rapid development of the industrial economy, but also led to the expansion of numerous environmental hazards. Nevertheless, the negative effects of the Industrial Revolution on the environment were not revealed until 1962 in the globally acclaimed book, *Silent Sprin*g, written by Rachel Carson. In this book, she took on the environmental and human dangers caused by indiscriminate use of pesticides. "Over increasingly large areas of the United States spring now comes unheralded by the return of birds, and the early mornings are strangely silent where once they were filled with the beauty of bird song" (from *Silent Spring*). The agro-environment is an important part of the natural environment and the basic material condition for agricultural production. Agro-environment degradation includes ecological destruction and environmental pollution, and the latter is the theme researched in this book. In addition to constraining the sustainable development of agriculture, agro-environmental deterioration also increases risks in food through material recycling. Nowadays, food safety and the agro-environment have become challenges around the world.

China, as a large agricultural country, has a 7,000-year history of sericulture and a 6,400-year history of rice farming. Over a long time, Chinese farmers have engaged in environmentfriendly agricultural production, which has had significant impact on present organic agriȬ culture. However, due to food shortages, China introduced industrial agriculture in the 1960s to improve crop yields. Over the last few decades, with a rising industry and economy, and promoted by incentive policies, chemical products have been increasingly put into agriculture in China (Figure 1.1).

Nowadays, China is the largest user of fertilizer, pesticides and plastic film in the world [2]. In 2010, Chinese agriculture consumed 55.61 million tons of chemical fertilizer, 1.75 million tons of pesticide and 2.17 million tons of plastic film [3-4], which is much higher than the world average. As nearly 60%-70% of chemical fertilizers, 60% of pesticides and 40% of plastic film

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Song and Nanseki; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Song and Nanseki; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

increasingly put into agriculture in China (Figure 1.1).

The Industrial Revolution which began in Britain in the mid-1700s, and spread to the rest of the world around the mid-1800s, not only brought the rapid development of the industrial economy, but also led to the expansion of numerous environmental hazards. Nevertheless, the negative effects of the Industrial Revolution on the environment were not revealed until 1962 in the globally acclaimed book, *Silent Sprin*g, written by Rachel Carson. In this book, she took on the environmental and human dangers caused by indiscriminate use of pesticides. "Over increasingly large areas of the United States spring now comes unheralded by the return of birds, and the early mornings are strangely silent where once they were filled with the beauty of bird song" (from *Silent Spring*). The agro-environment is an important part of the natural environment and the basic material condition for agricultural production. Agro-environment degradation includes ecological destruction and environmental pollution, and the latter is the theme researched in this book. In addition to constraining the sustainable development of agriculture, agro-environmental deterioration also increases risks in food through material recycling. Nowadays, food safety and the agro-environment have become challenges

China, as a large agricultural country, has a 7,000-year history of sericulture and a 6,400-year history of rice farming. Over a long time, Chinese farmers have engaged in environment-friendly agricultural production, which has had significant impact on present

Over the last few decades, with a rising industry and economy, and promoted by incentive policies, chemical products have been

Figure 1. Figure 1.1 Use of fertilizer, pesticides and plastic material in China Source: China Ministry of Agriculture (2012) [1]

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Source: China Ministry of Agriculture (2012) [1] **Figure 1.1** Use of fertilizer, pesticides and plastic material in China

DUMMY TEXT, KOJI NIJE IME KNJIGE

**Chapter — Introduction** 

Min Song and Teruaki Nanseki

**1.1. Problem statement** 

around the world.

gets in the environment [5-6], excessive use of chemical products is bound to have a notable influence on the ecological environment. In addition, with the development of animal husbandry, animal manure has become another source of agro-environment pollution. According to the data of the 1st national census on pollution sources in China (2010), agriculȬ tural production has seriously polluted the air, soil and especially the water environment. Agricultural sources provided 40% of chemical oxygen demand (COD), making a higher contribution than industrial and domestic sources. Simultaneously, agriculture is the major source of emissions of total nitrogen and total phosphorus, accounting for 57.2% and 67.4% of the total emissions, respectively. Agro-environment pollution has become an urgent challenge in China. Nowadays, China is the largest user of fertilizer, pesticides and plastic film in the world [2]. In 2010, Chinese agriculture consumed 55.61 million tons of chemical fertilizer, 1.75 million tons of pesticide and 2.17 million tons of plastic film [3-4], which is much higher than the world average. As nearly 60%-70% of chemical fertilizers, 60% of pesticides and 40% of plastic film gets in the environment [5-6], excessive use of chemical products is bound to have a notable influence on the ecological environment. In addition, with the development of animal husbandry, animal manure has become another source of agro-environment pollution. According to the data of the 1st national census on pollution sources in China (2010), agricultural production has seriously polluted the air, soil and especially the water environment. Agricultural sources provided 40% of chemical oxygen demand (COD), making a higher contribution than industrial and domestic sources. Simultaneously, agriculture is the major source of emissions of total nitrogen and total phosphorus, accounting for 57.2% and 67.4% of the total emissions, respectively. Agro-environment pollution has become an urgent challenge in China.

As is known to all, an unhealthy agro-environment cannot supply safe crops. With the degradation of the agro-environment in China, a series of food safety incidents broke out in recent years, such as the Hainan drug cowpea incident of 2011, the Qingdao drug leak incident of 2010 and the Guangdong drug watermelon incident of 2007. But agro-environmental pollution is not the only reason, abuse of addictives, microbe contamination and illegal processing operations also gave rise to successive food safety crises in China, e.g., excessive melamine in milk powder, salted duck egg containing Sudan Red and clenbuterol poisoning. As is known to all, an unhealthy agro-environment cannot supply safe crops. With the degradation of the agro-environment in China, a series of food safety incidents broke out in recent years, such as the Hainan drug cowpea incident of 2011, the Qingdao

The food safety crises, which have been listed in the top 10 of issues concerning peoples' livelihoods in China since 2005, have brought about a series of severe consequences as follows: first of all, significant outbreaks of food borne disease fundamentally undermine public trust. According to the Report on Chinese Food Safety in 2011-2012 released by the Chinese magazine *Well-off* and Tsinghua University, it was reported that 63.7% of the respondents believed that food safety in China is bad and 80.4% of the respondents thought food is not safe at all in China. Secondly, successive food safety incidents have a negative impact on food exports, weakening the international competitiveness and market reputation of food made in China. The massive media exposure of these food safety incidents has greatly reduced foreign consumers' confidence in food made in China. A survey carried out in Korea shows that nearly 90% of respondents think food imported from China is not safe [7], resulting in the international market raising barriers to limit food imports from China. For example, a number of countries took strong inspection measures on related products made in China after media exposure of a poison capsule event. Thirdly, food safety accidents often result in significant economic losses. Take the melamine incident as an example, it led to the bankruptcy of Sanlu milk enterprise, and at the same time, the entire milk industry suffered a big shock. During the incident, vast amounts of milk were poured away and a great number of dairy cows were slaughtered.

The object of the book is to reveal the challenges of the agro-environment and food safety in China from different perspectives, and try to find some solutions by analysing vast amounts of data gathered by large-scale surveys.

#### **1.2 Literature review**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

gets in the environment [5-6], excessive use of chemical products is bound to have a notable influence on the ecological environment. In addition, with the development of animal husbandry, animal manure has become another source of agro-environment pollution. According to the data of the 1st national census on pollution sources in China (2010), agriculȬ tural production has seriously polluted the air, soil and especially the water environment. Agricultural sources provided 40% of chemical oxygen demand (COD), making a higher contribution than industrial and domestic sources. Simultaneously, agriculture is the major source of emissions of total nitrogen and total phosphorus, accounting for 57.2% and 67.4% of the total emissions, respectively. Agro-environment pollution has become an urgent challenge

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

use of fertilizer use of pesticide use of plastic material

Figure 1. Figure 1.1 Use of fertilizer, pesticides and plastic material in China

Source: China Ministry of Agriculture (2012) [1]

**Figure 1.1** Use of fertilizer, pesticides and plastic material in China

has become an urgent challenge in China.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The Industrial Revolution which began in Britain in the mid-1700s, and spread to the rest of the world around the mid-1800s, not only brought the rapid development of the industrial economy, but also led to the expansion of numerous environmental hazards. Nevertheless, the negative effects of the Industrial Revolution on the environment were not revealed until 1962 in the globally acclaimed book, *Silent Sprin*g, written by Rachel Carson. In this book, she took on the environmental and human dangers caused by indiscriminate use of pesticides. "Over increasingly large areas of the United States spring now comes unheralded by the return of birds, and the early mornings are strangely silent where once they were filled with the beauty of bird song" (from *Silent Spring*). The agro-environment is an important part of the natural environment and the basic material condition for agricultural production. Agro-environment degradation includes ecological destruction and environmental pollution, and the latter is the theme researched in this book. In addition to constraining the sustainable development of agriculture, agro-environmental deterioration also increases risks in food through material recycling. Nowadays, food safety and the agro-environment have become challenges

China, as a large agricultural country, has a 7,000-year history of sericulture and a 6,400-year history of rice farming. Over a long time, Chinese farmers have engaged in environment-friendly agricultural production, which has had significant impact on present organic agriculture. However, due to food shortages, China introduced industrial agriculture in the 1960s to improve crop yields. Over the last few decades, with a rising industry and economy, and promoted by incentive policies, chemical products have been

DUMMY TEXT, KOJI NIJE IME KNJIGE

**Chapter — Introduction** 

Min Song and Teruaki Nanseki

**1.1. Problem statement** 

around the world.

0

Source: China Ministry of Agriculture (2012) [1]

1000000

2000000

3000000

4000000

5000000

6000000

unit: ton

increasingly put into agriculture in China (Figure 1.1).

2 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

As is known to all, an unhealthy agro-environment cannot supply safe crops. With the degradation of the agro-environment in China, a series of food safety incidents broke out in recent years, such as the Hainan drug cowpea incident of 2011, the Qingdao drug leak incident of 2010 and the Guangdong drug watermelon incident of 2007. But agro-environmental pollution is not the only reason, abuse of addictives, microbe contamination and illegal processing operations also gave rise to successive food safety crises in China, e.g., excessive melamine in milk powder, salted duck egg containing Sudan Red and clenbuterol poisoning.

The food safety crises, which have been listed in the top 10 of issues concerning peoples' livelihoods in China since 2005, have brought about a series of severe consequences as follows: first of all, significant outbreaks of food borne disease fundamentally undermine public trust. According to the Report on Chinese Food Safety in 2011-2012 released by the Chinese magazine *Well-off* and Tsinghua University, it was reported that 63.7% of the respondents believed that

in China.

Nowadays, China is the largest user of fertilizer, pesticides and plastic film in the world [2]. In 2010, Chinese agriculture consumed 55.61 million tons of chemical fertilizer, 1.75 million tons of pesticide and 2.17 million tons of plastic film [3-4], which is much higher than the world average. As nearly 60%-70% of chemical fertilizers, 60% of pesticides and 40% of plastic film gets in the environment [5-6], excessive use of chemical products is bound to have a notable influence on the ecological environment. In addition, with the development of animal husbandry, animal manure has become another source of agro-environment pollution. According to the data of the 1st national census on pollution sources in China (2010), agricultural production has seriously polluted the air, soil and especially the water environment. Agricultural sources provided 40% of chemical oxygen demand (COD), making a higher contribution than industrial and domestic sources. Simultaneously, agriculture is the major source of emissions of total nitrogen and total phosphorus, accounting for 57.2% and 67.4% of the total emissions, respectively. Agro-environment pollution As is known to all, an unhealthy agro-environment cannot supply safe crops. With the degradation of the agro-environment in China, a series of food safety incidents broke out in recent years, such as the Hainan drug cowpea incident of 2011, the Qingdao The agro-environment plays a decisive role in developing sustainable agriculture and providing safe food. With increasing industrialization, urbanization and agricultural modȬ ernization in China, excessive chemical materials are being widely applied to agricultural production [8]. Thus, the agro-environment is getting worse, which has hampered the development of sustainable agricultural. Economist first focused on the environment in 1920 when Pigou analysed pollution problems in his book on welfare economics. He pointed out that externality was the root of pollution and sewage enterprises should be taxed. For a long time "externality theory" was the overwhelmingly mainstream theory in environmental science. Later, with the development of institutional economics and information economics, economic theories about the environment were greatly promoted. Domestic scholars studied the agro-environment mainly based on the natural mechanisms, and from the technical and engineering perspective. Only a few researchers analysed the agro-environment from the view of economics, which could be divided into following: (1) the microeconomics perspectives. some studies revealed that externality resulted in agro-environmental problems. On the one hand, it is difficult to define the property rights of agro-environment, which will cause "market failure". On the other hand, a lack of incentive and restraint mechanisms with regard to agroenvironmental protection will lead to "government failure" [9]. Other scholars proved that in addition to the lack of protection systems, incentive and restraint mechanisms were the root of agro-environment problems through game analysis [10-11]. (2) Institutional economics perspective. Institutional economists considered that the dual control system had resulted in the generation and degradation of rural non-point source pollution [12]. (3) Farmers' behaviour analysis. Zhu (2000) [13] carried a survey in rural Beijing and the results showed that farmers lacked knowledge, awareness and motivation to environmentally protect. Zhu et al (2009) [14] studied farmers' perceptions on the environment through a survey in Hunan province and authors found that awareness among farmers in various regions is correlative with the region's development, and farmers' awareness is not in accordance with their behaviour. According to the results of a survey in Hunan province, Yan (2011) [15] pointed to the fact that farmers' awareness of protecting the environment was increasing.

Researching on food safety using economic methods began in the 1960s. During this period, economists brought forward a series of models to study realities in economic theory. FortunateȬ ly, food safety is one of the subjects investigated [16]. In the late 1980s, due to the occurrence of mad-cow disease (BSE), increasing numbers of people began to pay attention to food safety issues. In the 1990s, following the publication of *Economics of Food Safety* at the proceedings of the American National Workshop on the same title held in Washington, a series of reports were published,thus layingthefoundationforeconomicanalysisoffoodsafety.Now,moreandmore economists are studying food safety issues using empirical analysis methods, especially in the study of consumer behaviours, awareness and willingness to pay for safe food. For example, Chern et al (2002) [17] studied consumers' willingness to pay for genetically modified vegetaȬ bleoils throughcarryingout a surveyinShikoku,Japan,Norway,TaiwanandtheUnitedStates. Georges et al(2006)[18] selectedobjects from 12European countries anddividedthem into four groups to understand their perception on the food traceability system by group discussion.

The vast majority of Chinese economic research on domestic food safety issues uses foreign theories and research methods. Wang (2003) [19] carried out a survey of 289 consumers in Tianjin and analysed the process and characteristics of selecting safe food. Chern et al (2002) [17] revealed that consumers surveyed in Zhejiang province were relatively concerned about vegetable safety and their attitudes were found to be negative. Consumers were very willing to pay the extra cost for safe vegetables, but the price of the safe vegetables should be higher than ordinary vegetables by no more than 10% to 20%. Zeng et al (2008) [20] studied consumers' willingness to pay for moon cake with safe additives through a survey of 396 consumers in 25 supermarkets in Beijing.

Research on the agro-environment and food safety respectively from the perspective of farmers and consumers has increased in recent years. Most researchers only analyse and expound survey data, however, in-depth and comprehensive analysis, and related solutions are scant.

#### **1.3 Theoretical framework and methodology**

#### **1.3.1** *Kanshokufuji***: The theoretical basis**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The authors of this book were driven to develop the contexts based on the important concept of *Kanshokufuji*, a Japanese term created by Prof. Nanseki Teruaki, one of the chief editors of this book. With the literal meaning of *integrated environment and food*, this concept advocates a sound food system in a sound agro-environment [23]. In modern society, food safety is increasingly becoming of global concern, due to asymmetric information on the processes, additives in the long industrial chain, etc [21-22]. In this key concept for establishing a safe and sustainable next

generationfoodsystem,foodisdefinedusingabroaderconceptandrecognizedasbeingsupplied via more systematic processes. In addition to processed products ready for eat and drink, food includes the rawmaterialswithplant or animal origins,i.e., anysubstance consumedtoprovide nutritional support for the human. Meanwhile, the processes start from the growing condition of the agricultural, livestock and marine production, and cover the following processing and circulations of food. The processes also include eating and drinking, intake and metabolism of food, management of the residues and wastes. Based on the expanded and consistent concepts offoodandsupplyprocesses,thesafetyandrisksoffoodneedtobestudiedfromvariousaspects, including food, agriculture and environment [23].

As shown in Figure 1.2, the concept of *Kanshokufuji* aims to establish a safe and sustainable food system, through demonstrating the key notion that a sound food system can only begin from a sound agro-system. Once the soil and water are contaminated by human activities, including agriculture, it is difficult to conduct safe agricultural and livestock production. In the process of agricultural and livestock production, pesticides and veterinary drugs should by applied properly. Nevertheless, safe production of food cannot be maintained if the living environment of the agricultural and livestock products is polluted by cadmium and other heavy metals or poisonous chemicals contained in industrial liquid waste, despite endeavours in the proper application of the means of production. Thus, with this awareness in mind, the assurance of food safety and environmental protection is interpreted as integrated in, or inseparable from, *Kanshokufuji*. In other words, it calls on the technical advances and instituȬ tional designs to prevent pollution of the living environment and food stuffs. Meanwhile, there is another important concept of *Ishokudogen* which targets the risks which occur after food enters the body. Based on the latest advances in studies on food functions and inner metaboȬ lism, this Japanese term argues that medicine and food are both important to maintain human health, and they are from the same source.

Hence together with *Ishokudogen*, *Kanshokufuji* provides a key notion of establishing a food system covering different generations. Considering the risks of food before entering the body, safety of soil and water constitutes the prerequisite of the overall food safety system. MeanȬ while, proper eating and drinking habits are essential in maintaining health. Thereafter, when designing the cross-generation food system, it is of great importance to reorganize the integral process starting from the living environment of agricultural and livestock products, followed by production, processing, circulation, eating habits, intake and metabolism, and the disposal of the residues and wastes.

#### **1.3.2 Data sources**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

studied farmers' perceptions on the environment through a survey in Hunan province and authors found that awareness among farmers in various regions is correlative with the region's development, and farmers' awareness is not in accordance with their behaviour. According to the results of a survey in Hunan province, Yan (2011) [15] pointed to the fact that farmers'

4 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Researching on food safety using economic methods began in the 1960s. During this period, economists brought forward a series of models to study realities in economic theory. FortunateȬ ly, food safety is one of the subjects investigated [16]. In the late 1980s, due to the occurrence of mad-cow disease (BSE), increasing numbers of people began to pay attention to food safety issues. In the 1990s, following the publication of *Economics of Food Safety* at the proceedings of the American National Workshop on the same title held in Washington, a series of reports were published,thus layingthefoundationforeconomicanalysisoffoodsafety.Now,moreandmore economists are studying food safety issues using empirical analysis methods, especially in the study of consumer behaviours, awareness and willingness to pay for safe food. For example, Chern et al (2002) [17] studied consumers' willingness to pay for genetically modified vegetaȬ bleoils throughcarryingout a surveyinShikoku,Japan,Norway,TaiwanandtheUnitedStates. Georges et al(2006)[18] selectedobjects from 12European countries anddividedthem into four groups to understand their perception on the food traceability system by group discussion.

The vast majority of Chinese economic research on domestic food safety issues uses foreign theories and research methods. Wang (2003) [19] carried out a survey of 289 consumers in Tianjin and analysed the process and characteristics of selecting safe food. Chern et al (2002) [17] revealed that consumers surveyed in Zhejiang province were relatively concerned about vegetable safety and their attitudes were found to be negative. Consumers were very willing to pay the extra cost for safe vegetables, but the price of the safe vegetables should be higher than ordinary vegetables by no more than 10% to 20%. Zeng et al (2008) [20] studied consumers' willingness to pay for moon cake with safe additives through a survey of 396 consumers in 25

Research on the agro-environment and food safety respectively from the perspective of farmers and consumers has increased in recent years. Most researchers only analyse and expound survey data, however, in-depth and comprehensive analysis, and related solutions are scant.

The authors of this book were driven to develop the contexts based on the important concept of *Kanshokufuji*, a Japanese term created by Prof. Nanseki Teruaki, one of the chief editors of this book. With the literal meaning of *integrated environment and food*, this concept advocates a sound food system in a sound agro-environment [23]. In modern society, food safety is increasingly becoming of global concern, due to asymmetric information on the processes, additives in the long industrial chain, etc [21-22]. In this key concept for establishing a safe and sustainable next

awareness of protecting the environment was increasing.

**1.3 Theoretical framework and methodology**

**1.3.1** *Kanshokufuji***: The theoretical basis**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

supermarkets in Beijing.

To illustrate status of the key phrases within *Kanshokufuji*, large volumes of data were collected through various sources. Firstly, authorized data from the China Statistical Yearbook, China Agricultural Bulletin and data published through government websites, etc., were used to review the general situation and problems of the rural environment, agriculture and food. Secondly, other data published in academic monographs and journal articles was recited to provide additional proving or comparable data in analyses from general advances to the empirical findings within this book.

**Figure 1.2** *Kanshokufuji* and *Ishokudogen*

In addition to the second-hand data mentioned above, much more first-hand data was collected through a variety of field surveys by the authors. The field surveys can be divided to two categories, both of which were conducted based on questionnaires and interviews. One category includes the rural and farmer surveys, while the other consists of consumer surveys. For instance, the general situation of the rural environment and agricultural production in different regions was studied through a survey of 21 villages in six provincial-level regions of eastern China. Using the data collected from a survey 560 household farms scattered across 21 villages of six provincial regions, we studied farmers' behaviours, perceptions and major affecting factors upon the application of fertilizers and pesticides, and farmers' confidence about the safety of their self-produced grain products. In light of the interviews of 168 sample dairy farmers of Inner Mongolia and Hebei Province, dairy farmers' perception of risk and their points of view on risk management strategies were studied. Based on a survey of 512 respondents from Beijing and Shanghai, we studied consumer perceptions on food safety and the major affecting factors. To examine consumers' attitude toward the traceability system, two interview surveys were conducted in Beijing and two samples with 209 and 214 respondȬ ents, respectively, were used to analyse consumers' risk awareness and willingness to pay for safety-certified food (Figure 1.3).

#### **1.3.3 Major analysis methods**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

To fulfil the analyses mentioned above, a variety of methods are applied in the following chapters. Firstly, a series of empirical models are adopted to explore the implications behind the data collected. (1) Cobb-Douglas production function. From the perspectives of inputs change, institutional transition and technological progress, this book conducts a factor analysis of Chinese Agriculture Development after 1983. The macro-analysis is based on the time-series data issued by the government. After comparing the expressions of the two basic models of

**Figure 1.3** Theoretical framework of this book

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In addition to the second-hand data mentioned above, much more first-hand data was collected through a variety of field surveys by the authors. The field surveys can be divided to two categories, both of which were conducted based on questionnaires and interviews. One category includes the rural and farmer surveys, while the other consists of consumer surveys. For instance, the general situation of the rural environment and agricultural production in different regions was studied through a survey of 21 villages in six provincial-level regions of eastern China. Using the data collected from a survey 560 household farms scattered across 21 villages of six provincial regions, we studied farmers' behaviours, perceptions and major affecting factors upon the application of fertilizers and pesticides, and farmers' confidence about the safety of their self-produced grain products. In light of the interviews of 168 sample dairy farmers of Inner Mongolia and Hebei Province, dairy farmers' perception of risk and their points of view on risk management strategies were studied. Based on a survey of 512 respondents from Beijing and Shanghai, we studied consumer perceptions on food safety and the major affecting factors. To examine consumers' attitude toward the traceability system, two interview surveys were conducted in Beijing and two samples with 209 and 214 respondȬ ents, respectively, were used to analyse consumers' risk awareness and willingness to pay for

6 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

To fulfil the analyses mentioned above, a variety of methods are applied in the following chapters. Firstly, a series of empirical models are adopted to explore the implications behind the data collected. (1) Cobb-Douglas production function. From the perspectives of inputs change, institutional transition and technological progress, this book conducts a factor analysis of Chinese Agriculture Development after 1983. The macro-analysis is based on the time-series data issued by the government. After comparing the expressions of the two basic models of

safety-certified food (Figure 1.3).

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**1.3.3 Major analysis methods**

Source: Nanseki (2008) [23]

**Figure 1.2** *Kanshokufuji* and *Ishokudogen*

Cobb-Douglas production function, the better performing model is adopted to demonstrate the relationships between economic growth and inputting factors. (2) Principal component analysis (PCA) is adopted to analyse the major factors affecting farmers' perceptions and strategies on food risk management. (3) Binary logit regression model is used to analyse the determinants of farmers' confidence on the safety of their self-produced grain products farmers' behaviours in the application of agricultural chemicals. (4) Integrating with the choice modelling (CM) technique, the multinomial logit model is applied to examine consumers' attitude toward the traceability system. (5) Multivariate regression models are used to identify the significant determinants of pesticide application.

Meanwhile, further analyses are conducted through the construction or introduction of several new statistics. (1) In line with the soil properties represented by the geographical location in the National Fertilization Regionalization, the new statistic of *Fertilization coefficient* is formuȬ lated to isolate effects of farms' geographical location and planting structure, hence capturing farmers' propensities on fertilizing. (2) According to the *Animal manure coefficient* recommendȬ ed by the National Environment Protection Bureau (NEPB), the total amount of animal manure and the major compositions of BOD5, NH3-N, TN and TP are calculated. (3) Through the *MWTP coefficient*, we managed to study consumers' marginal willingness to pay (MWTP) on the information provided by the traceability system, and to examine which factors affect consumȬ ers' willingness to pay for the traceability system.

In addition, descriptive statistical analysis methods, including the one-way T-test, Chi-square test, statistics of coefficient of variation, mean, std. D, max, min, etc., were widely used in this book to provide general scenarios or comparisons. In particular, the one-way T-test is used in identifying the major factors of consumer perceptions towards food safety. The analysis is conducted from the perspective of variables' significance in identifying the discrepancies among most of perceptions. Moreover, further analysis is conducted on the impact of demoȬ graphic variables significant at the level of 0.01.

#### **1.4 Organization of the book**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Within the theoretical framework developed based on the conception of *Kanshokufuji*, the contents of the book are organized as shown in Figure 1.4 and Table 1.1. Chapter 2 studies the critical issues of agriculture, environment and food in China, including a factor analysis of the gross agricultural economy over recent decades, and a review of the food and environment. Based on the survey of 560 farmers within 21 villages of six eastern provincial-level regions, Chapters 3, 4 and 5 study the situation and determinants of rural environment, farmers' confidence about their self-produced agro-products, farmers' behaviours and perceptions and determinants of agro-chemicals' application, respectively. Chapter 6 analyses perceptions of dairy farmers on risk source and risk management, Chapter 7 studies consumer awareness and determinants in the top two Chinese metropolises upon food safety, while Chapter 8 aims to investigate consumers' risk awareness with regard to dairy products and their willingness to pay for certified safety food based on data from other surveys. As the concluding chapter, Chapter 9 includes an awareness comparison on environmental problems between farmers and consumers, followed by concluding and policy recommendations in light of the foregoing chapters.

**Chapter 2:** From the perspectives of inputs change, institutional transition and technological progress, this chapter conducted a factor analysis of Chinese agriculture development after 1983, when it began to develop as an independent industry by and large. This macro-analysis was based on the time-series data issued by the government, and the main model adopted was Cobb-Douglas production function. Through the application of SPSS, although several more independent variables were included into the model, the most effective factors eventually decided upon only contained the increment of chemical fertilizer, fixed agricultural assets, financial supports and the reduction of agricultural labour force, with the contribution rate of 53.70%, 15.57%, 4.77% and 1.66%, respectively. Furthermore, being the residual of the four variables above, it was calculated that technological progress contributed 24.30% to Chinese agriculture development in this period.

As a main conclusion, material inputs, including chemical fertilizer in the first place, composed the most important factor in agricultural development. As for the second factor, technical progress also promoted agricultural development to a considerable degree, while the contriȬ bution rate from institutional transition was comparatively low. Finally, a variety of suggesȬ tions were made on the topics such as secure application of chemical fertilizer, popularization of agro-technology, the increase of agro-capital, reduction of agro-labour, etc.

**Figure 1.4** Organization of this book

book to provide general scenarios or comparisons. In particular, the one-way T-test is used in identifying the major factors of consumer perceptions towards food safety. The analysis is conducted from the perspective of variables' significance in identifying the discrepancies among most of perceptions. Moreover, further analysis is conducted on the impact of demoȬ

8 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Within the theoretical framework developed based on the conception of *Kanshokufuji*, the contents of the book are organized as shown in Figure 1.4 and Table 1.1. Chapter 2 studies the critical issues of agriculture, environment and food in China, including a factor analysis of the gross agricultural economy over recent decades, and a review of the food and environment. Based on the survey of 560 farmers within 21 villages of six eastern provincial-level regions, Chapters 3, 4 and 5 study the situation and determinants of rural environment, farmers' confidence about their self-produced agro-products, farmers' behaviours and perceptions and determinants of agro-chemicals' application, respectively. Chapter 6 analyses perceptions of dairy farmers on risk source and risk management, Chapter 7 studies consumer awareness and determinants in the top two Chinese metropolises upon food safety, while Chapter 8 aims to investigate consumers' risk awareness with regard to dairy products and their willingness to pay for certified safety food based on data from other surveys. As the concluding chapter, Chapter 9 includes an awareness comparison on environmental problems between farmers and consumers, followed by concluding and policy recommendations in light of the foregoing

**Chapter 2:** From the perspectives of inputs change, institutional transition and technological progress, this chapter conducted a factor analysis of Chinese agriculture development after 1983, when it began to develop as an independent industry by and large. This macro-analysis was based on the time-series data issued by the government, and the main model adopted was Cobb-Douglas production function. Through the application of SPSS, although several more independent variables were included into the model, the most effective factors eventually decided upon only contained the increment of chemical fertilizer, fixed agricultural assets, financial supports and the reduction of agricultural labour force, with the contribution rate of 53.70%, 15.57%, 4.77% and 1.66%, respectively. Furthermore, being the residual of the four variables above, it was calculated that technological progress contributed 24.30% to Chinese

As a main conclusion, material inputs, including chemical fertilizer in the first place, composed the most important factor in agricultural development. As for the second factor, technical progress also promoted agricultural development to a considerable degree, while the contriȬ bution rate from institutional transition was comparatively low. Finally, a variety of suggesȬ tions were made on the topics such as secure application of chemical fertilizer, popularization

of agro-technology, the increase of agro-capital, reduction of agro-labour, etc.

graphic variables significant at the level of 0.01.

**1.4 Organization of the book**

agriculture development in this period.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

chapters.


Note: ӣ refers to the main topics, while ئ denotes sub-topics of each chapter

**Table 1.1** Topics of each chapter

**Chapter 3:** Based on a survey of 21 villages in six provincial-level regions of eastern China, this chapter studies that the situation of agricultural production in different regions and the relation of the rural environment to economic level. We found that 1and type, crop acreage, irrigation water source and irrigation methods have shown a certain difference between the north and south of China. The proportion of the collective rubbish and sewage disposal in rural areas is higher than industrial sewage and animal manure, the agricultural production and farmers' life guidance services are primarily supplied by government and agricultural extension centres. Within the surveyed villages, about 50% of them got subsidies from the government for constructing methane tanks, reducing the application of fertilizers, while only 10% of the surveyed villages were subsidized for the adoption of biodegradable plastic sheets. Meanwhile, empirical analysis revealed that the proportion of the collective rural rubbish and sewage disposal positively relate to the income level of farmers, while there is a negative correlation with distance from the nearest town.

**Chapter 4:** Based on the same survey of 560 household farms in six eastern provincial regions, this chapter studies farmer behaviours on the application of fertilizer, including the total amounts, main components of chemical fertilizer and the use of organic fertilizer. Then, it summarizes the farmers' perceptions, ranging from fertilizer choosing, field application, disposal of the used packages and awareness on the possible consequences of over fertilization. Nine indicators are adopted as the predictors, including information on the householders, land use and planting structure, household income and geographical location. The Fertilization coefficient is formulated to isolate the effects of farms' geographical location and planting structure, hence capture farmers' propensities on fertilizing. Through the adoption of binary logistic regression models, this chapter identifies significant determinants behind farmers' behaviours. As to the use of organic fertilizer, although demonstrated as statistically insignifȬ icant, possible impacts of chemical fertilization and breeding of livestock and poultry are included as predictors, in addition to the above indicators. Finally, a variety of policy recomȬ mendations are put forward, from increasing the fertilization efficiency of both chemical and organic fertilizer, to improving farmers' capability and awareness of scientific fertilization.

Meanwhile, this chapter studies farmers' application of pesticides, including the amounts of chemical pesticides, use of toxic pesticides and biological pest-control methods. Similarly, it summarizes the farmers' perceptions, ranging from choosing pesticides and field application to the awareness on the withdrawal period, possible consequences of overdosing and disposal of the containers. Thereafter, nine demographic indicators are incorporated as the candidate determinants, including information on the householders, land use and cropping structure, household income and geographical location. Through the adoption of multivariate OLS and logistic regression models, this chapter identifies significant determinants affecting farmers' behaviours. Finally, several policy recommendations are put forward, including the counterȬ measures to increase pesticidal efficiency, decreasing the use of toxic pesticides and improving farmers' capability and awareness on scientific application of pesticides.

**Chapter 5:** Following the publicity of a series of food safety incidents and the asymmetry of food information, consumers' confidence on Chinese food safety dropped dramatically. Approximately 70% of consumers are not confident of food safety. Compared with consumers,

farmers, as managers and producers in agriculture production, have a lot of information of agricultural products. Based on a survey to 560 samples in six eastern provinces in China, this study selected 346 grain crop farmers to analyse farmers' confidence on their products. The data shows that more than 80% of farmers are confident on their products. In order to better understand farmers' confidence, we analyse influencing factors through the binary logit egression model. The result indicated use (or not) of manure, location and sowing area significantly affected farmers' confidence on their own agro-products. According to the result, some recommendations were proposed at the end of the chapter.

**Chapter 3:** Based on a survey of 21 villages in six provincial-level regions of eastern China, this chapter studies that the situation of agricultural production in different regions and the relation of the rural environment to economic level. We found that 1and type, crop acreage, irrigation water source and irrigation methods have shown a certain difference between the north and south of China. The proportion of the collective rubbish and sewage disposal in rural areas is higher than industrial sewage and animal manure, the agricultural production and farmers' life guidance services are primarily supplied by government and agricultural extension centres. Within the surveyed villages, about 50% of them got subsidies from the government for constructing methane tanks, reducing the application of fertilizers, while only 10% of the surveyed villages were subsidized for the adoption of biodegradable plastic sheets. Meanwhile, empirical analysis revealed that the proportion of the collective rural rubbish and sewage disposal positively relate to the income level of farmers, while there is a negative

10 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Chapter 4:** Based on the same survey of 560 household farms in six eastern provincial regions, this chapter studies farmer behaviours on the application of fertilizer, including the total amounts, main components of chemical fertilizer and the use of organic fertilizer. Then, it summarizes the farmers' perceptions, ranging from fertilizer choosing, field application, disposal of the used packages and awareness on the possible consequences of over fertilization. Nine indicators are adopted as the predictors, including information on the householders, land use and planting structure, household income and geographical location. The Fertilization coefficient is formulated to isolate the effects of farms' geographical location and planting structure, hence capture farmers' propensities on fertilizing. Through the adoption of binary logistic regression models, this chapter identifies significant determinants behind farmers' behaviours. As to the use of organic fertilizer, although demonstrated as statistically insignifȬ icant, possible impacts of chemical fertilization and breeding of livestock and poultry are included as predictors, in addition to the above indicators. Finally, a variety of policy recomȬ mendations are put forward, from increasing the fertilization efficiency of both chemical and organic fertilizer, to improving farmers' capability and awareness of scientific fertilization.

Meanwhile, this chapter studies farmers' application of pesticides, including the amounts of chemical pesticides, use of toxic pesticides and biological pest-control methods. Similarly, it summarizes the farmers' perceptions, ranging from choosing pesticides and field application to the awareness on the withdrawal period, possible consequences of overdosing and disposal of the containers. Thereafter, nine demographic indicators are incorporated as the candidate determinants, including information on the householders, land use and cropping structure, household income and geographical location. Through the adoption of multivariate OLS and logistic regression models, this chapter identifies significant determinants affecting farmers' behaviours. Finally, several policy recommendations are put forward, including the counterȬ measures to increase pesticidal efficiency, decreasing the use of toxic pesticides and improving

**Chapter 5:** Following the publicity of a series of food safety incidents and the asymmetry of food information, consumers' confidence on Chinese food safety dropped dramatically. Approximately 70% of consumers are not confident of food safety. Compared with consumers,

farmers' capability and awareness on scientific application of pesticides.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

correlation with distance from the nearest town.

**Chapter 6:** The field survey was carried out in Inner Mongolia and Hebei Province in April and June 2010, respectively. A sample totalling 168 dairy farmers was available for analysis in this study. In this chapter, dairy farmers' perception of risk and their points of view on risk management strategies are studied. Risk is uncertainty that affects an individual's welfare, and is often associated with adversity and loss. In response to risky situations, farmers should be involved in risk management, making choices among alternatives so as to reduce the effects of the risks. The main research objectives are: to examine the dairy farmers' perception of risk and to examine the risk management strategies of dairy farmers.

**Chapter 7:** Based on a survey of 512 respondents from Beijing and Shanghai, this chapter studies consumer perceptions on food safety and the major affecting factors. In addition to the basic individual information of gender, age, educational background and employment, the demographic variables include professional experience of the respondent, family composition and also annual income. The perceptions consist of overall awareness about food safety, major sources of information and subjective reliability, understanding of the impact of environmenȬ tal protection, main threats to food safety, the top sources of agro-pollution, most risky procedure or stage, viewpoints on the major responsibility bearer of agro-pollution and the best way to control agro-pollution. After the descriptive analysis on demographic characterȬ istics and perception variables, the one-way T-test reveals that all the nine demographic variables are significant in identifying the discrepancies among most perceptions. Moreover, further analysis is conducted on the impact of demographic variables significant at the level of 0.01. Finally, a variety of policy recommendations are put forward, from strengthening the supervisory responsibility of the government, ensuring the all-round and effective supervision of food safety by the mass media and consolidating the supervision of key sectors, to accelerȬ ating the extension of environment-friendly technology.

**Chapter 8:** Consumers' risk awareness on dairy products and willingness to pay for certified safe food are studied based on related field surveys by the authors. In order to examine consumers' attitude toward the traceability system, an interview survey was conducted from September to October 2008 in Beijing, and 209 samples were collected in this survey. Data from another self-survey conducted in Beijing July 2008 by the authors is also used for analysis in this chapter. In this survey, 214 consumers were interviewed and applied as valid samples. The analysis of consumers' willingness to pay for certified safe food was carried out based on another survey. The analysis includes data of 209 respondents that correspond to 100% of the interviewed consumers in the survey site - Beijing.

**Chapter 9:** An awareness comparison on environmental problems between farmers and consumers was conducted in the first instance. Analysis of this section is based on two field surveys. The first one is the survey of farmers from six provincial regions as introduced in Chapter 3. While the other is a consumer survey held in Beijing, 2008. Consumers' attitude towards rice, vegetables, meat and milk were included in the questionnaire. There are 186 samples available from July and 209 samples available from September. The consumers' survey in July is mainly used in this case study. This survey includes respondents both from a supermarket survey and a home survey. In succession, based on the findings and conclusions from the foregoing chapters, a series of policy recommendations are put forward on risk management in China. Finally, perspectives on international cooperation between East Asia and the world are previewed from the standing of managing risks among food, environment and agriculture.

**Chapter 10:** As the last chapter, this chapter aims to establish an academic basis for the development of a risk governance system for food safety in East Asia for the cases of Japan and China. First, short histories of the food safety policy in Japan and China are reviewed. Secondly, the current statuses of the food traceability system in both countries are clarified. Thirdly, consumer perception on food safety is analysed from various perspectives. Fourthly, the current statuses of risk management at farm level (e.g., GAPs) in both countries are overviewed. Finally concluding remarks is given about further research. Various survey data, including several original surveys by the author, and government statistics are used for analysis in this chapter. The author's original survey on consumer awareness of food safety was done in Japan and China, 2008. The respondents of the preliminary surveys are 297 in total in China. The survey in Japan was an indoor group investigation using a survey slip. The survey in China was conducted by means of an individual interview. The author's original survey on consumer awareness of pork and milk traceability was conducted in China.

#### **References**


[6] Hou B., Hou J., Wang Z. Farmers' perception on pesticide residue and its influence on pesticide application. Heilongjiang Agricultural Sciences, 2010(2):99-103.

**Chapter 9:** An awareness comparison on environmental problems between farmers and consumers was conducted in the first instance. Analysis of this section is based on two field surveys. The first one is the survey of farmers from six provincial regions as introduced in Chapter 3. While the other is a consumer survey held in Beijing, 2008. Consumers' attitude towards rice, vegetables, meat and milk were included in the questionnaire. There are 186 samples available from July and 209 samples available from September. The consumers' survey in July is mainly used in this case study. This survey includes respondents both from a supermarket survey and a home survey. In succession, based on the findings and conclusions from the foregoing chapters, a series of policy recommendations are put forward on risk management in China. Finally, perspectives on international cooperation between East Asia and the world are previewed from the standing of managing risks among food, environment

12 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Chapter 10:** As the last chapter, this chapter aims to establish an academic basis for the development of a risk governance system for food safety in East Asia for the cases of Japan and China. First, short histories of the food safety policy in Japan and China are reviewed. Secondly, the current statuses of the food traceability system in both countries are clarified. Thirdly, consumer perception on food safety is analysed from various perspectives. Fourthly, the current statuses of risk management at farm level (e.g., GAPs) in both countries are overviewed. Finally concluding remarks is given about further research. Various survey data, including several original surveys by the author, and government statistics are used for analysis in this chapter. The author's original survey on consumer awareness of food safety was done in Japan and China, 2008. The respondents of the preliminary surveys are 297 in total in China. The survey in Japan was an indoor group investigation using a survey slip. The survey in China was conducted by means of an individual interview. The author's original survey on consumer awareness of pork and milk traceability was conducted in China.

[1] China Ministry of Agriculture.China Statistical Yearbook 2012. Shenyang: Liaoning

[2] Liu G.,et al. Current situation and measures of agricultural pollution in China. StudȬ

[3] Liu G.,et al. Current situation and measures of agricultural pollution in China. StudȬ

[4] CNSB (China National Statistical Bureau).Using of chemical fertilizers in different reȬ

[5] Sun J. Review of agricultural pollution and preventive technology in China. Journal

ies in International Technology and Economy, 2006; 9(4):17-21.

ies in International Technology and Economy, 2006; 9(4):17-21.

of Jishou University (Natural Science Edition), 2008; 29(5): 99-128.

gions in China, http://www.stats.gov.cn/.

and agriculture.

**References**

Education Press, 2012.5.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/



**Chapter 2**

## **Critical Issues of Food Safety and the Agro-Environment in China**

Dongpo Li, Hui Zhou, Min Song and Teruaki Nanseki

**2.1 Food safety in China**

#### **2.1.1 General situation**

[21] Tonsor G. T. Consumer inferences of food safety and quality. European Review of

[22] Li D., Nanseki T. Takeuchi S., Song M., Chen T., Zhou H. Consumer perceptions upon food safety and demographic determinants in China: empirical analysis based on a survey of 512 respondents. Journal of Faculty of Agriculture, Kyushu UniversiȬ

[23] Nanseki T. Perspective of information technology for food safety. Japanese Journal of

Agricultural Information Research 2008; 17 (4): 161-170 (in Japanese).

14 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Agricultural Economy 2011; 38 (2): 213-235.

ty, Japan 2012; 57 (2): 517-525.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Agricultural pollution and environmental problems directly influence the food safety situaȬ tion. Safe and healthy food only can be produced in a sound environment. In this chapter, agricultural pollution will be defined and environmental problems in China will be described.

With China's agro-production shifting to large scale and intensive operations, the huge amount of waste produced by the industry has not only worsened the circumstances where animals live, but also have adverse impacts on human health. The livestock industry has been urged to minimize pollutions caused by livestock production, explore effective waste manȬ agement policies and technologies, and promote sustainable development of the industry.

#### **2.1.2 Food safety issues**

Food safety is a global issue nowadays. Food safety problems have caused many losses to consumers, producers and governments. Food safety problems have many causes such as economic problems, lack of technology and policy. In order to control the food hazard and food safety problems, the EU, Japan and the USA have been conducting research and have made great progress. According to the Chinese Food Safety Situation Report, the rate of certified food has been increasing since 2006 and the certified high-quality foods are becoming leading products in the markets. Chemical input residue, such as those from pesticides, chemical fertilizers, feed addictive, veterinary processes and others, is decreasing.

The current food safety situation is optimistic on the whole, but there are still some risks: new food safety problems may appear along with technology development. Checking several reports on food safety issues, we found that most food safety problems happened at either the production stage or at the processing stage. Additionally, the reasons could be: 1) chemical

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Li et al.; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Li et al.; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

fertilizer or pesticide residue; 2) veterinary residue; 3) illegal addictive and 4) bacteria exceeding the standard.


Source: internet survey

**Table 2.1** Several food safety issues have occurred in China over recent years

#### **2.2 Agricultural pollution and environmental problems**

#### **2.2.1 Definition of agricultural pollution**

In the wider meaning, agricultural pollution is an ecological term that considers agricultural production as both the source and object of environmental deterioration. As shown in Figure 2.1, agricultural pollution is caused mainly by emissions of industrial wastes, including wasted water, gas and solid, urban-rural life sewage and rubbish, and pollution of agricultural production itself. Meanwhile, agricultural pollution can contaminate, farm land, water, etc., and result in quality risks for agricultural products [1].

**Figure 2.1** Mechanism of agricultural pollution in China

Compared with the other two sources, the non-point sources (NPSs) from rural sewage and rubbish and agricultural production come from scattered individual households and farms, which is difficult to control and thus of particular concern [2]. Simultaneously, they may result in serious environmental problems, including pollutions to soil and water due to overapplication of chemical fertilizers and pesticides, pollution from wasted mulching film and other agricultural wastes, environmental pollution and deficiency of soil NPK (nitrogen, phosphorus and potassium) caused by burning straw, water contamination by livestock and poultry manure, etc. [3]. Moreover, it leads to over residues of nitrate, nitrite, heavy metals and even poisonous substances, resulting in direct impacts on the safety of agricultural products [4].

Agricultural pollution problems have happened often in China. There are 23 typical agriculȬ tural environmental problems which have occurred in China during 2000 to 2010. Agricultural environmental problems are mainly divided into three categories: agriculture is one of the sources which pollute the environment, agriculture is polluted by industry and other departȬ ments, and agro-resources exhausted by improper use and development of natural resources. More than half of the pollution problems are caused by agricultural itself. The main reasons are over use of pesticide, fertilizer, plastic sheeting, improper animal waste treatment and rural solid waste.

#### **2.2.2 Chemical agro-input problems**

fertilizer or pesticide residue; 2) veterinary residue; 3) illegal addictive and 4) bacteria

16 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

the standard

addictives

2008 Milk powder melamine issue Illegal addictive http://news.qq.com/zt/2008/juchong/ 2008 Date poisoned issue Illegal addictive http://www.guyu.cn/content/48610.shtml

In the wider meaning, agricultural pollution is an ecological term that considers agricultural production as both the source and object of environmental deterioration. As shown in Figure 2.1, agricultural pollution is caused mainly by emissions of industrial wastes, including wasted water, gas and solid, urban-rural life sewage and rubbish, and pollution of agricultural production itself. Meanwhile, agricultural pollution can contaminate, farm land, water, etc.,

Pesticide residue http://www.gov.cn/jrzg/2010-02/28/ content\_1543980.htm

Pesticide residue http://news.sina.com.cn/c/

Pesticide residue http://news.xinhuanet.com/legal/2010-04/01/ c\_1212625.htm

2010-04-10/033120043336.shtml

06/0916/04/2R45E8S70001124J.html

http://news.sina.com.cn/c/ 2009-02-02/043517130520.shtml

http://news.163.com/

**Year Issue Reasons Source**

2009 Milk powder bacteria issues Bacteria exceeding

2006 Clenbuterol problems Illegal feed and

**2.2.1 Definition of agricultural pollution**

**Figure 2.1** Mechanism of agricultural pollution in China

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 2.1** Several food safety issues have occurred in China over recent years

and result in quality risks for agricultural products [1].

**2.2 Agricultural pollution and environmental problems**

exceeding the standard.

2010 Vegetable poisoned issue in Hainan Province

2010 Vegetable poisoned issue in Guangxi Province

2010 Vegetable poisoned issue in Qingdao city

Source: internet survey

Chemical fertilizer and pesticide residue is a big problem negatively affecting the agroenvironment in China. Since 2000, China has become the biggest chemical fertilizer producer in the world. The rapid growth in China's per hector chemical fertilizer application has contributed significantly to the growth in grain production, but also caused many environment problems such as groundwater and underground water pollution. The improper use of chemical fertilizer and its residue is becoming one of the biggest problems in food safety negatively affecting the environment [5-6].


Source: China Statistical Press [8]

Note: the requested ratio by MOA among N, P and K is 1:0.37:0.25

**Table 2.2** Use of chemical fertilizers in China (Unit: million *ton*)

According to the Ministry of Agriculture in China, the proper use ratio of chemical fertilizer among N, P and K is 1:0.37:0.25, but the reality is 1:0.34:0.24, which implies that the chemical use in China is not in proper balance, especially with regard to nitrogen overuse (Table 2.2). Based on research data from Henan Agricultural Bureau, only 1/3 of the chemical fertilizer is absorbed by plants, 1/3 gets into the air and 1/3 gets into the soil, and the chemical fertilizer pollutes both soil and air [7].

With the steady increase of agricultural production in China, extensively used pesticides have increased crop yields and produced high quality products over recent decades. Up to the end of 2010, the total amount of chemical pesticides produced in China amounted to 2.34 million tons, maintaining an average annual growth rate of 10.32% since 1985 [8]. China has become the largest producer, user and exporter of pesticides in the world. Meanwhile, the improper use of pesticides has become a major source of food safety incidents, which have resulted in serious threats (and losses) to the ecological environment, human health and economic development.

#### **2.2.3 Animal manure waste problems**

With the rapid development of the livestock industry, animal waste and hazardous residues of feed cause more and more pollutants. Some statistics suggest that in 2007 there were 439.895 million pigs, 105.948 million cattle, 285.657 million sheep and 9578.67 million poultry in China. According to the Animal Manure Coefficient recommended by the National Environment Protection Bureau (NEPB) (Table 2.3), we can calculate the total amount of animal manure at 2.147 billion tons, 42.445 million tons of BOD5, 0.418 million tons of CODcr, 4.935 million tons of NH3-N, 11.743 million tons of TN and 3.045 million tons of TP (Table 2.4), far greater amounts of solid waste from industry, which amounted to 1 billion tons in the same period of time.


Source: China Statistical Press [8]

Note: poultry one is the average of chickens and ducks.

**Table 2.3** Animal manure coefficient of NEPB (Unit: kg)


Source: calculation based on the data of China Statistical Press [8]

**Table 2.4** Animal waste discharged in 2007 (Unit: million *ton*)

According to the Ministry of Agriculture in China, the proper use ratio of chemical fertilizer among N, P and K is 1:0.37:0.25, but the reality is 1:0.34:0.24, which implies that the chemical use in China is not in proper balance, especially with regard to nitrogen overuse (Table 2.2). Based on research data from Henan Agricultural Bureau, only 1/3 of the chemical fertilizer is absorbed by plants, 1/3 gets into the air and 1/3 gets into the soil, and the chemical fertilizer

18 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

With the steady increase of agricultural production in China, extensively used pesticides have increased crop yields and produced high quality products over recent decades. Up to the end of 2010, the total amount of chemical pesticides produced in China amounted to 2.34 million tons, maintaining an average annual growth rate of 10.32% since 1985 [8]. China has become the largest producer, user and exporter of pesticides in the world. Meanwhile, the improper use of pesticides has become a major source of food safety incidents, which have resulted in serious threats (and losses) to the ecological environment, human health and economic

With the rapid development of the livestock industry, animal waste and hazardous residues of feed cause more and more pollutants. Some statistics suggest that in 2007 there were 439.895 million pigs, 105.948 million cattle, 285.657 million sheep and 9578.67 million poultry in China. According to the Animal Manure Coefficient recommended by the National Environment Protection Bureau (NEPB) (Table 2.3), we can calculate the total amount of animal manure at 2.147 billion tons, 42.445 million tons of BOD5, 0.418 million tons of CODcr, 4.935 million tons of NH3-N, 11.743 million tons of TN and 3.045 million tons of TP (Table 2.4), far greater amounts of solid waste from industry, which amounted to 1 billion tons in the same period of time.

**Animal/poultry Manure Urea BOD5 CODcr NH3-N TN TP**

Pig 398.00 656.70 25.98 26.61 2.07 4.51 1.70

Cattle 7300.00 3650.00 193.70 248.20 25.15 61.10 10.07

Sheep 950.00 -- 2.70 4.40 0.57 2.28 0.45

Poultry 26.30 -- 1.015 1.16 0.125 0.27 0.11

pollutes both soil and air [7].

**2.2.3 Animal manure waste problems**

Source: China Statistical Press [8]

Note: poultry one is the average of chickens and ducks.

**Table 2.3** Animal manure coefficient of NEPB (Unit: kg)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

development.

#### **2.2.4 Greenhouse gas emissions from the livestock sector**

With rapid economic development, large amounts of greenhouse gases, such as carbon dioxide, CFCs, methane nitrous oxide and others, were produced. These gases in the atmosȬ phere continue to accumulate and the average temperature of the mainland over the past century has increased significantly. The global average temperature increased 0.3-0.6 degrees, and the sea level rose 10-25 cm - indicating the fastest period of climate warming. After the Copenhagen Climate Change Conference, people began to pay more attention to greenhouse gases and climate change, and it became the most serious challenge facing the human race. The livestock sector is a major player in greenhouse gas emissions. According to the statistics from the Food and Agriculture Organization (FAO), the livestock sector is responsible for 18% of global greenhouse gas emissions measured in CO2 equivalent.

Livestock production can result in methane emission from enteric fermentation and both CH4 and nitrous oxide emissions from livestock manure management systems. Ruminants, such as cattle, are important sources of CH4 because of their large population and high CH4 emission rate as a result of their digestive system. Swine are also an important source of greenhouse gas, but less so than cattle. Poultry produce the least greenhouse gas. Animal waste will decompose in anaerobic conditions and produce methane, while the compost will produce large amounts of nitrous oxide.


Source: calculated based on the data of China Statistical Press [8]

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Note: 1. emission factors of livestock in sub-regions could not be available, so national emission factors are used to substitute.

**Table 2.5** Top 10 provinces of methane emissions from the livestock sector

#### **2.2.5 Veterinary residues in the livestock sector**

The drug residues in livestock products contain drugs for animals and humans, disinfection chemicals, pesticides and other chemicals, the most prominent is residues of clenbuteral. Besides these, abuse of antibacterial drugs, hormones, vitamins and trace elements can cause residues of drugs. Drugs are not used properly, animals are slaughtered before the end of the drug-free period, or unapproved drugs are used as additives - all factors which can lead to contamination of the livestock produce. Once they are discharged with the manure, they will pollute the soil and water, and the environment. The potential threats of drug residues to human health include: allergy and abnormal response, taretogenicity, mutagenicity, carcinoȬ genicity and bacterial infection. For example, dosages of hormones, such as cortisone and hydrocortisone which are widely used by veterinary surgeons, can cause few residues. But large scale over dosage can lead to large amounts of residues. However, their impacts on humans might be visible only after several generations, e.g., femininity in men and masculinity in women has some linkages with the abuse of hormones.

#### **2.3 Development of Chinese agriculture and previous studies**

#### **2.3.1 Agricultural development over latest decades**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

At the end of 1978, China launched *Reforms and Opening-up*, thus breaking up the highlyplanned economic institutions, and revitalizing the whole economy including those in rural areas. Up to the mid-1980s, as a prelude to the *Reform and Opening-up*, the *Household Contract Responsibility System* was expanded in rural areas across the country, where production teams were subsumed into 99% of villages and major production resources, symbolized by farmland, which were divided into household farms. After the reform, farmers could retain the rest of the goods and revenues as private property, once a certain amount of agricultural products or taxes were paid to the state as contracted. It released the long-term bound farming organizaȬ tions and increased farmers' motivation with regard to agricultural productivity, hence agricultural development reached a high level within a few years. As a fundamental measȬ urement, total grain yields amounted to 379 million tons in 1985, from 305 million tons in 1978. With this growth rate of some 25%, the problem of food security, which had puzzled China for a long time, was resolved by and large. In addition to benefiting national life and industrial development, it brought new opportunities to the overall economic reforms. At the same time, thanks to the effects of non-agricultural reforms, agriculture gradually developed as an industry capable of self-reliance.

By the mid-1980s, the rapid development of agriculture was realized primarily due to the powerful potential released by institutional reforms. By contrast, in subsequent periods, agriculture maintained its high-speed growth, under the progress of overall economic reforms. From 1983 to 2010, China's Gross Agricultural Output (GAO) rose from 275 billion yuan to 69319.76 billion yuan (current prices). Accounting for the influences of inflation, the GAO was 1242.70 billion yuan (using the constant prices of 1983), with an average annual growth rate of some 6.29 % being maintained in this period.

As shown below, factor analysis of China's agricultural development has been conducted in Lin et al. (1992) [9], Wang (2009) [10], Zhang (2008) [11] and other studies. However, these studies focused mainly on the specific changes of institutions, technology or inputs up to the early 21st century or even much earlier periods. That is, conducting comprehensive analysis and policy recommendations on the reasons behind agricultural development in recent periods remains a challenge to scholars when considering all the factors proposed above. Therefore, this chapter aims to clarify these issues, through factor analysis of China's agriculȬ tural development from the perspectives of inputs change, institutional transition and technological progress since 1983, when it began to develop as an independent industry by and large. Within this macro analysis on the national time-series data, the approaches adopted are mainly production functions.

**Figure 2.2** Annual growth rate of agricultural output in 1983-2010

#### **2.3.2 Review of previous studies**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**2.2.5 Veterinary residues in the livestock sector**

in women has some linkages with the abuse of hormones.

**2.3.1 Agricultural development over latest decades**

industry capable of self-reliance.

of some 6.29 % being maintained in this period.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**2.3 Development of Chinese agriculture and previous studies**

At the end of 1978, China launched *Reforms and Opening-up*, thus breaking up the highlyplanned economic institutions, and revitalizing the whole economy including those in rural areas. Up to the mid-1980s, as a prelude to the *Reform and Opening-up*, the *Household Contract Responsibility System* was expanded in rural areas across the country, where production teams were subsumed into 99% of villages and major production resources, symbolized by farmland, which were divided into household farms. After the reform, farmers could retain the rest of the goods and revenues as private property, once a certain amount of agricultural products or taxes were paid to the state as contracted. It released the long-term bound farming organizaȬ tions and increased farmers' motivation with regard to agricultural productivity, hence agricultural development reached a high level within a few years. As a fundamental measȬ urement, total grain yields amounted to 379 million tons in 1985, from 305 million tons in 1978. With this growth rate of some 25%, the problem of food security, which had puzzled China for a long time, was resolved by and large. In addition to benefiting national life and industrial development, it brought new opportunities to the overall economic reforms. At the same time, thanks to the effects of non-agricultural reforms, agriculture gradually developed as an

By the mid-1980s, the rapid development of agriculture was realized primarily due to the powerful potential released by institutional reforms. By contrast, in subsequent periods, agriculture maintained its high-speed growth, under the progress of overall economic reforms. From 1983 to 2010, China's Gross Agricultural Output (GAO) rose from 275 billion yuan to 69319.76 billion yuan (current prices). Accounting for the influences of inflation, the GAO was 1242.70 billion yuan (using the constant prices of 1983), with an average annual growth rate

The drug residues in livestock products contain drugs for animals and humans, disinfection chemicals, pesticides and other chemicals, the most prominent is residues of clenbuteral. Besides these, abuse of antibacterial drugs, hormones, vitamins and trace elements can cause residues of drugs. Drugs are not used properly, animals are slaughtered before the end of the drug-free period, or unapproved drugs are used as additives - all factors which can lead to contamination of the livestock produce. Once they are discharged with the manure, they will pollute the soil and water, and the environment. The potential threats of drug residues to human health include: allergy and abnormal response, taretogenicity, mutagenicity, carcinoȬ genicity and bacterial infection. For example, dosages of hormones, such as cortisone and hydrocortisone which are widely used by veterinary surgeons, can cause few residues. But large scale over dosage can lead to large amounts of residues. However, their impacts on humans might be visible only after several generations, e.g., femininity in men and masculinity

20 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

In the study period, the Chinese government deemed rural areas as regions with great potential to expand domestic demands, with agriculture as the primary industry in the rapid and stable economic growth. Therefore, in order to stabilize the *Household Contract Responsibility System*, further reforms were conducted on the institutions of pricing agricultural products, agriculȬ tural taxation, etc. In addition, to increase agricultural productivity and farmers' incomes, more funds were inputted to the development of agricultural sciences and technology, especially the innovation and extension of advanced agricultural production resources and new breeds. Thus, modernization of agriculture has been promoted, thanks to these policies beneficial to agriculture and technological advances. However, agriculture did not develop continuously and at a fast speed, and significant differences still existed between the annual growth rates. In particular, after a minus growth rate of 4.78% in 1989, an upheaval of 21.83% was revealed in 1990 (Figure 2.2).

With regard to the zigzagged growth curve of GAO mixed with increases and decreases, many scholarshave exploredthe causes fromdifferentperspectives.Lin(1992)[9] analysedtheoutput elasticityofeachfactorinagriculturaldevelopmentfrom1978to1984,usingprovince-levelpanel data. According to the conclusion, as the most important factors in the first half of the period, rural economic institutional reforms from production teams to the *Household Contract Responsi*Ȭ *bility System* supported the increase of agricultural production. Meanwhile, the significance of institutional reforms diminished sharply in the latter half of the period. In succession, fertilizȬ er applicationandtechnologicalprogress (throughtheproxyvariable of*T*) aremeasuredas also contributing greatly. To analyse significant factors behind the development of Chinese agriculȬ ture, this study includes three areas of analysis: factor inputs, institutional changes and technologicalprogress.However, as the studyperiodisupto1987,itnecessarytoconductfactor analysis of China's agricultural development within the following 20 years.

Huang et al (2005) [13] conducted an empirical analysis on the impact of changes in land ownership1 on agricultural growth in the period 1949-1978, from the founding of the People's Republic of China to the *Reforms and Opening-up*. The conclusions of this study illustrate the different effects of each factor on gross outputs of agriculture at different stages of land ownership. This research has a long but demoded study period, and did not include the variable of technological progress. Aiming at understanding the impacts of the agricultural innovation system, Qiao et al (2006) [14] analysed the significant factors of Chinese agricultural development in the period of 1978-2004 (sub-divided into five periods), based on the model specified by Griliches (1963) [15]. However, insignificant variables were included in some models for different periods such as *labour* and *power* in 1978-1984 and 1996-2002. Meanwhile, the study periods were divided into so many stages, especially including a two-year stage of 2003-2004, this reduced the accuracy of statistical analyses with models of multivariate regression, etc., thus blocking the accurate measurement of the whole study period from 1983 to 2006. In addition, this study did not include the contribution of technological progress.

Based on the panel data of provincial-level regions, Zhang et al. (2008) [11] analysed the developmentofChineseagricultureintheperiod1949-2005.Theresultindicatedthatthephysical inputs, particularly fertilizers and machinery, made a significant contribution to the total agricultural output, while farmland and labour contributed with lower or even negative ratios and large fluctuations. In this comprehensive empirical study, only the input elements were incorporated as determinants to agricultural development, while variables of technological progress and institutional change were excluded. In addition, Wang (2009) [10] studied the relationshipbetweentechnologicalprogress andeconomicdevelopmentinagriculture,withan extended Cobb-Douglas production function. It concluded that agricultural development increased the funds inputted on agricultural technical progress, while the latter needs to feed

<sup>1</sup> In the study period of this paper, land ownership in rural China passed through the stages of private ownership (1949-52), transition from private to collective ownership (1953-58), collective ownership through the people's commune (1959-62), and collective ownership of three subjects (people's commune, production brigade, production team), with the basis of production team (1963-78).

back the former mainly through the scientific conversion of concerning production elements and their organizations. Although this study was based on a long time period (1986-2004), the impacts of institutional change were not incorporated into the model (Table 2.6).

In previous studies, the development of Chinese agriculture was primarily attributed to three kinds of factors. The elemental inputs were the quantity of farmland, labour and agricultural assets, in addition to the liquid capitals of chemical fertilizers, etc. Institutional transitions referred to the changes of land ownership, agricultural price system, rural finance taxing forms, etc. Technology progresses included advances in farming methods related to increased production capacity of agricultural machinery and chemical fertilizers, and improved varieties of agricultural products. However, as noted above, there is still gap in research covering the period since 1983, when agriculture began to develop as an independent industry, with the adoption of the aforementioned factors to an integrated model, thus measuring the respective impacts on the development of Chinese agriculture. Meanwhile, further explorations are necessary in terms of the most appropriate indicator models to reflect the impacts from capital, land or other factors.

Therefore, with this in mind, based on data from the period 1983-2006 and production functions, and after thorough examination of the significance of each factor, this chapter selects a variety of indices with the availability of credible data, to demonstrate the impact of agricultural development. In detail, taking the 24-year period as a whole2 , the introduced time series data cover all the three types of variables as summarized above, i.e., inputs changes, technological progress and institutional transitions.

#### **2.4 Production mode of agricultural development**

#### **2.4.1 Theoretical model of production function**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

growth rates. In particular, after a minus growth rate of 4.78% in 1989, an upheaval of 21.83%

22 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

With regard to the zigzagged growth curve of GAO mixed with increases and decreases, many scholarshave exploredthe causes fromdifferentperspectives.Lin(1992)[9] analysedtheoutput elasticityofeachfactorinagriculturaldevelopmentfrom1978to1984,usingprovince-levelpanel data. According to the conclusion, as the most important factors in the first half of the period, rural economic institutional reforms from production teams to the *Household Contract Responsi*Ȭ *bility System* supported the increase of agricultural production. Meanwhile, the significance of institutional reforms diminished sharply in the latter half of the period. In succession, fertilizȬ er applicationandtechnologicalprogress (throughtheproxyvariable of*T*) aremeasuredas also contributing greatly. To analyse significant factors behind the development of Chinese agriculȬ ture, this study includes three areas of analysis: factor inputs, institutional changes and technologicalprogress.However, as the studyperiodisupto1987,itnecessarytoconductfactor

Huang et al (2005) [13] conducted an empirical analysis on the impact of changes in land ownership1 on agricultural growth in the period 1949-1978, from the founding of the People's Republic of China to the *Reforms and Opening-up*. The conclusions of this study illustrate the different effects of each factor on gross outputs of agriculture at different stages of land ownership. This research has a long but demoded study period, and did not include the variable of technological progress. Aiming at understanding the impacts of the agricultural innovation system, Qiao et al (2006) [14] analysed the significant factors of Chinese agricultural development in the period of 1978-2004 (sub-divided into five periods), based on the model specified by Griliches (1963) [15]. However, insignificant variables were included in some models for different periods such as *labour* and *power* in 1978-1984 and 1996-2002. Meanwhile, the study periods were divided into so many stages, especially including a two-year stage of 2003-2004, this reduced the accuracy of statistical analyses with models of multivariate regression, etc., thus blocking the accurate measurement of the whole study period from 1983 to 2006. In addition, this study did not include the contribution of technological progress.

Based on the panel data of provincial-level regions, Zhang et al. (2008) [11] analysed the developmentofChineseagricultureintheperiod1949-2005.Theresultindicatedthatthephysical inputs, particularly fertilizers and machinery, made a significant contribution to the total agricultural output, while farmland and labour contributed with lower or even negative ratios and large fluctuations. In this comprehensive empirical study, only the input elements were incorporated as determinants to agricultural development, while variables of technological progress and institutional change were excluded. In addition, Wang (2009) [10] studied the relationshipbetweentechnologicalprogress andeconomicdevelopmentinagriculture,withan extended Cobb-Douglas production function. It concluded that agricultural development increased the funds inputted on agricultural technical progress, while the latter needs to feed

1 In the study period of this paper, land ownership in rural China passed through the stages of private ownership (1949-52), transition from private to collective ownership (1953-58), collective ownership through the people's commune (1959-62), and collective ownership of three subjects (people's commune, production brigade, production team), with the

analysis of China's agricultural development within the following 20 years.

was revealed in 1990 (Figure 2.2).

basis of production team (1963-78).

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In studies about sources of economic development, Cobb-Douglas production functions3 are widely used to demonstrate the relationships between economic growth and inputting factors. As the original theoretical model, Cobb-Douglas production function is represented by the following formula:

$$Y = \beta\_0 e^{\theta t} \prod\_{n=1}^{N} \mathbf{x}\_n^{\ \beta\_n} \tag{2-1}$$

<sup>2</sup> To illustrate the impact of institutional changes, the author introduced several dummy variables and estimated the study period in different phases. However, the results did not show significant trends in terms of institutional changes, due to the short periods, thus illustrating the statistical insignificance of each model.

<sup>3</sup> In the studies of relationships between economic growth and inputting factors, in addition to production function, cost function and profit function are often also used. Nevertheless, independent price variables are needed in both of the latter two functions. In addition, cross-sectional data were used in many prior studies on cost and profit functions [16]. In China, only part of the price data of production factors has been published. Therefore, this chapter conducts factor analysis of Chinese agricultural development with the adoption of production function.

Here, *x*n represents the inputting factors of capital, labour, etc.; Άn is the elasticity of each factor; Ά<sup>0</sup> includes all the other factors as Total Factor Productivity (TFP); *t* is a proxy variable for the time trend variable of technological progress; Ά0, Άn and Ό are unknown parameters to be estimated.

Taking the natural logarithm on both sides of Eq.2-1, and calculating the partial differential of ln*Y* with *t*:

$$\frac{\partial \ln Y}{\partial t} = \theta \tag{2-2}$$

where Ό represents the rate of technological progress [17]. Hence, the Cobb-Douglas specifiȬ cation of production function implicitly assumes the technological change effect is constant to the output Y [18]. Meanwhile, as the Cobb-Douglas specification of production function is homothetic, thus we assume that the substitute elasticity between different factors' constant to be 14 [16].

In theory, using the model described above, we can compute the contribution of technological progress over time. However, as the rate of technological change is not constant every year, there are extreme possible difficulties in observing the contribution over time. Therefore, another specification of production function is needed as follows:

$$Y = \beta\_0 \prod\_{n=1}^{N} \mathbf{x}\_n^{\beta\_n} \tag{2-3}$$

Here, contribution of technological progress can be calculated with:

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

$$\mathbf{M}\_{Tech} = \mathbf{1} - \sum\_{n=1}^{N} \mathbf{M}\_n \tag{2-4}$$

That is, contribution of technological progress (*M*Tech) is obtained as the residual of subtracting the contribution of other factors (*M*n) from the growth rate of *Y* [19, 20]. This thus provides another basic method to estimate the contribution of technological progress, based on the Cobb-Douglas production function [21]. In this study, after comparing the results of the two models, the better performing Eq.2-4 is adopted.

<sup>4</sup> Despite the single homogeneous assumption of substitutability between the elements, similar homogeneity (i.e., constant returns to scale) is not assumed for the returns to scale, which is determined by the parameters to be estimated. For example, if ̕Άn= 1 means constant returns to scale; ̕Άn<1 indicates the diminishing returns to scale; ̕Άn>1 denotes increasing returns to scale.


Note: the agro-power is the sum of the energy used in agriculture; price ratio is the ratio of price index of agricultural products and production resources; price reform is represented by ratio of product prices determined by the government; tax and cost reform is represented by the proportion of agricultural tax in total agricultural production; financial support refers to the proportion of fiscal inputs to agriculture in total public budgets

**Table 2.6** Factor estimations of China's agricultural development in previous studies

#### **2.4.2 Indicators and data**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Here, *x*n represents the inputting factors of capital, labour, etc.; Άn is the elasticity of each factor; Ά<sup>0</sup> includes all the other factors as Total Factor Productivity (TFP); *t* is a proxy variable for the time trend variable of technological progress; Ά0, Άn and Ό are unknown parameters to be

24 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Taking the natural logarithm on both sides of Eq.2-1, and calculating the partial differential of

T

where Ό represents the rate of technological progress [17]. Hence, the Cobb-Douglas specifiȬ cation of production function implicitly assumes the technological change effect is constant to the output Y [18]. Meanwhile, as the Cobb-Douglas specification of production function is homothetic, thus we assume that the substitute elasticity between different factors' constant

In theory, using the model described above, we can compute the contribution of technological progress over time. However, as the rate of technological change is not constant every year, there are extreme possible difficulties in observing the contribution over time. Therefore,

*n*

E

1

=

That is, contribution of technological progress (*M*Tech) is obtained as the residual of subtracting the contribution of other factors (*M*n) from the growth rate of *Y* [19, 20]. This thus provides another basic method to estimate the contribution of technological progress, based on the Cobb-Douglas production function [21]. In this study, after comparing the results of the two

4 Despite the single homogeneous assumption of substitutability between the elements, similar homogeneity (i.e., constant returns to scale) is not assumed for the returns to scale, which is determined by the parameters to be estimated. For example, if ̕Άn= 1 means constant returns to scale; ̕Άn<1 indicates the diminishing returns to scale; ̕Άn>1 denotes

0 1

1 *N Tech n n M M*

E

*N n n Y x*

=

another specification of production function is needed as follows:

Here, contribution of technological progress can be calculated with:

models, the better performing Eq.2-4 is adopted.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

increasing returns to scale.

<sup>w</sup> <sup>=</sup> <sup>w</sup> (2-2)

<sup>=</sup> (2-3)

¦ (2-4)

ln*Y t*

estimated.

ln*Y* with *t*:

to be 14

[16].

In this chapter, to describe the development of Chinese agriculture and the factors over the period, indicators shown in Table 2.7 are adopted. The data sources include Bulletin of Chinese Agricultural Development (2007) and China Statistical Yearbook (relevant years), published by China's Ministry of Agriculture and State Statistical Bureau. Considering the impacts of time trend, all the monetary values are calculated in the constant prices of 1983.


Note: a As the prime currency unit, 7.97 yuan = 1 US\$ (middle exchange rate of 2006) and all the monetary values are calculated in the constant prices of 1983

Source: China Statistical Press [8] & China Ministry of Agriculture (2007) [12]

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 2.7** Summary statistics of Chinese agricultural development in the period 1983-2006

In the first place, as the dependent variable, Gross Agricultural Output (*Y*) is the total output value of the final products of agricultural activities, including farming, forestry, animal husbandry and fishery. The gross output of each agricultural product (*Y*<sup>t</sup> ) is obtained by multiplying the price and physical volumes of production, and then converting to the constant prices of 1983.

Due to the existence of multiple cropping in agricultural production, Sowing Area of AgriȬ cultural Land (*Ld*), rather than the areas of arable land, is adopted [14]. In origin, Labour Force (*Lb*) should be represented with of total working days or hours in a year, etc. However, viewing from the perspective of the real status of Chinese farmers, it is difficult to accurately measure their labouring times. Meanwhile, relevant data is not found from the *China Statistical Year*Ȭ *book*, *China Agricultural Yearbook*, and other sources. Therefore, referring to the earlier literature [9, 14], annual number of agricultural labours (10 thousand persons per year) is adopted in this study.

Agricultural capitals are divided into fixed and liquid capitals. The *value of fixed assets* (*Ats*) is the monetary expression of objects, tools and equipment directly used in agricultural producȬ tion, borrowed or owned by farms over a relatively long period of several years. Power Agricultural Machineries (*Pw*) is the sum of energy with machineries used in ploughing, irrigation, harvesting and transportation, etc., within the agricultural activities of farming, forestry, animal husbandry and fishery. In order to identify appropriate variables to represent the fixed capitals, *Pw* and *Ats* are incorporated into the model simultaneously. At the same time, as the most important liquid capital, Amounts of Chemical Fertilizer (*Fert*) refers to the standardized quantity of nitrogen, phosphorus, potash and compound fertilizers used in agricultural production. Here, the standardization depends on the content of nitrogen, phosphorus pentoxide, potassium, etc., in different types of fertilizers.

Additionally, three indicators are included to reflect the impact and effectiveness of instituȬ tional reforms in the study period, concerning agricultural commodity prices, agricultural taxation, aids and assistance to agricultural production, etc. Price Indices Ratio of Agro-Products and Inputting Materials (price ratio, *Rp*) is the ratio of price index and producer price indices for agricultural materials in each year. The series of reforms carried out in the field of agriculture began from institutions of commodity prices in the early 1980s. Thereafter, the price system once generally controlled by the state is gradually being reformed over a long period. By 2004, fixed purchase prices were completely abolished and grain prices began to be fully determined by the market. Meanwhile, Ratio of Agricultural Taxes (*Rt*) is the percentage of agricultural taxes of national fiscal revenues in each year. The agro-supporting funds are mainly used to finance agricultural production, irrigation, climate forecasting, infrastructure, R&D, etc. The Ratio of Fiscal Agro-Supporting Funds (*Rf*) refers to the percentage of public funds within Gross Agricultural Output (GAO). Using these two indicators, we intend to evaluate the impact arising from the reforms of agricultural taxation and fiscal institutions. From 2000, rural tax reforms were included in the unified reforms managed by the central government and from 2006, agricultural taxes were fully abolished nationwide and subsidies supporting agricultural production began to be directly distributed to farmers. At the same time, reform of the budgetary expenditure on agriculture finance came into force. In 2004, to balance the socio-economic development of urban and rural areas, the No.1 document issued by the top authorities proposed the key guideline for the rural policies as 'Giving More, Taking Less and Loosening Control', stressing that the government would increase its input to rural areas and agriculture, and reduce taxes and fees collected from farmers. Meanwhile, in the same document, another policy agenda committed to transforming the lack of financial agriculture input5 .

#### **2.4.3 Results of the estimation**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Var. Description Unit <sup>1983</sup> <sup>2006</sup> Annual growth**

Y Gross Agricultural Output (GAO) billion yuan a 275.00 1242.70 6.78 Ld Sowing area of agricultural plants million ha 143.99 157.02 0.38 Ats Value of fixed agricultural assets billion yuan 53.40 237.82 6.71 Pw Power of agricultural machineries million kw 180.22 726.36 6.25 Fert Amounts of chemical fertilizer million ton 16.60 48.34 4.76 Lb Number of agricultural labours 10000 person 316.45 294.05 -0.32

26 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Rt Ratio of agricultural taxes in fiscal revenue % 4.25 0.95 -6.31 Rf Ratio of fiscal agro-aiding funds in GAO % 4.83 7.48 1.92

In the first place, as the dependent variable, Gross Agricultural Output (*Y*) is the total output value of the final products of agricultural activities, including farming, forestry, animal

multiplying the price and physical volumes of production, and then converting to the constant

Due to the existence of multiple cropping in agricultural production, Sowing Area of AgriȬ cultural Land (*Ld*), rather than the areas of arable land, is adopted [14]. In origin, Labour Force (*Lb*) should be represented with of total working days or hours in a year, etc. However, viewing from the perspective of the real status of Chinese farmers, it is difficult to accurately measure their labouring times. Meanwhile, relevant data is not found from the *China Statistical Year*Ȭ *book*, *China Agricultural Yearbook*, and other sources. Therefore, referring to the earlier literature [9, 14], annual number of agricultural labours (10 thousand persons per year) is adopted in

Agricultural capitals are divided into fixed and liquid capitals. The *value of fixed assets* (*Ats*) is the monetary expression of objects, tools and equipment directly used in agricultural producȬ tion, borrowed or owned by farms over a relatively long period of several years. Power Agricultural Machineries (*Pw*) is the sum of energy with machineries used in ploughing, irrigation, harvesting and transportation, etc., within the agricultural activities of farming, forestry, animal husbandry and fishery. In order to identify appropriate variables to represent the fixed capitals, *Pw* and *Ats* are incorporated into the model simultaneously. At the same time, as the most important liquid capital, Amounts of Chemical Fertilizer (*Fert*) refers to the standardized quantity of nitrogen, phosphorus, potash and compound fertilizers used in

As the prime currency unit, 7.97 yuan = 1 US\$ (middle exchange rate of 2006) and all the monetary values are

Rp Price indices ratio of agro-products and inputting materials

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Source: China Statistical Press [8] & China Ministry of Agriculture (2007) [12]

**Table 2.7** Summary statistics of Chinese agricultural development in the period 1983-2006

husbandry and fishery. The gross output of each agricultural product (*Y*<sup>t</sup>

calculated in the constant prices of 1983

Note: a

prices of 1983.

this study.

**(%)**

) is obtained by

% 101.36 99.70 -0.07

In this study, factors analysis of agricultural is conducted through the development of an econometric model without the inclusion of time variable, based on the log-linear Cobb-Douglas production function as:

lnY=Constant+΅ln*Ld*+Ά1ln*Ats*+Ά2ln*Pw* + Ά3ln*Fert* + ·ln*Lb* + Έ1ln*Rp* + Έ2ln*Rt* + Έ3ln*Rf* + Ή (2-5)

where *Constant* is the intercept, ΅, Ά<sup>i</sup> , · and Έ<sup>i</sup> are unknown parameters to be estimated, and Ή is the random item.

Although we can include all the above variables into the final model and obtain higher fitness, it is better to develop models by omitting redundant variables which hardly contribute the

<sup>5</sup> In terms of the agricultural financial inputs, the total sum draws much more attention than the proportion of annual government expenditure.In recent years, the fiscal inputs to support agriculture have increased, while the proportion ofannual government expenditure decreased. In 1983-2006, the proportion decreased from 9.43% to 7.85% (China Statistical Yearbook).

total fitness. In econometric models, the change of determinant coefficient ̇*R*<sup>2</sup> , the change of *F* (̇*F*) and the probability significance of *p*̇ F are referential in selecting the variables [22]. After removing the insignificant variables according to the probability significance of *p*̇ F obtained by the software SPSS, the combination of the explanatory variables in the final model include four significant variables as shown in Table 2.8.

All the significant *F* and *t*-test at the level of 5%, the Adj.*R*<sup>2</sup> of 0.99 and Durbin-Watson value of 2.285 indicate good statistical fitness. In addition, fixed assets, chemical fertilizers and ratio of fiscal agro-supporting funds are all estimated with positive elasticity. Although the negative elasticity of agricultural labour is adverse to the general economic assumption, it meets with existence of surplus amounts of labour in Chinese agricultural production. In previous studies, both Lin (1992) [9] and Zhang et al (2008) [11] measured the negative elasticity of labour productivity. Therefore, this model estimates well China's agricultural and economic growth in the study period.


Note: \*\*\*, \*\*and \* represent statistical significance in the level of 1%, 5% and 10% respectively

Software: SPSS 13.0

#### **Table 2.8** Estimation of the production elasticity

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

With regard to the causes of the significant ratio of fiscal agro-supporting funds and insignifȬ icant pricing factors, these may be due to lower prices of agricultural products compared with the prices of fertilizers and other inputting industrial products, thus farmers find it difficult to attain positive agricultural productivity. In terms of the sowing area of agricultural plants, insignificance may result mainly from multicollinearity, as high relation coefficients of 0.95 and 0.87 exist between this variable and the value of fixed assets and fertilizer, respectively. Meanwhile, viewing from changes of inputs over the study period, when sowing area of agricultural plants increased 9.05%, the value of fixed assets and amount of fertilizer increased 345.38% and 191.26%, respectively. With respect to the major crops, acreage of grains and cotton declined 7.50% and 11%, respectively, thanks to the increased per unit yields of 38.88% and 63.45% - total yields eventually increased by 28.45% and 45.48%, respectively. Similarly, total yields of oil crops increased by 189.99%, due to the increased per unit yields of 77.12%. To sum up, in the study period, comparing with the physical inputs of fertilizers and fixed assets, etc., hence the increased in yield per unit, sowing area exerted slightly smaller effects, thus the insignificant result in the quantitative model is plausible.

#### **2.4.4 Contribution of each factor**

total fitness. In econometric models, the change of determinant coefficient ̇*R*<sup>2</sup>

28 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

four significant variables as shown in Table 2.8.

in the study period.

Note: \*\*\*,

Software: SPSS 13.0

**Table 2.8** Estimation of the production elasticity

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

*F* (̇*F*) and the probability significance of *p*̇ F are referential in selecting the variables [22]. After removing the insignificant variables according to the probability significance of *p*̇ F obtained by the software SPSS, the combination of the explanatory variables in the final model include

All the significant *F* and *t*-test at the level of 5%, the Adj.*R*<sup>2</sup> of 0.99 and Durbin-Watson value of 2.285 indicate good statistical fitness. In addition, fixed assets, chemical fertilizers and ratio of fiscal agro-supporting funds are all estimated with positive elasticity. Although the negative elasticity of agricultural labour is adverse to the general economic assumption, it meets with existence of surplus amounts of labour in Chinese agricultural production. In previous studies, both Lin (1992) [9] and Zhang et al (2008) [11] measured the negative elasticity of labour productivity. Therefore, this model estimates well China's agricultural and economic growth

**Variable Constant Lb Ats Fert Rf** Elasticity 7.672\*\* -0.824\*\* 0.159\*\*\* 0.988\*\*\* 0.306\*\*\* *t* (2.436) (-2.665) (2.848) (12.265) (4.824)

With regard to the causes of the significant ratio of fiscal agro-supporting funds and insignifȬ icant pricing factors, these may be due to lower prices of agricultural products compared with the prices of fertilizers and other inputting industrial products, thus farmers find it difficult to attain positive agricultural productivity. In terms of the sowing area of agricultural plants, insignificance may result mainly from multicollinearity, as high relation coefficients of 0.95 and 0.87 exist between this variable and the value of fixed assets and fertilizer, respectively. Meanwhile, viewing from changes of inputs over the study period, when sowing area of agricultural plants increased 9.05%, the value of fixed assets and amount of fertilizer increased 345.38% and 191.26%, respectively. With respect to the major crops, acreage of grains and cotton declined 7.50% and 11%, respectively, thanks to the increased per unit yields of 38.88% and 63.45% - total yields eventually increased by 28.45% and 45.48%, respectively. Similarly, total yields of oil crops increased by 189.99%, due to the increased per unit yields of 77.12%. To sum up, in the study period, comparing with the physical inputs of fertilizers and fixed assets, etc., hence the increased in yield per unit, sowing area exerted slightly smaller effects,

Indicator Sample size F Adj.R2 D-W Value 24 586.085\*\*\* 0.99 2.285

\*\*and \* represent statistical significance in the level of 1%, 5% and 10% respectively

thus the insignificant result in the quantitative model is plausible.

, the change of

In the study period, although the gross agricultural output increased by 351.89%, the amount of agricultural labour decreased by 7.08%, from 316 million to 294 million. In addition, the value of agricultural fixed assets increased from 53.4 billion yuan to 237.8 billion yuan, increasing more than four times; the three-fold increase of fertilizers meant an increase of 48.34 million tons (rising by 16.60 million tons). The ratio of fiscal agro-supporting funds in gross agricultural output also rose from 4.83% to 7.48%. Based on the multiplication of these changes on each factor and the corresponding elasticity, the contribution rate of agricultural growth can be assessed for each factor using percentage within total agricultural output. Furthermore, as shown in Eq.2-4, contribution of agricultural technological progress can be estimated by subtracting the contributions of the other factors (Table 2.9). In addition, as investment on agricultural R&D is already included in financial support for agriculture, the investment on agricultural R&D from the government is not included in the technological progress in this context.


a Contribution of technological progress (*Tech*) is calculated based on Eq.2-4

Software: Excel 2007

**Table 2.9** Contribution of each factor (1983-2006)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

According to the results of Table 2.9, within the growth rate of 351.89% of gross agricultural production in the study period, the increased amount of fertilizers, value of fixed assets, amount of financial support and the reduction of agricultural labour force, contributed 53.70%, 15.57%, 4.77% and 1.66%, respectively. Meanwhile, being the residual of the four variables, technological progress contributed 24.30% to Chinese agriculture development. Among the factors, increase in the amount of fertilizer inputs is the most significant factor, followed by technological progress. These two factors accounted for 78% of gross agricultural output growth, constituting major causes of Chinese agricultural development over the study period.

In succession, the value of fixed assets indicates great increases but small elasticity, thus the contribution remained 15.57%. As a proxy of institutional changes, the ratio of fiscal agrosupporting funds in GAO makes a small contribution of 4.77% in the study period. In terms of the minus and small elasticity of agro-labour, 1.66% is contributed due to the decreased numbers over the study period.

These results are in line with the conclusions of prior studies. Firstly, with regard to the basic agricultural production resources, the detected significant effects of chemical fertilizer are similar to those found in Lin (1992) [9], Zhang et al (2008) [11] and Qiao (2006) [14] and many other studies. As a key factor in second place, the importance of technological progress is measured in Wang (2009) [10] and other prior literature. As for the negative elasticity of agrolabour, which is in line with Lin (1992) [9] and Qiao (2006) [14], this indicates that transferring of agro-labour numbers have contributed to Chinese agricultural development. The major reasons behind this include engaging in other sectors enabled the farmers to obtain more funds to invest in fertilizers and fixed agricultural assets. Meanwhile, the non-agricultural experiȬ ences are beneficial in improving farm management and the trade of agro-products.

#### **2.4.5 Major conclusions and recommendations**

#### **1.** Findings and conclusions

In this chapter, a factor analysis of Chinese agriculture development in the period 1983-2006 is conducted, from the perspectives of inputs change, institutional transition and technological progress. As a result, we did not ascertain new findings and similar results to prior studies were obtained, using comprehensive perspectives, overall and long-term modelling and a comparison of different models in measuring the contribution of technological advances, etc.

With regard to the statistical significance of each factor, with the increment of chemical fertilizer was in first place, fixed agricultural assets next, followed by financial support and the reduction of agricultural labour force – these all constituted the major factors supporting China's agricultural development in the study period. In previous literature, different factors were detected as the first factor in different stages, such as agricultural technology in Lin (1992) [9], agricultural machinery and financial assistance in Qiao (2006) [14], agricultural machinery and labour force in Zhang et al (2008) [11], etc. (Table 2.6). In contrast, increased input of fertilizer is measured as the most important factor for China's agricultural development in the period 1983-2006, with an overwhelming contribution share. In addition, as the second factor, technological progress is concluded as supporting agricultural development with a considerȬ able share of contribution. Different from the models in Zhang et al (2008) [11] and Qiao (2006) [14], Wang J. (2009) [10] considered the significance of agricultural technology, although it was measured as contributing the lowest share among three types of factors. Inaccurate measureȬ ment of the contribution of agricultural technological progress will inevitably lead to a misunderstanding of the significant factors and thus policy recommendations with regard to agricultural development. Finally, although Lin J. Y. (1992) [9] and Qiao Z., et al (2006) [14] have suggested that institutional changes is a key factor for China's agricultural growth since the middle of the 1980s, this study shows that compared with the other factors, institutional changes holds a relatively low contribution share. It suggests that since the mid-1980s, despite a series of policies in favour of agriculture, few fundamental institutional changes, like the household contracting system, were carried out or institutions were not implemented effecȬ tively.

#### **2.** Policy recommendations

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Based on the above conclusions, the following policy recommendations can be made to accelerate agricultural development. First, the increased input of chemical fertilizers has contributed significantly and the amount is expected to keep increasing in the future. NeverȬ theless, the realization of sustainable and environmentally friendly agriculture has become an important issue. Therefore, for the safe application of chemical fertilizers, the government needs to extend soil surveying techniques, and promote the proper classification and approȬ priate amounts of fertilizers. Thereby, increase agricultural productivity while savings fertilizer costs and protecting the environment.

Meanwhile, as another important factor, advances in agricultural technology should be accelerated from the three perspectives of R&D, extension and funds. At present, although agricultural technology has developed rapidly in China, problems still remain in transferring and spreading the techniques to the fields. When analysing the causes, this can be attributed to the overwhelming ratio of household management, thus the small sizes of farmland in agriculture. Individual farms are limited in willingness and ability to introduce new agriculȬ tural technologies. In the governmental institutions specializing in extending agricultural technologies, there is poor connection between staff incomes and their achievements in extending agricultural technology, thus there is a lack of initiative and this can be pointed to as another reason. Thus, in addition to enlarging the managerial scales through encouraging the transfer of farmland use rights, the adoption of more marketing mechanisms to the agricultural extension institutions at the grassroots level, simultaneously constitutes an urgent measure.

At the same time, countermeasures are needed to serve the needs of agricultural labour, the number of which is estimated with a significant but minus production elasticity as shown above. To reduce the amount of agricultural labour, further endeavours are necessary to strengthen the non-agricultural vocational training of rural labour through the *Sunshine Project*<sup>6</sup> , accelerating the reform of the family registration system, so as to shift surplus labour to both the urban areas and local non-agricultural sectors.

In addition, as an important factor of institutional change, fiscal agro-supporting funds need to be increased. Despite the relatively low capital elasticity, the increase of fiscal agrosupporting funds is highly beneficial to increase the value of fixed assets and the promotion of agricultural technology advances. Thus, to strengthen the reforms of the financial budget on agriculture and transform the lack of agricultural fiscal inputs, it necessary to ensure sources of funds are channelled to agriculture, from a series of sources including governmental departments and financial institutions. Specifically, the main roles of government include the investment in agricultural infrastructure construction, and subsidies on the purchase of agricultural machinery and good seeds. Meanwhile, financial institutions are expected to create preferential prerequisites to provide more loans to farmers.

**3.** Further discussion

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

similar to those found in Lin (1992) [9], Zhang et al (2008) [11] and Qiao (2006) [14] and many other studies. As a key factor in second place, the importance of technological progress is measured in Wang (2009) [10] and other prior literature. As for the negative elasticity of agrolabour, which is in line with Lin (1992) [9] and Qiao (2006) [14], this indicates that transferring of agro-labour numbers have contributed to Chinese agricultural development. The major reasons behind this include engaging in other sectors enabled the farmers to obtain more funds to invest in fertilizers and fixed agricultural assets. Meanwhile, the non-agricultural experiȬ

In this chapter, a factor analysis of Chinese agriculture development in the period 1983-2006 is conducted, from the perspectives of inputs change, institutional transition and technological progress. As a result, we did not ascertain new findings and similar results to prior studies were obtained, using comprehensive perspectives, overall and long-term modelling and a comparison of different models in measuring the contribution of technological advances, etc.

With regard to the statistical significance of each factor, with the increment of chemical fertilizer was in first place, fixed agricultural assets next, followed by financial support and the reduction of agricultural labour force – these all constituted the major factors supporting China's agricultural development in the study period. In previous literature, different factors were detected as the first factor in different stages, such as agricultural technology in Lin (1992) [9], agricultural machinery and financial assistance in Qiao (2006) [14], agricultural machinery and labour force in Zhang et al (2008) [11], etc. (Table 2.6). In contrast, increased input of fertilizer is measured as the most important factor for China's agricultural development in the period 1983-2006, with an overwhelming contribution share. In addition, as the second factor, technological progress is concluded as supporting agricultural development with a considerȬ able share of contribution. Different from the models in Zhang et al (2008) [11] and Qiao (2006) [14], Wang J. (2009) [10] considered the significance of agricultural technology, although it was measured as contributing the lowest share among three types of factors. Inaccurate measureȬ ment of the contribution of agricultural technological progress will inevitably lead to a misunderstanding of the significant factors and thus policy recommendations with regard to agricultural development. Finally, although Lin J. Y. (1992) [9] and Qiao Z., et al (2006) [14] have suggested that institutional changes is a key factor for China's agricultural growth since the middle of the 1980s, this study shows that compared with the other factors, institutional changes holds a relatively low contribution share. It suggests that since the mid-1980s, despite a series of policies in favour of agriculture, few fundamental institutional changes, like the household contracting system, were carried out or institutions were not implemented effecȬ

Based on the above conclusions, the following policy recommendations can be made to accelerate agricultural development. First, the increased input of chemical fertilizers has

ences are beneficial in improving farm management and the trade of agro-products.

30 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**2.4.5 Major conclusions and recommendations**

**1.** Findings and conclusions

tively.

**2.** Policy recommendations

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In recent years, with China's rapid economic development, agricultural inputs are being increased, and the extension of agricultural technology is being enhanced. In addition, with

<sup>6</sup> *The Sunlight Project is a series of technical and vocational training programmes targeting the rural labour, carried out by the Chinese government since 2004. The project aims to improve the quality and skills of rural labours, thus promoting their employment in rural non-agricultural sectors and urban areas.*

the overall abolition of agricultural taxes and direct aid to agricultural production, etc., a series of agro-supporting policies were carried out by 2006. Therefore, although this study could not fully grasp the impact of these policies, further studies are necessary to assess China's overall agricultural growth factors, using annual data since 2007.

#### **References**


[12] CMA (China Ministry of Agriculture). Bulletin of Chinese Agricultural Development 2007: http://www.agri.gov.cn/sjzl/baipshu.htm.

the overall abolition of agricultural taxes and direct aid to agricultural production, etc., a series of agro-supporting policies were carried out by 2006. Therefore, although this study could not fully grasp the impact of these policies, further studies are necessary to assess China's overall

32 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

[1] Zhu L., Liang S., et al. Investigation on the agricultural pollution and prevention modes in different regions, by Institute of Agricultural Economics and Development, Chinese

[2] Edwin D. O., Zhang X., et al. Current status of agricultural and rural non-point source pollution assessment in China, Environmental Pollution 2010; (158): 1159–1168.

[3] Jia R., Lu Q., et al. Current Situation, causes and countermeasures of the agricultural pollution in China, Chinese Journal of Review of China Agricultural Science and

[4] Lin H., Li X. Effects and countermeasures of agricultural pollution to the quality safety of agricultural products, Chinese Journal of Ecological Economy 2009; (9): 146-149, 153.

[5] Wu S. 1991, Preliminary investigation of water N concentration within rural area in lower reach of Yangtze in China. Paper presented at the International conference on Agriculture and environment. Ohio State University, Columbus, Ohio, 10-13 NovemȬ

[6] Liu M., Du L., Zhang X., Farmers' willingness on organic fertilizer application based on logit model and influencing factors Journal of Anhui Agricultural science 2010; 38

[7] Wang Z. Analysis on the influencing factors of fertilization behavior of farmers: Also on scientific fertilization and environmental pollution. Annual Report of Economic and Technological Development in Agriculture. China Agricultural Press 2011: 211-219.

[8] CSP (China Statistical Press). China Statistical Yearbook: http://www.stats.gov.cn/tjsj/

[9] Lin J. Y. Rural Reforms and Agricultural Growth in China, the American Economic

[10] Wang J. Study on the relationship between inputs on R&D and economic growth in Chinese Agriculture, Chinese Journal of Journal of Agritechnical Economics 2009; (1):

[11] Zhang H., Chen Z. Analysis on the factor contribution to Chinese agriculture: study based on unstable panel data model, Chinese Journal South Economy 2008; (1): 61-75.

agricultural growth factors, using annual data since 2007.

Academy of Agricultural Sciences, August, 2009.

Technology2006; 8(1): 59-63.

ber, 1991; 53(2): 569-574.

Review 1992; 82 (1): 34-51.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

(9): 4827-4829.

ndsj/.

103 -109.

**References**


### **Situation and Determinants of Agro-Environment**

Tinggui Chen and Min Song

### **3.1 Introduction**

Along with China's fast economic development, environmental problems in rural areas are becoming increasingly severe. To tackle this problem in the rural environment, the Chinese government proposed the building of a 'Socialist New Countryside' since 2005, promoting the capacity and management to create clean villages as one target in this nationwide project. In April2005,ruralrubbishbegantobeincludedintherevisedversionoftheSolidWastesPollution PreventionLaw. "Adviceonstrengtheningthe rural environmentalprotection",releasedbythe State Council on November 13, 2007, put forward specific targets to improve the quality of rural drinkingwater, andcontrol soil andagriculturalpollutionsources.Furthermore, "ImplementaȬ tion of awarding for promoting the governance and settlement of rural environment outstandȬ ingissues",releasedbyStateCouncilonFebruary27,2009,raisedtheobjectivesandeffectiveness requirements of rural environment governance. Now Beijing, Shanghai, Zhejiang and other developedareasareimplementingrubbishtreatmentprojectsandtheawarenessofenvironmenȬ tal protection among rural residents is becoming stronger and stronger [1-2].

Based on a nationwide survey of 141 villages, Tang et al (2008) [3] analyses the problem of environmental pollution in rural China. The study shows that 49.53% of the villagers think civil rubbish is the main source resulting in environmental pollution in rural areas, while 32.71% of the villagers attribute the most important sources to industrial pollution, and in the third place, 9.35% of them selected chemical fertilizers as the major source of environmental pollution in rural areas. In order to have a better understanding of the current situation of rural environmental pollution and its major determinants, Huang et al (2010) [4] conducts a nationwide survey, with a representative sample of 101 villages from five provinces. The result indicates that about 44% of the villagers think that their environment became worse over the past 10 years. The following econometric analysis shows that rural enterprise development and local township expansion (or rising population density) has negatively affected the rural environment, while government's efforts have a positive effect on slowing down the pace of environmental pollution.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Chen and Song; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Chen and Song; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Based on the survey of 21 villages of eastern China's six provincial regions, this chapter aims to analyze the situation of agricultural production, the public service in rural regions and the relationship between the rural environment and economic level.

#### **3.2 Samples of the survey**

To understand the present situation and farmers' perceptions on agricultural pollution, we conducted the survey with questionnaire-based personal interviews to collect first-hand data. In the first section, our questionnaire contains the basic characteristics of each village, including general economic situation, land and land use. In the second section of the questionnaire, we investigate the situation of agricultural production, including crops, animal products, aquatic products, certificated agricultural production and irrigation. The following questions are designed to ascertain information about the rural environment, including rubbish collection, disposal of liquid and solid waste, construction of methane tanks, etc. In the final section, we collect rural public services data, such as agricultural production, rural life, government subsidies and village public expenditure. Simultaneously, we designed another questionnaire to understand farmers' perceptions on agricultural pollution, including basic characteristics of each farm, disposal of household rubbish, farmers' selection and application of fertilizers, pesticides and veterinary drugs, farmers' perceptions on agricultural pollution, and informaȬ tion about and recognition of safe agricultural production.

In January to March, 2011, we surveyed 21 villages and their 560 farms in eastern China's six provincial regions, including Beijing, Hebei, Shandong, Shanghai, Jiangsu and Zhejiang (Figure 3.1). The sampled area covers three major gain growing provincial regions and rural regions affiliating to the top two metropolises of China. The former three regions represent the northern mode of agricultural production in the Yellow River Basin, while the latter three demonstrate the characteristics of agricultural production in south China's Yangtze River Basin. With regard to topographic types, farms located on plains, hills and mountainous regions, and villages in inland areas, seaside areas and areas adjoining the metropolises are sampled. In addition to the staple grain crops of wheat, rice and corn, the other major agriȬ cultural products, including cotton, vegetables, fruit, oil crops, etc., and the main livestock, poultry, aquaculture products are being grown and cultivated in the sampled farms.

#### **3.3 Basic situation of the sampled areas**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In six sampled areas, the GDP of Jiangsu Province in 2010 is the largest, accounting for 10.33% of the national share. Beijing's GDP is the smallest, accounting for 3.52% of the national value. Being respectively the capital andthe largest city in China,Beijing andShanghai serve as central metropolises of the economy, and hence their GDP per capita is higher than in other provinȬ ces.Shanghai'sGDPpercapitaamountsto74,548yuan,beingthehighestinthesixareas,followed by Beijing with 71,938 yuan per capita. Hebei Province's is the lowest, less than the national

Source: revised based on http://www.chinamapxl.com/

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Figure 3.1** Location of the sampled areas

Based on the survey of 21 villages of eastern China's six provincial regions, this chapter aims to analyze the situation of agricultural production, the public service in rural regions and the

36 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

To understand the present situation and farmers' perceptions on agricultural pollution, we conducted the survey with questionnaire-based personal interviews to collect first-hand data. In the first section, our questionnaire contains the basic characteristics of each village, including general economic situation, land and land use. In the second section of the questionnaire, we investigate the situation of agricultural production, including crops, animal products, aquatic products, certificated agricultural production and irrigation. The following questions are designed to ascertain information about the rural environment, including rubbish collection, disposal of liquid and solid waste, construction of methane tanks, etc. In the final section, we collect rural public services data, such as agricultural production, rural life, government subsidies and village public expenditure. Simultaneously, we designed another questionnaire to understand farmers' perceptions on agricultural pollution, including basic characteristics of each farm, disposal of household rubbish, farmers' selection and application of fertilizers, pesticides and veterinary drugs, farmers' perceptions on agricultural pollution, and informaȬ

In January to March, 2011, we surveyed 21 villages and their 560 farms in eastern China's six provincial regions, including Beijing, Hebei, Shandong, Shanghai, Jiangsu and Zhejiang (Figure 3.1). The sampled area covers three major gain growing provincial regions and rural regions affiliating to the top two metropolises of China. The former three regions represent the northern mode of agricultural production in the Yellow River Basin, while the latter three demonstrate the characteristics of agricultural production in south China's Yangtze River Basin. With regard to topographic types, farms located on plains, hills and mountainous regions, and villages in inland areas, seaside areas and areas adjoining the metropolises are sampled. In addition to the staple grain crops of wheat, rice and corn, the other major agriȬ cultural products, including cotton, vegetables, fruit, oil crops, etc., and the main livestock,

poultry, aquaculture products are being grown and cultivated in the sampled farms.

In six sampled areas, the GDP of Jiangsu Province in 2010 is the largest, accounting for 10.33% of the national share. Beijing's GDP is the smallest, accounting for 3.52% of the national value. Being respectively the capital andthe largest city in China,Beijing andShanghai serve as central metropolises of the economy, and hence their GDP per capita is higher than in other provinȬ ces.Shanghai'sGDPpercapitaamountsto74,548yuan,beingthehighestinthesixareas,followed by Beijing with 71,938 yuan per capita. Hebei Province's is the lowest, less than the national

relationship between the rural environment and economic level.

tion about and recognition of safe agricultural production.

**3.3 Basic situation of the sampled areas**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**3.2 Samples of the survey**

average value and only accounts for 2/5 of that in Shanghai. Overall, the economic level of the three areas in the south is higher than in the three northern areas. Except for Hebei, the samȬ pled areas possess significantly higher GDP per capita than the national average (Table3.1).

The largest agricultural output value in 2010 is from Shandong Province, amounting to 367 billion yuan, and accounting for 9.93% of the country's agricultural output value. Meanwhile, its arable land area is the largest, accounting for 6.17% of the total area of the country. The agricultural output value is almost the same in Beijing and Shanghai, accounted for only 0.42% of the national value. Their arable land areas are also very close, 0.19% and 0.20% of the country, respectively. When calculating the output per hectare, we find that Beijing is the highest, followed by Shanghai, with the lowest being Hebei, lower than the national average value.

Analyzing the areas of arable land per capita, the values of the six provincial areas are lower than the national average. Of these, Shanghai possesses the smallest arable land area per capita of only 0.28 mu, far less than the national average of over 2 *mu* per capita. By contrast, Hebei Province has the largest value of 1.98 mu, being close to the national average. Moreover, comparing values of the six provincial regions, we can conclude that the resource endowments of arable land per capita of the three northern regions, i.e., Beijing, Hebei and Shandong, are better than that of the three southern regions of Shanghai, Jiangsu and Zhejiang (Figure 3.2).


Note: 100 dollar=634 yuan, http://www.boc.cn/sourcedb/whpj/, 2012.9.21

Source: China Statistical Press (2011) [5]

#### **Table 3.1** Basic information of the surveyed areas in 2010

Note: 1 ha=15 mu.

Source: China Statistical Press (2011) [5]

**Figure 3.2** Arable farmland area per capita in 2008

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Per capita income of rural households and per capita GAP in the six regions show the same structure. Per capita income of all the areas is higher than the national average. Shanghai and Beijing's are significantly higher than the other four areas. Overall, the incomes of the southern areas are higher than that of the northern areas (Figure 3.3).

Note: 100 dollar=634 yuan, http://www.boc.cn/sourcedb/whpj/, 2012.9.21. Source: China Statistical Press (2011) [5]

#### **Figure 3.3** Net income of rural household in 2010

The irrigation area proportions of the total arable land area in the six regions are all higher than the national average. Among them, the highest proportion is Beijing reaching 91.25%, 1.8 times the national average. Shanghai's is 82.39%, the second highest, and the lowest is in Shandong Province, being 65.94%.

Source: China Statistical Press (2011) [5]

0.0

Source: China Statistical Press (2011) [5]

**Figure 3.2** Arable farmland area per capita in 2008

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

areas are higher than that of the northern areas (Figure 3.3).

**GDP Total population**

Note: 100 dollar=634 yuan, http://www.boc.cn/sourcedb/whpj/, 2012.9.21

**Million person**

**(end of year)**

38 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**%**

**GDP per capita**

> **1000 yuan**

China 40.12 100.00 1340.91 100.00 29.92 3.69 100.00 121.72 100.00 Beijing 1.41 3.52 19.62 1.46 71.94 0.02 0.42 0.23 0.19 Hebei 2.04 5.08 71.94 5.36 28.35 0.25 6.69 6.32 5.19 Shandong 3.92 9.76 95.88 7.15 40.85 0.37 9.93 7.52 6.17 Shanghai 1.72 4.28 23.03 1.72 74.55 0.02 0.42 0.24 0.20 Jiangsu 4.14 10.33 78.69 5.87 52.64 0.23 6.14 4.76 3.91 Zhejiang 2.77 6.91 54.47 4.06 50.90 0.10 2.82 1.92 1.58

**Total agricultural output**

**Trillion yuan**

**Arable land area**

**% Million ha %**

Note: 1 ha=15 mu.

Country Beijing Hebei Shandong Shanghai Jiangsu Zhejiang

Per capita income of rural households and per capita GAP in the six regions show the same structure. Per capita income of all the areas is higher than the national average. Shanghai and Beijing's are significantly higher than the other four areas. Overall, the incomes of the southern

0.5

1.0

1.5

2.0

2.5

mu/person

Source: China Statistical Press (2011) [5]

**Table 3.1** Basic information of the surveyed areas in 2010

**Trillion yuan**

**%**

**Figure 3.4** Proportion of irrigated arable farmland in 2010



#### **Table 3.2** Amount of chemical fertilizers used in 2010

The amounts of chemical fertilizer application in the six surveyed areas are all higher than the national average. The highest is in Jiangsu Province, up to 716 kg per hectare, much higher than the national average of 457 kg, 632 kg in Shandong Province followed. Shanghai and Zhejiang are almost the same. Specifically, the amounts of nitrogen fertilizer application in the six surveyed areas are all higher than the national average, but the case in relation to phosphate, potash and compound fertilizer varies. The amount of phosphorus fertilizer application in Beijing and Shanghai is less than the national average. The amount of potassium fertilizer application in only Shandong is above the national average. The amount of compound fertilizer application in only Zhejiang is below the national average. The differences of fertilizer application may be mainly due to crop differences and soil differences.


**Table 3.3** Yields of major agricultural products in 2010 (Unit: kg/ha)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The average yield of rice in Shanghai, Shandong and The average yield of rice in Shanghai, Shandong and Jiangsu is more than 8000 kg per hectare, significantly higher than in other regions and the national average. Only Beijing's is lower than the national average of 6,553 kg per hectare. There are very small areas of rice cultivation in the three northern regions, and hence it is unnecessary to compare yield of rice in the north and the south. Wheat yields are above 5000 kg per hectare in Shandong and Hebei, while they are less than 4000 kg per hectare in Zhejiang and Shanghai, being lower than the national average. Overall the wheat yield in the three northern regions is lower than that of the three southern regions. The maize yields in Shanghai and Shandong are higher than the other regions and the national average, and the difference between the south and the north is not significant. Meanwhile, yields of cotton in Shanghai and Zhejiang are higher than in the other regions and the national average, and overall the three northern regions are higher than the three regions in the south. The average peanut yield in Shandong is much higher than that in other regions and the national average. The yield of rapeseed in Shandong is the highest, but the yields in Hebei and Beijing are very low, so overall the yield in the three northern regions is higher than the three regions in the south. The yield differences are likely to be influenced by regional climate, soil and irrigation.

#### **3.4 Agricultural production, rural environment and public services**

#### **3.4.1 Demographic characteristics**

**Total <sup>N</sup> <sup>P</sup> <sup>K</sup> Compound**

China 456.94 193.38 66.19 48.18 147.76 Beijing 589.99 296.50 37.98 31.07 224.43 Hebei 511.08 242.30 74.89 42.48 151.40 Shandong 632.47 216.39 66.37 61.74 287.97 Shanghai 485.33 253.73 41.40 24.18 166.01 Jiangsu 716.05 376.86 100.19 43.66 195.33 Zhejiang 479.99 273.32 62.32 37.33 107.04

40 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

The amounts of chemical fertilizer application in the six surveyed areas are all higher than the national average. The highest is in Jiangsu Province, up to 716 kg per hectare, much higher than the national average of 457 kg, 632 kg in Shandong Province followed. Shanghai and Zhejiang are almost the same. Specifically, the amounts of nitrogen fertilizer application in the six surveyed areas are all higher than the national average, but the case in relation to phosphate, potash and compound fertilizer varies. The amount of phosphorus fertilizer application in Beijing and Shanghai is less than the national average. The amount of potassium fertilizer application in only Shandong is above the national average. The amount of compound fertilizer application in only Zhejiang is below the national average. The differences of fertilizer

China 6553.03 4748.44 5453.68 1229.42 3455.45 1775.10 Beijing 6333.33 4609.51 5620.50 1150.00 2990.46 440.00 Hebei 6805.12 5084.51 5014.71 979.26 3517.36 1311.57 Shandong 8294.34 5779.55 6537.68 944.85 4211.78 2792.13 Shanghai 8327.65 3896.76 6659.14 1453.72 2761.80 2195.15 Jiangsu 8091.90 4816.37 5412.00 1106.70 3646.00 2444.00 Zhejiang 7020.99 3729.79 4455.45 1412.16 2826.74 1803.00

The average yield of rice in Shanghai, Shandong and The average yield of rice in Shanghai, Shandong and Jiangsu is more than 8000 kg per hectare, significantly higher than in other

**Rice Wheat Maize Cotton Peanut Rapeseed**

application may be mainly due to crop differences and soil differences.

Source: China Statistical Press (2011) [5]

Source: China Statistical Press (2011) [5]

**Table 3.3** Yields of major agricultural products in 2010 (Unit: kg/ha)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 3.2** Amount of chemical fertilizers used in 2010

**fertilizer**

The population sizes of villages are quite different, with the smallest value of 655 in Beijing, while the largest value of 3804 is sampled from Jiangsu Province. The villages in Hebei have the highest number of non-agricultural enterprises, followed by those in Shanghai Province. However, in terms of revenue from non-agricultural enterprises, the villages in Shanghai derive the largest value, followed by Zhejiang. To the nearest location of the local government from the village, the shortest is Shanghai, only 1 km; the farthest is Shandong, 29 km. The distance to the nearest market town from all villages is less than 7 km (Table 3.4).


Note: - means no data; \$100 exchange 644 yuan on July 29, 2011, http://www.boc.cn/sourcedb/whpj/

Source: field survey by the authors

**Table 3.4** Characteristics of each sampled village on average

The farmers in 2010 with the greatest cash income reside in Shanghai, where the average per capita is 11092 yuan, followed by Zhejiang, Jiangsu, Beijing, Shandong and Hebei (Table 3.5). Farmers' cash income in our survey data is much lower than the data from the China Statistical Yearbook 2010. In Beijing in particular, it is only about half of the Yearbook's data. NevertheȬ less, except for Beijing, we can get the same order of the other regions on data of annual cash income from the two sources. Therefore, the survey can be deemed as being representative of farmers in present day China. In Table 3.5, we have listed the data of cash income per household from the farmers' survey, where farmers from Jiangsu Province are recorded as having most cash income.


\* Note: 1=less than 10 thousand yuan, 2=10-30 thousand yuan, 3=30-50 thousand yuan, 4=more than 50 thousand yuan

**Table 3.5** Annual cash income of the farmers (yuan per capita)

#### **3.4.2 Agricultural production**

According to Table 3.6, the paddy field area and the orchard area in the south are greater than those in the north. The dry land area in the south is smaller than it in the north and five southern villages possess fish ponds, while there are no fish ponds in the northern villages. In addition, the average area of greenhouse in the south is found to be larger than that in the north.


Note: north means Beijing, Hebei and Shandong; south means Shanghai, Jiangsu and Zhejiang; 1 mu equals to 1/15 hectare

Source: field survey by the authors

**Table 3.6** Areas of agricultural land in each village (Unit: mu)

Rice is widely planted in the south, but none is planted in the sampled northern villages, despite there being paddy fields. The acreage of wheat in the south and north is almost the same. The acreage of corn in the south is much less than in the north. The vegetable acreage in the south is more than three times that of the north (Table 3.7).


**Table 3.7** Average crop acreage in each village (Unit: mu)

The farmers in 2010 with the greatest cash income reside in Shanghai, where the average per capita is 11092 yuan, followed by Zhejiang, Jiangsu, Beijing, Shandong and Hebei (Table 3.5). Farmers' cash income in our survey data is much lower than the data from the China Statistical Yearbook 2010. In Beijing in particular, it is only about half of the Yearbook's data. NevertheȬ less, except for Beijing, we can get the same order of the other regions on data of annual cash income from the two sources. Therefore, the survey can be deemed as being representative of farmers in present day China. In Table 3.5, we have listed the data of cash income per household from the farmers' survey, where farmers from Jiangsu Province are recorded as having most

42 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Data source Year Beijing Hebei Shandong Shanghai Jiangsu Zhejiang Total**

Farmer survey 2010 2.00\* 2.14 2.40 2.99 3.27 3.07 2.65 Counts of farmers 2010 88 120 60 88 89 110 555

Note: 1=less than 10 thousand yuan, 2=10-30 thousand yuan, 3=30-50 thousand yuan, 4=more than 50 thousand yuan

According to Table 3.6, the paddy field area and the orchard area in the south are greater than those in the north. The dry land area in the south is smaller than it in the north and five southern villages possess fish ponds, while there are no fish ponds in the northern villages. In addition, the average area of greenhouse in the south is found to be larger than that in the north.

Mean 788 2321 1276 457 181 385 0 258 61 90 Max 1200 8000 4000 2000 300 1000 0 1200 300 220 Min 550 50 26 10 30 30 0 4 6 6 Std. D. 295 2450 1290 864 138 535 0 527 117 95 Sample 4 9 9 5 3 3 0 5 6 4

Note: north means Beijing, Hebei and Shandong; south means Shanghai, Jiangsu and Zhejiang; 1 mu equals to 1/15

**Paddy field Dry land Orchard Fish ponds Greenhouse North South North South North South North South North South**

**Table 3.5** Annual cash income of the farmers (yuan per capita)

**3.4.2 Agricultural production**

China Statistic Yearbook 2009 14198 6266 7445 15189 9738 12177 Village survey 2010 7019 3583 4100 11092 9400 9794

cash income.

\*

hectare

Source: field survey by the authors

**Table 3.6** Areas of agricultural land in each village (Unit: mu)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The villages in the north did not plant rice in 2010. The yield of wheat in the north is a little more than in the south. The yield of corn in the north is much more than in the south. The yield difference of the survey data is the same as in the statistical data, according to Table 3.4. We cannot compare the yield of vegetables in the north and south because no data are available recording the vegetable varieties (Table 3.8).


**Table 3.8** Average yields of the major crops (Unit: kg/mu)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In the north only three villages have dairy cow farmers, while no villages sampled in the south answeredas raising cows.As for swine,the numberin thenorth is a littlemore than inthe south. Meanwhile, the number of broiler in the north is almost twice that of the south (Table 3.9).


**Table 3.9** Numbers of major livestock in the north and south

Similarly, although the maximum scales of dairy cow farmers are small in the north, no cow is raised in the southern villages interviewed. Scales of swine farmers (or companies) in the north are much larger than in the south. Scales of broiler farmers (or companies) in north are smaller than in the south (Table 3.10).


**Table 3.10** Max. number of livestock raised in one company or farmer

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

There are six villages where pollutant-free agricultural products are planted, eight villages where green agricultural products are planted, only one for organic agricultural products, and 11 villages where nothing of certification of agricultural products applies. There are no significant differences in certification of agricultural products between the south and north (Table 3.11). Meanwhile, the main methods and sources of agricultural irrigation are surveyed as quite different between the south and north. The main sources of agricultural irrigation in the south are rivers, lakes and ponds. There is a lot of water in the south, so the main methods of agricultural irrigation are soil channels and concrete channels. The main sources of agriȬ cultural irrigation in the north are wells, rivers and rain. There is not water enough in the south, so the main methods of agricultural irrigation are pipes and soil channels. The irrigation of soil channel is not efficient. The reason why the soil channel is one of the main methods of agricultural irrigation in the north may be that there is not enough money to build concrete channels (Table 3.12).


Note: two villages did not answer; numbers in parentheses indicate number of villages

Source: field survey by the authors

**Dairy cow Swine Broiler**

**North South North South North South**

Mean 36 0 1343 1247 6257 3386 Max 70 0 4000 5000 16500 10000 Min 4 0 140 180 200 200 Std.D. 33 0 1300 1876 6849 3534 Sample 3 0 7 6 7 7

44 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Similarly, although the maximum scales of dairy cow farmers are small in the north, no cow is raised in the southern villages interviewed. Scales of swine farmers (or companies) in the north are much larger than in the south. Scales of broiler farmers (or companies) in north are

Mean 22 0 372 194 2804 3645 Max 35 0 1000 400 6000 20000 Min 2 0 110 20 20 10 Std.D. 18 0 334 116 2274 7290 Sample 3 0 6 7 5 7

There are six villages where pollutant-free agricultural products are planted, eight villages where green agricultural products are planted, only one for organic agricultural products, and 11 villages where nothing of certification of agricultural products applies. There are no significant differences in certification of agricultural products between the south and north (Table 3.11). Meanwhile, the main methods and sources of agricultural irrigation are surveyed as quite different between the south and north. The main sources of agricultural irrigation in the south are rivers, lakes and ponds. There is a lot of water in the south, so the main methods of agricultural irrigation are soil channels and concrete channels. The main sources of agriȬ

**Dairy cow Swine Broiler**

**North South North South North South**

Source: field survey by the authors

Source: field survey by the authors

**Table 3.10** Max. number of livestock raised in one company or farmer

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 3.9** Numbers of major livestock in the north and south

smaller than in the south (Table 3.10).

**Table 3.11** Certification of agricultural products


Source: field survey by the authors

**Table 3.12** The main methods and sources of agriculture irrigation

#### **3.4.3 Rural environmental situation**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In the 21 sampled villages, there are 7 villages surveyed as having rubbish collection points (RCPs) in all the focal living points, accounting for 33% in total. Meanwhile, 8 villages have RCPs in most of their focal living points, with the proportion of 38% in total. In addition, there are 5 villages (24% in total) surveyed as having no RCPs all of which locate in Hebei and Shandong Province (Figure 3.5). Thus in general, situation in the south is better than the north. In 68% of the villages, it is the village committee who manages the rubbish collection. The two villages where nobody manages to collect rubbish are located in Shandong Province. Two villages in Hebei Province did not answer (Figure 3.6).

**Figure 3.5** Situation of rubbish collection points

**Figure 3.6** Managers of rubbish collection

The percentage of the collective rubbish disposal is almost consistent with the ranking of RCP. The five villages where the proportion of the collective rubbish disposal is lower than 10% are in Hebei and Shandong Province (Table 3.13).


**Table 3.13** The situation of rubbish collection

The proportion of the collective disposal of domestic sewage is 100% in six villages located in Beijing, Shanghai, Jiangsu and Zhejiang, while the ratio is less than 10% in the six villages of Hebei, Shandong and Zhejiang. In the 20 surveyed villages, there are 13 villages where more than 50% of domestic sewage is collectively disposed of. This illustrates that domestic sewage is handled well in survey areas. Of 19 surveyed villages, there are six villages where more than 70% of company wastewater is collectively disposed of, but less than 10% in 10 villages. In 20 surveyed villages, there are six villages where more than 70% of animal manure is collectively disposed of, but less than 10% in 10 villages. The villages where less than 10% of wastewater and manure are collectively disposed of are located across the six surveyed areas, i.e., not concentrated in one area. The pollution from company wastewater and animal manure is serious in the survey (Table 3.14).


#### **Table 3.14** Percentages of collective wastewater disposal

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The percentage of the farmers who had built methane tanks is less than 10% in 17 villages. As of 2009 the rural population stands at 34300 in Jiangsu Province. Given 3.5 persons per family there would be 9800 rural households. According to data there are 547 rural households who built methane tanks up to June 2009. So the percentage of village farmers who had built methane tanks would be 5.6% (Figure 3.7).

#### **3.4.4 Rural public services**

**Figure 3.5** Situation of rubbish collection points

**Figure 3.6** Managers of rubbish collection

Source: field survey by the authors

**Table 3.13** The situation of rubbish collection

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

in Hebei and Shandong Province (Table 3.13).

The percentage of the collective rubbish disposal is almost consistent with the ranking of RCP. The five villages where the proportion of the collective rubbish disposal is lower than 10% are

**All Most A few Nothing**

46 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Village 7 8 1 5

% of the collective rubbish disposal Percentage 100% 90-100% 70-90% <10% Village 8 4 3 5

**Whether there are RCPs?**

There are 10 villages where the agricultural service stations guide fertilizer use methods, of which seven are in the south. There are seven villages where the agricultural service stations guide pesticide use methods, of which five are in the south. Clearly, the agricultural service stations are dominant in fertilizer and pesticide use guidance, while the agricultural service stations are playing a greater role in the south than in the north. The government plays an important role in guiding livestock quarantining, and rubbish and domestic sewage disposal. Some villages choose others to supply the guidance on production and farmer's living, either because these villages have no corresponding agricultural production and living activities, or because there was nobody to provide these services on agricultural production and living (Table 3.15).

#### **Figure 3.7** Ratio of villages with methane tanks


Note: a section of the villages did not answer

Source: field survey by the authors

**Table 3.15** Guidance on production and farmers' living in 2008-2010

As to the question of how many times guidance on agricultural production and farmers' living in 2008-2010 were carried out, there are only a few answers. From these limited answers, it appears that the guidance on agricultural production and farmers' living in 2008-2010 was carried out at least once a year, including the use of methane tanks (Table 3.16).


Source: field survey by the authors

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 3.16** Number of guidance on agricultural production and farmers' living in 2008-2010

In terms of getting fiscal subsidies among the 19 villages, eight villages received subsidies to promote the use of methane tank; nine villages received subsidies to advocate the reduced use of fertilizer, while four villages were subsidized for reduction of pesticides. This indicates that fiscal subsidies are helpful in improving the rural environment, from the perspectives of adopting methane tanks, reducing the use of fertilizers and pesticides. Meanwhile, Table 3.17 shows that only two villages received fiscal subsidies for using biodegradable plastic sheets; four villages did not receive any subsidies - located in Jiangsu, Zhejiang and Shandong Provinces. Now that many farmers use plastic sheets in agricultural production, further fiscal subsidies on biodegradable plastic sheet usage will play a more important role in improving the rural environment.


#### **Table 3.17** Objectives of the government subsidies

**Use of the fertilizer**

**Figure 3.7** Ratio of villages with methane tanks

Note: a section of the villages did not answer

**Use of the fertilizer**

Source: field survey by the authors

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 3.15** Guidance on production and farmers' living in 2008-2010

**Use of the pesticides**

Source: field survey by the authors

Agricultural service station **Use of the pesticides**

**Livestock quarantine**

48 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Government 2 2 5 1 7 5 Sales company 1 1 0 0 0 0 Farmer technician 3 3 3 3 0 0 Others 4 4 5 6 4 6

As to the question of how many times guidance on agricultural production and farmers' living in 2008-2010 were carried out, there are only a few answers. From these limited answers, it appears that the guidance on agricultural production and farmers' living in 2008-2010 was

carried out at least once a year, including the use of methane tanks (Table 3.16).

**Table 3.16** Number of guidance on agricultural production and farmers' living in 2008-2010

**Livestock quarantine**

Mean 3.4 3.5 7.5 2 7 4.5 Max 6 6 12 3 8 6 Min 2 2 3 1 6 3 Village 5 4 2 2 2 2

**Use of methane tank**

10 7 1 3 1 1

**Use of methane tank**

**Rubbish classification**

**Rubbish classification**

**Sewage discharge**

**Sewage discharge** As shown in Table 3.18, transfer payment, collective economy income and agricultural subsidies are the main three sources of public expenditure in villages. In this survey, six villages get transfer payments from a higher level of government which is more than 60% of their total expenditure; seven villages have collective economy income which is more than 60% of their total expenditure; 14 villages have transfer payments and collective economy income which is more than 50% of their total expenditure. Meanwhile, the agricultural subsidies are revealed as accounting for only a smaller proportion of total expenditure. The transfer payment could improve the rural environment, but we are not sure that if the collective economy income can do the same. In addition, the village's companies are interviewed as providing funds to the village, although some of them may pollute the environment simultaneously.


**Table 3.18** Source of public expenditure in villages

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **3.4.5 Relation of rural environment and economic level**

In Figure 3.8, 1 means all, 2 means most, 3 means a few, and 4 means nothing. The correlation coefficient of GCP and cash income is -0.38. There are RCPs in all or most of the focal living points of villages when the annual cash income of per farmer is over 6000 yuan. The correlation coefficient of GCP and the distance from village to the nearest county is 0.49.

**Figure 3.8** Relationship of GCP and cash income

In Figure 3.9, we use a similar system of 1means 100%, 2 means 90-100%, 3 means 70-90%, 4 means 50-70%, 5 means 30-50%, 6 means 10-30%, 7 means <10%. The correlation coefficient of cash income and the collective domestic sewage disposal is -0.30. The correlation coefficient of the collective domestic sewage disposal and the distance from village to the nearest county is 0.42.

**Figure 3.9** Relationship of domestic sewage and cash income

#### **3.5 Conclusion**

Land type, crop acreage, irrigation water source and irrigation methods have shown a certain difference between the north and the south. The proportion of the collective disposal of rubbish and sewage in rural areas is higher than industrial sewage and animal manure. The agricultural production and farmers' life guidance services are primarily supplied by government and agricultural extension centres. In about 50% of surveyed villages the government supplies subsidies for methane tanks and reduction of fertilizers, but only 10% in the case of biodeȬ gradable plastic sheets. In the 21 surveyed villages, 14 of them have transfer payments and collective economy income which is more than 50% of their total expenditure. The ratio of the collective rural rubbish and sewage disposal is positively correlated with income level of the farmers, and is negatively correlated with the distances from the towns.

#### **References**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Figure 3.8** Relationship of GCP and cash income

**Figure 3.9** Relationship of domestic sewage and cash income

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

is 0.42.

In Figure 3.9, we use a similar system of 1means 100%, 2 means 90-100%, 3 means 70-90%, 4 means 50-70%, 5 means 30-50%, 6 means 10-30%, 7 means <10%. The correlation coefficient of cash income and the collective domestic sewage disposal is -0.30. The correlation coefficient of the collective domestic sewage disposal and the distance from village to the nearest county

50 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers


## **Farmer Behaviours Toward Food Safety and the Agro-Environment**

Dongpo Li, Tinggui Chen, Hui Zhou and Teruaki Nanseki

As revealed in the above chapters, increasing application of agricultural chemicals, includȬ ing chemical fertilizers and pesticides, has been demonstrated as a key factor in improving China's agricultural productivity over the last few decades. Meanwhile, the excessive use of chemicals has resulted in serious threats and losses to the ecological environment, human health and economic development. As household farms are the overwhelming managerial units in Chinese agriculture, this chapter tries to accelerate the safe application of pesticides, through understanding the behaviours and determinants of the farmers.

### **4.1 Application of agricultural chemicals over the last few decades**

In 2009, the chemical fertilizer applied to agriculture amounted to 54.04 million *ton*s in China and maintained an average annual growth rate of 6.01% since 1978 [1]. Meanwhile, fertilizer consumption was 467.98 kg per hectare of arable land in 2008, much larger than the average amount of 134.93 kg per hectare amongst 175 countries [2]. A field test has revealed the low fertilization efficiency in China: the average nitrogen absorption efficiency of wheat, corn and rice are 28.3%, 28.2% and 26.1%, far lower than that of 40-60% in European and American countries [3]. Furthermore, even lower nitrogen absorption efficiency of only 10 % exists in vegetables, fruits and flowers [4]. According to the Bulletin of the First National Census on Pollution Sources issued in 2010, the non-point pollution (NPP) of agriculture has become the first source of water contamination in China, while chemical fertilizer applied in crop proȬ duction constitutes the main source of agricultural NPP. The large volume of fertilizer residues has become a major source of environmental pollution and food safety incidents, thus proper application of fertilizer is drawing unprecedented public attention. The Chinese government has adopted the control of agricultural NPP into the twelfth Five-year Plan (2011-2015), with strengthening regulations on fertilizers.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Li et al.; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Li et al.; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

By the end of 2010, the total amount of chemical pesticides produced in China amounted to 2.34 million tons, maintaining an average annual growth rate of 10.32% since 1985 [1]. China has become the largest producer, user and exporter of pesticides in the world [5]. Meanwhile, the improper use of pesticides has become a major source of food safety incidents, which have resulted in serious threats and losses to the ecological environment, human health and economic development. Therefore, safe application of pesticides is drawing unprecedented public attention and Chinese government is strengthening regulations on the production, marketing and use of pesticides [6].

#### **4.2 Theoretical models**

#### **4.2.1 Model on fertilizer application**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In this study, only those farms that answered as using fertilizer in 2010 are included, thus the sample consists of 294 valid responses from this survey. Based on the theoretical model specified above, we include 10 indicators to represent the demographic characteristics of each farm (Table 4.1). Simultaneously, these indicators will be used as candidate determinants to interpret farmers' behaviours on fertilizer application.

(1) Considering the importance of householders in making productive decisions within household farms, many studies included concerning variables as determinants in the analysis of safe agricultural production. In this study, we include three variables to describe attribȬ utes of the householders, i.e., human resources, as *gender* [7], *age* [8] and education level (*edu*) [9]. (2) In agrarian societies, land is not used solely as a means by which to generate a livelihood, it is often also used for accumulating wealth and transferring it between generaȬ tions [10]. Thus two continuous variables on land cultivation are introduced: the sowing area of total agricultural products (*scale*), rather than total area of farmland is adopted with the consideration of multiple cropping [11]; sowing ratio of grain crops (*grainr*) is included to identify the effects of land use structure. (3) Meanwhile, another two variables are introȬ duced to measure impacts of discrepancies in household income: total annual cash income (*income*) affects household budgets and thus inputs to agriculture, including the purchase of fertilizer [8-9]; ratio of income from migrant job (*mir*) shows the main sourcing structure of household income, which affects the relative importance of agriculture and also the correȬ sponding inputs [12]. (4) To model the influence of geographic location on farmers' applicaȬ tion of fertilizer [13], two dichotomous dummy variables are included with *north* (north or south of China) equal to 1 if a farm is from Beijing, Hebei or Shandong, and *metro* (metropolȬ ises or not) coded as 0 if a farm affiliates to neither Beijing nor Shanghai. The statistical summary of each variable is shown in Table 4.1. Finally, according to the China National Fertilization Regionalization1 [14-15], the sampled areas cover four sub-regions as shown in the statistics following the characteristic variable of *fregion*.

<sup>1</sup> The China National Fertilization Regionalization is drafted by the Soil and Fertilizer Institute, Chinese Academy of Agricultural Sciences. According to the soil condition and fertilization characteristics, this national planning divides China's farmland into 31 sub-divisions within eight divisions.


Note: a as a main unit of land measurement in China, 1 mu=666.67*m*2; c the income sources contain migrant jobs and sales of agricultural products; c the bracketed numerals denote counts of farms.

Source: field survey by the authors

By the end of 2010, the total amount of chemical pesticides produced in China amounted to 2.34 million tons, maintaining an average annual growth rate of 10.32% since 1985 [1]. China has become the largest producer, user and exporter of pesticides in the world [5]. Meanwhile, the improper use of pesticides has become a major source of food safety incidents, which have resulted in serious threats and losses to the ecological environment, human health and economic development. Therefore, safe application of pesticides is drawing unprecedented public attention and Chinese government is strengthening regulations on the production,

54 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

In this study, only those farms that answered as using fertilizer in 2010 are included, thus the sample consists of 294 valid responses from this survey. Based on the theoretical model specified above, we include 10 indicators to represent the demographic characteristics of each farm (Table 4.1). Simultaneously, these indicators will be used as candidate determinants to

(1) Considering the importance of householders in making productive decisions within household farms, many studies included concerning variables as determinants in the analysis of safe agricultural production. In this study, we include three variables to describe attribȬ utes of the householders, i.e., human resources, as *gender* [7], *age* [8] and education level (*edu*) [9]. (2) In agrarian societies, land is not used solely as a means by which to generate a livelihood, it is often also used for accumulating wealth and transferring it between generaȬ tions [10]. Thus two continuous variables on land cultivation are introduced: the sowing area of total agricultural products (*scale*), rather than total area of farmland is adopted with the consideration of multiple cropping [11]; sowing ratio of grain crops (*grainr*) is included to identify the effects of land use structure. (3) Meanwhile, another two variables are introȬ duced to measure impacts of discrepancies in household income: total annual cash income (*income*) affects household budgets and thus inputs to agriculture, including the purchase of fertilizer [8-9]; ratio of income from migrant job (*mir*) shows the main sourcing structure of household income, which affects the relative importance of agriculture and also the correȬ sponding inputs [12]. (4) To model the influence of geographic location on farmers' applicaȬ tion of fertilizer [13], two dichotomous dummy variables are included with *north* (north or south of China) equal to 1 if a farm is from Beijing, Hebei or Shandong, and *metro* (metropolȬ ises or not) coded as 0 if a farm affiliates to neither Beijing nor Shanghai. The statistical summary of each variable is shown in Table 4.1. Finally, according to the China National

1 The China National Fertilization Regionalization is drafted by the Soil and Fertilizer Institute, Chinese Academy of Agricultural Sciences. According to the soil condition and fertilization characteristics, this national planning divides

[14-15], the sampled areas cover four sub-regions as shown in

marketing and use of pesticides [6].

**4.2.1 Model on fertilizer application**

interpret farmers' behaviours on fertilizer application.

**4.2 Theoretical models**

Fertilization Regionalization1

the statistics following the characteristic variable of *fregion*.

China's farmland into 31 sub-divisions within eight divisions.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 4.1** Demographic characteristics of the sampled farms applying fertilizer

#### **4.2.2 Model on pesticides application**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Here, only farms that answered as using pesticides in 2010 are included in this chapter. From this survey, a total sample size of 220 valid responses is used in this study. We include nine indicators to represent the demographic characteristics of each farm (Table 4.9). In the following sections, these indicators will be used as candidate determinants to interpret farmers' behaviours.

(1) Due to the key role of the householder in making productive decisions within a family farm, many studies included relevant variables in the analysis of safe agricultural production. In this study, we include three variables to describe characteristics of the householders, i.e., human resources, such as *gender*, *age* and education level. (2) At the same time, to model the impacts of land cultivation to safe agricultural production as in Song et al. (2010) [16], two continuous variables are introduced: the sowing area of total agricultural products (*scale*), rather than total area of farmland is adopted with the consideration of multiple cropping [14]; sowing ratio of grain crops (*grainr*) is included to identify the significance of land use structure. (3) Meanwhile, another two variables are used to measure the impacts of discrepancies in household income: total annual cash income (*income*) affects household budget and thus the inputs to agricultural production, including those on pesticides and spraying apparatuses [17-18]; ratio of income from migrant job (*mir*) shows the main sourcing structure of family income, which affects the relative importance of agriculture and the inputs [12]. (4) Finally, two dichotomous dummy variables are incorporated to show the importance of geographic location as in Zhang et al. (2004) [19], with *north* equal to 1 if a farm is from Beijing, Hebei or Shandong, and *metro* coded as 0 for farms locating in neither Beijing nor Shanghai.

#### **4.3 Analysis on fertilizer application**

#### **4.3.1 Behaviours on fertilizer application**

To capture the major behaviours of chemical fertilizer application in a farm, in addition to an aggregate amount, quantities of nitrogen, phosphate, potash and compound fertilizers used in each agro-product are included.

In the sampled farms, the nitrogen fertilizers mainly include carbamide, ammonium bicarȬ bonate, etc.; the major phosphate fertilizer used is calcium superphosphate; potash fertilizers consist of potassium sulphate, etc. Amongst the three types of macro-element fertilizers, nitrogen fertilizers are most widely used by 278 (94.56%) farms, while potash fertilizers are used only in four (1.36%) farms. Although many compound fertilizers contain all the macro elements, the general fertilizing trend of rich nitrogenous and poor potash nutrients is testified to from the survey data (see [7, 14]). Meanwhile, the application of organic fertilizer (mainly including manure and compost) is represented in terms of the counts of farms amongst both the total sample and those who used chemical fertilizer simultaneously (Table 4.2).


Note: the bracketed numerals denote counts of farms

Source: field survey by the authors

**Table 4.2** Application of fertilizer in the sampled farms


**Table 4.3** Application of fertilizer in each agricultural plant

area of farmland is adopted with the consideration of multiple cropping [14]; sowing ratio of grain crops (*grainr*) is included to identify the significance of land use structure. (3) Meanwhile, another two variables are used to measure the impacts of discrepancies in household income: total annual cash income (*income*) affects household budget and thus the inputs to agricultural production, including those on pesticides and spraying apparatuses [17-18]; ratio of income from migrant job (*mir*) shows the main sourcing structure of family income, which affects the relative importance of agriculture and the inputs [12]. (4) Finally, two dichotomous dummy variables are incorporated to show the importance of geographic location as in Zhang et al. (2004) [19], with *north* equal to 1 if a farm is from Beijing, Hebei or Shandong, and *metro* coded

56 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

To capture the major behaviours of chemical fertilizer application in a farm, in addition to an aggregate amount, quantities of nitrogen, phosphate, potash and compound fertilizers used

In the sampled farms, the nitrogen fertilizers mainly include carbamide, ammonium bicarȬ bonate, etc.; the major phosphate fertilizer used is calcium superphosphate; potash fertilizers consist of potassium sulphate, etc. Amongst the three types of macro-element fertilizers, nitrogen fertilizers are most widely used by 278 (94.56%) farms, while potash fertilizers are used only in four (1.36%) farms. Although many compound fertilizers contain all the macro elements, the general fertilizing trend of rich nitrogenous and poor potash nutrients is testified to from the survey data (see [7, 14]). Meanwhile, the application of organic fertilizer (mainly including manure and compost) is represented in terms of the counts of farms amongst both

**Unit N Mean Min Max Std. D. C.V.**

the total sample and those who used chemical fertilizer simultaneously (Table 4.2).

Organic fertilizer used in total farms Dummy 300 1=used (206); 0=unused (94)

fertilizers Dummy 224 1=used (137); 0=unused (87)

Chemical fertilizer kg/mu 294 58.489 8.890 285.710 45.609 0.780 Nitrogen kg/mu 278 34.589 2.140 285.710 34.276 0.991 Phosphate kg/mu 12 29.748 12.820 85.710 20.664 0.695 Potash kg/mu 4 132.500 10.000 200.000 89.954 0.679 Compound kg/mu 194 36.714 4.440 200.000 28.065 0.764

as 0 for farms locating in neither Beijing nor Shanghai.

**4.3 Analysis on fertilizer application**

**4.3.1 Behaviours on fertilizer application**

in each agro-product are included.

Farms that used organic and chemical

Source: field survey by the authors

Note: the bracketed numerals denote counts of farms

**Table 4.2** Application of fertilizer in the sampled farms

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The surveyed agro-products include wheat, corn, rice, cotton, fruits, vegetables, oilseed and peanut. The average fertilizer used in the three main grain crops of wheat, corn and rice is 55.31 kg per mu, which is much less than that of the other products at 91.60 kg per mu (Table 4.3). Within the three main grain crops, wheat is applied with the largest amounts of fertilizer, while vegetable is mostly fertilized amongst all the other categories of agricultural plants. As to the organic fertilizer, it is much more widely used in the three main grain crops than in the other agricultural plants.

#### **4.3.2 Perceptions on fertilizer application**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In the questionnaire, four questions concern farmers' perceptions on fertilizer application, from choosing, applying and determining the amounts of chemical fertilizer, to the conseȬ quences of over fertilization. Moreover, as most of the fertilizer bags are made from PVC, containing a variety of toxic cancer-causing substances, long-term storage of food can easily bring about damp mildew and produce a strong carcinogen of aflatoxin [20]. Thus the improper disposal of fertilizer containers may endanger environmental safety and human health, and farmers' disposal of the used fertilizer packages is required simultaneously. For each question, the number of valid responses, counts and proportions of responses to each choice are shown in Table 4.4.

For most of the farmers, productive effects are the first determining factors in choosing and using fertilizer, less attention is paid to the environmental effects and sprayers' health. When determine the mounts of fertilizer, more than 50% of farmers answered that they followed the


Source: field survey by the authors

**Table 4.4** Perceptions concerning fertilizer application

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

package instructions, while one third of them relied on their own experiences. In terms to the disposal of used fertilizer packages, almost 60% of farmers answered that they rinsed and reused, thus posing threats to the environment and human health. In many rural areas, farmers are storing their grains and other food stuffs in the used fertilizer bags, hence putting their food at high risk of being contaminated. Some farmers even rinse the used fertilizer bags in rivers, lakes, etc., hence constituting public water contaminations [21]. On the possible consequences of over fertilization, as a multiple-choice question, soil compaction is chosen by farmers accounting for an overwhelming ratio of 68.94%, followed by another choice of crop lodging with 45.78%. As to water contamination, this is chosen by less than one third of the respondents.

Thus, the proper and traditional perceptions coexist amongst the farmers, such as applying fertilizer according to the package instructions, concern about possible soil compaction due to over fertilization, while rinsing and reusing the packages for food storage, etc.

#### **4.3.3 Analysis of the behaviour determinants**

#### *4.3.3.1 Calculating the Fertilization coefficient*

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

package instructions, while one third of them relied on their own experiences. In terms to the disposal of used fertilizer packages, almost 60% of farmers answered that they rinsed and reused, thus posing threats to the environment and human health. In many rural areas, farmers are storing their grains and other food stuffs in the used fertilizer bags, hence putting their food at high risk of being contaminated. Some farmers even rinse the used fertilizer bags in rivers, lakes, etc., hence constituting public water contaminations [21]. On the possible consequences of over fertilization, as a multiple-choice question, soil compaction is chosen by farmers accounting for an overwhelming ratio of 68.94%, followed by another choice of crop lodging with 45.78%. As to water contamination, this is chosen by less than one third of the

**1. Determinants on choosing of fertilizer (Single-choice with 546 valid responses)**

**2. Determinants of using fertilizer (Single-choice with 546 valid responses)**

**3. Determinants of fertilizing amounts (Single-choice with 546 valid responses)**

**4. Disposal of the fertilizer packages (Single-choice with 555 valid responses)**

**5. Consequences from over fertilization (Multiple-choice with 557 valid responses)**

**Price Productivity Sellers Peer practices Follow-up services Environment**

**Costs Productivity Environmental effect Sprayers' health Quality of agro-product**

120 (21.98%) 343 (62.82%) 13 (2.38%) 7 (1.28%) 63 (11.54%)

**Container instructions Private experience Extension instruction Peer practices**

**Rinsing and recycling Burning up Littering Collective recycling Others**

326 (58.74%) 33 (5.95%) 57 (10.27%) 133 (23.96%) 6 (1.08%)

**Crop lodging Soil compaction Water contamination Yields increasing Unknown Others**

255 (45.78%) 384 (68.94%) 148 (26.57%) 85 (12.56%) 39 (7.00%) 17 (3.05%)

Note: numerals are the counts of valid responses and the bracketed numbers are the corresponding %s of responses

278 (50.92%) 191 (34.98%) 42 (7.69%) 35 (6.41%)

103 (18.86%) 380 (69.60%) 16 (2.93%) 31 (5.68%) 1 (0.18%) 15 (2.75%)

58 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Thus, the proper and traditional perceptions coexist amongst the farmers, such as applying fertilizer according to the package instructions, concern about possible soil compaction due to

over fertilization, while rinsing and reusing the packages for food storage, etc.

respondents.

Source: field survey by the authors

**Table 4.4** Perceptions concerning fertilizer application

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Application of fertilizers is mainly affected by three factors: soil properties represented by the geographical location in the National Fertilization Regionalization, agricultural planting structure and farmers' propensities. This study aims to identify the discrepancies amongst farmers in terms of their propensities and thus behaviours with regard to fertilizer application. Hence, for further analysis, it is necessary to isolate the impacts of the former two factors. In this survey, average amounts of fertilizer applied per mu in each agro-product vary amongst different areas in the China National Fertilization Regionalization (Table 4.5).

To show the pure effect of farmers' propensities on determining amounts of chemical fertilizer, an indicator of *FC* (Fertilization Coefficient) for the *i*-th farm is formulated as:

$$F\mathbf{C}\_{i} = \sum\_{j=1}^{8} \left( \frac{s\_{ij}}{s\_{i}} \cdot \frac{f\_{ij}}{f\_{\overline{k}^{j}}} \right) \text{ ( $i = 1, \text{ $ \cdots $, $ 294; k = 1, \text{  $\cdots$ ,  $4}$ )}\tag{4-1}$$


**Table 4.5** Average amounts of fertilizer applied to each agro-product in different regions (Unit: kg/mu)

where *sij* is the sowing scale of the *j*-th agricultural product in the *i*-th farm; *si* is the total sowing scale of agricultural plants in the *i*-th farm; *fij* is the fertilizer applied per mu to the *j*-th agricultural product in the *i*-th farm; *<sup>f</sup> kj*¯is the average amount of fertilizer applied per mu to the *j*-th agricultural product in the *k*-th region.

The summary statistics of the *FC*s for the 294 valid responses are shown in Table 4.6. To differentiate farmers' behaviours of fertilization driven by their propensities, they are divided into three groups in terms of their *FC*s, and the summary statistics for each group are provided in the same table. Group II embraces *FC*s fluctuating within 50% around 1, which represents the moderate amount of fertilizer determined by certain location and planting structure. Meanwhile, farms falling into Groups I and III indicate propensities of applying fertilizer with 50% under and over the moderate amounts, respectively. Statistics in this table show that Group II includes 180 farms (61.22%) with smallest Coefficient of Variance (CV) compared to the other two groups.


Source: field survey by the authors

**Table 4.6** Summary statistics of FC in different groups

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### *4.3.3.2 On the total amounts of fertilizer*

To model the factors significant for the *FC* of a farm falling to any of the groups above, the dependent variable is a dichotomous indicator coded 1 if belonging to a certain group and 0 if not. As the OLS models are inappropriate for the discrete and limited dependent variables [22], a binary logit regression model is adopted and formulated as [23]:

$$\log\left[\frac{P(Y\_1)}{P(Y\_0)}\right] = \beta\_0 + \sum\_{i=1}^{9} \beta\_i \mathbf{x}\_i + \varepsilon \tag{4-2}$$

where *P*(*Y*1) denotes the odds of *FC* belonging to a certain group, while *P*(*Y*0) represents being in other groups; *x*1, *x*2, …, *x*9 are the variables except for *fregion* in Table 4.1; Ά0 and Ά*<sup>i</sup>* are coefficients to be estimated; Ή is the random error.

Estimation of the model is carried out through application of the binary logistic regression procedure in SPSS 13.0. The backward approach is adopted to remove the statistically insignificant variables, (*p*-valueǃ0.1), from the initial model with all the candidate determiȬ nants as independent variables. The final model selected includes predictors embracing a *p*value less than 0.01 (Table 4.7). The column of B estimates log-odds coefficients of Ά*<sup>i</sup>* in Eq. 4-2, for predicting the dependent variable from the independent variables. The last column lists the exponentiation of B, the ratio of *P*(*Y*1) and *P*(*Y*0), thus called *odds ratios* simultaneously. In this case, an *odds ratio* over 1 denotes that the farm is more probable to fall into the group, while an *odds ratio* less than 1 implies that the farm is more likely to fall out of the group [24].

The results show that (1) sowing area (*scale*) is an essential factor occurring in all the three groups, as negative within both I and III, while positive in Group II. This reveals the existence of scale economy in terms of fertilizer application in the sampled farms, thus the increase of managerial scale is favourable for appropriate fertilization [15]. (2) As another significant determinant, total annual *income* is beneficial for the probability of using fewer amounts of fertilizer [13]. In this survey, an apparent positive relationship exists between annual *income* and non-agricultural ratios. Within the farms included in this model, no migrant income occurred in the farms with annual cash income less than 10000 yuan, while this ratio in the other three income levels of Table 4.1 are 26.71%, 51.36% and 55.54%, respectively. The more non-agricultural income usually results in less farming time and attention to agricultural yields, thus the application of fertilizer may be decreased. (3) Meanwhile, effect from income ratio of migrant job (*mir*) is measured as being negatively correlated with amounts of fertilizer. Due to the instability and high expenditure of living away from homeland, most migrant farmers have to leave their families at home and engage in agriculture. As most of the left family members are women, children and the elderly, they are prone to improving agricultural productivity through the use of chemical fertilizer. The negative effect of *mir* in Group I may reveal that the more they get from migrant jobs, the more they can afford to use fertilizer [7]. Meanwhile, due to the lack of prime labour, most of them are not over-fertilizing, thus being positive in Group II. (4) The positive effect of *age* in Group I reveals that farms headed by the elderly are less prone to using fertilizer than the average amounts. This may be interpreted as due to limitation of physical power, disposable income, etc. (5) As analysed above, the three types of staple grain crops are supplied with less fertilizer than the other agricultural products. Therefore, their sowing ratios (*grainr*) go positively in Group I and are hence negatively correlated with the total amounts of fertilizer.


Omnibus tests of coefficients for model I: Chi-square (5)=23.941, Sig.=0.000\*\*\*

Omnibus tests of coefficients for model II: Chi-square (1)=5.002, Sig.=0.025\*\*

Omnibus tests of coefficients for model III: Chi-square (3)=25.191, Sig.=0.000\*\*\*

Note: \*\*\* and \*\*represents statistical significance in the level of 1% and 5%, respectively

Software: SPSS 13.0

50% under and over the moderate amounts, respectively. Statistics in this table show that Group II includes 180 farms (61.22%) with smallest Coefficient of Variance (CV) compared to

60 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Group Range N Mean Min Max Std. D. C.V.** I (0, 0.50) 84 0.310 0.080 0.496 0.107 0.343 II [0.50, 1.5) 180 0.949 0.500 1.486 0.252 0.266 III [1.5, +Ƙ) 30 2.133 1.514 3.804 0.588 0.276 Total 294 0.887 0.080 3.804 0.577 0.650

To model the factors significant for the *FC* of a farm falling to any of the groups above, the dependent variable is a dichotomous indicator coded 1 if belonging to a certain group and 0 if not. As the OLS models are inappropriate for the discrete and limited dependent variables

9

=

where *P*(*Y*1) denotes the odds of *FC* belonging to a certain group, while *P*(*Y*0) represents being in other groups; *x*1, *x*2, …, *x*9 are the variables except for *fregion* in Table 4.1; Ά0 and Ά*<sup>i</sup>*

Estimation of the model is carried out through application of the binary logistic regression procedure in SPSS 13.0. The backward approach is adopted to remove the statistically insignificant variables, (*p*-valueǃ0.1), from the initial model with all the candidate determiȬ nants as independent variables. The final model selected includes predictors embracing a *p*value less than 0.01 (Table 4.7). The column of B estimates log-odds coefficients of Ά*<sup>i</sup>*

4-2, for predicting the dependent variable from the independent variables. The last column lists the exponentiation of B, the ratio of *P*(*Y*1) and *P*(*Y*0), thus called *odds ratios* simultaneously. In this case, an *odds ratio* over 1 denotes that the farm is more probable to fall into the group, while an *odds ratio* less than 1 implies that the farm is more likely to fall out of the group [24]. The results show that (1) sowing area (*scale*) is an essential factor occurring in all the three groups, as negative within both I and III, while positive in Group II. This reveals the existence of scale economy in terms of fertilizer application in the sampled farms, thus the increase of managerial scale is favourable for appropriate fertilization [15]. (2) As another significant determinant, total annual *income* is beneficial for the probability of using fewer amounts of

 EH

« » ¬ ¼ ¦ (4-2)

are

in Eq.

0 0 1

E

« » =+ +

( ) *i i i*

[22], a binary logit regression model is adopted and formulated as [23]:

1

*P Y Log <sup>x</sup>*

( )

ª º

*P Y*

the other two groups.

Source: field survey by the authors

**Table 4.6** Summary statistics of FC in different groups

coefficients to be estimated; Ή is the random error.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

*4.3.3.2 On the total amounts of fertilizer*

**Table 4.7** Binary logistic regression on FC of different groups

#### *4.3.3.3 On the application of organic fertilizer*

With the same binary logistic regression procedure in SPSS 13.0, we measure significant factors for the application of organic fertilizer in the sampled farms. Besides the aforementioned nine variables, we add three variables into the candidate determinants: amount of chemical fertilizer (*fert*), quantity of *livestock* and *poultry* to capture the possible impacts from these predictors, with the hypothesis that these variables affect farmers' application of organic fertilizer.

As shown in Table 4.8, through the predictor selection backward method, six variables are included in the final model. Judging from the *odds ratio* of each variable, impact of each variable can be identified.

(1) Farms from the north (*north*=1) are less likely to use organic fertilizer. Further investigations are necessary to explore the possible reasons for this from planting structure, habits and awareness on the function of organic fertilizer. (2) *Age* of farm head is positively correlated with farmyard application. This may be interpreted as the accumulation of social experiences, farmers are more confident about the effectiveness of organic fertilizer, or the significance of properly disposing of faeces and urine. (3) Similar with the findings of Zheng (2010) [25], annual cash *incomes* is also positively correlated with farmyard application. With the increase of income, farmers need cleaner environments and safer food supplies, thus they are apt to fertilize their farmland with organic fertilizer, rather than chemical fertilizer (as analysed above). (4) Farms with larger sowing *scale*s are less prone to use organic fertilizer, probably due to the fact that they are pursuing higher production efficiency and tend to use chemical fertilizer. In addition, the collection and application of organic fertilizer enough for their large sowing scales is consuming labour and funds. (5) The sowing ratio of grain crops (*grainr*) is negatively correlated with the application of organic fertilizer, which can be interpreted as that most of the grains are sold while the economic agro-products will be consumed by the farmers themselves. Hence, they are tending to fertilize the economic crops with organic fertilizer which is labour-consuming, but deemed as salubrious by the farmers [26]. (6) Income the ratio of migrant job (*mir*) is found to be negatively correlated with the application of organic fertilizer. The main reason behind this may be the fact that farms which are more reliant on the non-agricultural income usually have less time for farming, much less fertilizing their farmland through organic fertilizer. Meanwhile, no significant relationships are detected between the application of organic fertilizer and chemical fertilizer (similar to Zheng (2010) [25]), breeding of livestock and poultry. This indicates the existence of a certain blindness about the application of organic fertilizer, which may bring about improper disposal of manure and compost, thus environmental pollutions.

#### **4.3.4 Conclusions and Recommendations**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### *4.3.4.1 Major conclusions*

Based on a survey of 560 household farms in six eastern provincial level regions of China, this study explores farmers' behaviours, perceptions and determinants of fertilizer application. The behaviours involve total amount of chemical fertilizer and the use of organic fertilizer;


Cases included in analysis: 267; Missing cases: 33; Total cases selected: 300

Dependent variable: whether organic fertilizer is used, with 178 cases = 1, and 89 cases = 0

Omnibus tests of model coefficients: Chi-square (6)=86.382, Sig.=0.000\*\*\*

Note: \*\*\*, \*\*and \*represent statistical significance in the level of 1%, 5% and 10%, respectively.

Software: SPSS 13.0

*4.3.3.3 On the application of organic fertilizer*

compost, thus environmental pollutions.

**4.3.4 Conclusions and Recommendations**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

*4.3.4.1 Major conclusions*

fertilizer.

can be identified.

With the same binary logistic regression procedure in SPSS 13.0, we measure significant factors for the application of organic fertilizer in the sampled farms. Besides the aforementioned nine variables, we add three variables into the candidate determinants: amount of chemical fertilizer (*fert*), quantity of *livestock* and *poultry* to capture the possible impacts from these predictors, with the hypothesis that these variables affect farmers' application of organic

62 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

As shown in Table 4.8, through the predictor selection backward method, six variables are included in the final model. Judging from the *odds ratio* of each variable, impact of each variable

(1) Farms from the north (*north*=1) are less likely to use organic fertilizer. Further investigations are necessary to explore the possible reasons for this from planting structure, habits and awareness on the function of organic fertilizer. (2) *Age* of farm head is positively correlated with farmyard application. This may be interpreted as the accumulation of social experiences, farmers are more confident about the effectiveness of organic fertilizer, or the significance of properly disposing of faeces and urine. (3) Similar with the findings of Zheng (2010) [25], annual cash *incomes* is also positively correlated with farmyard application. With the increase of income, farmers need cleaner environments and safer food supplies, thus they are apt to fertilize their farmland with organic fertilizer, rather than chemical fertilizer (as analysed above). (4) Farms with larger sowing *scale*s are less prone to use organic fertilizer, probably due to the fact that they are pursuing higher production efficiency and tend to use chemical fertilizer. In addition, the collection and application of organic fertilizer enough for their large sowing scales is consuming labour and funds. (5) The sowing ratio of grain crops (*grainr*) is negatively correlated with the application of organic fertilizer, which can be interpreted as that most of the grains are sold while the economic agro-products will be consumed by the farmers themselves. Hence, they are tending to fertilize the economic crops with organic fertilizer which is labour-consuming, but deemed as salubrious by the farmers [26]. (6) Income the ratio of migrant job (*mir*) is found to be negatively correlated with the application of organic fertilizer. The main reason behind this may be the fact that farms which are more reliant on the non-agricultural income usually have less time for farming, much less fertilizing their farmland through organic fertilizer. Meanwhile, no significant relationships are detected between the application of organic fertilizer and chemical fertilizer (similar to Zheng (2010) [25]), breeding of livestock and poultry. This indicates the existence of a certain blindness about the application of organic fertilizer, which may bring about improper disposal of manure and

Based on a survey of 560 household farms in six eastern provincial level regions of China, this study explores farmers' behaviours, perceptions and determinants of fertilizer application. The behaviours involve total amount of chemical fertilizer and the use of organic fertilizer; **Table 4.8** Binary logistic regression on application of organic fertilizer

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

farmers' perceptions range from choosing and field application, the consequences of over fertilization and disposal of the used packages. Logistic regression models are used to identify the significant determinants of their behaviours.

The survey shows that most farms are using nitrogen fertilizers, while potash fertilizers are used in a few farms. Compared with the other plants, less chemical fertilizers are used in the main grain crops of wheat, corn and rice. Judging from the Fertilization Coefficient, more than 60% of farms are using fertilizer in amounts deviating by no more than 50% from the average amounts in each fertilizing region. Perceptions of proper fertilization are held by some farmers, including applying the fertilizer by instructions, recycling the packages collectively and concern about the possible crop lodging and soil compaction due to over fertilization. SimulȬ taneously, traditional conceptions are still held by many farmers, such as the over emphasized production effects and private experiences, reusing the packages for food storage, etc.

According to the empirical analyses, sowing area and ratio of migrant income is measured as positively correlated, while annual income is negatively correlated with appropriate fertilizaȬ tion. As to the odds of using organic fertilizer, no significant effect is detected from chemical fertilizer application and the breeding of livestock and poultry. Nevertheless, cash income and age of householders are positively correlated with the application of organic fertilizer, while location in the north, sowing scale, ratio of grain crops and migrant income are measured as negatively correlated with the same behaviour.

#### *4.3.4.2 Policy recommendations*

(1) As shown above, the fertilizing elements are not well balanced, and amounts of fertilizer used in many farms deviate much from the moderate levels. Therefore, it is an urgent task for the government to provide prompt, accurate and convenient soil testing techniques, and recommend referential standardized fertilizing amounts to farmers with different land properties and planting structures [3]. (2) Enlarging the managerial scales of agriculture. As analysed above, larger scale is positively correlated with maintaining appropriate fertilizing amounts. Managerial scales of the farms can be expanded either through the concentration of land based on the farms' own wishes or joining into the Specialized Farmers' Cooperatives as demonstrated by Sun (2008) [27] and Dai (2010) [12]. (3) Promoting migrant employment of rural labour, as ratio of migrant income is positively correlated with appropriate use of fertilization and application of organic fertilizer. As for accelerating the transfer of surplus labour from agriculture to the other sectors, thus increasing rural household incomes, the main tasks include promoting vocational training, perfecting the employment information networks and protecting the legal rights of the migrant workers. (4) Strengthening social education on scientific fertilization. This survey reveals that behaviours, including fertilizing by private practices, misusing the used packages, etc., still exist amongst many farmers, and their perceptions on safe application of fertilizers need to be improved. Hence education on appropriate amounts of fertilizer, balancing the elements, proper recycling of the used packages, etc., needs to be strengthened [26].

#### **4.4 Analysis on pesticide application**

#### **4.4.1 Behaviours on pesticide application**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

To capture the major behaviours of pesticide application in a farm, three aspects are included in our questionnaire. As usually a variety of pesticides, with different pest control and environmenȬ tal effects, are used in a farm, weights of pesticides applied in each agricultural product are summed to constitute the total amounts. Meanwhile, as the control of toxic pesticides and promotion of biological pest controls are of great importance for safe agricultural production, relevant characteristics are also included. The toxic pesticides incorporate methamidophos, furadan (carbofuran) and folimat [28]. According to the No. 199 Bulletin of the China AgriculturȬ al Ministry (2002), it is prohibited to apply methamidophos in agriculture and Furadan cannot be used on vegetables, fruit, tea and medicinal herbs. As another major toxic pesticide, FoliȬ mat has been banned in some regions including Zhejiang [29], Jiangsu [30], etc.

The bio-control methods of pests in agriculture are measures to eliminate insects, mites, weeds and plant diseases, etc., that rely on certain biological mechanisms of secretion, smell, predaȬ tion, parasitism, herbivory, etc., thus reducing the use of chemical pesticides. For example, using the smell of onions to kill microbes causing black spike of wheat, intercropping beans in corn fields to attract beneficial insects that prey upon pests, raising ducks and fish in rice fields to control weeds, etc [31]. The application of pesticides and bio-control measures in the sampled farms are shown in Table 4.10.


Note: a as a main unit of land measurement in China, 1 mu=666.67*m*2; b the income sources contain migrant jobs and sales of agricultural products; c the bracketed numerals denote counts of farms

Source: field survey by the authors

*4.3.4.2 Policy recommendations*

packages, etc., needs to be strengthened [26].

**4.4 Analysis on pesticide application**

**4.4.1 Behaviours on pesticide application**

sampled farms are shown in Table 4.10.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

(1) As shown above, the fertilizing elements are not well balanced, and amounts of fertilizer used in many farms deviate much from the moderate levels. Therefore, it is an urgent task for the government to provide prompt, accurate and convenient soil testing techniques, and recommend referential standardized fertilizing amounts to farmers with different land properties and planting structures [3]. (2) Enlarging the managerial scales of agriculture. As analysed above, larger scale is positively correlated with maintaining appropriate fertilizing amounts. Managerial scales of the farms can be expanded either through the concentration of land based on the farms' own wishes or joining into the Specialized Farmers' Cooperatives as demonstrated by Sun (2008) [27] and Dai (2010) [12]. (3) Promoting migrant employment of rural labour, as ratio of migrant income is positively correlated with appropriate use of fertilization and application of organic fertilizer. As for accelerating the transfer of surplus labour from agriculture to the other sectors, thus increasing rural household incomes, the main tasks include promoting vocational training, perfecting the employment information networks and protecting the legal rights of the migrant workers. (4) Strengthening social education on scientific fertilization. This survey reveals that behaviours, including fertilizing by private practices, misusing the used packages, etc., still exist amongst many farmers, and their perceptions on safe application of fertilizers need to be improved. Hence education on appropriate amounts of fertilizer, balancing the elements, proper recycling of the used

64 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

To capture the major behaviours of pesticide application in a farm, three aspects are included in our questionnaire. As usually a variety of pesticides, with different pest control and environmenȬ tal effects, are used in a farm, weights of pesticides applied in each agricultural product are summed to constitute the total amounts. Meanwhile, as the control of toxic pesticides and promotion of biological pest controls are of great importance for safe agricultural production, relevant characteristics are also included. The toxic pesticides incorporate methamidophos, furadan (carbofuran) and folimat [28]. According to the No. 199 Bulletin of the China AgriculturȬ al Ministry (2002), it is prohibited to apply methamidophos in agriculture and Furadan cannot be used on vegetables, fruit, tea and medicinal herbs. As another major toxic pesticide, FoliȬ

The bio-control methods of pests in agriculture are measures to eliminate insects, mites, weeds and plant diseases, etc., that rely on certain biological mechanisms of secretion, smell, predaȬ tion, parasitism, herbivory, etc., thus reducing the use of chemical pesticides. For example, using the smell of onions to kill microbes causing black spike of wheat, intercropping beans in corn fields to attract beneficial insects that prey upon pests, raising ducks and fish in rice fields to control weeds, etc [31]. The application of pesticides and bio-control measures in the

mat has been banned in some regions including Zhejiang [29], Jiangsu [30], etc.

**Table 4.9** Demographic characteristics of the sampled farms applying pesticides


Note: the bracketed numerals denote counts of farms

Source: field survey by the authors

**Table 4.10** Application of pesticides in the sampled farms


**Table 4.11** Application of pesticides in each agricultural product

The agricultural products we surveyed include wheat, corn, rice, cotton, oilseed, soy and fruits, and application of pesticides per mu of each product is presented in Table 4.11. The average pesticides used in the three main grain crops of wheat, corn and rice is 0.51 kg per mu, which is much less than that of the other products at 1.79 kg per mu. Meanwhile, judging from the Coefficient of Variance (CV), amounts of pesticides used in these main grain crops show larger discrepancies than that of the other products. According to the survey, toxic pesticides are used in all the products except for fruit, amongst which methamidophos is used in rice and soy, Folimat is used in wheat, cotton, cole and cotton, while Furadan is used in cotton. Finally, bio-control methods are used on far fewer farms, covering most of the products other than cotton and soy.

#### **4.4.2 Perceptions on pesticide application**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Within this questionnaire, five questions concern farmers' perceptions on pesticide applicaȬ tion, from choosing and field application to the withdrawal periods, and the possible conseȬ quences of overdosing. Moreover, as pesticide containers may be toxic and improper disposal may threaten environmental safety and human health [32], another question is adopted in this topic. For each question, the number of valid responses, counts of responses and the correȬ sponding percentages to each choice are shown in Table 4.13.

This shows that for most of the farmers, productive effects are the most determining factors in choosing and using pesticides, less attention is paid to the environmental effects and sprayers' health. When determining the doses, almost 50% of the farmers answer that they follow the container instructions, while one third of them rely on their own experiences. Although more than 80% of the farmers have heard of the withdrawal period of pesticides, less than 20% of the farmers answered as being well known. In the disposal of pesticide containers, almost 40% of the farmers answered as littering, thus threatening the environment and human health. On the possible consequences of overdosing, the negative effects on sprayers' health, food safety and environment are recognized by more than half of the respondents simultaneously. Hereby, the coexistence of the proper and traditional perceptions is shown amongst the farmers.

#### **4.4.3 Analysis on the behaviour determinants**

#### *4.4.3.1 On the total amount of pesticides*

**Application of chemical pesticides Number of farms using**

**Toxic pesticides**

**Bio-control**

**Unit N Mean Min Max Std. D. C.V.**

66 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Source: field survey by the authors

cotton and soy.

**Table 4.11** Application of pesticides in each agricultural product

**4.4.2 Perceptions on pesticide application**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

sponding percentages to each choice are shown in Table 4.13.

Wheat kg/mu 95 0.37 0.01 3.00 0.56 1.53 48 9 Corn kg/mu 48 0.23 0.02 1.25 0.27 1.21 32 6 Rice kg/mu 46 1.09 0.02 5.00 1.31 1.20 15 2 Cotton kg/mu 28 1.44 0.02 6.00 1.69 1.17 17 0 Fruiter kg/mu 26 5.29 0.40 11.67 2.50 0.47 0 27 Oilseed kg/mu 34 0.58 0.15 2.13 0.43 0.75 33 4 Soy kg/mu 27 0.32 0.10 0.50 0.09 0.28 26 0

The agricultural products we surveyed include wheat, corn, rice, cotton, oilseed, soy and fruits, and application of pesticides per mu of each product is presented in Table 4.11. The average pesticides used in the three main grain crops of wheat, corn and rice is 0.51 kg per mu, which is much less than that of the other products at 1.79 kg per mu. Meanwhile, judging from the Coefficient of Variance (CV), amounts of pesticides used in these main grain crops show larger discrepancies than that of the other products. According to the survey, toxic pesticides are used in all the products except for fruit, amongst which methamidophos is used in rice and soy, Folimat is used in wheat, cotton, cole and cotton, while Furadan is used in cotton. Finally, bio-control methods are used on far fewer farms, covering most of the products other than

Within this questionnaire, five questions concern farmers' perceptions on pesticide applicaȬ tion, from choosing and field application to the withdrawal periods, and the possible conseȬ quences of overdosing. Moreover, as pesticide containers may be toxic and improper disposal may threaten environmental safety and human health [32], another question is adopted in this topic. For each question, the number of valid responses, counts of responses and the correȬ

This shows that for most of the farmers, productive effects are the most determining factors in choosing and using pesticides, less attention is paid to the environmental effects and sprayers' health. When determining the doses, almost 50% of the farmers answer that they follow the container instructions, while one third of them rely on their own experiences. Although more than 80% of the farmers have heard of the withdrawal period of pesticides, less than 20% of the farmers answered as being well known. In the disposal of pesticide containers, almost 40% of the farmers answered as littering, thus threatening the environment and human health. On the possible consequences of overdosing, the negative effects on

In the prior studies, multivariate OLS regression models are used to identify the significant determinants of pesticide application, as in Wang (2004) [11], Li et al. (2007) [17], etc. In this study, the model used to find the important factors of total chemical pesticide amount is formulated as:

$$Y = \beta\_0 + BX + \mu \tag{4-3}$$

where *Y* is the total amount of pesticides applied per mu, *X*=(*x*1, *x*2, …, *x*9) T is a vector containing the nine variables listed in Table 4.12, Ά0 and *B*=(Ά1, Ά2, …, Ά9) are coefficients that need to be estimated, while *u* is the random error.

Through the multivariate linear regression process, with the independent variable selection backward method in the statistical software of SPSS 13.0, four significant determinants are chosen in the final model. The relevant statistics of the model are shown in Table 4.12. The significant values of *F* and *t* (*p*-value < 0.1) indicate the good fitness of this model2 .


Note: dependent variable: pesticides used per mu; \*\*\*, \*\*and \*represent statistical significance in the level of 1%, 5% and 10%, respectively.

Software: SPSS 13.0

**Table 4.12** Statistics of the significant determinants on total pesticide used per mu

<sup>2</sup> Although a not very high R2 value of 0.406 is given in the table, it should not be used to judge the fitness of a model. The fact that R2 never decreases when any variable is added to a regression makes it a poor tool for deciding whether one or several variables should be added to a model. Low R2s in regression equations are not uncommon, especially for cross-sectional analysis. Thus, using R2 as the main gauge of success for an econometric analysis can lead to inaccuracies [33].

The results show that farms affiliating to the two metropolises of Beijing and Shanghai (*metro*=1), or headed by males (*gender*=1) are positively correlated, while ratios of income from migrant jobs (*mir*) and grains sowing scales (*grainr*) are negatively correlated with the amount of chemical pesticides applied per mu. (1) The coefficient of *metro* can be explained by the comparison of average pesticides used per mu and other indicators of the farms. Within the 199 farms included in this model, farms affiliating to the metropolises applied 2.64 kg of pesticides per mu with the sowing area of 3.84 mu on average, while the corresponding indicators in non-metropolis farms are 0.38 kg per mu and 6.70 mu*,* respectively, thus the former may have to maintain high yields through more application of pesticides. SimultaneȬ ously, the higher annual cash incomes in farms affiliating to the metropolises3 enable them to input more in pesticides. However, we should notice that this discrepancy may threaten the environmental and food safety of the metropolises.


a Note: numerals are the counts of valid farm, and the bracketed numbers are the corresponding %s of farms

Source: field survey by the authors

**Table 4.13** Perceptions concerning pesticide application

<sup>3</sup> Using the codes of 1, 2, 3, 4 to denote the ascending income levels of Table 1, within the 199 farms included in this model, the mean in farms affiliating to the metropolises is 3.02, while that in the other farms is 2.48.

(2) As to the finding that male headed farms are applying more pesticides, this indicates that males are more concerned about the productive effects of farming activities and more able to spray large volume of pesticides because of pure physical power, as investigated by Li et al. (2007) [32]. (3) The negative effect of income ratio of migrant job is in line with Li et al. (2007) [17]. The greater the amount of non-agricultural income, usually results in less farming time and attention to agricultural yields, thus the application of pesticides may be decreased. (4) As analysed above, the three types of staple grain crops are supplied with less pesticides than the other agricultural products. Therefore, their sowing ratio goes negatively correlated with the total amount of pesticides.

#### *4.4.3.2 On toxic pesticide application*

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The results show that farms affiliating to the two metropolises of Beijing and Shanghai (*metro*=1), or headed by males (*gender*=1) are positively correlated, while ratios of income from migrant jobs (*mir*) and grains sowing scales (*grainr*) are negatively correlated with the amount of chemical pesticides applied per mu. (1) The coefficient of *metro* can be explained by the comparison of average pesticides used per mu and other indicators of the farms. Within the 199 farms included in this model, farms affiliating to the metropolises applied 2.64 kg of pesticides per mu with the sowing area of 3.84 mu on average, while the corresponding indicators in non-metropolis farms are 0.38 kg per mu and 6.70 mu*,* respectively, thus the former may have to maintain high yields through more application of pesticides. SimultaneȬ

68 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

input more in pesticides. However, we should notice that this discrepancy may threaten the

**Price Productivity Sellers Peer practices Follow-up services Environment** 103 (18.86%) a 380 (69.60%) 16 (2.93%) 31 (5.68%) 1 (0.18%) 15 (2.75%)

**Costs Productive effect Environmental effect Sprayers' health Quality of agro-product**

120 (21.98%) 343 (62.82%) 13 (2.38%) 7 (1.28%) 63 (11.54%)

**Container instructions Private experience Extension Instruction Peer practices** 278 (50.92%) 191 (34.98%) 42 (7.69%) 35 (6.41%)

**Individual recycling Burning up Littering Collective recycling Others** 79 (14.36%) 73 (13.27%) 212 (38.55%) 182 (33.09%) 4 (0.73%) **6. Consequences from overdosing of pesticides (Multiple-choice with 557 valid responses)**

**Sprayers' health Food security Pollution Higher effectivity Unknown Others** 337 (60.50%) 423 (75.94%) 316 (56.73%) 105 (18.85%) 16 (2.87%) 9 (1.62%)

Note: numerals are the counts of valid farm, and the bracketed numbers are the corresponding %s of farms

the mean in farms affiliating to the metropolises is 3.02, while that in the other farms is 2.48.

3 Using the codes of 1, 2, 3, 4 to denote the ascending income levels of Table 1, within the 199 farms included in this model,

enable them to

ously, the higher annual cash incomes in farms affiliating to the metropolises3

**1. Determinants on choosing of pesticides (Single-choice with 546 valid responses)**

**2. Determinants of using pesticides (Single-choice with 546 valid responses)**

**3. Determinants of pesticides dose (Single-choice with 546 valid responses)**

**4. Withdrawal period of pesticides (Single-choice with 557 valid responses) Knows very well Knows fairly well Knows a little Unknown** 97 (17.41%) 248 (44.5%) 105 (18.85%) 107 (19.21%) **5. Disposal of the pesticide containers (Single-choice with 550 valid responses)**

a

Source: field survey by the authors

**Table 4.13** Perceptions concerning pesticide application

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

environmental and food safety of the metropolises.

As to the model factors significant for application of toxic pesticides defined above, the dependent variable is a dichotomous indicator coded 1 if applied and 0 if not. As the OLS modes like Eq.4-3 are inappropriate for discrete and limited dependent variables [22], a binary logit regression model is adopted [34], [12] and defined as [23]:

$$\log\left[\frac{P(Y\_1)}{P(Y\_0)}\right] = \beta\_0 + \sum\_{i=1}^{9} \beta\_i x\_i + \varepsilon \tag{4-4}$$

where *Y* is the application of toxic pesticides with *P*(*Y*1) denoting the probability of being applied, while *P*(*Y*0) means that toxic pesticides are unapplied, *xi* (*i*=1, 2, …, 9) are the nine variables listed in Table 4.14, Ά0 and Ά*<sup>i</sup>* (*i*=1, 2, …, 9) are coefficients that need to be estimated, and Ή is the random error.

Estimation of this model is carried out through application of the binary logistic regression procedure in SPSS 13.0. The backward approach is adopted to remove the statistically insignificant variables, (*p*-valueǃ0.1), from the initial model with all the candidate determiȬ nants as independent variables. The final model includes four predictors, all of which embrace *p*-values less than 0.01 (Table 4.14). Column B estimates log-odds coefficients of Ά*<sup>i</sup>* in Eq.4-4, for predicting the dependent variable by the independent variables. The last column lists the exponentiation of B, the ratio of *P*(*Y*1) and *P*(*Y*0), thus called *odds ratios* simultaneously. In this case, an *odds ratio* over 1 denotes that the toxic pesticides are more probably used, while an *odds ratio* less than 1 implies that the toxic pesticides are less likely to be used [24].

Within the four significant variables listed in Table 4.14, *mir* is positively correlated with the odds of toxic pesticides being applied, while the other three variables are negatively correlated with the application probability of toxic pesticides on a farm. (1) Being the capital and largest city in China, respectively, and especially thanks to the hosting of the Olympic Games and World Expo, Beijing and Shanghai have adopted stringent regulations to prevent the use of highly toxic pesticides [35-36]. Therefore, the less probability of applying toxic pesticides and negative effect of *metro* can be interpreted. (2) For a farmer, the greater the income from migrant jobs means less time and attention for farming in general. However, due to unstable conditions and high living expenditure outside of the native place (which is usually defined using the county level), most migrant farmers have to leave their families at home and engage in agriculture [19]. As most of the left family members are women, children and the elderly, they are prone to control pests through the more efficient toxic pesticides. The positive effect of *mir* may reveal that the greater the income obtained from migrant jobs, the more they can afford to buy and use the toxic pesticides. (3) However, when were observe farms' cash *income* with units of dozens of thousand yuan as shown in Table 4.14, farms with an upper level of income tend to use less toxic pesticides as their major income comes from non-agricultural sectors4 . Through tradeoffs with the probable efficient pest control by toxic pesticides, most of them may prefer to conserve the environment and food security. (4) Finally, as the three types of staple grain crops need less pesticide in general, the application of toxic pesticides is negatively correlated with the *grainr* simultaneously.


Cases included in analysis: 199; missing cases: 21; total cases selected: 220

Dependent variable: whether toxic pesticides are used, with 93 cases = 1, and 106 cases = 0

Omnibus tests of model coefficients: *Chi*-square (4)=71.642, Sig.=0.000\*\*\*

Note: \*\*\* represents statistical significance in the level of 1%

Software: SPSS 13.0

**Table 4.14** Binary logistic regression on whether toxic pesticides used

#### *4.4.3.3 On the adoption of biological pest-controls*

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

With the same binary logistic regression procedure in SPSS 13.0 and the nine variables as the candidate determinants, we measure the significant factors for the implementation of biologȬ ical pest control in the sampled farms.

As shown in Table 4.15, using the backward predictor selection method, three variables are included in the final model. Judging from the *odds ratio* of each variable, (1) farms from the north (*north*=1) or (2) affiliating to the two metropolises (*metro*=1) are likely to adopt biological measures. Within the farms answered as conducting biological pest control, 36 are from the north and 31 from the two metropolises, accounting for 78.26% and 67.39% with the valid number of 46, respectively. As to the positive significance of *metro*, this may be because that

<sup>4</sup> Within the 199 farms included in this model, no migrant income occurred in the farms with annual cash income less than 10000 yuan, while this ratio in the other three income levels of Table 1 are 22.70%, 54.58% and 60.56%, respectively.

as previously mentioned, being the capital and largest city in China respectively, Beijing and Shanghai are making full use of their solid industrial foundation and advantages in technolȬ ogy, trade, information - making greater efforts to promote the research and production of low toxicity and environmentally friendly pesticides (see [37, 31]). As to the difference between the north and south, further investigations are necessary to explore the possible reasons for variations in cropping structure, farming habits, the degree of pest damage, etc. [18], hence searching for suitable countermeasures to extend biological pest controls in different regions. Meanwhile, (3) income ratio of migrant job (*mir*) is found negatively correlated with the introduction of biological pest controls. this may be because those farms rely more on nonagricultural incomes, thus usually have less time to attend to farming, much less controlling pests through biological methods.


Cases included in analysis: 274; Missing cases: 286; Total cases selected: 560

Dependent variable: whether biological pest controls are implemented, with 27 cases = 1, and 247 cases = 0

Omnibus tests of model coefficients: *Chi*-square (3)=47.607, Sig.=0.000\*\*\*

Note: \*\*\*, \*\*and \*represent statistical significance in the level of 1%, 5% and 10%, respectively

Software: SPSS 13.0

.

county level), most migrant farmers have to leave their families at home and engage in agriculture [19]. As most of the left family members are women, children and the elderly, they are prone to control pests through the more efficient toxic pesticides. The positive effect of *mir* may reveal that the greater the income obtained from migrant jobs, the more they can afford to buy and use the toxic pesticides. (3) However, when were observe farms' cash *income* with units of dozens of thousand yuan as shown in Table 4.14, farms with an upper level of income tend to use less toxic pesticides as their major income comes from non-agricultural sectors4

70 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Through tradeoffs with the probable efficient pest control by toxic pesticides, most of them may prefer to conserve the environment and food security. (4) Finally, as the three types of staple grain crops need less pesticide in general, the application of toxic pesticides is negatively

Metropolis or not (*metro*) -2.507\*\*\* 0.607 17.051 1 0.000 0.082 Income ratio of migrant job (*mir*) 0.018\*\*\* 0.005 11.975 1 0.001 1.081 Total cash income in 2010 (*income*) -0.755\*\*\* 0.251 9.019 1 0.003 0.470 Ratio of grain sowing scale (*grainr*) -0.027\*\*\* 0.006 18.828 1 0.000 0.974 (Constant) 2.515 0.640 15.458 1 0.000 12.363

With the same binary logistic regression procedure in SPSS 13.0 and the nine variables as the candidate determinants, we measure the significant factors for the implementation of biologȬ

As shown in Table 4.15, using the backward predictor selection method, three variables are included in the final model. Judging from the *odds ratio* of each variable, (1) farms from the north (*north*=1) or (2) affiliating to the two metropolises (*metro*=1) are likely to adopt biological measures. Within the farms answered as conducting biological pest control, 36 are from the north and 31 from the two metropolises, accounting for 78.26% and 67.39% with the valid number of 46, respectively. As to the positive significance of *metro*, this may be because that

4 Within the 199 farms included in this model, no migrant income occurred in the farms with annual cash income less than 10000 yuan, while this ratio in the other three income levels of Table 1 are 22.70%, 54.58% and 60.56%, respectively.

**Variables B S.E. Wald df Sig.** *odds ratio*

correlated with the *grainr* simultaneously.

Cases included in analysis: 199; missing cases: 21; total cases selected: 220

Omnibus tests of model coefficients: *Chi*-square (4)=71.642, Sig.=0.000\*\*\*

**Table 4.14** Binary logistic regression on whether toxic pesticides used

Note: \*\*\* represents statistical significance in the level of 1%

*4.4.3.3 On the adoption of biological pest-controls*

ical pest control in the sampled farms.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Software: SPSS 13.0

Dependent variable: whether toxic pesticides are used, with 93 cases = 1, and 106 cases = 0

**Table 4.15** Binary logistic regression on implementation of biological pest control

#### **4.4.4 Conclusions and recommendations**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### *4.4.4.1 Major conclusions*

The survey shows that pesticides used on the three staple grain crops are less than those used on other products, but there is more discrepancy amongst the farms. The toxic pesticides are applied in most of the products and some 50% of the sampled farms, while bio-control methods are used in only about one sixth of the farms. Perceptions on proper application of pesticides exist amongst some of the farmers, including applying according to the instructions on the containers, awareness on the withdrawal periods, collective recycling of the containers, concern about sprayers' health and food security. Simultaneously, traditional conceptions still influence many of them, such as the over emphasized importance of productive effects and private experiences, littering by incorrect disposal of pesticide containers, etc.

According to the empirical analyses, farms in the two metropolises and those headed by males are positively correlated, while ratios of income from migrant jobs and grain sowing scales are negatively correlated with the amounts of pesticides applied. With respect to the application probability of toxic pesticides, the income ratio of migrant job is positively correlated, while the other three variables of *metro*, *income* and *grainr* embrace negative effects. Farms' location, whether north or affiliating to the metropolises, is measured as positively correlated, while ratio of migrant income is negatively correlated with the odds of adopting biological pest controls.

#### *4.4.4.2 Policy recommendations*

(1) Extending advanced techniques to improve pesticidal efficiency and guarantee safe application of pesticides. In addition to the alternative techniques and products of toxic pesticides, biological pest-controlling techniques, techniques on efficient pesticide spraying, monitoring the residues, decomposing rubbish including pesticide containers, etc., are urgently needed by the farmers. (2) Severe inspection on the production, circulation and use of highly toxic pesticides, including the improvement of the licensing, registration and classification systems of pesticide production, establishing tracing back systems and cracking down on the illegal production and trafficking of highly toxic pesticides. (3) According to the foregoing analysis, the ratio of migrant income is negatively correlated with the amount of pesticides; total income is negatively correlated with the use of toxic pesticides. Therefore, continuing transfer of surplus labour from agriculture to the other sectors is still necessary, which can simultaneously improve the total income of rural households. The main tasks include promoting vocational training, perfecting the employment information networks and protecting the legal rights of migrant workers. (4) This survey reveals that behaviours like littering from the incorrect disposal of containers and spraying pesticides by private practice still exist amongst many farmers, and their perceptions on the safe application of pesticides need to be improved. Hence, education on the scientific application of pesticides, which is limited in traditional education, needs to be strengthened [5].

#### **References**


[5] Wei Q., Tao L.,Song X. Pesticide safety management in China: opportunity and deȬ velopment strategy, Chinese Journal of Quality and Safety Supervision 2011; 1: 11-14.

negatively correlated with the amounts of pesticides applied. With respect to the application probability of toxic pesticides, the income ratio of migrant job is positively correlated, while the other three variables of *metro*, *income* and *grainr* embrace negative effects. Farms' location, whether north or affiliating to the metropolises, is measured as positively correlated, while ratio of migrant income is negatively correlated with the odds of adopting biological pest

72 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

(1) Extending advanced techniques to improve pesticidal efficiency and guarantee safe application of pesticides. In addition to the alternative techniques and products of toxic pesticides, biological pest-controlling techniques, techniques on efficient pesticide spraying, monitoring the residues, decomposing rubbish including pesticide containers, etc., are urgently needed by the farmers. (2) Severe inspection on the production, circulation and use of highly toxic pesticides, including the improvement of the licensing, registration and classification systems of pesticide production, establishing tracing back systems and cracking down on the illegal production and trafficking of highly toxic pesticides. (3) According to the foregoing analysis, the ratio of migrant income is negatively correlated with the amount of pesticides; total income is negatively correlated with the use of toxic pesticides. Therefore, continuing transfer of surplus labour from agriculture to the other sectors is still necessary, which can simultaneously improve the total income of rural households. The main tasks include promoting vocational training, perfecting the employment information networks and protecting the legal rights of migrant workers. (4) This survey reveals that behaviours like littering from the incorrect disposal of containers and spraying pesticides by private practice still exist amongst many farmers, and their perceptions on the safe application of pesticides need to be improved. Hence, education on the scientific application of pesticides, which is

[1] CNSB (China National Statistic Bureau) The production and growth rate of main inȬ

[2] World Bank World Development Indicators Database (2011). http://data.worldȬ

[3] Zhang F., Wang J., Zhang W. Nutrient use efficiencies of major cereal crops in China and measures for improvement. Chinese Journal of Soil Sciences 2008; 45(9): 915-924.

[4] Zhang W., Xu A., Ji H. Estimation of agricultural non-point source pollution in China and the alleviating strategies III: a review of policies and practices for agricultural non-point source pollution control in China, Chinese Journal of Agricultural Sciences

limited in traditional education, needs to be strengthened [5].

dustrial products (2011). http://www.stats.gov.cn/.

controls.

**References**

bank.org/data-catalog.

2004; 37(7): 1026-1033.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

*4.4.4.2 Policy recommendations*



[32] Li H., Fu X., Wu X. The wishes and influencing factors of farmers' safe use of pestiȬ cides: survey and analysis on 214 farmers of Guanghan Prefecture, Sichuan Province, China, Chinese Journal of Agricultural and Technological Economy 2007; 5: 99-104.

[18] Zhu Y., Wu L. Comparison of pesticide application behaviors amongst farmers in different acreages, Chinese Journal of Zhejiang Agricultural Sciences 2010; 5:

74 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

[19] Zhang Y., Ma J., Kong X., Zhu Y. Determinants on Farmers' use of green pesticides: empirical analysis on 15 counties of the Shanxi, Shaanxi and Shandong Provinces,

[20] Han W. Survey and analysis on environmental problems in the poverty rural areas of

[21] Zhang J., Li R. Rural environmental pollution and the countermeasures for sustainaȬ ble development, Chinese Journal of Anhui Agricultural Sciences 2007; 35(15):

[22] Jack J., John D. Econometric Methods (Fourth Edition). The McGraw-Hill Companies,

[23] Seddighi H. R., Lawler K. A., Katos A. V. Econometric: A Practical Approach. RoutȬ

[24] Bruin J. Newtest: Command To Compute New Test. UCLA: Academic Technology Services, Statistical Consulting Group 2006. http://www.ats.ucla.edu/stat/stata/ado/

[25] Zheng X. Analysis of the influencing factors on the farmers' use of manures in DanȬ jiangkou reservoir area, Chinese Journal of Hunan Agricultural University (Social

[26] Yin C., Wu P., Zhang Y. Research on farmers' will to reduce the amount of crop ferȬ

[27] Sun Q. Analysis on the determinants of production of safety agricultural products by

[28] Zhang W. Analysis on the market conditions of Folimat and some other highly toxic pesticides in China, Chinese Journal of Pesticides Marketing Bulletin 2008; 20: 27.

[29] Tao K., Zhou H. Furadan, folimat and some other pesticides are prohibited since toȬ

[30] SCSC (Thirteenth Standing Committee of Suzhou People's Congress). Bulletin of suȬ pervision and management regulations on food safety in Suzhou Prefecture 2007.

[31] Zhou W., Song X. Safeguarding the shopping baskets of the Capital through strengthened supervision and innovative services to: Interview with Zhang Lingjun, director of Beijing Pesticide Inspection Institute, Chinese Journal of Pesticide Science

tilizer. Chinese Journal of Jiangsu Agricultural Sciences 2010; 1: 384-387.

farms, Chinese Journal of Food and Nutrition in China 2008; 1: 15-17.

http://www.szfzb.gov.cn/news/fzb/2007/1/25/fzb-15-12-49-3707.shtml.

morrow, Chinese Newspaper of Jiaxing Daily 2005; 6(30): 2.

and Administration 2009; 30(12): 24-26.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Sichuan Province, Chinese Journal of Rural Economics 2005; 11: 99-101.

Chinese Journal of Rural Economy 2004; 1: 44-52.

1024-1029.

4588-4617.

analysis/.

Inc., New York 1997: 415-418.

Sciences) 2010; 11: 11-15.

ledge Press, London 2000: 105-106.


#### **Chapter 5**

### **Farmer Perceptions on the Safety of Agricultural Products**

Xiaoou Gao and Min Song

### **5.1 Introduction**

With the accelerating process of global integration and the increasingly fierce competition in the international market, the safety of agricultural products, as the most basic material for human survival and development, has come into global focus. China is a large agricultural country, in which the output and the export of agricultural products have constantly increased over recent years. In 2011, the agricultural output in China was 737.113 billion US dollars, accounting for 17.3% of global agricultural output, which is far ahead of other countries in the world. Since 2008, China has become the fifth largest exporter of agricultural products [1]. Statistics show the import and export of agricultural products amounted to 154.03 billion US dollars in 2011 and the export volume was 60.13 billion US dollars. However, China is facing greater challenges due to the outbreak of a series of agro-food safety scandals in recent years.

With the development of industry and the economy, promoted by incentive polices, chemical products have been increasingly put into agriculture in China over recent decades. China has become the largest user of fertilizers, pesticides and plastic film in the world [3]. In 2010, Chinese agriculture consumed 55.61 million tons of chemical fertilizer, 1.75 million tons of pesticide and 2.17 million tons of plastic film [4-5] - all far higher than the world average. Nearly 60-70% of the chemical fertilizer, 60% of the pesticide and 40% of the plastic film are exposed to the environment directly [6-7], posing ultimate threats to the safety of agroproducts, food and human health. Recently, multiple agro-food safety incidents occurred, e.g., the Qingdao drug leak incident of 2010, the Hainan drug cowpea incident of 2011 and the Guangdong drug watermelon incident of 2007. Disclosure of these agro-food safety scandals has led to a series of severe consequences. First of all, considerable outbreaks of food borne disease have fundamentally undermined public trust. According to the Report on Chinese Food Safety in 2011-2012 released by the Chinese magazine of *Well-off* and Tsinghua UniverȬ sity, it was reported that 63.7% of the respondents believed that food safety in China is bad and that 80.4% of the respondents thought food is not safe at all in China. Secondly, food safety

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Gao and Song; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Gao and Song; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

**Chapter 5 — Farmer Perceptions on the Safety of Agricultural Products** 

With the accelerating process of global integration and the increasingly fierce competition in the international market, the safety of agricultural products, as the most basic material for human survival and development, has come into global focus. China is a large agricultural country, in which the output and the export of agricultural products have constantly increased over recent years. In 2011, the agricultural output in China was 737.113 billion US dollars, accounting for 17.3% of global agricultural output, which is far ahead of other countries in the world. Since 2008, China has become the fifth largest exporter of agricultural products [1].

was 60.13 billion US dollars. However, China is facing greater challenges due to the outbreak of a series of agro-food safety

China. A survey carried out in Korea shows that nearly 90% of respondents think food imported from China is not safe [8]. On the

Figure 1. Figure 5.1 Volume of exported agro-products in China Source: China Ministry of Agriculture [2]

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Xiaoou Gao and Min Song

**5.1. Introduction** 

scandals in recent years.

scandals have serious negative impacts on food exports. On the one hand, they undermine the international competitiveness and market reputation of food exported from China. Massive exposure of food safety incidents in China greatly reduces foreign consumers' confidence in food labelled as made in China. A survey carried out in Korea shows that nearly 90% of respondents think food imported from China is not safe [8]. On the other hand, the internaȬ tional market will raise barriers to limit food imported from China. For example, a number of countries applied stricter standards on products made in China after the disclosure of the poisoned capsule event in early 2012. Thirdly, food safety accidents often result in significant economic losses. Take the Sanlu incident as an example, it led to the bankruptcy of a famous enterprise and brand, simultaneously causing the entire milk industry to suffer. During the incident, a vast amount of milk was poured away, and a great number of dairy cows were slaughtered. With the development of industry and the economy, promoted by incentive polices, chemical products have been increasingly put into agriculture in China over recent decades. China has become the largest user of fertilizers, pesticides and plastic film in the world [3]. In 2010, Chinese agriculture consumed 55.61 million tons of chemical fertilizer, 1.75 million tons of pesticide and 2.17 million tons of plastic film [4-5] - all far higher than the world average. Nearly 60-70% of the chemical fertilizer, 60% of the pesticide and 40% of the plastic film are exposed to the environment directly [6-7], posing ultimate threats to the safety of agro-products, food and human health. Recently, multiple agro-food safety incidents occurred, e.g., the Qingdao drug leak incident of 2010, the Hainan drug cowpea incident of 2011 and the Guangdong drug watermelon incident of 2007. Disclosure of these agro-food safety scandals has led to a series of severe consequences. First of all, considerable outbreaks of food borne disease have fundamentally undermined public trust. According to the Report on Chinese Food Safety in 2011-2012 released by the Chinese magazine of *Welloff* and Tsinghua University, it was reported that 63.7% of the respondents believed that food safety in China is bad and that 80.4% of the respondents thought food is not safe at all in China. Secondly, food safety scandals have serious negative impacts on food exports. On the one hand, they undermine the international competitiveness and market reputation of food exported from China. Massive exposure of food safety incidents in China greatly reduces foreign consumers' confidence in food labelled as made in

In the previous literature, more and more economists have studied food safety issues using empirical analysis methods, especially on the topics of consumer behaviours, awareness and willingness to pay for certified safe food. For example, Chern, W. S., et al (2002) [9] studied consumers' willingness to pay for genetically modified vegetable oils through carrying out a survey in Shikoku of Japan, Norway, Taiwan and the United States. Georges G, et al (2006) [10] selected interviewed consumers from 12 European countries, divided them into four groups, and managed to understand their perception on the food traceability system. Wang (2003) [11] carried out a survey of 289 consumers in Tianjin, China, and analysed their decision process and characteristics when choosing safe food. Through a survey in Zhejiang Province, China, Zhou (2004) [12] revealed that consumers are concerned about vegetable safety - illustrating their negative attitudes about the current situation. They are very willing to pay the extra cost for certified safe vegetables, but the price of the safe vegetable should be no more 10% to 20% higher than the ordinary vegetable. Zeng et al (2008) [13] studied consumers' willingness to pay for moon cake with safe additives, through a survey of 396 consumers sampled from 25 other hand, the international market will raise barriers to limit food imported from China. For example, a number of countries applied stricter standards on products made in China after the disclosure of the poisoned capsule event in early 2012. Thirdly, food safety accidents often result in significant economic losses. Take the Sanlu incident as an example, it led to the bankruptcy of a famous enterprise and brand, simultaneously causing the entire milk industry to suffer. During the incident, a vast amount of milk was poured away, and a great number of dairy cows were slaughtered. In the previous literature, more and more economists have studied food safety issues using empirical analysis methods, especially on the topics of consumer behaviours, awareness and willingness to pay for certified safe food. For example, Chern, W. S., et al (2002) [9] studied consumers' willingness to pay for genetically modified vegetable oils through carrying out a survey in Shikoku of Japan, Norway, Taiwan and the United States. Georges G, et al (2006) [10] selected interviewed consumers from 12 European countries, divided them into four groups, and managed to understand their perception on the food traceability system. Wang (2003) [11] carried out a survey of 289 consumers in Tianjin, China, and analysed their decision process and characteristics when choosing safe food. Through a survey in Zhejiang Province, China, Zhou (2004) [12] revealed that consumers are concerned about supermarkets in Beijing. But at least until now, studies on food safety from the farmers' perspective are extremely rare.

The objective of this chapter is to understand the situation of agro-food in China through analysing farmers' confidence and determinants on their own agro-products safety. Different from consumers, farmers are both managers and producers of agro-products, and they possess much information about agricultural production - from sowing, fertilizing and controlling pests to harvesting. Therefore, their confidence on the safety of their agro-products provides a new perspective to studying the real quality of agro-products and ensuring food safety.

The remaining sections will be organized as follows: Section 2 represents the field survey and characteristics of the sampled farmers, including their demographic information, farming condition and knowledge of agro-products' quality. Section 3 describes farmers' confidence on their own agro-products and analyses impacting factors by using the binary logit regression model. In Section 4, some conclusions and recommendations are proposed.

#### With the development of industry and the economy, promoted by incentive polices, chemical products have been increasingly put **5.2 Data source and demographic characteristics**

#### into agriculture in China over recent decades. China has become the largest user of fertilizers, pesticides and plastic film in the world [3]. In 2010, Chinese agriculture consumed 55.61 million tons of chemical fertilizer, 1.75 million tons of pesticide and 2.17 **5.2.1 Data source**

scandals have serious negative impacts on food exports. On the one hand, they undermine the international competitiveness and market reputation of food exported from China. Massive exposure of food safety incidents in China greatly reduces foreign consumers' confidence in food labelled as made in China. A survey carried out in Korea shows that nearly 90% of respondents think food imported from China is not safe [8]. On the other hand, the internaȬ tional market will raise barriers to limit food imported from China. For example, a number of countries applied stricter standards on products made in China after the disclosure of the poisoned capsule event in early 2012. Thirdly, food safety accidents often result in significant economic losses. Take the Sanlu incident as an example, it led to the bankruptcy of a famous enterprise and brand, simultaneously causing the entire milk industry to suffer. During the incident, a vast amount of milk was poured away, and a great number of dairy cows were

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

78 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Figure 1. Figure 5.1 Volume of exported agro-products in China

Source: China Ministry of Agriculture [2]

**Figure 5.1** Volume of exported agro-products in China

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Chapter 5 — Farmer Perceptions on the Safety of Agricultural Products** 

With the accelerating process of global integration and the increasingly fierce competition in the international market, the safety of agricultural products, as the most basic material for human survival and development, has come into global focus. China is a large agricultural country, in which the output and the export of agricultural products have constantly increased over recent years. In 2011, the agricultural output in China was 737.113 billion US dollars, accounting for 17.3% of global agricultural output, which is far ahead of other countries in the world. Since 2008, China has become the fifth largest exporter of agricultural products [1]. Statistics show the import and export of agricultural products amounted to 154.03 billion US dollars in 2011 and the export volume was 60.13 billion US dollars. However, China is facing greater challenges due to the outbreak of a series of agro-food safety

exports. On the one hand, they undermine the international competitiveness and market reputation of food exported from China.

Xiaoou Gao and Min Song

**5.1. Introduction** 

scandals in recent years.

unit: billion of us dallors

In the previous literature, more and more economists have studied food safety issues using empirical analysis methods, especially on the topics of consumer behaviours, awareness and willingness to pay for certified safe food. For example, Chern, W. S., et al (2002) [9] studied consumers' willingness to pay for genetically modified vegetable oils through carrying out a survey in Shikoku of Japan, Norway, Taiwan and the United States. Georges G, et al (2006) [10] selected interviewed consumers from 12 European countries, divided them into four groups, and managed to understand their perception on the food traceability system. Wang (2003) [11] carried out a survey of 289 consumers in Tianjin, China, and analysed their decision process and characteristics when choosing safe food. Through a survey in Zhejiang Province, China, Zhou (2004) [12] revealed that consumers are concerned about vegetable safety - illustrating their negative attitudes about the current situation. They are very willing to pay the extra cost for certified safe vegetables, but the price of the safe vegetable should be no more 10% to 20% higher than the ordinary vegetable. Zeng et al (2008) [13] studied consumers' willingness to pay for moon cake with safe additives, through a survey of 396 consumers sampled from 25

slaughtered.

Source: China Ministry of Agriculture [2]

million tons of plastic film [4-5] - all far higher than the world average. Nearly 60-70% of the chemical fertilizer, 60% of the pesticide and 40% of the plastic film are exposed to the environment directly [6-7], posing ultimate threats to the safety of agro-products, food and human health. Recently, multiple agro-food safety incidents occurred, e.g., the Qingdao drug leak incident of 2010, the Hainan drug cowpea incident of 2011 and the Guangdong drug watermelon incident of 2007. Disclosure of these agro-food safety scandals has led to a series of severe consequences. First of all, considerable outbreaks of food borne disease have fundamentally undermined public trust. According to the Report on Chinese Food Safety in 2011-2012 released by the Chinese magazine of *Welloff* and Tsinghua University, it was reported that 63.7% of the respondents believed that food safety in China is bad and that 80.4% of the respondents thought food is not safe at all in China. Secondly, food safety scandals have serious negative impacts on food As described in the Chapter 3, the data in this chapter is obtained through a field survey of 560 samples from 21 villages carried out in January to March 2011, which covered six provinces of eastern China (Figure 3.1). In addition to the number of sampled farmers among the six provinces, their distribution between the north and south, the two metropolises and other areas are also included. In this research, the main research objective is to analyse farmers' confidence and determinants on their own agro-products safety; there are 346 respondents available to apply to this research, therefore the sample size for the current research is 346 (Table 5.1).


on the topics of consumer behaviours, awareness and willingness to pay for certified safe food. For example, Chern, W. S., et al (2002) [9] studied consumers' willingness to pay for genetically modified vegetable oils through carrying out a survey in Shikoku **Table 5.1** Distribution of sampled farmers

#### of Japan, Norway, Taiwan and the United States. Georges G, et al (2006) [10] selected interviewed consumers from 12 European countries, divided them into four groups, and managed to understand their perception on the food traceability system. Wang **5.2.2 Demographic characteristics**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

(2003) [11] carried out a survey of 289 consumers in Tianjin, China, and analysed their decision process and characteristics when choosing safe food. Through a survey in Zhejiang Province, China, Zhou (2004) [12] revealed that consumers are concerned about The results of this survey show that farmers' average age is 48.92 years. Most of them are middle-aged and aged, a half of them are older than 45 years old and younger than 60 years old, while 17.34% of them are older than 60 years old. The number of male farmers is much greater than that of female, accounting for 84.10% in total. Farmers' general education level is low, as 80.92% of them are educated to no higher than high school. There are 98 farmers who participated in off farming works in 2010, accounting for 28.32 in total. From the perspective of household income, 16.23% of the farmers' household income in 2010 is less than 10,000 yuan, 11.30% of the farmers' household income is more than 50,000 yuan, and 72.47% of the farmers' household income is between 10000 yuan and 50000 yuan (Table 5.2).


Source: field survey by the authors in 2011

**Table 5.2** Demographic characteristics of farmers in 2010

#### **5.2.3 Farming condition**

As shown in Table 5.3, sowing scales of grain crops are quite different among sample farmers, where the maximum sowing area is 112 mu, much more than the minimum sowing area of 0.2 mu. More than 80% of farmers sell their agro-products. In agricultural production, a great number of farmers used chemical fertilizer by spreading it on the surface of the land, while 35.55% of them used manure as an auxiliary fertilizer. Some farmers still used prohibited highly toxic pesticides, such as dichlorvos (DDVP). Meanwhile, very few farmers responded as having adopted biological methods to control pests.


Source: field survey by the authors in 2011

**Table 5.3** Farming condition of the sampled farmers

concerned about yield, 17.58% of farmers are most concerned about cost, and 7.49% of farmers are most concerned about the

As shown in Table 5.3, sowing scales of grain crops are quite different among sample farmers, where the maximum sowing area is 112 mu, much more than the minimum sowing area of 0.2 mu. More than 80% of farmers sell their agro-products. In agricultural production, a great number of farmers used chemical fertilizer by spreading it on the surface of the land, while 35.55% of them used manure as an auxiliary fertilizer. Some farmers still used prohibited highly toxic pesticides, such as dichlorvos (DDVP).

#### **5.2.4 Farmers' perception on the safety of agro-products** Sowing area 346 mu 5.20 0.2 112 7.40 Selling ratio 346 % 61.78 0 100 133.89

**5.2.3. Farming condition** 

low, as 80.92% of them are educated to no higher than high school. There are 98 farmers who participated in off farming works in 2010, accounting for 28.32 in total. From the perspective of household income, 16.23% of the farmers' household income in 2010 is less than 10,000 yuan, 11.30% of the farmers' household income is more than 50,000 yuan, and 72.47% of the farmers'

80 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Characteristics Count Unit Mean Min Max Std. D**

Education 346 dummy Illiteracy=1 (4.05%); Primary=2 (28.03%); Middle=3 (48.84%); High=4 (14.74%); Advanced=5 (4.34%)

As shown in Table 5.3, sowing scales of grain crops are quite different among sample farmers, where the maximum sowing area is 112 mu, much more than the minimum sowing area of 0.2 mu. More than 80% of farmers sell their agro-products. In agricultural production, a great number of farmers used chemical fertilizer by spreading it on the surface of the land, while 35.55% of them used manure as an auxiliary fertilizer. Some farmers still used prohibited highly toxic pesticides, such as dichlorvos (DDVP). Meanwhile, very few farmers responded

**Characteristics Count Unit Mean Min Max Std. D** Sowing area 346 mu 5.20 0.2 112 7.40 Selling ratio 346 % 61.78 0 100 133.89 Manure using 346 dummy Using manure = 1 (35.55%); Not using manure = 0 (64.45%)

Less than 10,000 yuan=1 (16.23%), 10,000-30,000 yuan=2 (39.13%), 30,000-50,000 yuan=3 (33.33%), more than 50,000 yuan=4 (11.30%)

Off farming work 346 dummy Do off farming work=1 (28.32%), Not do off farming work=0 (71.68%)

**Age 346 years 48.92 21 85 11.26**

household income is between 10000 yuan and 50000 yuan (Table 5.2).

Gender 346 dummy Male=1 (84.10%); Female=0 (15.90%)

Income 346 dummy Annual income of a household

Source: field survey by the authors in 2011

Source: field survey by the authors in 2011

**Table 5.3** Farming condition of the sampled farmers

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**5.2.3 Farming condition**

**Table 5.2** Demographic characteristics of farmers in 2010

as having adopted biological methods to control pests.

In the farming behaviours section of our questionnaire, there are two questions about agroproduct safety. The first one deals with what is the prior determinant for farmers when using fertilizers and pesticides. The result shows that 71.47% of farmers are most concerned about yield, 17.58% of farmers are most concerned about cost, and 7.49% of farmers are most concerned about the quality of agricultural products. The second question deals with what farmers think are the consequences of overdosing. As a multiple-choice question, farmers choose the answer of imperilling food safety which accounts for an overwhelming ratio of 71.10%, followed by another choice of imperilling sprayers' health at 51.73%. Next is the choice of environment pollution at 50.58% (Table 5.4). Manure using 346 dummy Using manure = 1 (35.55%); Not using manure = 0 (64.45%) Table 3. Table 5.3 Farming condition of the sampled farmers Source: field survey by the authors in 2011 **5.2.4. Farmers' perception on the safety of agro-products**  In the farming behaviours section of our questionnaire, there are two questions about agro-product safety. The first one deals with what is the prior determinant for farmers when using fertilizers and pesticides. The result shows that 71.47% of farmers are most


**Table 5.4** Farmers' perception on agro-product safety Consequence of overdose 346 dummy Food quality =1 (71.10%); others = 0

Table 4. Table 5.4 Farmers' perception on agro-product safety

Source: field survey by the authors in 2011

Figure 2. Figure 5.2 Farmers' information channels Source: field survey by the authors in 2011

**Figure 5.2** Farmers' information channels

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Source: field survey by the authors in 2011

Figure 3. Figure 5.3 Farmers' reliance on different information sources Source: field survey by the authors in 2011

As to the farmers' conception on agricultural pollution, the data shows that TV and radio are their most common ways of obtaining food safety information and related news, followed by government notices, friends and relatives, labels on food packages, **Figure 5.3** Farmers' reliance on different information sources

Source: field survey by the authors in 2011

newspapers and magazines, research reports and others (Figure 5.2). However, they have the mostly reliance on information provided by the government, while TV and radio are ranked in second palace, with labels on food packages ranked last. In another words, most of the farmers have no trust at all in the information disclosed by the food manufacturers (Figure 5.3). **5.3. Analysis on farmers' confidence and determinants 5.3.1. Farmers' confidence on the safety of agro-products**  As to the farmers' conception on agricultural pollution, the data shows that TV and radio are their most common ways of obtaining food safety information and related news, followed by government notices, friends and relatives, labels on food packages, newspapers and magaȬ zines, research reports and others (Figure 5.2). However, they have the mostly reliance on information provided by the government, while TV and radio are ranked in second palace, with labels on food packages ranked last. In another words, most of the farmers have no trust at all in the information disclosed by the food manufacturers (Figure 5.3).

As shown in Figure 5.4, as to farmers' confidence on safety of their own agro-products, 19.65% of them think that they are very

As to the perception on agro-product safety, farmers' confidence is influenced by their demographic characteristics [14], farming condition and knowledge on agro-production safety [15]. In our study, gender, age, education level, the doing off-farming work or not and household income are included as demographic factors. The sowing area of field crop, selling ratio of agro-production and use of manure or not are included as farming condition factors. Answers to the questions about agro-products safety are included

as farming condition factors. In addition, location and city scale are included to examine the significance of geography.

4, 1.16%

#### safe, 63.78% of them think that they are safe, 11.27% of them think that they are not safe, while 1.16% of them think that they are not safe at all, and 4.05% of them responded as being neutral. **5.3 Analysis on farmers' confidence and determinants**

Figure 4. Figure 5.4 Farmers' confidence on agro-product safety

14, 4.05%

very safe safe neutral not safe not safe at all

**5.3.2. Factors affecting farmers' confidence** 

Source: field survey by the authors

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

68, 19.65%

0

50

100

250

**5.3.3. Model specification** 

#### **5.3.1 Farmers' confidence on the safety of agro-products**

221, 63.87% 150 200 As shown in Figure 5.4, as to farmers' confidence on safety of their own agro-products, 19.65% of them think that they are very safe, 63.78% of them think that they are safe, 11.27% of them think that they are not safe, while 1.16% of them think that they are not safe at all, and 4.05% of them responded as being neutral.

39, 11.27%

words, most of the farmers have no trust at all in the information disclosed by the food manufacturers (Figure 5.3).

9.25% 6.65%

reseach report

As to the farmers' conception on agricultural pollution, the data shows that TV and radio are their most common ways of obtaining food safety information and related news, followed by government notices, friends and relatives, labels on food packages, newspapers and magazines, research reports and others (Figure 5.2). However, they have the mostly reliance on information provided by the government, while TV and radio are ranked in second palace, with labels on food packages ranked last. In another

3.18%

Label on food package 0.29%

Others

not safe at all, and 4.05% of them responded as being neutral.

Figure 3. Figure 5.3 Farmers' reliance on different information sources

17.05%

TV and radio Friends,

relatives

4.34%

Newspaper and magazine

**5.3. Analysis on farmers' confidence and determinants** 

**5.3.1. Farmers' confidence on the safety of agro-products** 

Source: field survey by the authors in 2011

59.25%

Government notice

0%

10%

20%

30%

40%

50%

60%

70%

Figure 4. Figure 5.4 Farmers' confidence on agro-product safety Source: field survey by the authors

Source: field survey by the authors

#### **5.3.2 Factors affecting farmers' confidence**

As to the farmers' conception on agricultural pollution, the data shows that TV and radio are their most common ways of obtaining food safety information and related news, followed by government notices, friends and relatives, labels on food packages, newspapers and magazines, research reports and others (Figure 5.2). However, they have the mostly reliance on information **5.3.2. Factors affecting farmers' confidence**  As to the perception on agro-product safety, farmers' confidence is influenced by their demographic characteristics [14], farming condition and knowledge on agro-production safety [15]. In our study, gender, age, education level, the doing off-farming work or not and household income are included as demographic factors. The sowing area of field crop, selling ratio of agro-production and use of manure or not are included as farming condition factors. Answers to the questions about agro-products safety are included As to the perception on agro-product safety, farmers' confidence is influenced by their demographic characteristics [14], farming condition and knowledge on agro-production safety [15]. In our study, gender, age, education level, the doing off-farming work or not and household income are included as demographic factors. The sowing area of field crop, selling ratio of agro-production and use of manure or not are included as farming condition factors. Answers to the questions about agro-products safety are included as farming condition factors. In addition, location and city scale are included to examine the significance of geography.

#### as farming condition factors. In addition, location and city scale are included to examine the significance of geography. **5.3.3 Model specification**

Figure 3. Figure 5.3 Farmers' reliance on different information sources

17.05%

82 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

TV and radio Friends,

relatives

As to the farmers' conception on agricultural pollution, the data shows that TV and radio are their most common ways of obtaining food safety information and related news, followed by government notices, friends and relatives, labels on food packages, newspapers and magaȬ zines, research reports and others (Figure 5.2). However, they have the mostly reliance on information provided by the government, while TV and radio are ranked in second palace, with labels on food packages ranked last. In another words, most of the farmers have no trust

4.34%

Newspaper and magazine

**5.3. Analysis on farmers' confidence and determinants** 

**5.3.1. Farmers' confidence on the safety of agro-products** 

at all in the information disclosed by the food manufacturers (Figure 5.3).

not safe at all, and 4.05% of them responded as being neutral.

**5.3 Analysis on farmers' confidence and determinants**

221, 63.87%

**5.3.1 Farmers' confidence on the safety of agro-products**

Figure 4. Figure 5.4 Farmers' confidence on agro-product safety

14, 4.05%

very safe safe neutral not safe not safe at all

39, 11.27%

As shown in Figure 5.4, as to farmers' confidence on safety of their own agro-products, 19.65% of them think that they are very safe, 63.78% of them think that they are safe, 11.27% of them think that they are not safe, while 1.16% of them think that they are not safe at all, and 4.05%

**5.3.2. Factors affecting farmers' confidence** 

Source: field survey by the authors

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

68, 19.65%

of them responded as being neutral.

0

50

100

150

200

250

**5.3.3. Model specification** 

provided by the government, while TV and radio are ranked in second palace, with labels on food packages ranked last. In another

3.18%

Label on food package 0.29%

Others

As shown in Figure 5.4, as to farmers' confidence on safety of their own agro-products, 19.65% of them think that they are very

As to the perception on agro-product safety, farmers' confidence is influenced by their demographic characteristics [14], farming condition and knowledge on agro-production safety [15]. In our study, gender, age, education level, the doing off-farming work or not and household income are included as demographic factors. The sowing area of field crop, selling ratio of agro-production and use of manure or not are included as farming condition factors. Answers to the questions about agro-products safety are included

as farming condition factors. In addition, location and city scale are included to examine the significance of geography.

4, 1.16%

words, most of the farmers have no trust at all in the information disclosed by the food manufacturers (Figure 5.3).

9.25% 6.65%

reseach report

Source: field survey by the authors in 2011

**Figure 5.3** Farmers' reliance on different information sources

59.25%

Government notice

0% 10% 20% 30% 40% 50% 60% 70%

Source: field survey by the authors in 2011

**5.3.3. Model specification**  A binary logit regression model is adopted to examine factors impacting on farmers' confiȬ dence, which is shown as:

$$\mathbf{y} = \boldsymbol{\beta}\_0 + \boldsymbol{\beta}\_1 \mathbf{x}\_1 + \boldsymbol{\beta}\_2 \mathbf{x}\_2 + \dots + \boldsymbol{\beta}\_k \mathbf{x}\_k + \boldsymbol{u} \tag{5-1}$$

safe, 63.78% of them think that they are safe, 11.27% of them think that they are not safe, while 1.16% of them think that they are while the distribution function of y is:

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

$$f\left(y\right) = P^y \left(1 - P\right)^{1 - y}, \ \left(y = 0, \ 1\right) \tag{5-2}$$

where when farmers are very confident or rather confident, *y*=1; otherwise, y=0; *X* (*x1*, *x2*, *…*, *xk*) represents a series variables affecting farmers' confidence on agricultural production safety; *u* is the random error. The model is calculated by using SPSS 18.0 for Windows.


Note: \*\*\*, \*\*and \*represent statistical significance in the level of 1%, 5% and 10%, respectively.

Software: SPSS 18.0

**Table 5.5** Effects of characteristics on farmers' confidence

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **5.3.4 Result and discussion**

As shown in Table 5.5, only using manure or not, location and sowing area are significant, while other factors are measured as being insignificant upon farmers' confidence on the safety of their own agro-products.

As manure is considered less harmful to agricultural environments and products, farmers using manure are more confident about their own products. In the north there are 86 farmers who use manure, accounting for 40.19% of northern farmers. Meanwhile, there are 37 farmers using it in the south, 28.03% of the sampled farmers from the south. At the same time, the data suggests that southern farmers focus more on the safety of agro-products, and they have more knowledge and better understanding of the harmful effects of agricultural chemicals. ThereȬ fore, southern farmers' attitudes on agro-product safety are more negative. Furthermore, farmers who own larger scales of land are likely to be professional. They not only have more knowledge on agro-product safety, but also tend to pursue more yields through applying more agro-chemicals. As a result, they lack confident on the safety of their own agro-products.

#### **5.4 Conclusion and recommendation**

#### **5.4.1 Major conclusions**

**Characteristics B S.E. Wald df Sig. odds ratio**

84 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Location \*\* 0.989 0.440 5.056 1 0.025 2.688

City 0.218 0.423 0.267 1 0.606 1.244

Gender -0.137 0.460 0.089 1 0.765 0.872

Age -0.019 0.016 1.344 1 0.246 0.981

Education level -0.240 0.220 1.187 1 0.276 0.787

Off-farming work -0.315 0.351 0.803 1 0.370 0.730

Income -0.205 0.223 0.839 1 0.360 0.815

Sowing area\* -0.054 0.033 2.800 1 0.094 0.947

Selling ratio 0.007 0.005 2.197 1 0.138 1.007

Manure using\*\*\* 1.062 0.408 6.773 1 0.009 2.892

Determinants of using fertilizer and pesticide 0.016 0.050 0.096 1 0.756 1.016

Consequence of overdose 0.447 0.354 1.588 1 0.208 1.563

As shown in Table 5.5, only using manure or not, location and sowing area are significant, while other factors are measured as being insignificant upon farmers' confidence on the safety

As manure is considered less harmful to agricultural environments and products, farmers using manure are more confident about their own products. In the north there are 86 farmers who use manure, accounting for 40.19% of northern farmers. Meanwhile, there are 37 farmers using it in the south, 28.03% of the sampled farmers from the south. At the same time, the data suggests that southern farmers focus more on the safety of agro-products, and they have more knowledge and better understanding of the harmful effects of agricultural chemicals. ThereȬ fore, southern farmers' attitudes on agro-product safety are more negative. Furthermore, farmers who own larger scales of land are likely to be professional. They not only have more

Omnibus tests of model coefficients Chi-square (12) = 25.948, Sig. =0.010\*\*\*

\*\*and \*represent statistical significance in the level of 1%, 5% and 10%, respectively.

Note: \*\*\*,

Software: SPSS 18.0

**5.3.4 Result and discussion**

of their own agro-products.

**Table 5.5** Effects of characteristics on farmers' confidence

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Farmers do not have sufficient awareness of agro-product safety during production. When using chemical fertilizers and pesticides, only 7.51% of the farmers considered the negative impacts on the safety of agro-products, while most of the farmers preferred to concentrate on increasing the yields.

Farmers' knowledge of agro-product safety is not very comprehensive. On the one hand, influenced by the publicity surrounding pesticide residues incidents, 71.10% of them think overdosing will influence agro-products' safety. However, 58.38% of farmers understand well or better understand the withdrawal period. On the other hand, farmers using manure are very confident about the safety of their agro-products. In fact, the safety of manure is not perfect, despite the fact that it is less harmful to agricultural environments and products than agro-chemicals.

Farmers who sow more farmland are less confident on their products, which may be because the farmers who own more land are more likely to be professional farmers. They tend to pursue productivity through inputting more agro-chemicals.

#### **5.4.2 Policy recommendations**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

To improve farmers' awareness on agro-product safety, more related information needs to be released by the mass media including TV, radio, etc. As revealed by the analyses in this chapter, the public media are the most commonly used channels by which the farmers gather inforȬ mation and the recommendations above must be effective in helping them to improve their behaviours.

This survey reveals many farmers have poor knowledge about the withdrawal period and outcome of overusing agro-chemicals. Hence, concerning public education on the withdrawal period and proper use of chemical fertilizers and pesticides, it needs to be strengthened.

Strengthening the supervision of farmers, especially farmers who own large-scale land, is also necessary. This is because, as professional farmers, they are concerned with agro-products yields rather than quality. Therefore, related government departments should strengthen their supervision of farming behaviours.

There are some limitations of this study, e.g., in the model, amounts of fertilizers and pesticides should be included as variables and location may not be significant in influencing farmers' confidence on agro-product safety. However, the data collected is not complete. In future studies, more scientific variables should be included, with the adoption of further optimized models.

#### **References**


### **Farmer Perceptions on Risk Sources and Management**

Hui Zhou and Teruaki Nanseki

#### **6.1 Introduction**

**References**

Education Press, 2012.5.

2004; (11):44-52.

of China 2006; (11):25-34.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

[1] Wu, T. China's agricultural export doubled in a decade after accession to the WTO (2011). http://www.cfqn.com.cn/Article/2011/2093q/2093a/19090786589501.htm. [2] China Ministry of Agriculture. China Statistical Yearbook 2012. Shenyang: Liaoning

[3] Liu G P, et al. Current situation and measures of agricultural pollution in China.

[4] CNSB (China National Statistical Bureau). Using of agricultural plastic film and pesȬ

[5] CNSB (China National Statistical Bureau).Using of chemical fertilizers in different reȬ

[6] Sun J L. Review of agricultural pollution and preventive technology in China. JourȬ

[7] Hou B, Hou J, Wang Z W. Farmers' perception on pesticide residue and its influence on pesticide application. Heilongjiang Agricultural Sciences 2010; (2):99-103.

[8] The survey shows that nearly 90% of Koreans distrust China's food safety http://

[9] Chern, W.S., Rickertsen, K., Tsuboi, N., Fu T. Consumer acceptance and willingness to pay for genetically modified vegetable oil and salmon: A multiple-country assessȬ

[10] Georges G, Rafia H. Consumers' Perception on Food Traceability in Europe. InternaȬ tional Food & Agribusiness Management Association World Food & Agribusiness

[11] Wang Z. Perception on food safety and consumption decides: empirical analysis of

[12] Zhou J. Analysis on consumers' attitude, perception and purchasing behaviors on vegetables safety: basis on a survey in Zhejiang Province. Chinese Rural Economy

[13] Zeng Y., Liu Y., Wang X. Analysis on willingness to pay for food safety with HierȬ archical model: in the case of consumers' willingness to pay for moon cake additives.

[14] Feng Z., Li Q. Analysis on farmers' perception and determinants of agro-products

[15] Zhou J. Analysis on farmers' behaviors and determinants of vegetable quality conȬ trolling: evidence from 369 vegetable farmers in Zhejiang Province, Rural Economy

individual consumers in Tianjin. Chinese Rural Economy 2003; 4:41-48.

ment. AgBioForum, 2002; 5(3): 105-112. From: http://www.agbioforum.org.

Studies in International Technology and Economy, 2006; 9(4):17-21.

86 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

nal of Jishou University (Natural Science Edition) 2008; 29(5):99-128.

ticide in different regions in China, http://www.stats.gov.cn/.

gions in China, http://www.stats.gov.cn/.

gb.cri.cn/27824/2012/08/16/6071s3812468.htm.

Symposium, Buenos Aires, Argentina, 2006; (7):1-11.

Journal of Agro-technical Economics 2008; 1:84-90.

safety. Agricultural Economic Problems 2007; (11):22-26.

Agricultural management comprises several parts, including risk management, informaȬ tion management, technology management, etc. Understanding and estimating potential risk quickly and properly is very important. This is a first step in risk management. Additionally, collecting, processing, providing and analysing information effectively is also necessary in risk management as well as in agricultural management. This is a first step in information management. In information management, new technologies such as information and communication technology (ICT) have increasingly been used on manageȬ ment strategies (Figure 6.1).

In this chapter, farmers' perceptions of risk and their points of view on risk management strategies are studied. China is the third largest milk producer in the world, right after the USA and India [2]. As a way to increase farmers' income and improve people's diet, the Chinese government encourages farmers to raise more cows and produce more milk. However, risks in both food and agriculture including food contamination by chemicals, animal diseases and price variability in China lead to threats, both to dairy farmers and consumers. Dairy farmers are most negatively affected by these risks, as they belong to the weakest part of the whole milk supply chain. It is important for policy makers to understand farmers' perceptions and their responses to the risk in order to help dairy farmers reduce their losses. In that, dairy farmers are used as an example, and dairy farmers' risks and risk management strategies are mainly examined in this research.

Risk is uncertainty that affects an individual's welfare, and is often associated with adversity and loss [3]. In response to risky situations, farmers should be involved in risk management, making choices among alternatives so as to reduce the effects of the risks. The main research objectives are: to examine the dairy farmers' perception of risk and to examine the risk management strategies of dairy farmers.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Zhou and Nanseki; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Zhou and Nanseki; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Source: Nanseki (2011) [1]

**Figure 6.1** Full picture of management in agriculture

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **6.2 Literature review**

Agriculture is inherently risky. Farm outputs depend on weather and biological processes over which producers have little control, and competition in domestic as well as international markets exposes agricultural producers to unanticipated price fluctuations [4]. However, some of the risks in the livestock sector are different from those in the crop sector. This chapter mainly focuses on dairy farmers' risk perception and risk management strategies.

There are several studies about farmers' risk perception. The Economic Research Service (ERS) has summarized American studies [3]. In the US, dairy farmers were most concerned about commodity price risk, production risk and changes in government laws and regulations. Dairy farmers in Arizona perceived the costs of operating inputs to be the greatest source of risk [5]. Dairy farmers in New Zealand viewed price risk and rainfall variability as the highest risks [6]. Meuwissen et al (2001) [7] found that Dutch livestock farmers considered price and production risks to be most important. In Japan, the biggest and the sometime occurring risks to livestock farmers are decrease of production by animal death and decrease of quality by equipment breakdown [1].

ERS also found that keeping cash at hand was the chief risk management strategy for every farm size, for every commodity specialty, and in every region studied; use of derivative and insurance markets was also considered important [3]. Maintaining animal health was viewed as the most effective strategy. The research in Holland found that, producing at the lowest possible costs and insurance were the most important risk management strategies [7]. A study among Finnish farmers found changes in agricultural policy to be the most important risk factor, while maintaining adequate liquidity and solidity was the most important management response [8].

In China, there are several risks that the whole dairy industry faces. The food safety problem is considered as one of the biggest risk for the dairy industry [9-10]. Lack of a complete law and regulation system, having a breed of cow which is not pure and low yield are considered important risks for the whole industry [9]. However, in China, most studies in this field focused on the macro level - from the point of view of the whole industry - and there are very few studies focusing on dairy farmers' perceptions presently. Therefore, this research focuses on uncovering dairy farmers' perceptions of risk and their risk management strategies in China.

#### **6.3 Data and method**

**6.2 Literature review**

**Figure 6.1** Full picture of management in agriculture

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Source: Nanseki (2011) [1]

breakdown [1].

Agriculture is inherently risky. Farm outputs depend on weather and biological processes over which producers have little control, and competition in domestic as well as international markets exposes agricultural producers to unanticipated price fluctuations [4]. However, some of the risks in the livestock sector are different from those in the crop sector. This chapter mainly

There are several studies about farmers' risk perception. The Economic Research Service (ERS) has summarized American studies [3]. In the US, dairy farmers were most concerned about commodity price risk, production risk and changes in government laws and regulations. Dairy farmers in Arizona perceived the costs of operating inputs to be the greatest source of risk [5]. Dairy farmers in New Zealand viewed price risk and rainfall variability as the highest risks [6]. Meuwissen et al (2001) [7] found that Dutch livestock farmers considered price and production risks to be most important. In Japan, the biggest and the sometime occurring risks to livestock farmers are decrease of production by animal death and decrease of quality by equipment

ERS also found that keeping cash at hand was the chief risk management strategy for every farm size, for every commodity specialty, and in every region studied; use of derivative and insurance markets was also considered important [3]. Maintaining animal health was viewed

focuses on dairy farmers' risk perception and risk management strategies.

88 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

#### **6.3.1 Questionnaire and analysis technique**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

In this study, a survey has been conducted to sample farmers in order to ascertain the risks they are most concerned about and their risk management strategies to cope with such risks. There are a number of sources of risk and risk management strategies listed in the questionȬ naire. Based on previous studies of ERS (1997) [3], Wilson et al. (1988) [5], Meuwissen et al. (2001) [7] and Abdi (2004) [11], 22 sources of risk and 18 risk management strategies were given in the questionnaire at first. However, after consulting with experts and considering the current situation in China, finally, 19 sources of risk and 15 risk management strategies were selected in the questionnaire using in this study (Table 6.1 and Table 6.2). In China, since 2006 farmers do not need to pay tax, so that all the risks associated with tax payment are deleted.

To understand which risk and risk management strategies have more weight, a 5-level Likert scale is used. Dairy farmers are requested to judge and label different levels of each risk and risk management strategies. Level 1 stands for the least important and level 5 stands for the most important.

Farmers' risk perception and risk management strategies were studied by descriptive analyses. When the number of variables is large, a multivariate analysis technique such as principal component analysis (PCA) is effective to reduce its dimensionality. PCA is one multivariate technique that analyses a data table in which observations are described by several intercorrelated quantitative dependent variables [12]. Its goal is to extract the important information from the data, to represent it as a set of new orthogonal variables called principal components, and to display the pattern of similarity of the observations and of the variables as points in maps. In this research, therefore PCA is used to analyse the survey data. PCA is suitable to extract the principal risk and strategy from many risks and risk management strategies. To reduce the dimensionality of the variables, PCA is a powerful tool to do such work, without losing much information. Additionally, this method is often used in similar pieces of research. PCA is a proper way to analyse data and is adopted in this case. All analyses including PCA were carried out by using SPSS 13.0 for Windows [13].


**Table 6.1** Risk sources as categorized by dairy farmers


Note: 1=Least Important, 5=Most Important

PCA is a proper way to analyse data and is adopted in this case. All analyses including PCA

1. Changes in consumer preferences 1 2 3 4 5

90 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

2. Production diseases such as mastitis 1 2 3 4 5

3. Domestic epidemic animal diseases such as para tuberculosis 1 2 3 4 5

4. Non-domestic epidemic animal diseases such as foot and mouth disease 1 2 3 4 5

5. Misuse of veterinary drugs and veterinary drug residues 1 2 3 4 5

6. Related food safety issues occurring 1 2 3 4 5

7. Food safety news in media 1 2 3 4 5

8. Milk yield variability 1 2 3 4 5

9. Milk price variability 1 2 3 4 5

10. Corn yield variability 1 2 3 4 5

11. Corn price variability 1 2 3 4 5

12. Crop yield variability 1 2 3 4 5

13. Crop price variability 1 2 3 4 5

14. Costs of operating inputs 1 2 3 4 5

15. Changes in technology 1 2 3 4 5

16. Changes in government support payments 1 2 3 4 5

17. Health problems among family members 1 2 3 4 5

18. Hard to get loan 1 2 3 4 5

19. Fire damage, flood, drought, or other damage 1 2 3 4 5

Note: 1=Least Important, 5=Most Important

**Table 6.1** Risk sources as categorized by dairy farmers

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Category Level**

were carried out by using SPSS 13.0 for Windows [13].

Note: keeping cash at hand is considered as one of the risk management strategies that can handle all kinds of risk and fits all sizes of farms; producing at the lowest cost means dairy farmers will do everything that could help them to keep the variable costs low, such as use the least amount of feed; preventing and reducing livestock diseases here means during the dairy feeding, farmers add some medicine to the feed to avoid certain diseases; using human medicine instead of animal medicine and so on.

**Table 6.2** Dairy farmers' risk management strategies

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

After PCA, cluster analysis-hierarchical clustering was used to separate the farmers into different groups. Cluster analysis or clustering is the task of assigning a set of objects into groups (called clusters) so that the objects in the same cluster are more similar (in some sense or another) to each other than to those in other clusters.

Cluster analysis itself is not one specific algorithm, but the general task to be solved. It can be achieved by various algorithms that differ significantly in their notion of what constitutes a cluster and how to efficiently find them. Popular notions of clusters include groups with low distances among the cluster members, dense areas of the data space, intervals or particular statistical distributions.

#### **6.3.2 Study site and respondent attributes**

The field survey was carried out in Inner Mongolia and Hebei Province in April and June 2010, respectively. A sample of 168 dairy farmers is available for analysis in this study. Both Hebei Province and Inner Mongolia Autonomous Region are the main grain-growing and important milk production areas in China. The borders of the two provinces are adjacent to each other and located in the north part of China. Inner Mongolia is the largest milk producer in China and Hebei Province is the third biggest milk producer in China (Figure 6.2). By the end of 2010, the total population of the two regions amounted to 24.72 million and 71.94 million, accounting for 1.84% and 5.37% of the national population. With regard to average disposable income per urban household, both Inner Mongolia (17.70 thousand yuan) and Hebei Province (16.26 thousand yuan) are lower than the national average level of 19.11 thousand yuan, ranking the 10th and 14th among 31 mainland provincial regions of China. Meanwhile, the average conȬ sumption expenditure per urban household in Inner Mongolia and Hebei were 13.99 and 10.32 thousand yuan, while the national mean was 13.47 in the same time period [14]. Thus, the two regions have sound representativeness with regard to the consuming capability of China and are suitable as the survey areas.

Source: revised based on http://www.chinamapxl.com/

**Figure 6.2** Location of Inner Mongolia and Hebei Province

In this survey, most respondents are male, and nearly 80% of the respondents obtained less than nine years of education (Table 6.3). Most of the dairy farmers own less than 20 cattle, more than half of the dairy farmers only own around 10 cattle in each household. Meanwhile most dairy farmers have their own lands to grow corn which is used as the main feed for cows. Compared to countries such as the USA and some European countries, the annual average yield of each cow is not as good as could be hoped for, being only around 4000 kg. In the USA, a pure Holstein cow can produce around 10000 kg of milk per year. This is because the cows in the study area are pure Holstein mixed with local cattle, meaning the yield of second generation mixed Holstein cows is lower than pure Holstein cows.


**Table 6.3** Demographic features of sample farmers

#### **6.4 Result and discussion**

**6.3.2 Study site and respondent attributes**

are suitable as the survey areas.

Source: revised based on http://www.chinamapxl.com/

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Figure 6.2** Location of Inner Mongolia and Hebei Province

The field survey was carried out in Inner Mongolia and Hebei Province in April and June 2010, respectively. A sample of 168 dairy farmers is available for analysis in this study. Both Hebei Province and Inner Mongolia Autonomous Region are the main grain-growing and important milk production areas in China. The borders of the two provinces are adjacent to each other and located in the north part of China. Inner Mongolia is the largest milk producer in China and Hebei Province is the third biggest milk producer in China (Figure 6.2). By the end of 2010, the total population of the two regions amounted to 24.72 million and 71.94 million, accounting for 1.84% and 5.37% of the national population. With regard to average disposable income per urban household, both Inner Mongolia (17.70 thousand yuan) and Hebei Province (16.26 thousand yuan) are lower than the national average level of 19.11 thousand yuan, ranking the 10th and 14th among 31 mainland provincial regions of China. Meanwhile, the average conȬ sumption expenditure per urban household in Inner Mongolia and Hebei were 13.99 and 10.32 thousand yuan, while the national mean was 13.47 in the same time period [14]. Thus, the two regions have sound representativeness with regard to the consuming capability of China and

92 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

#### **6.4.1 Farmers' perception of risk**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Table 6.4 shows the dairy farmers' perception of risk in both Inner Mongolia and Hebei Province. The table shows that, on average, the highest scores are given to variability of milk price in both places. In general, price and production risks are considered as the biggest risk. In Inner Mongolia, milk price variability, non-domestic epidemic animal diseases, such as foot and mouth disease, and corn price variability are considered as the three biggest risks. However, the situation in Hebei Province is slightly different. Besides milk price variability and non-domestic epidemic animal diseases such as foot and mouth disease, other food safety issues are also considered important risks for dairy farmers. In 2008, the milk powder scandal begun in Hebei Province at first, and the scandal finally made the local dairy company bankrupt. It subsequently became difficult for some local farmers to sell fresh milk. This shows that the direct experience of a milk and food safety issue in past still negatively affects dairy farmers in Hebei Province.


Source: calculation based on survey 2010

Note: the order of risk is based on mean score of each one (Column 2)

**Table 6.4** Mean score and rank for source of risk

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

These results in China are similar to those obtain in related research carried out in some other countries, such as price variability in products, animal diseases such as foot and mouth disease and others. However, in those developed countries, mastitis is not such a serious disease to cows, but in China, it is still a big problem which significantly affects Chinese dairy farmers. This is because in most developed countries, better and more modern methods are used in raising cows, while in China old-fashioned and traditional ways still predominate among farmers allowing this kind of production disease to occur more easily. For example, in most developed countries, the total mixed ration (TMR) technique is used to raise cows. However, those small scale dairy farmers are unaware of TMR in China. Inappropriate ways of raising cows may cause mastitis and other diseases. In most developed countries, only machines are used to milk cows, but in China, some farmers still use hands to milk cows. Additionally, the disinfection technique used on the cows in China is not good enough, helping to aggravate the problem of mastitis for Chinese dairy farmers. Meanwhile, because of this kind of disease, large amounts of antibiotics are used in China, and this causes antibiotic residues to occur more often than in developed countries.

**Category**

Non-domestic epidemic animal diseases such as

Misuse of veterinary drugs and veterinary drug

Domestic epidemic animal diseases such as Para

Source: calculation based on survey 2010

**Table 6.4** Mean score and rank for source of risk

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Note: the order of risk is based on mean score of each one (Column 2)

foot and mouth disease

residues

tuberculosis

Milk price variability 4.86 1 4.89 1 4.82 1 Food safety issue news in media 4.47 2 4.78 2 4.01 4

94 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Related food safety issues occurring 4.41 4 4.74 3 3.93 6 Corn price variability 4.15 5 4.19 5 4.09 3 Production diseases such as mastitis 3.83 6 3.82 7 3.84 7

Crop price variability 3.67 8 4.04 6 3.22 11 Corn yield variability 3.65 9 3.41 10 3.99 5 Crop yield variability 3.31 10 3.37 11 3.12 13

Fire damage, flood, dry, or other damage 3.23 12 3.76 8 2.46 18 Health problems among family members 3.21 13 3.07 13 3.41 10 Milk yield variability 3.10 14 3.06 14 3.16 12 Costs of operating inputs 2.62 15 2.38 15 2.97 15 Changes in government support payments 2.58 16 2.31 16 2.99 14 Changes in technology 2.38 17 2.28 17 2.53 16 Changes in consumer preferences 2.36 18 2.26 18 2.51 17 Hard to get loan 1.73 19 1.61 19 1.91 19

These results in China are similar to those obtain in related research carried out in some other countries, such as price variability in products, animal diseases such as foot and mouth disease and others. However, in those developed countries, mastitis is not such a serious disease to cows, but in China, it is still a big problem which significantly affects Chinese dairy farmers. This is because in most developed countries, better and more modern methods are used in raising cows, while in China old-fashioned and traditional ways still predominate among farmers allowing this kind of production disease to occur more easily. For example, in most developed countries, the total mixed ration (TMR) technique is used to raise cows. However,

**Total Hebei Province Inner Mongolia Mean Rank Mean Rank Mean Rank**

4.42 3 4.31 4 4.57 2

3.70 7 3.67 9 3.74 8

3.30 11 3.11 12 3.59 9

Table 6.5 shows the Varimax rotated factor loadings for source of risk. Before starting PCA, sampling adequacy was checked to detect if the data will factor well. In SPSS, sampling adequacy is measured by using the Kaiser-Meyer-Olkin (KMO) criterion. KMO varies from 0 to 1 and a KMO higher than 0.5 implies that PCA can be applied to the data [15]. The KMO value of source of risk is 0.517 and is acceptable for PCA. The number of risk was reduced by applying PCA. This resulted in seven factors with eigenvalues larger than 1 and in total accounted for 77.4% (which can be regarded as satisfactory in social sciences).

In addition, Table 6.5 shows the factor loadings for the source of risks. According to the loadings, the factors can be described as 'production risk', 'institutional risk', 'animal disease', 'input market risk', 'milk contamination risk' and 'personal risk' and 'output market risk', respectively.

On the first factor 'production risk', high loading goes to input especially feed variability and some changes in farm operating input. However, those risks from outside of the farm business, such as fire damage, flood damage and other food safety issues, show a high negative loading. With regard to food safety issues occurring, this is also included in production risk. It also can be considered as some risk outside of the dairy business itself, so that the loading is negative but not very high. High loading of government support policy change and changes in technology are related to the second factor 'institutional risk'. In China, most technology improvements are done by national research institutes. Change in technology is also considȬ ered as an institutional risk. Animal diseases, such as foot and mouth disease, mastitis and para tuberculosis, veterinary residue and misuse of veterinary medicine also related to the third factor 'animal disease risk'. However, the government provides the vaccine for foot mouth disease twice a year for farmers free of charge, so mouth and foot disease shows a high negative loading.

On the other hand, mastitis and para tuberculosis are not included in the government free vaccine programme. Feed price, such as corn and other crop prices, show a high loading and contributes to the fourth factor of 'input market risk'. The factor 'milk contamination risk' mainly explains the situation in Hebei Province. The factor 'personal risk' is likely to reflect changes in the family, only health problems among household members shows a high loading in this factor. In addition, the milk price variability and change in consumer preferences are related to the last factor 'output market risk'. 'Changes in consumer preference' refers to the situation where, for example, consumers choose to consume more soybean milk instead of cow milk. However, milk yield is negatively correlated with milk price, so it shows a negative sign in this factor.


Source: calculation based on data of survey in 2010

Extraction method: principal component analysis and loadings larger than 0.5 are in bold

**Table 6.5** Varimax rotated factor loading for source of risk

#### **6.4.2 Farmers' risk management strategies**

**Category**

Fire, flood, drought, or other

Related food safety issues

Changes in government support

Domestic epidemic diseases such

Misuse of veterinary drugs and

Non-domestic epidemic animal diseases such as foot and mouth

Production diseases such as

Health problems among family

Cumulative % of the variance

Source: calculation based on data of survey in 2010

**Table 6.5** Varimax rotated factor loading for source of risk

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Changes in consumers

as para tuberculosis

veterinary residues

damages

occurring

payments

disease

mastitis

members

preferences

explained

**Production**

**Institution**

96 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Crop yield variability **0.834** 0.286 -0.019 0.102 -0.029 -0.004 -0.012 Costs of operating inputs **0.810** 0.347 -0.088 0.218 0.247 0.061 -0.009

Corn yield variability **0.715** 0.260 -0.182 -0.09 -0.083 -0.139 -0.426

Changes in technology 0.406 **0.578** 0.362 0.032 -0.079 0.400 -0.033

Corn price variability 0.186 0.000 0.084 0.785 0.059 -0.243 0.197 Crop price variability -0.396 0.325 0.012 **0.724** -0.220 -0.061 0.034

Food safety issue news in media -0.479 0.176 0.443 0.111 **-0.553** -0.092 0.149

Milk yield variability 0.093 0.263 0.171 0.321 0.414 0.087 **-0.565** Milk price variability -0.133 0.225 -0.031 0.247 -0.087 -0.014 **0.753**

24.329 37.776 48.647

Extraction method: principal component analysis and loadings larger than 0.5 are in bold

Hard to get load 0.133 0.372 0.298

**Component**

**-0.796** 0.057 -0.029 -0.055 -0.118 -0.151 0.008

**-0.507** 0.208 0.436 0.173 -0.426 -0.218 0.123


0.080 0.101 **0.866** -0.078 0.135 -0.058 -0.094


0.297 0.242 **-0.725** 0.008 -0.181 0.223 -0.179

0.012 -0.017 **0.644** -0.242 0.04 -0.146 -0.118

**-0.55 9**

0.044 0.061 -0.482 -0.092 0.147 **0.692** 0.210

0.061 -0.161 0.096 -0.209 -0.063 -0.129 **0.799**

57.69 3

**Input market**

**Milk contami -nation**

**Pers onal**

0.373 0.153 0.332

65.344 71.933 **77.419**

**Output Market**

**Animal disease** Producingatthe lowest cost andpreventingorreducinglivestockdiseases are consideredas the most useful risk management strategies in both Inner Mongolia and Hebei Province (Table 6.6).


Source: calculation based on the data of survey in 2010

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Note: the order of the risk management is based on the mean score of each one (Column 2)

**Table 6.6** Mean score and rank for the risk management strategies

However, some strategies such as production contract, farmer corporation group and agriȬ cultural insurance, which are viewed as very important in other countries, were not popular in China, since most dairy farmers in China are still separate and small-scale. In the USA, Patrick and Musser showed that the large-scale US farmers viewed liability insurance as important managerial response to risk [16]. Additionally, the 1996 USDA survey found that keeping cash at hand was the chief risk management strategy for every farm size, for every commodity specialty. This result is also proved in the current research and shows that in China, farmers also consider keeping cash at hand to be a way to deal with all kinds of risk. Quitting the business is listed as the last thing the farmers would do, and it indicates that the farmers would not easily to give up this business when they meet any risk.

As with the sources of risk, the number of risk management strategies was also reduced by applying PCA. The KMO value of risk management strategies is 0.512 and this is acceptable for PCA. Table 6.7 shows the Varimax rotated factor loadings for risk management strategies. This resultedin6 factorswitheigenvaluesgreaterthan1, andtotalvariance explainedbythe 6 factors accounts for more than 80% of all the variables, which is also considered as satisfactory in social science. Based on the concentration of factor loadings, the six factors can be described as 'cost decrease', 'income stabilization', 'income increase', 'farmer group', 'insurance' and 'consultanȬ cy', respectively.


**Table 6.7** Varimax rotated factor loadings for risk management strategies

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Controlling fixed costs through shared ownership of equipment and partnership loads high on Factor 1 which is 'cost decrease'. However, compared with others, risk reducing technology is outside of the farm business itself, so it shows a negative loading in this factor. Factor 2 and Factor 3 are related to farmers' income. Factor 2, 'income stabilization', has a high loading of product contract which makes a certain income from the farm business itself. Factor 3, 'income increase', has a high loading of off-farm work which keeps a certain income out of the current farm business. Factor 4 is named 'farmer group', only one component is included in this factor. It has a high loading of joining farmers' corporation associations. Factor 5, 'insurance', has high loadings of purchasing agricultural insurance. In addition, preventing or reducing animal disease also shows a high loading on this component. We believe proper treatment and conducting animal health check-up's can reduce the possibility of animal disease, and it can be viewed as a kind of insurance for farmers. Factor 6, 'consultancy', has high loadings of consultant service and extension training in many aspects.

The cluster analysis is based on risk management strategies. Four clusters were assessed by hierarchical cluster analysis using the Ward method. The dendrogram in Figure 6.3 shows the principle and process of hierarchical cluster analysis. All the original cases are sorted into different groups in a way that the distance between two objects is maximal if they belong to the same group and minimal otherwise. Among the methods measuring distances between the objects, the Ward method is regarded as very efficient because it uses an analysis of variance approach to evaluate the distances between clusters. In short, this method attempts to miniȬ mizethesumofsquares(SS)ofanytwo(hypothetical)clustersthatcanbeformedateachstep[17].

#### Software: SPSS 13.0

As with the sources of risk, the number of risk management strategies was also reduced by applying PCA. The KMO value of risk management strategies is 0.512 and this is acceptable for PCA. Table 6.7 shows the Varimax rotated factor loadings for risk management strategies. This resultedin6 factorswitheigenvaluesgreaterthan1, andtotalvariance explainedbythe 6 factors accounts for more than 80% of all the variables, which is also considered as satisfactory in social science. Based on the concentration of factor loadings, the six factors can be described as 'cost decrease', 'income stabilization', 'income increase', 'farmer group', 'insurance' and 'consultanȬ

98 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Factor 2 Income stabilization**

Risk reducing technology **-0.719** 0.210 -0.189 0.320 0.108 -0.208

Production contracts 0.005 **0.893** 0.008 0.028 -0.115 -0.084

Off-farm work -0.024 -0.070 **0.909** 0.074 -0.055 -0.039 Collecting information 0.443 -0.254 **0.641** 0.039 0.156 0.283

Buying agricultural insurance -0.041 0.017 -0.210 -0.106 **0.729** 0.205

Joining extension training 0.430 -0.468 -0.216 -0.016 0.092 **-0.665**

Produce at the lowest cost 0.464 -0.156 -0.477 0.096 0.365 0.141

Controlling fixed costs through shared ownership of equipment and partnership loads high on Factor 1 which is 'cost decrease'. However, compared with others, risk reducing technology is outside of the farm business itself, so it shows a negative loading in this factor. Factor 2 and Factor 3 are related to farmers' income. Factor 2, 'income stabilization', has a high loading of

**Component**

**0.843** 0.086 0.056 -0.145 0.223 -0.100

**0.815** -0.143 -0.019 -0.008 -0.057 0.116

**0.582** 0.462 **0.588** -0.010 0.055 0.080



0.119 -0.155 0.211 0.065 **0.814** -0.195

0.439 -0.216 -0.089 0.005 -0.009 **0.778**

0.156 -0.467 -0.189 0.190 -0.071 **-0.684**

27.622 43.917 56.236 65.842 73.529 **80.429**

**Factor 4 Farmer group**

**Factor 5 Insurance** **Factor 6 Consultancy**

**Factor 3 Income increase**

cy', respectively.

Shared ownership of equipment, joint operations

Asset flexibility - farm building with multiple use

machinery

hand

diseases

business

explained

**Category**

Keeping fixed costs low - rent

Liquidity - keeping cash at

Joining the farmers' corporation

Prevent/reduce livestock

Using consultant service or consultant extension workers

Kill or sell the cows, quit the

Cumulative % of the variance

**Table 6.7** Varimax rotated factor loadings for risk management strategies

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Factor 1 Cost decrease**

**Figure 6.3** Dendrogram of hierarchical cluster analysis

As shown in Figure 6.3, with the increase of distances between the clusters, they are sorted into fewer clusters, until all the cases merge into one cluster with the largest distance of 25. In most of the cases, we need to balance the numbers of clusters and the distances, to analyse the features of the original objects with appropriate scales. Therefore, we decided to choose four clusters and marked them A, B, C and D in Figure 6.3. Table 6.8 shows the different opinions on new risk management strategies among different clusters. As analysed in the last chapter, joining farmers' corporation and production contract are considered new risk management strategies, and others are considered traditional risk management strategies. Farmers in Cluster A are more interested in new strategies such as joining farmers' corporations and using production contracts than farmers in other clusters. Meanwhile, it seems that they are less interested in training programmes compared with farmers in Cluster B, C and D. This implies that farmers in Cluster A are more self-reliant. They rely less on help from training and consultant services, and are more self-reliant.


**Table 6.8** Opinion on risk management strategies among different clusters

Table 6.9 shows the comparison among different clusters. Farmers in Cluster A are slightly younger but have a higher education level than other clusters. Meanwhile, farmers in Cluster A usually own more cows than other clusters. All of these facts imply that farmers in Cluster A are younger, better educated and have a larger scale of farm. Farmers with these features are more self-reliant and more interested in new risk management strategies. Put another way, those older, less well educated farmers with smaller scale farms rely more on outside help, training programmes and traditional ways of risk management response.


**Table 6.9** Comparison of socio-characteristics among four groups

Based on above analysis, Cluster B, Cluster C and Cluster D indicate similar opinions about risk management strategies, and farmers in these three groups show a indicate different ideas than those farmers in Cluster A. Therefore, finally, Cluster B, Cluster C and Cluster D are combined into one large group, called Group 2, Cluster A is called Group 1.


**Table 6.10** Different features of the two groups

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

As shown in Figure 6.3, with the increase of distances between the clusters, they are sorted into fewer clusters, until all the cases merge into one cluster with the largest distance of 25. In most of the cases, we need to balance the numbers of clusters and the distances, to analyse the features of the original objects with appropriate scales. Therefore, we decided to choose four clusters and marked them A, B, C and D in Figure 6.3. Table 6.8 shows the different opinions on new risk management strategies among different clusters. As analysed in the last chapter, joining farmers' corporation and production contract are considered new risk management strategies, and others are considered traditional risk management strategies. Farmers in Cluster A are more interested in new strategies such as joining farmers' corporations and using production contracts than farmers in other clusters. Meanwhile, it seems that they are less interested in training programmes compared with farmers in Cluster B, C and D. This implies that farmers in Cluster A are more self-reliant. They rely less on help from training and

100 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Shared ownership of equipment, joint operations 2.80 2.93 3.87 3.88 3.52 Join the training programme 2.00 3.96 4.00 3.86 3.63 Joining farmers' corporation 3.96 2.63 2.40 3.13 2.98 Asset-flexibility - farm building with multiple uses 2.00 2.21 2.87 3.05 2.68 Production contracts 3.92 2.36 3.70 2.13 2.82

Table 6.9 shows the comparison among different clusters. Farmers in Cluster A are slightly younger but have a higher education level than other clusters. Meanwhile, farmers in Cluster A usually own more cows than other clusters. All of these facts imply that farmers in Cluster A are younger, better educated and have a larger scale of farm. Farmers with these features are more self-reliant and more interested in new risk management strategies. Put another way, those older, less well educated farmers with smaller scale farms rely more on outside help,

**Cluster A B C D All**

**Cluster A (26) Cluster B (33) Cluster C (40) Cluster D (69)**

consultant services, and are more self-reliant.

**Table 6.8** Opinion on risk management strategies among different clusters

Source: calculation based on the data of survey in 2010

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Table 6.9** Comparison of socio-characteristics among four groups

training programmes and traditional ways of risk management response.

Age 41.5 42.9 42.5 41.7 Educated more than 9 years 19.20% 9.20% 10.30% 13.20% Number of Cows 25 12 14 19

#### **6.5 Conclusions and policy recommendations**

Price and production risks are considered as the most important risk to dairy farmers. Milk price variability is the most important risk to dairy farmers in both Hebei Province and Inner Mongolia. A milk safety issue in the media is a very important risk for dairy farmers in Hebei Province and negatively affects them. Non-domestic epidemic animal diseases, such as foot and mouth disease, are considered as the most serious problems among all kinds of disease.

Based on the above results, we found that the risks to dairy farms can be categorized as 'production risk', 'institutional risk', 'animal disease', 'input market risk', 'milk contamination risk', 'personal risk' and 'output market risk'. Among these risks, price risk and the animal diseases risk are the most serious in dairy farmers' perception. Risk management strategies are categorized as 'cost decrease', 'income stabilization', 'income increase', 'farmer group', 'insurance' and 'consultancy'. Among these strategies, dairy farmers feel that producing at the lowest cost and preventing livestock diseases are the most effective risk management strategies as a single strategy. Dairy farmers in China largely rely on extension or veterinary services, so developing the veterinary extension services to guide farmers to avoid livestock diseases is a good way to reduce these related risks. Extension and veterinary services are good example of 'consultancy'.

Farmer groups and agricultural insurance becomes a way to help farmers to manage risk in crops and some other animal products, however, it is not widely accepted by dairy farmers. Promotion of these strategies might be useful for dairy farmers as well and might help them to avoid many risks and reduce losses.

For the government, more consultant services should be offered; some new risk management strategies should be extended to farmers such as production contracts. For dairy farmers, new risk management strategies, such as agricultural insurance, should be more accepted. The milk powder scandal in China happened in 2008 having a significant negative effect on the local milk industry (mainly in reference to Hebei Province). It totally destroyed the local business and seriously affected public trust in the industry nationwide. With that in mind, even though two years have passed, farmers are still having difficulty selling their milk. A set of effective risk management strategies for dairy farmers is needed.

Although milk price variability is considered as the top risk to dairy farmers, the milk price data is not available in the current research and this is a limitation of the research. In the future, milk market price data should be collected and how milk prices change should be also examined. New risk management strategies are more popular in younger and better educated farmers such as farmers in Group 1 - they usually own more cows and show more self-reliance. These younger and better educated farmers are more interested in joining farm associations, using production contracts with diary enterprises, but are less dependent on extension workers.

Cluster B C and D are similar to each other, so these three clusters are combined into one big group; they rely on training programmes and consultant services more than Group 1. HowȬ ever, those older and less well educated farmers with fewer cattle are not willing to use these new risk strategies. Controlling fixed costs through shared ownership of equipment and partnership loads high on Factor 1 which is 'cost decrease'. However, compared with others, risk reducing technology is outside of the farm business itself, so it shows a negative loading in this factor. Factor 2 and Factor 3 are related to farmers' income. Factor 2, 'income stabilizaȬ tion', has a high loading of production contracts which makes a certain income from the farm business itself. Factor 3, 'income increase', has a high loading of off-farm work which keeps a certain income out of the current farm business. Factor 4 is named as 'farmer group', only one component is included in this factor. It has a high loading of joining farmers' corporation associations. Factor 5, 'insurance', has a high loading of purchasing agricultural insurance. Additionally, preventing or reducing animal diseases also shows a high loading on this component. We believe that by using proper treatment and undertaking animal health checkups can reduce the possibility of animal diseases, and it can be viewed by farmers as a kind of insurance. Factor 6, 'consultancy', has high loadings of consultant service and extension training in many aspects.

#### **References**


[4] Fleisher, B. (1990). Agricultural Risk Management. Lynne Rienner Publishers Inc. USA.

powder scandal in China happened in 2008 having a significant negative effect on the local milk industry (mainly in reference to Hebei Province). It totally destroyed the local business and seriously affected public trust in the industry nationwide. With that in mind, even though two years have passed, farmers are still having difficulty selling their milk. A set of effective

102 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Although milk price variability is considered as the top risk to dairy farmers, the milk price data is not available in the current research and this is a limitation of the research. In the future, milk market price data should be collected and how milk prices change should be also examined. New risk management strategies are more popular in younger and better educated farmers such as farmers in Group 1 - they usually own more cows and show more self-reliance. These younger and better educated farmers are more interested in joining farm associations, using production contracts with diary enterprises, but are less dependent on extension

Cluster B C and D are similar to each other, so these three clusters are combined into one big group; they rely on training programmes and consultant services more than Group 1. HowȬ ever, those older and less well educated farmers with fewer cattle are not willing to use these new risk strategies. Controlling fixed costs through shared ownership of equipment and partnership loads high on Factor 1 which is 'cost decrease'. However, compared with others, risk reducing technology is outside of the farm business itself, so it shows a negative loading in this factor. Factor 2 and Factor 3 are related to farmers' income. Factor 2, 'income stabilizaȬ tion', has a high loading of production contracts which makes a certain income from the farm business itself. Factor 3, 'income increase', has a high loading of off-farm work which keeps a certain income out of the current farm business. Factor 4 is named as 'farmer group', only one component is included in this factor. It has a high loading of joining farmers' corporation associations. Factor 5, 'insurance', has a high loading of purchasing agricultural insurance. Additionally, preventing or reducing animal diseases also shows a high loading on this component. We believe that by using proper treatment and undertaking animal health checkups can reduce the possibility of animal diseases, and it can be viewed by farmers as a kind of insurance. Factor 6, 'consultancy', has high loadings of consultant service and extension

[1] Nanseki, T. Management of Risk and Information in Agriculture. Agriculture and

[2] FAO (Food and Agriculture Organization)FAOSTAT Agriculture Database, Accessed on November (2008). available at http://faostat.fao.org/site/339/default.aspx

[3] ERS (Economic Research Service)USDA. Agricultural Economic Report Managing Risk in Farming: Concepts, Research (1997). http://www.nal.usda.gov/ref/USDAȬ

Forestry Statistics Publishing Inc. Japan (2011). (in Japanese)

risk management strategies for dairy farmers is needed.

workers.

training in many aspects.

pubs/aer.htm(774)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**References**


#### **Chapter 7**

### **Consumer Perceptions on Food Safety and Demographic Determinants**

Dongpo Li, Tinggui Chen, Hui Zhou and Teruaki Nanseki

With the fast development of science and technology in food production and processing, food is being supplied to satisfy increasingly diversified tastes, nutritional requirements, etc. Nevertheless, food safety is growing to be a global concern among consumers simultaneousȬ ly, due to their asymmetric information on the processes, additives in the long industrial chain, and also influence of the flourishing public media. Consumers are demanding the reȬ inforced assurance of food safety and even one isolated event may cause major market disȬ ruptions. In addition to the endeavours by governmental agencies and enterprises, food safety, especially from the perspective of promoting consumer confidence, has received the considerable attention of scholars. Thus, in this chapter, the research perspective will be changed from interviewing farmers to consumers, with the further analysis of their percepȬ tions and determining factors towards food safety.

#### **7.1 Introduction**

According to the latest research revealed by the international food and grocery expert IGD, China surpassed the US to become the world's largest food and grocery retail market at the end of 2011 [1]. In recent years, especially after the melamine milk powder incident occurred in September 2008 (for more details, refer to Qiao et al. (2012) [2]), many scholars have conducted empirical studies on food safety, based on consumer surveys. In general, the study topics include: (1) consumers' overall perceptions on food safety. Wang et al (2009) [3] and Xu et al (2010) [4] examine consumers' willingness to pay (WTP) for safe fishery products and certified and traceable food, respectively. Qiao et al (2010) [5] studies the changes of consumers' confidence in the domestic dairy industry after melamine milk powder incident, while Zhang et al (2010) [6] divides the sampled consumers into four groups, in respect to their perceptions about and attitudes toward GM food. (2) Consumers' behaviours on choosing safe food. Zhang

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Li et al.; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Li et al.; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

et al (2010) [7] examines consumers' identification of safe dairy products; Ortega et al (2011) [8] measures the heterogeneity in consumers' preferences for selecting safe pork; Kim (2009) [9] conducts factor analysis on consumers' purchase of GM food. (3) Integrated study of perceptions and behaviours towards food safety. Han et al (2012) [10] compares the consistency of consumers' stated and revealed preferences to certified pork.

Although the existing studies have covered many essential aspects and provided instructive recommendations, there are still a variety of topics that need to be researched further. For instance, (1) consumers' overall awareness of the food safety situation, which constitutes the basis for analysis of individual behaviours; (2) inclusion of questions covering the whole industrial food supply chain, from agricultural production as the origin process; (3) empirical analysis of the relationship between consumers' demographic characteristics and perceptions, etc. Therefore, based on the survey of 512 respondents from Beijing and Shanghai, the top two metropolises in China, this study analyses consumer perceptions towards food safety, including general concern and valuation; major information sources and the subjective reliabilities; awareness about the causes and countermeasures of food safety risks. To explore significant determinants behind the perceptions, a variety of demographic variables with regard to the respondents are included, from gender, age, employment, education backȬ ground, to the member composition and annual income of each household. The remainder of the chapter is organized as follows: Section 2 briefly describes the questionnaire, sampling and demographic characteristics; Section 3 illustrates the major perceptions of the surveyed consumers; Section 4 analyses the major determinants behind consumer perceptions; Section 5 presents the conclusions and policy recommendations, followed by further discussion.

#### **7.2 Field survey**

#### **7.2.1 Questionnaire and sampling**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

To understand the present situation and farmers' food safety perceptions, we conduct the survey using questionnaire-based personal interviews, to collect first-hand data. The quesȬ tionnaire consists of 30 multi-choice questions, which are divided into the following four sections, according to different topics of information we intend to collect.

As shown in Table 7.1, (1) we enquire about consumers' overall awareness towards food safety with four questions in the first section. The main topics include their concern about food safety, ranking from *very much* to *not at all*; subjective valuation of the safety of the current food supply, with the main candidate answers ranging from *very safe* to *very risky*; the major source of information, where the candidate answers incorporate not only the traditional mass media of radio and television, newspapers and magazines, and newly developing media of the Internet, but also the other routes such as relatives, experts, commercial advertisements, package information, etc; the information sources deemed to be most reliable vary from government, experts, Internet, relatives, producers and vendors, and commercial advertisement. In addition, with a question for reference, we asked about consumers' awareness of three kinds of certified food: nuisance-free, green and organic. (2) Section 2 is composed of seven questions, within which the first five questions include consumers' awareness on the relationship between environmental pollution and food safety, with 5-level ordinal options from *very intimate* to *does not exist*; most risky substance to food safety, with the multiple choice options of industrial, agricultural and civil pollution; top source of agro-pollution, to be chosen from industrial pollutants, urban or rural civil pollutants, and agricultural chemicals; most risky procedure or stage within the food supply chain, covering agro-production to consumption. In addition, we asked two further questions about consumers' perceptions on the major responsibility bearers of agro-pollution, with the options of the government, farmers, producȬ ers of agro-inputting materials (chemicals in the main), consumers, etc.; best ways to control agro-pollution, including the optional answers of legislative perfection, extending environȬ mental technology, etc. (3) Although not adopted in this study, we asked consumers' percepȬ tions on the safety of dairy products with 10 questions in Section 3. The topics vary from overall awareness; subjective valuation on the risky substances and processes; acquaintance of the food certificating systems of GAP (Good Agricultural Practices) and HACCP (Hazard Analysis Critical Control Point); purchasing and consuming behaviours. (4) In the final section, our questionnaire contains eight questions on the demographic characteristics of the respondents. In addition to gender, age, employment and education level, we asked about their family scale, member composition and annual income of the household.

In January to March, 2012, we surveyed consumers in the two metropolises of Beijing and Shanghai. The survey was completed thanks to the kind cooperation of the China Agricultural University and Shanghai Ocean University, from where altogether 30 students were selected and trained as surveyors. All of them are undergraduates or postgraduates majoring in food economics or similar fields. The respondents were determined in two ways: on one hand, they are relatives of the surveyors or people who live near to them; on the other hand, random surveys were conducted through interviewing consumers, encountered mainly near the major supermarkets. In principle, one surveyor can interview no more than 20 consumers. Because some of the authors participated in this survey as well, the initial sample size amounted to 617. Nevertheless, screened mainly by rationality and completion of data, 512 samples are accepted as valid and used in the final analysis, thus the ratio of valid samples is 82.89 %.

#### **7.2.2 Demographic characteristics**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

et al (2010) [7] examines consumers' identification of safe dairy products; Ortega et al (2011) [8] measures the heterogeneity in consumers' preferences for selecting safe pork; Kim (2009) [9] conducts factor analysis on consumers' purchase of GM food. (3) Integrated study of perceptions and behaviours towards food safety. Han et al (2012) [10] compares the consistency

106 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Although the existing studies have covered many essential aspects and provided instructive recommendations, there are still a variety of topics that need to be researched further. For instance, (1) consumers' overall awareness of the food safety situation, which constitutes the basis for analysis of individual behaviours; (2) inclusion of questions covering the whole industrial food supply chain, from agricultural production as the origin process; (3) empirical analysis of the relationship between consumers' demographic characteristics and perceptions, etc. Therefore, based on the survey of 512 respondents from Beijing and Shanghai, the top two metropolises in China, this study analyses consumer perceptions towards food safety, including general concern and valuation; major information sources and the subjective reliabilities; awareness about the causes and countermeasures of food safety risks. To explore significant determinants behind the perceptions, a variety of demographic variables with regard to the respondents are included, from gender, age, employment, education backȬ ground, to the member composition and annual income of each household. The remainder of the chapter is organized as follows: Section 2 briefly describes the questionnaire, sampling and demographic characteristics; Section 3 illustrates the major perceptions of the surveyed consumers; Section 4 analyses the major determinants behind consumer perceptions; Section 5 presents the conclusions and policy recommendations, followed by further discussion.

To understand the present situation and farmers' food safety perceptions, we conduct the survey using questionnaire-based personal interviews, to collect first-hand data. The quesȬ tionnaire consists of 30 multi-choice questions, which are divided into the following four

As shown in Table 7.1, (1) we enquire about consumers' overall awareness towards food safety with four questions in the first section. The main topics include their concern about food safety, ranking from *very much* to *not at all*; subjective valuation of the safety of the current food supply, with the main candidate answers ranging from *very safe* to *very risky*; the major source of information, where the candidate answers incorporate not only the traditional mass media of radio and television, newspapers and magazines, and newly developing media of the Internet, but also the other routes such as relatives, experts, commercial advertisements, package information, etc; the information sources deemed to be most reliable vary from government, experts, Internet, relatives, producers and vendors, and commercial advertisement. In addition, with a question for reference, we asked about consumers' awareness of three kinds of certified food: nuisance-free, green and organic. (2) Section 2 is composed of seven questions,

sections, according to different topics of information we intend to collect.

of consumers' stated and revealed preferences to certified pork.

**7.2 Field survey**

**7.2.1 Questionnaire and sampling**


45.9% of the respondents have received a university education and 14.8% are postgradȬ uates; 28.1% having the experience of attending a high school, while respondents with a middle school and less educational background account for only 11.2%.

**3.** Household information. As to the dichotomous questions on whether a family includes preschool children, primary or middle school students, or people over 60 years, positive answers are 10.2%, 40.8% and 31.1%, respectively. Finally, with respect to the annual household income, one third of the respondents answered as 70-150,000 yuan per year, followed by 35-70,000 yuan per year (28.6%), less than 35,000 yuan per year (17.2%) and 150-300,000 yuan per year (16.4%), while only 4.5 responded as over 300,000 yuan per year (Table 7.1).


Note: a public institution refers to the institution of public interests, i.e., hospital, educational institutions, academy, etc.; b yuan is the major currency unit in China, and 1 US\$ equals to 6.30 yuan at the end of 2011.

Source: field survey by the authors

**Table 7.1** Demographic characteristics of the surveyed consumers

#### **7.3 Perceptions on food safety**

45.9% of the respondents have received a university education and 14.8% are postgradȬ uates; 28.1% having the experience of attending a high school, while respondents with a

**% Characteristic Valid N %**

Yes 209 40.8

500000 Ɖ 4 0.8

493 100.0

505 100.0 No 303 59.2

512 100.0 70000-150000 164 33.3

**3.** Household information. As to the dichotomous questions on whether a family includes preschool children, primary or middle school students, or people over 60 years, positive answers are 10.2%, 40.8% and 31.1%, respectively. Finally, with respect to the annual household income, one third of the respondents answered as 70-150,000 yuan per year, followed by 35-70,000 yuan per year (28.6%), less than 35,000 yuan per year (17.2%) and 150-300,000 yuan per year (16.4%), while only 4.5 responded as over 300,000 yuan per year

**1. Basic individual information** 2.3 Education level of the respondent (*d*5) 499 100.0 1.1 Gender of the respondent (*d*1) 512 100.0 Primary school and less 10 2.0 Male 208 40.6 Junior middle school 46 9.2 Female 304 59.4 High school 140 28.1 1.2 Age of the respondent (*d*2) 504 100.0 Undergraduate 229 45.9 <20 30 6.0 Postgraduate 74 14.8

30-39 131 26.0 3.1 Preschool Child (*d*6) 512 100.0 40-49 111 22.0 Yes 52 10.2 50-59 55 10.9 No 460 89.8 60 Ɖ 26 5.1 3.2 Primary and middle school student (*d*7) 512 100.0

Government 31 6.1 3.3 Elderly over 60 years (*d*8) 512 100.0 Public institution a 117 23.2 Yes 159 31.1 Enterprise 145 28.7 No 353 68.9

*d*9)

Jobless 34 6.7 < 35000 85 17.2 Other (student, retired, etc) 109 21.6 35000-70000 141 28.6

Yes 76 14.8 150000-300000 81 16.4 No 436 85.2 300000-500000 18 3.7

Note: a public institution refers to the institution of public interests, i.e., hospital, educational institutions, academy, etc.;

b yuan is the major currency unit in China, and 1 US\$ equals to 6.30 yuan at the end of 2011.

**Table 7.1** Demographic characteristics of the surveyed consumers

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Self-employed 69 13.7 3.4 Annual household income (yuan b /year,

middle school and less educational background account for only 11.2%.

108 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Valid N**

20-29 151 30.0 **3. Household information**

(Table 7.1).

**2. Additional individual information**

(*d*3)

2.1 Employment of the respondent

2.2 Background in agriculture, food and medicine (*d*4)

Source: field survey by the authors

**Characteristic**

#### **7.3.1 Overall awareness on current situation**


#### **7.3.2 On the major affecting factors**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**1.** The relationship of environment and food safety. According to the data collected in this survey, 70.8% of the interviewed consumers believe that there is a *very intimate* relationȬ ship between the environment and food safety; 20.3% admit the existence of an *intimate* relationship; while very few respondents deny the influence of the environment on food safety. This indicates that the concept of environmental protection has gained wide acceptance, even from the perspective of food safety.


#### **7.3.3 On the risk management**



Source: field survey by the authors

**2.** Top threat to food safety. Among the optional answers, industrial pollution is selected by 57.5% of consumers, which is the largest proportion among respondents; followed by agricultural pollution with the proportion of 29.4%; while only 9.1% of the consumers selected civil pollution as the top threat to food safety. In addition to industrial pollution, this result reveals that impacts of agricultural pollution are drawing public attention

110 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**3.** Top source of agro-pollution. Agricultural chemicals and industrial pollutants are the most significant source of agro-pollution to 45.7% and 39.7% of the interviewed consumȬ ers, respectively. Many of the food safety problems in China can be traced back to the farm level, as some farmers still rely heavily on the use of highly toxic pesticides to maintain the output of agro-products [11-12]. Meanwhile, both urban and rural civil pollutants are

**4.** Most risky procedure or stage of food supply. Among consumers' answer, processing and agricultural production are most risky to food safety, with the proportion of 46.3% and 42.6%, respectively. For food processing, the high ratio may due to asymmetric informaȬ tion on the operations, additives and the frequent disclosure of related scandals by the public media [2, 8]. Simultaneously, far fewer respondents selected the other procedures or stages, i.e., harvest of agro-products (2.4%), transportation (4.8%), marketing (1.4%),

**1.** Major responsibility bearer of agro-pollution. Most of the surveyed consumers ascribe the responsibility to government, followed by producers of agro-chemicals, etc., with the proportions at 54.0% and 33.6%, respectively. In previous literature, Dellios et al (2009) [13] introduces the concept of corporate social responsibility (CSR) into the food industry, and explores the important role of government in tackling food safety problems. Similarly, Qiao et al (2010) [5] analyses the responsibilities of government and corporate agencies, in taking efficient countermeasures to ensure food safety and maintain consumer confiȬ dence. In addition, far fewer respondents believe that this responsibility should be borne

**2.** Best ways to control agro-pollution. The answers to which can be divided into three groups. The first group includes perfecting the legislation, extending environmental technology, supported by 33.2 and 26.9 % of respondents, respectively. As surveyed by Li et al. (2012) [2], quite a few (12.4%) consumers have knowledge about the Act of Food Safety that came into effect on June 1, 2009. According to our prior survey, toxic pesticides are still being widely used, while bio-controls of pests are not yet well extended to agricultural production [12]. The second group comprises strengthening the penalties, enlarging farm managerial scales, with the largest proportion of 19.0% and 10.6%, respectively. The significance of the larger farming scales lies in the easier adoption of advanced and environmental technology, through initiatives and capital power. The third group consists of subsiding environmental behaviours and others, as viewed by 7.3% and

simultaneously.

selected by no more than 8% of the respondents.

consumption (0.2%), etc. (Table 7.2).

by farmers (5.7%) and consumers (1.4%), etc.

3.0% of surveyed consumers, respectively (Table 7.2).

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**7.3.3 On the risk management**

**Table 7.2** Summary of consumer perceptions on food safety

#### **7.4 Impact of demographic characteristics**

#### **7.4.1 Significance of demographic effects**

To measure the relative importance of the candidate perceptions, we calculate their ratios within each demographic variable. Furthermore, similar to Steiner et al. (2005) [14] and Gacula et al. (2006) [15], coefficient of variation (CV) of these ratios is computed to showcase a discrepancy of consumer perceptions towards a certain optional answer about food safety, crossing different features within each demographic characteristic (Table 7.3). Taking the first value of 0.17 as an instance, this is the CV of percentages of male and female respondents concerning food safety *very much*. In general, a smaller value of CV indicates less variation of responding ratios, hence less influence from the difference of this demographic characteristic. Another instance relates to the ratios of concern about food safety *very much*, where the CVs from different gender and age are 0.17 and 0.14, respectively, thus the affect of gender is larger than that of the age. In succession, to identify the significance of the CVs, a one-way T-test is conducted with the application of SPSS 13.0 [14]1 . The null hypothesis is that each population mean of CVs is not significantly different from 0, when respondents' choices are unaffected by this demographic characteristic. If the null hypothesis is rejected with a smaller *p*-value than the thresholds of 0.01, 0.05 or 0.1, then there is evidence that a significant discrepancy exists among different features within a demographic characteristic, and vice versa [16].


Note: a responses of *no idea* are excluded in this table; b the same coefficient is computed within other perceptions

Software: SPSS 13.0

**Table 7.3** Coefficient of variation of responding ratios within each characteristic

<sup>1</sup> The R2 value of 0.406 should not be used to judge fitness of a model. The fact that R2 never decreases when any variable is added to a regression makes it a poor tool for deciding whether one or several variables should be added to a model. Low R2s in regression equations are not uncommon, especially for cross-sectional analysis. Thus using R2 as the main gauge of success for an econometric analysis can lead to difficulties [17].


Note: a numerals are the T-values of the one-sample T-test on whether the mean of CVs within each perception significantly differs from 0; b \*\*\*, \*\* and \* denote statistical significance in the level of 0.01, 0.05 and 0.1 respectively

Software: SPSS 13.0

**7.4 Impact of demographic characteristics**

within a demographic characteristic, and vice versa [16].

Extending environmental

Subsiding environmental

**Table 7.3** Coefficient of variation of responding ratios within each characteristic

success for an econometric analysis can lead to difficulties [17].

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

technology

behaviours

To measure the relative importance of the candidate perceptions, we calculate their ratios within each demographic variable. Furthermore, similar to Steiner et al. (2005) [14] and Gacula et al. (2006) [15], coefficient of variation (CV) of these ratios is computed to showcase a discrepancy of consumer perceptions towards a certain optional answer about food safety, crossing different features within each demographic characteristic (Table 7.3). Taking the first value of 0.17 as an instance, this is the CV of percentages of male and female respondents concerning food safety *very much*. In general, a smaller value of CV indicates less variation of responding ratios, hence less influence from the difference of this demographic characteristic. Another instance relates to the ratios of concern about food safety *very much*, where the CVs from different gender and age are 0.17 and 0.14, respectively, thus the affect of gender is larger than that of the age. In succession, to identify the significance of the CVs, a one-way T-test is conducted with the

112 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

significantly different from 0, when respondents' choices are unaffected by this demographic characteristic. If the null hypothesis is rejected with a smaller *p*-value than the thresholds of 0.01, 0.05 or 0.1, then there is evidence that a significant discrepancy exists among different features

**Perception and options a** *d1 d2 d3 d4 d5 d6 d7 d8 d9*

…b … ………………………

Very much 0.17 0.14 0.17 0.08 0.26 0.02 0.03 0.07 0.32 Much 0.09 0.12 0.07 0.10 0.18 0.03 0.04 0.03 0.21 A little 0.20 0.39 0.52 0.07 0.59 0.00 0.15 0.13 0.54 Not at all 0.99 1.22 1.08 0.82 0.96 1.41 1.41 0.41 1.11

Perfecting the legislation 0.12 0.13 0.23 0.19 0.39 0.16 0.13 0.03 0.49

Strengthening the penalties 0.17 0.42 0.56 0.09 0.32 0.22 0.21 0.15 0.55 Enlarging farm managerial scales 0.07 0.46 0.28 0.12 0.36 0.08 0.34 0.13 0.63 Other 0.24 1.31 1.31 0.70 1.13 0.42 0.24 0.51 2.06

responses of *no idea* are excluded in this table; b the same coefficient is computed within other perceptions

1 The R2 value of 0.406 should not be used to judge fitness of a model. The fact that R2 never decreases when any variable is added to a regression makes it a poor tool for deciding whether one or several variables should be added to a model. Low R2s in regression equations are not uncommon, especially for cross-sectional analysis. Thus using R2 as the main gauge of

. The null hypothesis is that each population mean of CVs is not

0.02 0.39 0.23 0.23 0.09 0.00 0.16 0.05 0.27

0.06 0.61 0.55 0.06 0.56 0.04 0.02 0.26 0.71

**7.4.1 Significance of demographic effects**

application of SPSS 13.0 [14]1

Concern about food safety

Best way to control agropollution

Note: a

Software: SPSS 13.0

**Table 7.4** T-values of the discrepancy of response with regard to different characteristics

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

As shown in Table 7.4, education level of the respondent (*d*5) and annual household income (*d*9) are significant, despite the differences in significance level, in capturing respondents' discrepancy on options towards all the 10 types of perceptions. Meanwhile, all the other seven demographic variables are measured as significant in identifying discrepancies among most of the perceptions. Thus, in our questionnaire, the adoption and classification of the demoȬ graphic variables are rational, in showcasing the diversified perceptions among surveyed consumers.


**Table 7.5** Percentages of responses to the major sources of information

#### **7.4.2 Effect on the overall awareness**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Based on the results of the T-test in Table 7.4, it makes sense to conduct further analysis on consumers' perception, taking into account the significant effects of demographic characterȬ istics. Nevertheless, to capture the major determinants of consumers' perceptions, only relationships at the level of 0.01 are included below.

**1.** Most used information source is significantly determined by the following variables. i) Gender of the respondents (*Gender*), where males get more information from the tradiȬ tional media of radio, television, newspapers and magazines, while females are relying more on information from the Internet and relatives. ii) Age of the respondent (*d*2) is found as being positively correlated with the use of radio and television, and negatively correlated with the Internet, respectively. iii) For education level of the respondent (*d*5), negative relationships are found with the use of radio and television, while there is a positive correlation with the Internet. iv) Annual household income (*d*9) is measured as being negatively correlated with radio and television, while positively correlated with the Internet, similarly (Table 7.5). These findings are in line with the reality that the Internet is usually used for gathering information by young people with more leisure time, who demand a fashionable lifestyle and who tend to be better educated. However, the traditional media are still important in affecting consumers' perception and behaviour.

**2.** Most reliable information source. i) On the significant variable of *Gender*, 40.4% of male respondents answer as being convinced by information released by the government in the first place, while 37.3% of the females believe experts to be the most reliable source of information concerning food safety. ii) In terms of the *Age* variable, consumers aged 50-59 are sampled as having greatest faith in government and relatives, while those aged 40-49 choose the experts. In addition, although proportioned only 10.1%, the 20-29 aged consumers responded as having most faith in the Internet. iii) Analysing from the perspective of employment, self-employed consumers and those working for the governȬ ment are identified as possessing the top proportions of those who believe in the inforȬ mation issued by the government and experts, respectively. iv) With regard to the impacts of education, consumers with middle-level educational background show most reliance on government, while consumers who have received higher levels of education have most confidence in the information given by experts. v) As to the effects of income, a negative relationship is found with the percentage of reliance on government. Meanwhile, households with the total annual income over 500,000 yuan have the largest proportion of trust in experts in related fields (Table 7.6).

#### **7.4.3 Effect on the perceptions of risk management**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**7.4.2 Effect on the overall awareness**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Primary and less**

**Junior middle**

**High school**

Commercial advertisement

Commercial advertisement

relationships at the level of 0.01 are included below.

**Table 7.5** Percentages of responses to the major sources of information

Based on the results of the T-test in Table 7.4, it makes sense to conduct further analysis on consumers' perception, taking into account the significant effects of demographic characterȬ istics. Nevertheless, to capture the major determinants of consumers' perceptions, only

**Gender (d1) Age of the respondent (d2)**

Radio & television 50.0 52.2 26.7 45.8 45.7 56.8 66.0 80.8 Newspaper & magazine 8.3 9.4 3.3 8.5 10.9 12.6 3.8 0.0 Internet 27.9 24.6 60.0 33.1 29.5 15.3 15.1 7.7 Relatives 5.2 4.4 0.0 3.5 6.2 3.6 9.4 7.7 Experts 2.1 3.0 6.7 2.1 2.3 2.7 1.9 0.0

114 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Package information 4.5 3.9 3.3 4.9 3.1 5.4 3.8 3.8 Other 0.7 1.0 0.0 1.4 0.8 0.0 0.0 0.0 #Valid N 290 203 30 142 129 111 53 26

> **Underg raduate**

**Female Male <20 20-29 30-39 40-49 50-59 60**Ɖ

1.4 1.5 0.0 0.7 1.6 3.6 0.0 0.0

**Education level of the respondent (d5) Annual household income (1000 yuan) (d9)**

**<35**

0.0 2.2 3.6 0.5 0.0 1.2 2.2 1.9 0.0 0.0 0.0

**35 -70**

**70 -150**

**150 -300**

**300 -500**

**500**Ɖ

**Postgraduate**

Radio & television 60.0 67.4 58.3 47.5 38.4 63.9 52.5 47.8 44.3 50.0 25.0 Newspaper & magazine 0.0 4.3 8.6 10.5 8.2 6.0 10.1 11.5 6.3 5.6 0.0 Internet 10.0 2.2 18.0 33.3 37.0 19.3 23.7 26.1 34.2 33.3 75.0 Relatives 20.0 8.7 5.0 2.7 6.8 3.6 4.3 5.7 6.3 5.6 0.0 Experts 0.0 2.2 1.4 2.3 5.5 0.0 2.2 3.2 2.5 5.6 0.0

Package information 10.0 10.9 5.0 2.3 4.1 4.8 4.3 3.8 6.3 0.0 0.0 Other 0.0 2.2 0.0 0.9 0.0 1.2 0.7 0.0 0.0 0.0 0.0 #Valid N 10 46 139 219 73 83 139 157 79 18 4

**1.** Most used information source is significantly determined by the following variables. i) Gender of the respondents (*Gender*), where males get more information from the tradiȬ tional media of radio, television, newspapers and magazines, while females are relying more on information from the Internet and relatives. ii) Age of the respondent (*d*2) is found


often buy food outside and intake it by themselves. Thus, their food safety is drawing significant attention from both the family and society. For the respondents with students in their family, the largest proportion of 37.6% supported the perfecting of related legislation, which is larger than that of negative respondents. Meanwhile, more consumers without students in the family answered in favour of extending environment-friendly technologies (29.8%) and strengthening the penalty on behaviours undermining the agricultural environment (21.6%), being 5.9% and 4.5% higher than that of the consumers with a student family member, respectively (Table 7.8).


**Table 7.6** Percentages of responses to the most reliable sources of information


Source: field survey by the authors

often buy food outside and intake it by themselves. Thus, their food safety is drawing significant attention from both the family and society. For the respondents with students in their family, the largest proportion of 37.6% supported the perfecting of related legislation, which is larger than that of negative respondents. Meanwhile, more consumers without students in the family answered in favour of extending environment-friendly technologies (29.8%) and strengthening the penalty on behaviours undermining the agricultural environment (21.6%), being 5.9% and 4.5% higher than that of the consumers

**Gender (d1) Age of the respondent (d2) Employment of the respondent (d3)**

5.1 6.7 0.0 6.0 5.4 8.1 3.7 7.7 6.5 6.8 6.3 2.9 11.8 3.8

**Education level of the respondent (d5) Annual household income (1000 yuan) (d9)**

0.0 4.4 5.1 5.8 9.5 4.8 4.3 6.9 7.4 5.6 25.0

**<35**

**35 -70**

**70 -150**

**150 -300**

**300 -500**

**500**Ɖ

**Postgraduate**

**60**Ɖ **Gov.**

**Public ins.**

**Enterprise**

**Self emplo yed**

**Jobless**

**Other**

**50 -59**

with a student family member, respectively (Table 7.8).

**20 -29**

**30 -39**

116 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**40 -49**

Government 30.8 40.4 37.9 25.5 37.2 40.5 40.7 38.5 35.5 34.2 33.6 42.6 35.3 32.1 Experts 37.7 27.9 34.5 33.6 34.1 37.8 24.1 30.8 38.7 41.0 33.6 27.9 26.5 29.2 Internet 6.8 5.3 3.4 10.1 5.4 0.9 9.3 7.7 0.0 3.4 4.9 10.3 2.9 11.3 Relatives 10.3 11.1 10.3 10.1 9.3 9.0 18.5 11.5 16.1 7.7 9.1 10.3 17.6 12.3 Producers & vendors 2.4 1.9 3.4 4.0 1.6 0.9 0.0 3.8 0.0 0.9 1.4 4.4 2.9 3.8

None 5.8 5.8 10.3 9.4 6.2 1.8 1.9 0.0 3.2 5.1 9.1 1.5 2.9 6.6 Other 1.0 1.0 0.0 1.3 0.8 0.9 1.9 0.0 0.0 0.9 2.1 0.0 0.0 0.9 #Valid N 292 208 29 149 129 111 54 26 31 117 143 68 34 106

> **Underg raduate**

Government 30.0 55.6 42.0 30.5 25.7 41.7 37.1 33.8 32.1 22.2 0.0 Experts 20.0 20.0 29.0 38.1 36.5 21.4 40.0 34.4 35.8 22.2 50.0 Internet 20.0 4.4 2.9 6.6 9.5 6.0 7.1 5.6 3.7 22.2 0.0 Relatives 20.0 11.1 14.5 7.5 10.8 13.1 6.4 10.6 9.9 27.8 0.0 Producers & vendors 10.0 4.4 1.4 1.8 2.7 6.0 1.4 0.6 3.7 0.0 0.0

None 0.0 0.0 3.6 8.4 5.4 7.1 2.1 6.3 7.4 0.0 25.0 Other 0.0 0.0 1.4 1.3 0.0 0.0 1.4 1.9 0.0 0.0 0.0 #Valid N 10 45 138 226 74 84 140 160 81 18 4

**F. M. <20**

**Primary and less**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Junior middle**

**Table 7.6** Percentages of responses to the most reliable sources of information

**High school**

Commercial advertisement

Commercial advertisement **Table 7.7** Percentages of responses to the major responsibility bearer of agro-pollution


Source: field survey by the authors

**Table 7.8** Percentages of responses to the best ways to control agro-pollution

#### **7.5 Conclusions and recommendations**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **7.5.1 Major conclusions**

Based on the survey of 512 respondents from Beijing and Shanghai, this chapter studies consumer perceptions on food safety. (1) Analyses on the overall awareness indicate that most of the interviewed consumers are concerned about food safety; more than half of the respondȬ ents think the current situation of food supply is risky or very risky; the television and Internet are the most important sources of information on food safety, while information from the government and experts are deemed most reliable. (2) In terms of perceptions on the major affecting factors, the significance of environmental protection in ensuring food safety has been accepted by more than 90% of the surveyed consumers. Industrial and agricultural pollution are thought to be the top threat to food safety by almost 90% of respondents, while agrochemicals and industrial pollutant are the top source of agro-pollution, at more than 80% of the sampled consumers. Processing and agricultural production are the most risky procedure or stage, with the proportion of 46.3% and 42.6%, respectively. (3) As to the risk management of food safety, most of the respondents think that the government and producers of the materials inputted to agriculture, especially those of agro-chemicals, should take the responȬ sibilities for agro-pollution at first. While 60% of the consumers believe perfecting the legislaȬ tion and extending environmental technology are the best ways to control agro-pollution.

The T-test reveals that all the nine demographic variables are significant in identifying discrepancies among most of the perceptions. (1) With respect to the most important inforȬ mation source, the Internet is more used for gathering information by young people with more leisure time, demands for a fashionable lifestyle and who tend to be better educated. MeanȬ while, the traditional media of television, newspapers and magazines are still important in affecting consumer perception and behaviour. (2) From the perspective of most reliable information source, the government is supported more by male, older and self-employed consumers, while experts are trusted by more female, middle-aged, public servants and rich consumers. (3) The government is attributed taking the responsibility for agro-pollution significantly by those with an annual income over 300,000 yuan. (4) Whether there is primary or middle school student in a household is measured as a high-significantly factor affecting respondents' attitude toward the best ways to control agro-pollution.

#### **7.5.2 Policy recommendations**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

(1) Since the government is chosen as the most reliable source of information and the major responsibility bearer of agro-pollution by most respondents, its supervisory obligations on food safety should be strengthened. Under the unified coordination and leadership of the National Food Safety Committee2 , responsibility for each department needs to be clarified, hence improving the supervisory efficiency through more initiatives and regular investigaȬ tions. (2) Considering the fact that most of the consumers responded as concerned and worried about food safety, the results of the supervisions should be disclosed to the public promptly. (3) Based on the findings of most risky procedure or stage, key sectors of food safety superviȬ sion include: appropriate sterilization, additives and labelling in the processing operations of food manufacturers, proper manufacturing and application of agro-chemicals. (4) As to the Internet which is found to be widely used but no so trusted, further inspections are needed to improve its reliability in releasing food safety information. On the other hand, the government should make full use of mass media in collecting information and communicating with the public. (5) Being one of the best ways to control agro-pollution, relevant legislations are necessary, including further provisions in the existing acts and formulation of an Act of Agricultural Pollution Prevention, etc. (6) Due to the intimate relation between the environȬ ment and agro-pollution, and the functions in controlling agro-pollution, extension of environment-friendly techniques needs to be accelerated.

<sup>2</sup> According to provision on the Act of Food Safety (2009), the National Food Safety Committee was established in the State Council on February 9 2010. As the top administrative agency of food safety, this committee is headed by the outstanding vice-premier and other two vice-premiers, and composed of 15 concerned ministries.

#### **7.5.3 Further discussion**

are thought to be the top threat to food safety by almost 90% of respondents, while agrochemicals and industrial pollutant are the top source of agro-pollution, at more than 80% of the sampled consumers. Processing and agricultural production are the most risky procedure or stage, with the proportion of 46.3% and 42.6%, respectively. (3) As to the risk management of food safety, most of the respondents think that the government and producers of the materials inputted to agriculture, especially those of agro-chemicals, should take the responȬ sibilities for agro-pollution at first. While 60% of the consumers believe perfecting the legislaȬ tion and extending environmental technology are the best ways to control agro-pollution.

118 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

The T-test reveals that all the nine demographic variables are significant in identifying discrepancies among most of the perceptions. (1) With respect to the most important inforȬ mation source, the Internet is more used for gathering information by young people with more leisure time, demands for a fashionable lifestyle and who tend to be better educated. MeanȬ while, the traditional media of television, newspapers and magazines are still important in affecting consumer perception and behaviour. (2) From the perspective of most reliable information source, the government is supported more by male, older and self-employed consumers, while experts are trusted by more female, middle-aged, public servants and rich consumers. (3) The government is attributed taking the responsibility for agro-pollution significantly by those with an annual income over 300,000 yuan. (4) Whether there is primary or middle school student in a household is measured as a high-significantly factor affecting

(1) Since the government is chosen as the most reliable source of information and the major responsibility bearer of agro-pollution by most respondents, its supervisory obligations on food safety should be strengthened. Under the unified coordination and leadership of the

hence improving the supervisory efficiency through more initiatives and regular investigaȬ tions. (2) Considering the fact that most of the consumers responded as concerned and worried about food safety, the results of the supervisions should be disclosed to the public promptly. (3) Based on the findings of most risky procedure or stage, key sectors of food safety superviȬ sion include: appropriate sterilization, additives and labelling in the processing operations of food manufacturers, proper manufacturing and application of agro-chemicals. (4) As to the Internet which is found to be widely used but no so trusted, further inspections are needed to improve its reliability in releasing food safety information. On the other hand, the government should make full use of mass media in collecting information and communicating with the public. (5) Being one of the best ways to control agro-pollution, relevant legislations are necessary, including further provisions in the existing acts and formulation of an Act of Agricultural Pollution Prevention, etc. (6) Due to the intimate relation between the environȬ ment and agro-pollution, and the functions in controlling agro-pollution, extension of

2 According to provision on the Act of Food Safety (2009), the National Food Safety Committee was established in the State Council on February 9 2010. As the top administrative agency of food safety, this committee is headed by the

outstanding vice-premier and other two vice-premiers, and composed of 15 concerned ministries.

, responsibility for each department needs to be clarified,

respondents' attitude toward the best ways to control agro-pollution.

environment-friendly techniques needs to be accelerated.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**7.5.2 Policy recommendations**

National Food Safety Committee2

Food safety will continue to be a public concern in China. In addition to the increased demand for healthy and safe foods with the rapidly growing economy, it can be attributed to the frequent occurrence of scandals. As we write this manuscript April 2012, large volumes of jellies and medicine capsules have been found to contain excessive amounts of chromium. In this wide-ranging incident, some famous companies are even involved, and the major substance concerned is suspected of being illegally added industrial gelatin. For the governȬ ment and domestic food industry, they still have to strive for food safety. Meanwhile, a variety of topics are open to academic research, in terms of exploring technologies and countermeasȬ ures to ensure food safety. For instance, comparative study on behaviours and perceptions between producer and consumers, on different foods or cultures, are beneficial for the administration of food safety.

#### **References**



**Chapter 8**

### **Consumers' Risk Awareness and Willingness to Pay for Certified Food**

Hui Zhou and Teruaki Nanseki

**8.1 Introduction**

[10] Han Q., Zhou H., Nanseki T., Wang J. The consistency of consumer's stated preferȬ ence and revealed preference: evidence from agricultural product market in China, J.

[11] Zhou J., Jin S. Safety of vegetables and the use of pesticides by farmers in China: EviȬ

[12] Li D., Nanseki T., Takeuchi S., Song M., Chen T., Zhou H. Farmers' behaviors, perȬ ceptions and determinants of pesticides application in China: evidence from six eastȬ

[13] Dellios R., Yang X., Yilmaz N. K. Food safety and the role of the government: impliȬ

[14] Steiner C. F., Darcy-Hall T. L., Dorn N. J., Garcia E. A., Mittelbach G. G, Wojdak J. M. The influence of consumer diversity and indirect facilitation on trophic level biomass

[15] Gacula M. Jr., Rutenbeck S. Sample size in consumer test and descriptive analysis. J.

[16] Bruin J. Newtest: Command To Compute New Test. UCLA: Academic Technology Services. Statistical Consulting Group 2006: http://www.ats.ucla.edu/stat/spss/

[17] Wooldridge J. M. Introductory Econometrics: A Modern Approach (2nd Edition).

ern provincial-level regions. J. Fac. Agr., Kyushu Univ. 2012; 57 (1): 255-263

dence from Zhejiang Province. Food Control 2008; 20: 1043-1048.

120 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

cations for CSR policies in China. i Business 2009; 1: 75-84

South-Western Thomson Learning, Mason 2003: 41, 81.

Fac. Agr., Kyushu Univ. 2012; 57 (1): 227-234.

and stability. OIKOS 2005; 110: 556-566

Sens. Stud. 2006; 21 (2): 129-145

output/Spss\_ttest.htm

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **8.1.1 Development of the dairy industry**

The dairy industry has huge potential in China. The production and consumption of milk in China have increased dramatically, especially since 2000. However, recent food safety problems which have occurred in the livestock sector in China have negatively affected consumers' confidence in purchasing foods. Food safety problems in the dairy industry have created lack of confidence among the public in buying dairy products.

New approaches to ensure food safety, such as traceability system, good agricultural practices (GAP), hazard analysis and critical control point (HACCP), are of concern to the government, producers and even consumers. In Japan, Aizaki has evaluated Japanese consumers' willingness to pay for GAP certificated milk [1]. The traceability system could be a new way to respect consumers' right to choose safe and quality food, and to provide production information to consumers which may help improve consumers' confidence in the food they consume. Basically, one of the functions of a traceability system is to monitor food producers in order to avoid food safety problems such as misuse of veterinary medicines and so on. Therefore, in this chapter, the traceability system is viewed as a main example, to examine consumers' awareness of food risk and willingness to pay for food safety certification.

At present, the traceability system can be found in some supermarkets and hypermarkets in China and it is mainly used on vegetables (Figure 8.1). The traceability system is a new thing to Chinese consumers and consumer attitudes toward the traceability system are not as yet clear. In this chapter, the main research objectives are to examine consumers' purchasing behaviour on milk, to study consumers' response to food safety issues and to investigate the consumers' awareness of the traceability system.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

© 2013 Zhou and Nanseki; licensee InTech. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Zhou and Nanseki; licensee InTech. This is a paper distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Source: photographs by the authors

**Figure 8.1** Food traceability system in Beijing

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **8.1.2 Food safety certification**

Food safety and food quality have increasingly come to the forefront of consumer concerns, industry strategies and government policy initiatives. In recent years, a number of serious food safety problems within China have negatively affected consumer confidence on both domestic and exported food products.

New approaches to ensure food safety are becoming more integrated, in order to uphold consumers' rights to know about the commodities they consume. A traceability system can be one efficient way to advocate the prompt and accurate movement of information covering all processes in the food supply, so that the public have confidence to choose and consume food. Basically, a traceability system is a tool to monitor food producers in order to avoid food safety problems such as illegal additives, misuse of veterinary drugs, occurrence of animal diseases, etc.

The traceability system is still new to Chinese consumers, whose attitudes towards the traceability system have been examined. In this chapter, the production history in the traceability system is mainly studied, due to the findings that most consumers are concerned about the production process in both Japan and China [2]. The main objecȬ tives are to study consumers' marginal willingness to pay (MWTP) on the information that the traceability system provides, and to determine factors affecting consumers' willingȬ ness to pay for a traceability system.

#### **8.2 Literature review**

As a hot topic, traceability system has been studied for several years. Several research groups have studied the global requirements and impacts of a traceability system [3, 4]. Meanwhile, some other authors, such as Sahin E., et al [5], considered information technology (IT) to be the fundamental tool to assist bringing about revolutionary changes in product traceability.

There is much research mainly focusing on consumer behaviour-consumer opinions on the traceability system applied to the meat supply chain in Europe, America and Canada [6]. US consumers were willing to pay \$0.23 more on beef hamburgers with traceability assurance, \$0.50 to add assurances on animal treatment, \$0.63 to add extra assurances of food safety, and \$1.06 to upgrade the sandwich to one which contains all three upgrades. For pork, the same respective upgrades were valued on the average at \$0.50, \$0.53, \$0.59 and \$1.1, respectively [6]. The consumers in Canada were willing to pay non-trivial amounts (around 0.33\$) for beef with a traceability assurance. However, the quality assurance with respect to food safety and onfarm production methods for beef were more valuable to consumers than simple traceability system assurance. Bundling traceability with both of these quality assurances yielded a \$0.83 premium over the traceability-only beef [7].

However, in China, there has only been limited research in terms of economics. After Bird Flu occurred in China, the consumers in Beijing were willing to pay 0.80-1.53 yuan/kg more for chicken with a traceability label if they had more knowledge about the traceability label while the price of chicken was 9.40 yuan/kg [8].

#### **8.3 Data collection and research method**

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **8.3.1 Research areas**

Source: photographs by the authors

**8.1.2 Food safety certification**

and exported food products.

occurrence of animal diseases, etc.

ness to pay for a traceability system.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Figure 8.1** Food traceability system in Beijing

Food safety and food quality have increasingly come to the forefront of consumer concerns, industry strategies and government policy initiatives. In recent years, a number of serious food safety problems within China have negatively affected consumer confidence on both domestic

122 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

New approaches to ensure food safety are becoming more integrated, in order to uphold consumers' rights to know about the commodities they consume. A traceability system can be one efficient way to advocate the prompt and accurate movement of information covering all processes in the food supply, so that the public have confidence to choose and consume food. Basically, a traceability system is a tool to monitor food producers in order to avoid food safety problems such as illegal additives, misuse of veterinary drugs,

The traceability system is still new to Chinese consumers, whose attitudes towards the traceability system have been examined. In this chapter, the production history in the traceability system is mainly studied, due to the findings that most consumers are concerned about the production process in both Japan and China [2]. The main objecȬ tives are to study consumers' marginal willingness to pay (MWTP) on the information that the traceability system provides, and to determine factors affecting consumers' willingȬ To examine consumers' attitude toward the traceability system, an interview survey was conducted from September to October 2008 in Beijing, with 209 valid samples collected. Meanwhile, data from another field survey in Beijing, July 2008, conducted by Nanseki et al. (2008) [9] is also used for analysis in this chapter. In this survey, 214 consumers were interȬ viewed and their answers collected as valid samples.

The total population of Beijing amounted to 17.15 million by 2010, within which some 11.9 million are registered permanent residents, and around five million are temporary residents. Beijing consists of 16 districts (Figure 8.2), and based on the location and socio-economic features, they are categorized into four regions (Table 8.1), according to the capital planning of main functional regions released in July, 2012 [10]. As the Capital and one of the biggest cities in China, Beijing has one of the highest averages of milk consumption zones in China (Figure 8.3). In particular, thanks to the Olympic Games held in the summer of 2008, the traceability system has been applied in some supermarkets in Beijing. Therefore, we chose Beijing as the research area in this chapter.

Source: revision based on http://www.chinahighlights.com/beijing/map.htm

Source: China Ministry of Agriculture (2012) [11]

**Figure 8.3** Consumption of dairy products in some regions


**Table 8.1** Population distribution in Beijing (2010) (Unit: 10000 persons)

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

#### **8.3.2 Survey and data**

Source: China Ministry of Agriculture (2012) [11]

**Figure 8.2** Location and districts of Beijing

**Figure 8.3** Consumption of dairy products in some regions

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Source: revision based on http://www.chinahighlights.com/beijing/map.htm

124 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

This is a chapter, in order to research consumers' attitudes towards milk purchasing, and milk safety and quality, an interview survey was conducted in September and October 2008. The study areas cover seven urban districts, which share 83% of the population of Beijing. AcȬ cording to the population allocation of each district, we choose a certain number respondents randomly from each district. The survey was carried out in the four main supermarkets or hypermarkets. According to the population distribution, also because of limited of transporȬ tation, seven districts which include 60% of the population were chosen to carry out the survey (Table 8.2).



Note: after the district adjustment of 2010, Xuanwu and Chongwen were merged into Xicheng and Dongcheng Districts, respectively.

Source: consumer survey in September, 2008

**Table 8.2** Data distribution in each district

To understand the factors affecting consumers' purchase of milk and other dairy products, it is significant to interview not only the dairy producers, but also the policy-makers.


Source: consumer survey in September, 2008

**Table 8.3** Socio-economic characteristics across treatments

In this survey, around 250 respondents are interviewed, but only 209 questionnaires are completely finished. The respondents are Beijing local citizens or people who have lived in Beijing longer than three years. Most respondents are female due to the fact that shopping is mainly done by housewives and in this survey, 69% of the correspondents are females while only 31% are males. The majority of the respondents are confined within the ages of 26-35, 19-25 and 36-45 years old. Additionally, the sample shows that the respondents have quite high levels of education. This is because Beijing is the capital city, where many famous universities are located. Hence in this survey, some 28% of the respondents answered as having college-level educational experience, and 48% of them responded as having attended univerȬ sities or postgraduate schools. Nearly 80% of the respondents have household income less than 10000 yuan per month, about 29% of them have monthly household incomes of 3000-6000 yuan*,* and 25% have monthly household incomes of 6000-10000 yuan on average per month. Around 20% of the respondents have a monthly household income of more than 10000 yuan, while only less than 4% have a household income of less than 1000 yuan per month.

#### **8.3.3 Research method**

**Location N %** Haidian 60 28.7 Chaoyang 53 25.3 Dongcheng 18 8.5 Xicheng 22 10.6 Xuanwu \* 17 8.2 Chongwen \* 10 4.8 Fengtai 29 13.8 Total 209 100.0

126 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

respectively.

Source: consumer survey in September, 2008

**Table 8.2** Data distribution in each district

Source: consumer survey in September, 2008

**Table 8.3** Socio-economic characteristics across treatments

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Note: after the district adjustment of 2010, Xuanwu and Chongwen were merged into Xicheng and Dongcheng Districts,

To understand the factors affecting consumers' purchase of milk and other dairy products, it

Gender Male 31.2 Age Under 18 1.4

Household income <1000 3.8 26-35 29.1

**Category % Category %**

Female 68.8 19-25 23.4

1000-3000 19.6 36-45 20.5 3000-6000 28.7 46-55 15.4 6000-10000 24.8 56-65 7.2 10000-15000 11.4 "/> 66 2.8 15000-20000 4.4 Education Primary school 0.9 "/>20000 3.8 Junior high school 5.3 No answer 2.8 Senior high school 16.7

> College 28.3 University 48.8

is significant to interview not only the dairy producers, but also the policy-makers.

This study also applies the choice modelling (CM) technique in examining which attributes are significant determinants of the values people place on non-market goods, i.e., the traceaȬ bility system. CM or stated preference (SP) that uses the attribute-based technique was first applied by Louviere et al (2000) [12], Louviere et al (1983) [13] and Adamowicz et al (1998) [14]. This technique originated in market research and transport literature, and has recently been applied to the valuation of non-market goods. In this survey, attributes and levels are used to create choice sets using a 33 x6 orthogonal effects design which produced 36 choice sets and were divided into six versions. CM techniques require respondents to compare and select one option out of three in all the choice sets as shown in Table 8.4.


#### **Table 8.4** Choice and set on traceability system

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

The multinomial logit model is used to analyse the data. The options chosen by the respondents in the CM can be modelled in a random utility framework which can be expressed as the sum of the systematic component. The utility obtained by individual *i* from choosing alternative *j* in a choice set which can be expressed as:

$$\mathcal{U}\_{i\circ} = V\_{i\circ} + \mathcal{e}\_i \tag{8-1}$$

where *Vij* denotes the observable portion of the utility and Ή*ij* indicates the error term. This study assumes that the utility for an option (*i*) depends on a vector of its observable attributes (*Z*) and a vector of the socio-economic characteristics of respondents (*S*) as:

$$\mathcal{U}\_{i\dot{\jmath}} = V\_{i\dot{\jmath}}(Z\_{i\dot{\jmath}}, S\_{i\dot{\jmath}}) + \varepsilon\_{i\dot{\jmath}}(Z\_{i\dot{\jmath}}, S\_{i\dot{\jmath}}) \tag{8-2}$$

Option *j* is chosen over alternative *h* of *Uij* > *Uih*. Probability if individual *i* choosing option *j* can be defined as follows:

$$\mathcal{L}\pi\_{ij} = \Pr\left\{V\_{ij} + \varepsilon\_{ij} \ge V\_{ih} + \varepsilon\_{ih} \colon \forall h \in \mathbb{C}\right\},\tag{8-3}$$

where *Ci* is the choice set for individual *i*; *Vij* is a conditional indirect utility function and has a linear form of:

$$\boldsymbol{V}\_{ij} = \boldsymbol{\beta}\_0 + \boldsymbol{\beta}\_1 \mathbf{X}\_1 + \boldsymbol{\beta}\_2 \mathbf{X}\_2 + \dots + \boldsymbol{\beta}\_n \mathbf{X}\_n \tag{8-4}$$

where Ά*<sup>1</sup> -* Ά*<sup>n</sup>* is the vector of coefficient attached to the vector of attributes *X*. While the socioeconomic characteristics impact on attributes, the function has a form:

$$V\_{ij} = ASC + \sum\_{k=1}^{K} \beta\_k Z\_{ik} + \sum\_{k=1}^{K} \sum\_{h=1}^{H} \gamma\_{kh} Z\_{ik} \mathbf{S}\_{ih} \tag{8-5}$$

where Ά and · is parameter, *Z* is the attribute associated with the alternative, *S* represents socioeconomic characteristics.

The marginal value of a change within a single attribute can be represented as a ratio of coefficients as follows:

$$\text{MWTP} = \frac{\beta\_{attribute}}{\beta\_{price}} \tag{8-6}$$

Option C was coded as zero value and alternative specific constants were equal to 1 with either option A and B being selected [15]. In this study the software package LIMDEP 9.0 NLOGIT4.0 was used to estimate the multinomial logit model [16].

#### **8.4 Results and discussion**

*U V ij ij i* = +

128 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

( ,) ( ,) *U VZS ZS ij ij ij ij ij ij ij* = + H

(*Z*) and a vector of the socio-economic characteristics of respondents (*S*) as:

EE

economic characteristics impact on attributes, the function has a form:

*V ASC Z*

can be defined as follows:

a linear form of:

economic characteristics.

coefficients as follows:

S

H

where *Vij* denotes the observable portion of the utility and Ή*ij* indicates the error term. This study assumes that the utility for an option (*i*) depends on a vector of its observable attributes

Option *j* is chosen over alternative *h* of *Uij* > *Uih*. Probability if individual *i* choosing option *j*

HH

0 11 22 ...... *V XX X ij n n* =+ + + +

 E

where *Ci* is the choice set for individual *i*; *Vij* is a conditional indirect utility function and has

where Ά*<sup>1</sup> -* Ά*<sup>n</sup>* is the vector of coefficient attached to the vector of attributes *X*. While the socio-

1 11

where Ά and · is parameter, *Z* is the attribute associated with the alternative, *S* represents socio-

The marginal value of a change within a single attribute can be represented as a ratio of

*attribute price*

E

E

Option C was coded as zero value and alternative specific constants were equal to 1 with either option A and B being selected [15]. In this study the software package LIMDEP 9.0 NLOGIT4.0

= ==

*K KH ij k ik kh ik ih k kh*

E

*MWTP*

was used to estimate the multinomial logit model [16].

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

(8-1)

(8-2)

(8-4)

*ij ij ij ih ih* t Pr{*V V hC* ; , } (8-3)

 E

 J*Z S*

=+ + ¦ ¦¦ (8-5)

<sup>=</sup> (8-6)

#### **8.4.1 Consumers' milk purchasing behaviour and reaction on food safety issues**

The survey was carried out in Beijing in 2008. As one of the highest average milk consumption zones in China, consumers' milk purchasing behaviour was also examined. As the modern supply chain developed, supermarkets became the most popular place for people to buy daily goods. So people in Beijing are used to buying milk and other dairy products in supermarkets. Besides buying at the supermarket, milk delivery services are also common.

Source: consumer survey in September 2008

#### **Figure 8.4** Main milk purchasing place

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

For many families, milk is consumed every day. Some people have milk at breakfast, while others have milk at night before sleeping, or both (about 250 ml -500ml per day). A few people consume milk as their main drink and drink milk very often (more than 500 ml per day). For old people and young children, milk, yogurt and other dairy products are the main source of calcium. Nearly 60% of families are used to consuming more than 250ml of milk per day per person, around 17% of the families are used to having more than 500ml of milk per day. Around 11% of the respondents rarely drink milk everyday (Figure. 8.5).

Source: consumer field survey in September, 2008

**Figure 8.5** Milk consumption per household in a day

However, a few food safety incidents have negatively affected consumers' confidence on the food they consume, for example, the milk powder incident which happened in China in 2008. Before the incident happened, most people felt the milk was safe to consume, but after the incident, the situation totally changed [17] (Figure 8.6).

Source: field survey in July and September, 2008

**Figure 8.6** Consumer's confidence on food safety in 2008

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

After the incident happened, respondents' views on safety become more important than price. Most respondents thought safety is much more important than milk price [9] (Figure 8.7).

Source: field survey in July,2008

Source: consumer field survey in September, 2008

**Figure 8.5** Milk consumption per household in a day

30%

very safe

**Figure 8.6** Consumer's confidence on food safety in 2008

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Source: field survey in July and September, 2008

0% 10% 20% 30% 40% 50% 60%

incident, the situation totally changed [17] (Figure 8.6).

48%

20%

14%

4% 4% 1%

safe neutral not

37%

31%

safe

After the incident happened, respondents' views on safety become more important than price. Most respondents thought safety is much more important than milk price [9] (Figure 8.7).

11%

July

September

not safe at all

However, a few food safety incidents have negatively affected consumers' confidence on the food they consume, for example, the milk powder incident which happened in China in 2008. Before the incident happened, most people felt the milk was safe to consume, but after the

130 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**Figure 8.7** Consumer perceptions on safety and price

After the incident happened, most respondents wanted more heavy supervision of milk production, some people wanted more production information, while some people stopped purchasing milk altogether (Figure 8.8).

**Figure 8.8** Consumers' requests on ensuring food safety

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

According to the survey conducted in July 2008, the respondents thought most milk safety problems happened on farms and in processing factories [16]. Very few respondents thought milk safety problems happened during transportation or at the wholesale stage (Figure 8.9).

**Figure 8.9** Consumers' awareness on the most risky stage in food supply

#### **8.4.2 Consumers' awareness of the traceability system**

As a new system to ensure food safety, the traceability system is only known by a few people. Only around 42% of the respondents had heard about the food traceability system before, and for more than half of the respondents this was the first time they had heard about this system. However, around 87% of the respondents thought the traceability system was necessary to ensure food safety. Even people who had never heard about traceability system before the survey, still felt it was necessary and would be a way to ensure food safety and food quality (Table 8.5).


**Table 8.5** Consumers' awareness of the traceability system

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

This result shows that most of the consumers accept the traceability system, whether or not they had heard of the traceability system. However, there were still around 13% of the respondents who did not approve of the traceability system. Distrusting the information was the main reason for this. Additionally, there were some respondents who did not care about the traceability system. They thought the traceability system was not perfect enough to ensure food safety and food quality. Only a few respondents viewed a possible higher price as the reason they would not choose a traceability system. At present, the traceability system is a hot topic in China; there is some research about consumers' attitude toward traceability systems, but on other products such as pork and beef. This result was similar to research on pork. According to Greene et al. (2002) [16], the reason for respondents not choosing pork with a traceability system was that people were worried about possible false information.



**Table 8.6** Reasons for not choosing the traceability system

Source: field survey in July, 2008

(Table 8.5).

Source: consumer survey in September, 2008

**Table 8.5** Consumers' awareness of the traceability system

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

**Figure 8.9** Consumers' awareness on the most risky stage in food supply

**8.4.2 Consumers' awareness of the traceability system**

As a new system to ensure food safety, the traceability system is only known by a few people. Only around 42% of the respondents had heard about the food traceability system before, and for more than half of the respondents this was the first time they had heard about this system. However, around 87% of the respondents thought the traceability system was necessary to ensure food safety. Even people who had never heard about traceability system before the survey, still felt it was necessary and would be a way to ensure food safety and food quality

132 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

Heard about traceability system 78 11 89 87.6

Not heard about traceability system 104 16 120 86.6

Total 182 27 209 87.0

This result shows that most of the consumers accept the traceability system, whether or not they had heard of the traceability system. However, there were still around 13% of the

% of heard about traceability system 42.8 40.7 42.5

**Necessary Unnecessary Total % of necessary**

Source: consumer survey in September, 2008

\* Note: we set the basic price of milk as 1.70 yuan/250 ml with the assumed constant price

**Table 8.7** Amounts of WTP for milk with the traceability system

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/

Of 182 respondents, 30% would accept the traceability system if the price did not change. Around 70% of the respondents would accept an increased milk price with the traceability system. However, most of them only accepted a limited price increase. Only approximately 4% of the respondents were willing to pay a high price for the milk with a traceability system (three or four times more expensive than the original price of normal milk). This result showed that consumers were willing to bear part of the cost of a traceability system, but to a limited extent, and most of the cost should be borne by the government and producers. The governȬ ment should take the responsibility to ensure the basic food safety (Table 8.7).

#### **8.4.3 Consumers' willingness to pay for certified safety food**

The respondents are mainly asked about their attitude toward the traceability system. Table 8.9 shows both the attribute variables and non-attribute variables used in choice modelling. Attribute variables are the information that the traceability system can provide to consumers, and non-attribute variables are mainly socio-economic information. The model estimated the price attributes which interacted with socio-economic characteristics and estimated how these socio-economic characteristics impact on price attribute. The variables are described as AGE\*PRICE, GENDER\*PRICE, and so on.

Table 8.10 showed the result of the estimation MWTP of information that the traceability system provides and the estimation of socio-economic characteristics which impact upon price attribute. According to Table 8.4, respondents preferred all the information except processing information with pictures. The reason may be that people are more familiar with these famous processing companies, they can get much information about these processing enterprises through many channels; however, most people do not have a clear idea about farm informaȬ tion, breeding information, feed information and animal medicine use which are all very important and affect food safety, especially feed use and animal medicine use.

With regard to farm information, consumers have a higher willingness to pay (WTP) on information with pictures than only information, that being 2.28 yuan and 2.03 yuan for 250ml of milk while the basic price of milk is 1.70 yuan. For consumers, the more information, the better. Additionally, consumers are willing to pay about 3.69 yuan for 250ml of milk with a traceability system which includes antibiotics records and only 2.95 yuan for all animal medicine records. This was a very high marginal willingness to pay, especially on antibiotics' usage. It was more than twice the original price. However, the survey was carried out right after the milk powder incident happened, so the result might have been biased and over estimated the real willingness to pay. Consumers are more concerned about animal medicine use, especially antibiotics' use. Processing factory information is viewed as least preferred, while processing factory information with pictures is not significant in the statistics. When asked about processing factory information, consumers have a lower WTP than other attribȬ utes at only 0.87 yuan*,* while the processing information with pictures is not significant in the statistics. The reason for this might be that consumers or the respondents already have enough information on processing factories, especially those famous brands compared with other information. They can get this kind of information through many channels such as news, the Internet, or see the factory for themselves. They might be more interested in some introductions to these processing factories, rather than pictures and people do not prefer the attribute of price through the coefficient.


**Table 8.8** Explanation of attribute and non-attribute variables in the choice models

4% of the respondents were willing to pay a high price for the milk with a traceability system (three or four times more expensive than the original price of normal milk). This result showed that consumers were willing to bear part of the cost of a traceability system, but to a limited extent, and most of the cost should be borne by the government and producers. The governȬ

The respondents are mainly asked about their attitude toward the traceability system. Table 8.9 shows both the attribute variables and non-attribute variables used in choice modelling. Attribute variables are the information that the traceability system can provide to consumers, and non-attribute variables are mainly socio-economic information. The model estimated the price attributes which interacted with socio-economic characteristics and estimated how these socio-economic characteristics impact on price attribute. The variables are described as

Table 8.10 showed the result of the estimation MWTP of information that the traceability system provides and the estimation of socio-economic characteristics which impact upon price attribute. According to Table 8.4, respondents preferred all the information except processing information with pictures. The reason may be that people are more familiar with these famous processing companies, they can get much information about these processing enterprises through many channels; however, most people do not have a clear idea about farm informaȬ tion, breeding information, feed information and animal medicine use which are all very

With regard to farm information, consumers have a higher willingness to pay (WTP) on information with pictures than only information, that being 2.28 yuan and 2.03 yuan for 250ml of milk while the basic price of milk is 1.70 yuan. For consumers, the more information, the better. Additionally, consumers are willing to pay about 3.69 yuan for 250ml of milk with a traceability system which includes antibiotics records and only 2.95 yuan for all animal medicine records. This was a very high marginal willingness to pay, especially on antibiotics' usage. It was more than twice the original price. However, the survey was carried out right after the milk powder incident happened, so the result might have been biased and over estimated the real willingness to pay. Consumers are more concerned about animal medicine use, especially antibiotics' use. Processing factory information is viewed as least preferred, while processing factory information with pictures is not significant in the statistics. When asked about processing factory information, consumers have a lower WTP than other attribȬ utes at only 0.87 yuan*,* while the processing information with pictures is not significant in the statistics. The reason for this might be that consumers or the respondents already have enough information on processing factories, especially those famous brands compared with other information. They can get this kind of information through many channels such as news, the Internet, or see the factory for themselves. They might be more interested in some introductions to these processing factories, rather than pictures and people do not prefer the attribute of price

important and affect food safety, especially feed use and animal medicine use.

ment should take the responsibility to ensure the basic food safety (Table 8.7).

134 Food Safety and the Agro-Environment in China: The Perceptions and Behaviours of Farmers and Consumers

**8.4.3 Consumers' willingness to pay for certified safety food**

AGE\*PRICE, GENDER\*PRICE, and so on.

through the coefficient.

AH Formatter V6.0 MR4a (Evaluation) http://www.antennahouse.com/


**Table 8.9** Explanation non-attribute variables in Choice models

Table 8.10 also shows the results of the impacts of socio-economic characteristics on price attribute (WTP). Only AGE\*PRICE and EDU\*PRICE are in 1% significant, INȬ COME\*PRICE is in 10% significant. Other variables are not significant in the statistics. AGE\*PRICE is negative, young people find it easier to accept the traceability system and are more willing to pay for a traceability system. EDU\*PRICE is positive, higher educatȬ ed people find it easier to accept the traceability system and have higher WTP. INȬ COME\*PRICE is positive, higher income people are willing to pay more money, but not very much. This may imply that income only impacts a little on WTP for the traceability system. No matter their level of income, people are concerned about food safety and the traceability system, and they need safe food whether rich or poor.


Source: consumer field survey 2008
