EMERGING MARKET S FOR U.S. PORK IN CHINA By Maolong Chen A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Agricultural, Food, and Resource Economics Master of Science 2015 ABSTRACT EMERGING MARKET S FOR U.S. PORK IN CHINA By Maolong Chen Economic globalization has opened up international markets for U.S. food products, especially new markets in emerging economies. While opportunities for increased demand for U.S. pork in China look promising, little is known about this emerging market. The objective of this thesis is to provide a n analysis of Chinese consumer demand for pork and evaluate the potential for U.S. pork in China. Two essays are developed to achieve this goal. The first essay explores the determinants of Chinese consumer perception of pork quality and identif ies the relationship between Chinese consumer perception of pork quality and their attit ude towards pork attributes including credence, experience and search attributes. The empirical model is applied to survey data from consumers in three major cities in mainland China and Hong Kong collected in 2014 summer , and is estimated using a seemingl y unrelated regression method. After investigating consumer's perception of pork quality in the first essay, the second essay aims to provide a closer look at factors that influence consumer purchasing decisions of pork from China and developed countries. A choice experiment approach is employed to examine consumer's willingness to pay (WTP) for select quality attributes in pork. Marketing and agribusiness implications of the findings are discussed. iii ACKNOWLEDGEMENTS First of all, I would like to express my sincere gratitude and thanks to my major professor, Dr. David L. Ortega , who lead me to the academic world, provide d me with constructive advice in my research, and offered support in my studies. Second, I want to thank our collaborators for their assistance with the field work in China. Special appreciation goes to Dr. H. Holly Wang and Dr. Laping Wu, who contribute d much to the success of this project. Also, I want to thank my committee members, Dr. Fran k Lupi and Dr. Chris topher Wolf, for their continued guidance and patience in every aspect of this work; all my MSU colleagues and friends who are very supportive and make m y life enjoy able at Michigan State. Furthermore, I would especially like to thank my wife Chaoran Hu, my family and friends in China for their support and encouragement . iv TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ......................... vi LIST OF FIGURES ................................ ................................ ................................ ...................... vii KEY TO ABBREVIATIONS ................................ ................................ ................................ ...... viii 1. Introduction ................................ ................................ ................................ ................................ . 1 1.1 The U.S. pork market ................................ ................................ ................................ ............ 1 1.2 China's domestic pork economy ................................ ................................ ........................... 2 1.2.1 Pork supply in China ................................ ................................ ................................ ...... 2 1.2.2 Domestic pork demand in China ................................ ................................ .................... 7 1.2.3 China as a pork importer ................................ ................................ .............................. 11 1.3 Thesis objective ................................ ................................ ................................ .................. 13 2. Determinants of consumers' preference for pork ................................ ................................ ...... 15 2.1 Definition and classification of pork attributes ................................ ................................ ... 15 2 .2 Differences in credence attributes between Chinese and U.S. pork ................................ ... 16 2.2.1 Different policies and regulations on food safety ................................ ........................ 16 2.2.2 Increasing awareness of animal welfare in China ................................ ........................ 20 2.2.3 Rising concerns on the environmental impacts of China's pork industry .................... 21 2.3 Difference in experience and search attributes of Chinese and U.S. pork .......................... 23 3. Data ................................ ................................ ................................ ................................ ........... 25 3.1 Data sources ................................ ................................ ................................ ........................ 25 3.2 Data description ................................ ................................ ................................ .................. 26 4. Chinese consumers' perception of pork quality ................................ ................................ ........ 31 4.1 I ntroduction ................................ ................................ ................................ ......................... 31 4.2 Methodology ................................ ................................ ................................ ....................... 35 4.3 Data summary ................................ ................................ ................................ ..................... 40 4.4 Results ................................ ................................ ................................ ................................ . 42 4.4.1 Pork Credence Attributes ................................ ................................ ............................. 45 4.4.2 Pork Search and E xperience Attributes ................................ ................................ ....... 46 4.4.3 Patriotism Effect ................................ ................................ ................................ .......... 47 4.4.4 Demographic Impacts ................................ ................................ ................................ .. 48 4.5 Conclusions ................................ ................................ ................................ ......................... 49 5. Consumer WTP for quality attribute s in domestic vs. imported pork ................................ ...... 51 5.1 Introduction ................................ ................................ ................................ ......................... 51 5.2 Theoretical framework ................................ ................................ ................................ ........ 59 5.2.1 The choice experiment ................................ ................................ ................................ . 63 5.2.2 Empirical model and willingness - to - pay estimation ................................ .................... 66 v 5.3 Empirical results ................................ ................................ ................................ ................. 69 5.3.1 RPL estimates and WTP for selected attributes ................................ ........................... 69 5.3.2 Guangzhou versus Hong Kong Consumers ................................ ................................ . 74 5.3.3 Patriotism effects on WTP for U.S. pork ................................ ................................ ..... 75 5.3.4 D ................................ .......................... 79 5.4 Conclusion and implications ................................ ................................ ............................... 80 5.4.1 Summary of the pertinent results ................................ ................................ ................. 80 5.4.2 Implications ................................ ................................ ................................ .................. 81 APPENDICES ................................ ................................ ................................ .............................. 83 Appendix A. Method Comparison ................................ ................................ ............................ 84 Appendix B. Original Survey ................................ ................................ ................................ ... 85 REFERENCES ................................ ................................ ................................ ............................. 97 vi LIST OF TABLES Table 1: Information on main pork suppliers in China ................................ ................................ ... 6 Table 2: Top 5 pork exporters to China (Values in thousands of US dollars) .............................. 11 Table 3: Demographi c statistics ................................ ................................ ................................ .... 27 Table 4: Purchasing behavior statistics ................................ ................................ ......................... 29 Table 5: Pe rception statistics ................................ ................................ ................................ ........ 41 Table 6: Mainland China regression ................................ ................................ ............................. 43 Table 7 : Hong Kong regression ................................ ................................ ................................ .... 44 Table 8: Attributes used in choice experiment ................................ ................................ ............. 64 Table 9: Choice experiment description ................................ ................................ ....................... 65 Table 10: Parameter estimates of Model I from RPL ................................ ................................ ... 71 Table 11: Parameter estimates of Model II from RPL ................................ ................................ .. 72 Table 12: Willingness to pay, mean values [95% confidence interval] ................................ ........ 73 Table 13: Comparison of WTPs for pork attributes in Guangzhou and Hong Kong ................... 74 Table 14: WTP for different types of U.S. pork ................................ ................................ ........... 77 Table 15: Determinants of patriotism ................................ ................................ ........................... 79 Table A - 1: Comparison between ordered logit regression and OLS ................................ ............ 84 vii LIST OF FIGURES Figure 1: Trend of pork trade in China from 1961 to 2011 ................................ ............................ 3 Figure 2: Food consumption shares in decades (KCal/Day/Capita) ................................ ............... 9 Figure 3: Annual consumption per capita of meat in China from 1975 to 2012, by meat product ................................ ................................ ................................ ................................ ....................... 10 Figure 4: Classification of WTP measurement methods ................................ .............................. 52 ................................ ................................ ............... 60 Figure 6: Frequency of patriotism scores in each city ................................ ................................ .. 76 viii KEY TO ABBREVIATIONS AHA Animal welfare standards APSRI Animal Product Safety Research Institute CN China EU European Union FAOSTAT Food and Agriculture Organization Corporate Statistical Database FSMA Food Safety Modernization Act GMO Genetically modified organism HACCP Hazard Analysis & Critical Control Points IO International Other ISO International Organization fo r Standardization KCAL Kilocalorie LCM Latent class model OLS Ordinary least square method RPL Random parameter logit RUM Random utility model SUR Seemingly unrelated regression UHT Ultra high temperature US United States USDA U.S. Department of Agriculture WSPA World Society for Protection of Animals WTP Willingness to Pay WTO World Trade Organization 1 1. I ntroduction 1.1 The U.S. pork market The p ork industry i n the U.S. has faced multiple threats. Increased feed price s and decreased buying power of U.S. meat consumers are threatening the competitiveness and sustainability of U.S. pork industries. Increased feed price s are primarily a result of increase s in corn price, which is the major feed component for hog production. In 2012, corn price reached to a record high of $8.43 per bushel ; prices were just $2 per bushel in 2007 (Hunt 2012). This drastic change was mainly a result of the su rging demand from corn - based ethanol industries, which has developed rapidly in recent years. Even though corn price s have decreased recently, price fluctuations still significantly and negatively affects the profitability of suppliers , as feed cost comprises approximately 60 percent of the cost of raising hogs . In addition, decreased buying power of U.S. consumers has suppressed domestic market growth, further contributing to low market prices. The w eakened buying power of U.S. consumers was mainly a result of the economic recession . Due to the financial crisis of 2008, median household income experienced a sharp dec rease, directly influencing power in the U.S. Although the U.S. econom y is incrementally recovering in recent years , many researchers believe that it will take a long period to bounce back; median household income will not get back to the level in 2007 until 2019 (Snyder 2014). Also, as substitutes for pork are available in the U.S. mar ket, such as chicken, consumer de sire of purchasing pork would be weakened if pork price s remain relatively unattractive (Hunt 2012). Therefore, serious questions arise as to how U.S. pork suppliers will remain profitable and seek growth when facing these market constraints. 2 Economic g lo balization has opened up international markets for U.S. goods , especially new market s in emerging economies. China, with one - fifth of the world's population and a significant share of its citizenry entering the middle class, has become a major customer for high quality products. Pork, as the staple meat in China, accounts for over 50% of total meat expenditures (Ortega et al. 2009). The Chinese p ork market was historically self - sufficient. However, due to the recent price fluctuation of domestic pork and in creased demand for safe and high quality pork, China has become a net pork importer since 2008. In 2011, China (including Hong Kong ) became the third largest export destination for U.S. pork accounting for 910 million USD and 483 thousand metric tons of pr oduct a new record. With increased production cost s as well as rising concerns over food safety, animal disease epidemics and environmental challenges, the role of China as a major pork importer will likely keep growing. While the Chinese market looks promising for U.S. pork suppliers seeking to expand business and growth , a dditional information is needed in order to assess the potential of China as an emerging market for U.S. pork . A good starting point is an understanding of domestic pork econ omy. 1.2 China's domestic pork economy 1.2.1 Pork supply in China China is the largest pork producer in the world, followed by the European Union (EU) and the United States. The Chinese mainland produced 49 million metric tons of pork in 2012 accounting f or about half of the world 's pork productio n . For comparison, the E . U . produced 27.2 million metric tons and the U.S. produced 10.6 million metric tons. China has a long history 3 of producing and consuming pork, and historically Chinese domestic supply has sufficiently fulfill ed domestic demand. B efore 2007, China was a net pork exporter , and became a net pork importer since 2008 (FAOSTAT) . s and import s during 1961 to 2011 can be found in Figure 1 . Figure 1 : Trend of pork trade in China from 1961 to 2011 Source: FAOSTAT T he high productivity and commercial value of raising pigs is giving rise to a thriving domestic pork industry in China. China's highly productive pork industry is characterized by low farming requirement s and a short production cycle ( McOrist et al. 2011 ) . In China, hog farmers can easily raise pigs in their backyards and make profi ts in the short run. The high commercial value is a result of Chinese diet, which is centered on pork. F or Chinese consumers , pork meat is not the only desirable product from the animal . They also consume organ meat and other parts such as ear, tail and fe et, which are common in traditional Chinese dishes. Therefore , processors are able to take full advantage of pigs , whose every part is valuable to Chinese consumers. 0 50000 100000 150000 200000 250000 300000 350000 400000 450000 500000 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 tonnes Meat of Swine Export Meat of Swine Import 4 Pork production is mainly carried out in three ways in China . Backyard farming was the primary method of production in the past, accounting for about 74 percent of total hog production in 2001 (Ramzy 2009; Pan and Nelson 2012 ). This number has experienced a sharp decrease recently due to concern s about food safety; hog s from scattered farmin g accounted for about 37 percent in 2012 ( Pan and Nelson 2012 ). W ith technological and econom ic development , intensive pig industries , including small and large commercial farms , are growing fast in China . McOrist et al. (2011) indicated that farms with more than 30 sows account for about 60% of the total pig population in China in 2009 . Large farms are mainly supplying in major cities in China such as Beijing and Shanghai , but small farms also supply a co nsiderable amount of pigs and pork in their own provinces. The m ain pork suppliers in China can be classified in several ways. From the perspective of reputation, Shuanghui, and Yurong are the most popular pork processing companies in China. Shuanghui , as one of the largest pork processor s in China, has attracted much attention from all over the world due to its acquisition of the world 's largest meat processor: Smithfield ,Inc . According to the financial report of Shuanghui in 2012 , their total revenue from pork business was about 31.6 billion Yuan (5 .0 billion US D ) , and fresh - chilled products contributed to over 40% of total revenues . Yurong , another leading pork processor and the largest slaughter ing company in China, also accounts for a large share of pork sales. Their main product is fresh - chilled pork, which made up about 90 % of total revenues in 2012 and earned about 21 billion HKD ( 2.7 billion USD ) . Compared to these large processors, wh o have expand ed their business throughout the world, several re latively smaller companies are mainly focusing on regional markets . For example, Shunxin had over 40% meat market share in Beijing in 2011 , Gaojin is the largest processed pork supplier in Sichuan , and Yihaotuzhu is a major 5 player in Guangdong province. G iven their nature and regional scope , it is rare to see these brands in other provinces. In terms of business region, China's pork companies can be classified as either international or domestic companies. International c ompanies such as Shuanghui, Yurong, Jinluo, Zhongpin and Gaojin, not only supply fresh pork in domestic market, but also export frozen pork to other countries. Domestic c ompanies , such as Yihaotuzhu and Shunxin, only focus on regional markets in China. Table 1 presents basic supply - chain information on the main Chinese pork suppliers . T wo points are worth highlighting. First, for most international companies who expa nd their business in other countries, their production process is governed by international food safety standard s . As a result, certifications are provided , such as HACCP and ISO9001 , aiming to meet the export requirement s and attract foreign consumers . Ho wever, domestic companies do not implement thes e standards in their production. I nstead, they provide products with informal Chinese certification such as "pollution - free label" to attract domestic consumers. Second, most international companies have multi ple breeder sources including domestic, foreign , and self - sourcing , whereas the domestic companies mainly rely on self - sourcing . These differences are consistent with the fact that international companies are trying to supply their products to various cons umers in different markets while the domestic companies are just focusing on consumers in regional and local market s . 6 Table 1 : Information on main pork suppliers in China Company Business region beside China mainland F ood s afety c ertification Other meat product Supply chain features Breeder source Self - slaught er Fresh chilled meat S elf - sourc e D omestic F oreign Rank in 2009 Rank in 2013 Shuang Hui Export to Japan, Singapore, Korea, etc . A nd acquired Smithfield ISO9001,HACCP ,GAP Chicken × × × No. 10 No.4 Yu Rong Export to Russia, North Korea, HK and South East Asia etc. HACCP,ISO1400 1,QS Cattle, sheep, chicken, duck, rabbit, goose, etc. × × × No. 1 No.6 Jin Luo Branch office and company in Russia, Singapore and Hong Kong , etc. for export ISO9001,ISO140 01,ISO22000,HA CCP Chicken × × No. 3 No.5 De Li Shi Export to Russia, HK and Singapore , etc . HACCP,SGS Bi - products from big bone, skin, blood and offal. × No. 5 No.7 Wen Shi Export to HK and Macau No International Food Safety Certification Major in c hicken and pork , minor in duck and cow × × He Mei Er Joint venture with Hormel(U.S.); Export to Japan, HK and Philippines , etc . HACCP, TQC Beef, chicken, turkey × Yi Hao Tu Zhu Mainly Guangdong and started to expand to other provinces since 2013. Pollution - free None × Shun Xin Mainly Beijing Pollution - free, HACCP None × Zhong Pin Export to southeast Asia and e ast Europe, etc. ISO9001 ,ISO140 00,GMP,GAP,SS OP, HACCP, P ollution - free Beef × No.2 No.3 Gao Jin Mainly eastern China ; Export to Russia, Korea, Japan, etc. ISO9001, ISO14001 HACCP, Pollution - free, Green . Processed beef, chicken × No.6 No.2 7 In addition to domestic pork production, pig meat from the E.U., the U.S., Canada and Brazil is known to have made its way to the mainland through a "grey channel" originating in Hong Kong to avoid high tariffs (Peter 2004 ; Collins and Sun 2010 ) ; imports into Hong Kong were estimated at over 266 thousand metric tons of pig meat in 2011 (FAOSTAT). This amount was significantly higher than the amount of annual pork consumption in this area. Therefore, Gale et al. (2012) indicated that this illegal trade contributed to a large amount of pork supply in China's pork market, although the actual amount of re - exported pig meat is uncertain. Hong Kong has minimal trade barriers for imports since there are no custom duties on food imports required in this region, whereas food imports into China are subject to 10 - 35 % custom duties and additional value - added tax es of approximately 17 % (Ngulu 2014). This policy may help explain why the illegal pork re - exportation from Hong K ong to the mainland is so rampant. Two main problems arise from the grey channel. First, meat transported through this unregulated channel may pose a danger to consumers since it is difficult to inspect quality, and poor transportation condition may lead to food safety problems (Cooper 2012). In addition, this grey channel would result in a loss of brand as the pork being imported in this way ends up in food service processed channels . T he Chinese consumers do not know they are eating pork from developed c ountries. As a result, foreign pork suppliers may miss out on price premium s (Collins and Sun 2010; Montlake 2013 ). 1. 2.2 Domestic p ork demand in China As the most populous country, China is also the world 's largest food consumer. M eat consumptio n in China grew rapidly in the past three decades and has beco me the most important category of food consumption in recent years. Data from FAO, presented in Figure 8 2 , reveals how the Chinese dinner plate has changed in the past half a century . In the 1960s, cereals 1 and starchy roots provided 84 % of total calories for Chinese people, while meat 2 contribut ed to only 4 %. As a result of economic development from 1980 to 2000, meat, poultry, fish, eggs and dairy became a key source of calories for Chinese people and provided 19 % of total calories in the 2000s . Among all kinds of meat, pork is the most favored animal protein ; its consumption increased rapi dly during this period. The data f rom U.S. D epartment of A griculture (USDA) shows that annual pork consumption increased from 16 million metric tons in 1985 to 52 million metric tons in 2012, while annual consu mption of meat increased from 16.7 to 71 million metric tons. The rising trend of meat consumption is highlighted in Figure 3 . 1 Cereals include cereals and pulses but exclude beer. 2 Meat includes meat and offal. 9 Figure 2 : Food consumption shares in decades (KCal/Day/Capita) Source : Figure is m ade by Author, using the d ata from FAOSTAT . 10 Figure 3 : Annual consumption per capita of meat in China from 1975 to 2012, by meat product S ource: USDA 2012 Multiple factors have led to increased consumption of meat, especial ly pork, in China. For example, due to the econo mic reform of the 1980s the economy grew rapidly , significantly improving the quality of life. Rising income s generated by the development of the economy allowed Chinese citizens to consume more high - value food. Also, urbanization , through its effect on food consumption patterns and food availability, is another driver that significantly spur red meat consumption. As the negative effect of urbanization on cereal grains in high income Asian countries has been noted by Huang and David (1993), the increasing demand for meat is not surprising in China, where the urbanization level has doubled in the past 30 years and h as reached about 50 percent (Chen 2007) . Additionally , better food availability in urban area s guarantees consumer's ability to purchase their desirable food as their income s increase . 0 5 10 15 20 25 30 35 40 45 1975 1980 1985 1990 1995 2000 2005 2010 2015 Kilograms Pork Chicken Beef 11 1. 2.3 China as a pork importer With the rising appetite for pork, Chi nese consumers are no longer only focusing on domestic pork products. China has been a net pork importer since 2008 . Net imports of swine meat increased from 291 thousand metric tons in 2008 t o 387 thousand metric tons in 2011 (FAOSTAT) . Main foreign pork suppliers for China are listed in Table 2 . As the largest pork exporter to China, U.S. pork suppliers increased their sales in China from 439.7 million to 1068.8 milli on USD. Multiple factors have led China to increase its pork imports . The first set of factors originate s from the domestic supply, including higher price s and a tarnished reputation due to the occurrence of several food safety incidents. The second set of factors is from domestic deman d, and changing consumer preferences and lifestyle . Table 2 : Top 5 pork export ers to China (Values in thousands of US dollars) Rank in 2012 2008 2009 2010 2011 2012 Changes from 10 - 12 United States 1 439,708 82,213 236,233 1,234,857 1,068,791 352% Denmark 2 259,858 196,369 357,902 316,367 364,439 2% Germany 3 69 176 42,949 78,547 287,514 569% Canada 4 111,381 129,121 222,532 204,677 237,720 7% Spain 5 234 41,426 54,633 124,313 186,198 241% Source: United Nations Commodity Trade Statistics, United Nations Statistics Division Factors from the supply side are mainly a result of the unstable nature of China's pork industry. F irst , given the positive relationship between China's pork price and its imports ( Gale et al. 2012) , the fluctuation of domestic pork price creates opportunities for foreign pork suppliers. Higher pr ices in China relative to western countries are a result of complementary preference for pork cuts and rising production cost s . Complementary preference s refers to the fact that less desirable pork cuts for western consumers , such as offal, pig ears and ta il , are given high er value by Chinese consumers. Additionally , Chinese 12 consumers prefer fat t i er meat , while western c onsumers prefer leaner pork . These complementary preference s have led to a lower price of imported pork and have encourage d trade between countries. In addition, rising cost s are also contributing to increased pork price s . Production cost is comprised of feed cost , labor and capital. In China, the rapidly rising feed cost, which accounts for over 50% of total cost , is the main driver behind rising pork production cost s (Gale et al. 2012). Second, the tarnished reputation of domestic pork suppliers may also help to incr e ase China's pork imports. As a result of recent pork safety scandals , such as Clenbuterol - tainted pork in 2011 and the Huangpu river dead pig incident in 2013 , Chinese consumers' trust on their domest ic pork suppliers has been weakened . This distrust may potentially increase western pork suppliers' sales due to their relatively well - established reputations . Despite the significant role of domestic pork suppliers, factors from the demand side may also impact pork imports. First, w ith the rising income generated by the development of the Chinese economy, consumers will demand higher quality and improved food sa fety, which in turn will possibly boost sales of imported pork in China. Ortega et al. (2009) found that food - safety - sensitive consumers in Beijing and Shanghai ha d a positive willingness to pay for U.S. pork. This implies that, as more and more people are jumping into the middle class in urban China, imported pork may be one of the options for pork consumers in the wealthier cities, who are seeking safer and higher quality food . Second , busier lifestyle s in China resulting from economic and social development imply that consumers have less time to purchase and prepare fresh food. Consequently , convenient foods such as frozen meat products and online retailing are becom ing more popular in urban China . As Ortega et al. (2009) noted that 13 consumers havi ng purchased frozen meat before are more likely to accept imported pork, increased demand for frozen pork arising from the trend of busy lifestyles would motivate pork imports in China. Given the drivers from both supply and demand side, China is becom ing promising for western pork suppliers . T he emergence of foreign imported pork will motivate the Chinese market to be more competitive and diverse, meaning that Chinese consumers will be confronted with various pork products from different countries and wit h differentiated quality attributes. Knowing h ow Chinese consumers perc eive imported pork, and what their preferences are , will allow for a better understanding of the potential of the Chinese pork market for western pork suppliers. 1.3 Thesis objective T he objective of this thesis is to provide a n analysis of Chinese consumer's preference for pork along with an evaluation of the potential for U.S. pork in China. Two essays are developed in order to achieve these goals. T he first essay is trying to explore the determinants of Chinese consumer perception of pork quality. Empirical model s are developed to capture the relationship between Chinese consumer perception of pork quality and their attitude towards pork quality characteristics including cr edence, experience and search attributes . The model is applied to survey data from consumers in four major cities in China, and estimated using a seemingly unrelated regression method in order to provide comparisons between pork from China and other countr ies. Results from the first essay will 14 broaden our understanding of how current Chinese co nsumers evaluate pork quality from domestic and international markets. After investigating consumer's perception of pork quality in the first essay , t he second essay in chapter 5 aims to provide a closer look at factors that influence consumer purchasing decisions of pork from China and developed countries. A choice experiment approach is employed to examine c onsumer's willingness to pay (WTP) for selec t quality attributes in pork . well as country of origin will be estimated. Also, another particular interest in this research is to explore how level of patri otism will affect their WTP for domestic and imported pork. The monetary importance of the product attributes for Chinese consumers generated in this essay will help quantify the result s from the first essay. This thesis is organized as follows: chapter 2 will present background information on determinants of consumer preference for pork and offers a comparison between pork from China and the U.S. in terms various attributes. Chapter 3 will describe the data used in this study. Chapter 4 will present the f irst essay entitled "Chinese consumers perception of pork quality". Chapter 5 will present the second essay entitled "Chinese consumers' willingness to pay for pork credence attributes". C opies of surveys and supplementary information for both essays are c ontained in the Appendices at the end of this thesis. 15 2 . Determinants of consumers' preference for pork 2 .1 Definition and classification of pork attributes Consumers no longer treat pork products as a single commodity since multiple att ributes are incorporated during the production process. Product - specific characteristics, such as appearance, food safety assurances , animal welfare and environmental certifications will play an increased role in determining consumers' perception of pork q uality and will affect their purchas ing decision s (Meuwissen et al. 2007). Researchers and economists traditionally categorize d product characteristics as search, experience and credence attributes (Nelson 1970; Caswell et al. 1996). A search attribute is one that consumers can recognize before purchase. For pork products, price, expiration date and appearance are examples of search attributes. Experience attributes such as taste and tenderness are evaluated after consumption . Researchers define credence at tributes as those that cannot be discern even after consuming the product; examples include food safety, animal welfare, environmental impact and country of origin. Nevertheless, information signaling such as food safety certification, animal welfare assur ance , environmental - friendly assurance and country of origin labeling, can help transform credence attributes in to search attributes ( Caswell et al. 1996). Previous research has investigated consumer preferences for pork - specific quality attributes ( Grunert 2005; Bernués et al. 2003; Cicia et al. 2010). According to these studies , consumers may rely on multiple attributes to judge product quality , but the relative importance of these attributes for consumers may differ between countries . For example, food safety attribute s have acquired price premium s from consumers in both developed and developing countries (Loureiro et al. 2007; Ubilava et a l. 2009), b ut the amount of premium consumers are willing 16 to pay is different depending on their trust and perc eption s . For instance, Loureiro et al. (2007) found that the willingness to pay of U.S. beef consumers for quality certification is four times higher than traceability. However, Ubilava et al. (2009) found a different result in the Republic of Georgia, whe re consumers are willing to pay about 48% more for prod u ct traceability compared to quality certification. This is a result of consumer mistrust o f their own country's quality certification system . Janssen and Hamm (2012) presented a similar finding based on their study in European countries, where consumers were capable of distinguish ing differentiat ed products ; only labels under strict standards re ceive d a relatively high premium. Given that pork pro duced in different countries could be composed of different quality attributes, a compar ison between pork from China and the U.S. with respect to quality attri butes is necessary. We select attributes that encompass food safety, animal welfare, environmental impact and country of origin to explore , as thes e are receiving increas ed attention from Chinese consumers and they have a potential to affect purchas ing decisions . 2 .2 Difference s in credence attributes between Chinese and U.S. pork 2.2.1 Different p olicies and regulations on food safety D ue to the important links between food and human health, such as food - borne illness , f ood safety concerns are shared by consumers in both developed and developing countries . F ood safety risks may arise at any point of the food production chain since the product goes through many stages to get from the farm to the consumer . For example, pathogen s can be introduced into the product if pig s are fed with contaminated water or feed. Also, pork may 17 be contaminated if the pig is slaughtered under poor sanitary conditions . T herefore, many countries have implemented different policies and regulations to enhance food safety throughout the whole supply chain . Moreover food producers are making efforts to differentiate their products from their competitors by providing food safet y assurances . For example, producers will present some food safety certification on the package to state that their pigs are produced under certain inspection and quality management system s . Also, some countries require producers to implement traceability systems to improve the e fficiency of product recall and ensure food safety. This type of information will be signaled through labels certified by legal organizations , and potent ially affect consumer purchasing decision. In accordance with previous studies , we specify food safety attributes as certification, traceability and additive - free labels ( McCarthy and Henson 2003; Ortega et al. 2011; Krystallis et al. 2006 ) , and compare difference s in standards, policies and impl ementation between China and the United States . Food safety certification will be provided for products that are produced under specific safety inspections and quality management systems. Several food safety systems and standards have been developed and i ntroduced in the food industry, in order to ensure food safety and quality. For example, Hazard Analysis and Critical Control Point (HACCP) is a system designed to prevent biological, chemical and physical hazards in production processes. Products produced under this system will be labeled "HACCP". Many o ther global certifications such as ISO certifications and World Trade Organization (WTO) standards are also used worldwide. These international food safety systems have been awarded and implemented in China 's food industries. In addition to these global food safety certifications, 18 many countries have developed their own standards and labels to certify food safety and quality . The Chinese government has introduced several food safety assurance schemes in re ce nt years. These programs established a number of standards that provide certification labels, including organic, green , pollution - free, and safe label s (Paull 2008; Calvin et al. 2006) . However, given that Chinese consumers' confidence in their domestic food was discouraged after the exposure of many high - profile foo d incidents, they have presented a level of doubt about whether the Chinese government is able to manage these standards well (Sun and Coll ins 2012). In the U.S., p ork produce r s can choose to participate in the certification programs granted Service (USDA - AMS), and their pork products will be labeled with USDA - backed certifications. I n contrast to China, these assurance labels certified by the U.S. government enjoy a good reputation and receive price premium s from U.S. consumers ( Schumacher and Tonsor 2012 ). Although Chinese consumers may not be familiar with these foreign labels or as surances, their potential distrust in domestic certifications may create opportunities for imported food products, which are produced under strict standards and are certified by governments or organizations in developed countries . Traceability is defined as the ability to track the history, location or use of an entity, by means of recorded identifications ( Baines and Davies, 1998 ) . As f ood safety risks may arise at any point in the food production process , traceability system s - assuring the transparency in food supply chain - are regarded as a useful technique to reinforce food safety. Pouliot et al. (2008) indicated that traceability is linked to food safety by in to supply safer food. In addition to the economical motivations to suppliers, both the U.S. 19 and China have implemented policies and regulations to incentivize producers to implement traceability system s . In the United States, The Food Safety Modernization Act (FSMA), signed int o law in January 2011, requires all food processors in compliance to enhance traceability systems. In China, the government also noticed the importance of traceability and published a new food safety law in 2009 that explicitly implement s traceability system s . However, due to the fragmented nature of Chinese agriculture, it is not certain whether this system can be implemented successfully (Ramzy 2009) . As traceability system s play an important role in en suring food safety, and has been well introduced in pork production in developed countries, foreign imported pork carrying such attribute could be attractive to food safety - sensitive consumers in China. Additive - free claim s are another characteristic that consumers use to evaluate food safety , as unfamiliar chemical terms have emerge d following the exposure of recent food safety scandals. For example, after the clenbuterol - tainted pork incident occurred in China, several kinds of lean - meat powder inclu ding clenbuterol and ractopamine were explicitly prohibited in pork production in China. However, as ractopamine is completely safe for human consumption , it is legally allowed in U.S. pork production, which gave rise to a sharp decline of U.S. pork export to China in the beginning of 2013 (USDA 2012). Antibiotic s are another example that has been hotly debated among researchers and policy makers for a long time . T he use of antibiotics can promote production efficiency and benefit suppliers by destroying or slowing down the growth of bacteria , but create s potential health risks to consumers (Lust et al. 2006). Both China and the U.S. are legally allowed to use these substances to keep animals healthy, but the overuse of antibiotic d r ugs in China's livestock production has given 20 rise to significant concerns amongst scientists and policy makers . Philpott (2013) indicated that half of China's antibiotics go es to livestock and massive antibiotic resistance strains have been found in Chinese pig farms ( Zhu, et al. 2013). Regarding Chinese consumer's food safety sensitivity after various food safety incidents, their attitudes towards these additive s use d in pork production can potentially influence their purchasing decision . Thus, further research is req uired to examine consumer's purchasing behavior regarding additive - free claims or other certifications that guarantee food safety. 2.2.2 Increasing awareness of animal welfare in China In recent years, animal welfare has gained attention and importance in China . L arge government - funded projects on farm animal welfare have been established to build Chinese standards for animal welfare (Nielsen et al. 2012). For example, a humane slaughter program was implemented by the World Society for Protection of Animals (WSPA) and the Beijing ChaoyangAnhua Animal Product Safety Research Institute (APSRI) in 2007. The aim of the program is to improve the welfare of animals slaughtered in China by developing and implementing p re - slaughter and slaughter animal welfare curricula for undergraduate veterinarians . A nimal welfare encompasses various aspects that include housing condition, feed, stock density, method of slaughter, transportation and farming method (Den Ouden et al.1 997; Harper et al.2001; Liljenstolpe 2008). For some consumers, who are trying to maximize their utility from c onsuming pork, a well - treated pig means better quality pork. But the following question arises : what type of information will consumers rely on i n order to evaluate this 21 credence attribute ? According to research conducted by Mayfield et al. (2007) in Europe, the most important information for consumers is the animal - welfare assurance label, followed by grading system, information on housing conditi on and information on feed. However, animal welfare certification s and grading system s have still not been esta blished in China. In contrast, t he U.S. has organizations that help to verify animal welfare standards. For example, the American Humane Associ ation is the first welfare certification program in the United States to ensure the humane treatment of farm animals and provide s third - party independent certification for producers who meet the animal welfare standards ( AHA). In addition, some other programs in the U.S. such as the National Organic Program and the non - profit Humane Farm Animal Care also provide animal welfare certifications. The standards of these programs include raising condition, humane slaughter, and certi fied animal transportation . As consumers may count on these characteristics to evaluate animal welfare and judge the quality of pork, animal welfare certifications can possibly become another tool for imported pork suppliers to differentiate their products in China. Therefore, additional research is needed to understand Chinese attitude towards animal welfare. 2.2.3 Rising c oncerns on the e nvironmental impacts of China's pork industry With the rapid economic development, many environment al prob lems such as air pollution, water pollution and soil contamination have been plaguing China in recent years . For example, recently, t he most hazardous pollution is PM 2.5 particles; which are small enough that they can penetrate into lungs and enter the bloodstream . The average level of PM 2.5 in China is 20 times higher than the safe limit, and over 800 million people in China are 22 affected (Peng et al. 2013). Heavy industrial dust, as well as coal burning and automobile exhaust, are the main contributors to this toxic air pollution. Water pollution is another big problem in China. According to China's State Environmental Protection Administration (SEPA) in 2006, 60% of the country's rivers are polluted and cannot be used as drinking water sources, and the main contributors includ e industrial emission and extensive use of pesticides and fertilizers. In addition, soil contamination is becoming increasingly prevalent while more than 2% of the farmland in China cannot be used to grow food because of pollution ( Osborne 2013). This farmland is polluted with heavy metals and some other chemicals released by various industries. Pork production , which highly depends on concentrated feed and land input s , is linked to these various environmental issues. S everal type s of environmental pollution associated with Chinese livestock industry have been observed (Zhang et al. 2010; Li et al. 2005). For example, heavy metal contents have increased in pig manure, which can cause metal pollution in the soil (Jiang et al. 2010; Cang et al. 2004 ; Wang et al. 2013 ). Water pollution , such as algae bloom , has also been noted , which is mainly a result from the nutrients in animal manure (Gu et al. 2008; Sun et al. 2012 ; Wang et al. 2010; Gao et al. 2013 ). Moreover, Fu and Yu (2013) indicated that the livestock industry in China has become a major source of methane, one kind of greenhouse gas , resulting in air pollution . Given the association between environmental pollution and the livestock industry, Chinese pork consumers w ho are concerned about environmental issues have two potential options to remedy the situation . One is purchasing environmental - friendly pork products, such as 23 pollution - free pork , and the other is to purchase imported pork; the latter having no direct imp act on the domestic environment. With this understanding, imported pork seems to be competitive since environmental - friendly attribute s will have a relatively high price. Thus , research on Chinese consumers' awareness of environmental issues and attitudes towards environmental ly friendly products is of great importanc e . Moreover, a dditional research on consumer attitudes towards environmentally friendly pork products is needed to help evaluate the potential of imported pork in China. 2.3 Difference in experience and search attributes of Chinese and U.S. pork S earch attributes of pork include appearance, packaging , and price characteristics that can be recognized before purchase. Examples of e xperience attributes of pork include fresh ness, tenderness, an d taste, which can only be perceived after consumption . One of the difference s between Chinese and U.S. pork in terms of these tangible attributes is that U.S. pork is leaner , or has less marbling than Chinese pork. T his is driven by both consumer preferences and producer's farming method . Western consumers prefer lean meat while Chinese consumers are drawn to pork with a certain fat content (Ortega et al. 2009) . In response to this demand, p roducers in the U.S. supply more lean meat by feeding hog s with a lean - meat powder called r actopamine. Currently over 80% of hogs raised in the U.S. are fed with this type of additive (Couch 2014). In contrast , Chinese pork used to involve more external fat, which was a result of the traditional backyard farming method , the primary production method in Chin a in the past . Most of the hogs raised in backyard s were fed with less strict diet s compared to modern industrialized production. As a result, the growth rates of hogs were slow and the amount of external fat w as high (Wang et al. 1998) . This situation has 24 changed recently as many farms have transferred into specialized or commercial farms with advanced equipment and modern production method . Some other attributes may also differ based on the fact that Chinese pork can be found in various retail channels such as wet markets , meat store s and supermarket s, while i mported pork is only accessible for Chinese consumers in high - end supermarket s . Chin a used to import frozen U.S. pork, which ended up in food service channel and was not able to be recognized directly by consumers (Montlake 2013). As retailers have realized the high margins of U.S. pork, chilled U.S. imported pork is starting to emerge in international supermarket s in China. Therefore, in terms of packaging , imported pork sold in the high - end supermarket must be more attractive than Chinese pork sold in the wet market. However, in terms of freshness, imported pork ( chilled or frozen ) in th e supermarket may possibly be less attractive for consumers who are used to purchase fresh pork in wet market s . P ork imported from the U.S. will be chilled and shipped a long distance ; whether transportation will significantly affect freshness is hard to s ay. Therefore, it is of interest to explore whether consumer s will perceive chilled imported pork sold in the supermar ket as fresh as the pork sold in the wet market , which is easier for consumers to judge the quality and freshness of . G iven the potential difference among Chinese and U.S. imported pork in terms of tangible characteristics, additional research is needed to examine how Chinese consumers evaluate the quality of pork from different countries. 25 3. Data 3.1 Data sources Because information on consumer preferences for the pork attributes of interest is not available in the marketplace, the collection of primary data was necessary. The data utilized in this study was collected via consumer interviews in four major cities in China including Beijing, Sh anghai, Guangzhou and Hong Kong in summer 2 0 14 . These four tier one cities are the wealthiest in China. Beijing being the capital , is the second largest city in China by urban population ( after Shanghai ) , and is the political an d educational center of China. Shanghai, with the most urban population, is the economic and financial center of China. Guangzhou , the third largest city in China, is the key transportation hub and trading center in southern China. Hong Kong, a special adm inistrative region of China, is famous for its low taxation and free trade and is viewed as a world city . The survey s were conducted through in - person interviews with consumers, and the questionnaires were executed at point of purchase. A multi - stage s ampling procedure was used to construct the sample. First, in each city, we randomly selected 3 to 5 districts that had supermarkets selling imported meat products. For example, in Shanghai 5 districts (Yangpu, Xuhui, Pudong, Minhang and Changning) were ra ndomly selected from a total of 18 districts. In the second stage, various food retail outlets were randomly selected from a roster compiled by the researchers that included domestic and international supermarkets in each district. Within each store, consu mers were selected using a quasi - random technique that entailed intercepting every 3 rd customer upon completion of a questionnaire. In each city, we 26 surveyed 200 respondents and local university students were hired and trained to conduct the interviews. 3.2 Data description Descriptive statistics of the s urvey data is shown in Table 3 and Table 4 . Table 3 presents demographic statistics of respondents by region . The average age of the respondents in mainland China ( Beijing, Shanghai and Guangzhou ) is 37.5, and the average age of the respondent s in Hong Kong is 41.7 years . About 30 percent of respondents are male in both regions ; this is expected, as women are the primary shopper . As for the education level, about 33% of mainland respondents complete d a university degree ; the most frequent level in our sample . Secondary school is the most frequent education level among Hong Kong respondents, accounting for 46.23% of responses . In both regions, over 60 per cent of respondents are married and the average household size is about 3.5 individuals . T he average monthly household income level of mainland respondents is in the range of 8000 to 10000 RMB ( 1282 to 16 0 3 USD 3 ) , and the average level of Hong Kong respondents is betwe e n of 20000 to 25000 HKD ( 2580 to 3225 USD 4 ) . 3 USD to CNY exchange rate in 2014 June was 6.239. 4 USD to HKD exchange rate in 20 14 June was 7.752. 27 Table 3 : Demographic statistics Variables D e sc r iption Mean or distribution Mainland China Hong Kong Observations N=600 N=200 Age Age as of J une 2014 37.51 (13.71) 41.74 (14.90) Gender ( %) S hare of male 34.57 31.13 Education (%) Primary school 3.44 14.62 Secondary school 29.56 46.23 Two - year college 25.97 8.49 University 33.00 27.83 Graduate school 8.03 2.83 Marital Status (%) Married 68.01 63.33 Single 30.70 32.86 Other 1.29 3.81 Household size Number of family member s 3.47 (1.30) 3.83 (1.14) Children Number of children < 6 years old 0.62 (0.79) 0.72 (0.80) Senior Number of persons > 60 years old 0.75 (0.94) 0.47 (0.74) Monthly household income (%) First range is for mainland and unit is in RMB. N umber in "()" is for Hong Kong and unit is in HKD < 2,000 (5,000) 2.16 1.89 2,000 - 4,000 (5,000 - 10,000) 12.50 7.08 4,000 - 6,000 (10,000 - 15,000) 16.38 13.68 6,000 - 8,000 (15,000 - 20,000) 13.22 21.70 8,000 - 10,000 (20,000 - 25,000) 14.94 15.57 10,000 - 12,0000 (25,000 - 30,000) 13.36 8.96 12,000 - 14,000 (30,000 - 35,000) 6.61 8.49 14,000 - 16,000 (35,000 - 40,000) 4.17 5.19 16,000 - 18,000 (40,000 - 45,000) 4.17 1.42 18,000 - 20,000 (45,000 - 50,000) 4.74 5.19 20,000 - 22,000 (50,000 - 55,000) 3.59 2.36 > 22,000 (55,000 - 60,000) 4.17 1.89 (>60,000) / 6.60 Note: S tandard d eviation in parentheses . 28 Information regarding food and pork shopping behavior s is presented in Table 4 . As expected, a lmost all of the respondents were pork consumers . Most of respondents in both mainland and Hong Kong chose to walk to purchase pork, and t he average travel time to a supermarke t and wet market for mainland respondents was 17 minutes, w hile Hong Kong respondents spent about 9.9 and 10.9 minutes to go to supermarket and wet market , respectively . For mainland respondents, 19 percent of them indicated having purchased imported pork and 6 percent of them noted having purchased U.S. pork in the past . For Hong Kong respondents, 51 percent of them ha d purchased imported pork and 27 percent of them ha d purchased U.S. pork. In both regions, most of the respondents f requented wet market s as the primary place to purchase pork, and the share of frozen pork purchased was relatively low ( in mainland it is 10% and in Hong Kong 7%). The average prices of fresh chilled pork loin that mainland and Hong Kong respondents recent ly purchased were 17.78 RMB and 35.13 HKD , respectively. Chinese pork was the most purchased product for both mainland and Hong Kong respondents in the last month, accounting for 94.55% and 54.25% of pork purchases, respectively. In terms of quality claims , 55.67 % of pork purchased by mainland respondents was known to contai n a food safety claim, followed by an environmental - friendly claim (20.63%) and an animal welfare claim (12.20%). Similarly, for Hong Kong respondents, 28.30% of pork possessed a food s afety claim, followed by an environmental - friendly claim (10.85%) and an animal welfare claim (8.49%). 29 Table 4 : Purchasing behavior statistics Variable s Desc r iption Mean or distribution Mainland China Hong Kong Observations N=600 N=200 Food purchaser (%) Equal to 1 if yes 72.43 55.19 Food preparer (%) Equal to 1 if yes 72.78 56.13 Pork eater (%) Equal to 1 if yes 99.28 99.53 Supermarket time Travel time to supermarket (min) 17.12 (14.18) 9.88 (8.40) Wet market time Travel time to wet market (min) 16.99 (15.12) 10.89 (5.60) Supermarket Mode (%) Walk 63.45 95.92 Bike 13.53 0.00 Car 8.06 2.04 Public transportation 14.96 2.04 Wet market Mode (%) Walk 68.70 93.94 Bike 16.38 0.00 Car 6.52 2.53 Public transportation 8.41 3.54 Imported pork purchase (%) Equal to 1 if purchased imported pork in past 18.71 51.18 US pork purchase (%) Equal to 1 if purchased U.S. pork in past 5.71 27.49 Purchase location (%) Wet market 48.35 56.73 Specialized Meat Store 8.44 9.62 Low - end domestic supermarket 10.30 29.33 High - end domestic supermarket 28.76 14.90 International supermarket 8.44 3.85 Other 2.26 0.96 Chilled (%) Share of chilled pork 51.94 25.67 Fresh (%) Share of fresh pork 39.86 67.38 Frozen (%) Share of frozen pork 8.20 6.95 Information on purchased pork Price Price of pork purchased last month 5 17.78 (5.94) 35.13 (11.23) Country of Origin (%) China 94.55 54.25 US 0.86 6.60 IO 6 0.14 2.83 Do not know 4.45 36.32 5 The price is in RMB in mainland China, and in HKD in Hong Kong. 6 IO stands for other international countries. 30 Food safety claim (%) Equal to 1 if purchased pork with food safety claim 55.67 28.30 Animal welfare claim (%) Equal to 1 if purchased pork with animal welfare claim 12.20 8.49 Environmental - friendly claim (%) Equal to 1 if purchased pork with environment - friendly claim 20.63 10.85 Note: Standard deviation in parentheses. 31 4. Chinese consumers' perception of pork quality 4.1 Introduction The definition of quality given by the International Organization of Standardization (ISO), is satisfy stated or implied needs" (ISO 8402). According to this definition, product quality is a complicated and multifaceted concept, which is determined by multiple characteristics of a product. These characteristics such as physical properties, special production method and quality control system developed throughou t the agricultural marketing system will objectively affect or influence product quality , regarded as objective quality . However, as the objective quality of a product will be judged subjectively by consumers relying on the available intrinsic and extrinsi c cues, the delivery of quality information from suppliers to consumers may not be efficient. Morgan (1985) indicated that there is a "quality perception gap" between supplier and consume r , by presenti ng an example that manufacturer ' s quality perception is . Therefore, a user - oriented approach, which emphasized the consumer perspective, was underlined. Researchers employing this approach believed that the valuation of quality by consumers, regarded as subjective quality, is dependent on their perceptions, needs and goals (Garvin 1984, Steenkamp 1990). Brunsø et al. (2005) concluded that there are two major approaches to study ing subjective quality : multidimensional and hierarchical approaches. The m ultidimensional approach as sumes that consumers evaluate product quality based on a combination of multiple product attributes. According to a summary made by Issanchou (1996), potential factors playing a role in influenc ing consumer's quality perception and purchasing decision incl ude convenience, animal welfare, safety, healthiness, intrinsic and extrinsic quality cues. By contrast, the 32 hierarchical approach postulates that quality itself is not the aim that consumers desire; rather, consumers are pursuing the value obtained by pur chasing the quality of the product. Many empirical studies following the multidimensional approach have been conducted. These works incorporate multiple product attributes categorized as search, experience and credence attributes, or intrinsic and extrins ic attributes, to investigate how these factors Becker et al. 2000; Verbeke and Ward 2006; Espejel et al. 2007). For example, Bernués et al. (2003) conducted surveys in Europe to investigate which extrinsic attributes of red meat could impact consumer's quality perception. Respondents were asked to report on the importance of multiple attributes to achieve quality in beef and lamb. Through a multi - attribute model esti mated by principal component analysis, they found that animal feeding is the most important attribute, followed by environmental - friendly production and animal welfare concern s . Similarly, Becker et al. (2000) asked consumers to rate both intrinsic and ext rinsic quality cues from the perspective of helpfulness of evaluating meat quality. They found that country of origin and place of purchase are the most helpful cues, followed by color, brand/label, leanness and marbling. Rather than incorporate attributes into one model, Verbeke and Ward (2006) examined consumer's interest in quality indicators separately. They conducted a survey in Belgium in 2000 and asked respondents Likert - scale questions such as "to what extent do you pay attention to the information cue" and "to what extent is this cue important for you when purchasing". The authors employed an ordered probit model to assess the impact of individual characteristics and a labeling information campaign. They 33 found that certified quality marks and seals of guarantee received the highest interests from consumers. As food quality is a subjective and multi - dimensional concept, consumers in different countries with different history, background and tradition, may present different attitudes towards quality a ttributes. With this understanding, several empirical studies have been conducted to investigate what type of quality attributes Chinese consumers may rely on to evaluate the quality of food product s (Ngapo et al. 2005; Balestrini and Gamble 2006; Wong et al. 2008). For example, Balestrini and Gamble (2006) conducted a questionnaire in Shanghai to investigate the type of cues Chinese wine consumers use to judge the wine quality. By using a paired - sample t - test to measure the difference among the L ikert - scale scores of the wine cues, they found that Chinese wine consumers are more likely to use extrinsic cues, such as country of origin, to evaluate wine quality. Ngapo et al. (200 7 ) examined which appearance attributes Chinese consumers may use to as sess pork quality, by asking consumers to select their preferred pork chop from multiple alternatives. The result showed that color and fat cover are the most important appearance characteristics for consumers. However, the quality of food is not only a m ulti - dimensional concept, but also a dynamic one where the importance of quality attributes to consumer is changing rapidly (Issanchou 1996; Grunert and Valli 2001). Consumers will rely more on food safety attributes to evaluate food quality after they bec ome exposed to food incidents (Latouche et al. 1998). Also, w hile consumers' concern s on animal welfare and environment rise, their quality perception of the 34 food that is not animal - or environmental - friendly will be negatively affected (Wandel and Bugge 1 997 ; Mayfield et al. 2007 ). Therefore, it is necessary to re - evaluate Chinese consumers' perception of pork quality, especially after recent food safety incidents were exposed in China. This may raise the following concerns: Will Chinese consumers pay more attention to food safety aspects while they judge the pork quality? Given that animal welfare is becoming a hot topic in China, is this trend related to the quality of pork products? W ill the increasing concern on pollution among Chinese consumers lead th em to care more about the environmental aspect of pork production? H ow will Chinese consumers evaluate imported pork from developed countries? The present study seeks to answer these questions. The focus of this chapter is on two objectives. The first obj ective is to evaluate the relationships between Chinese consumers' pork quality perception and their preferences f or meat credence attributes including food safety, animal welfare and the environment . The second objective is to examine how specific pork ch aracteristics influence consumer's perception of pork quality in China. Pork characteristics including appearance, taste, price and origin, are classified as search and experience attributes. The e ffects of other key factors including patriotism, past expe rience on import ed pork will also be evaluated. Given that the socio - political nature is much different between China and Western countries, the effects dome stic products. In addition, as imported pork has been scarce in China, past purchasing since most of them are only familiar with domestic pork. 35 Both mainland China and Hong Kong data are applied to examine regional effects on quality perceptions . As the special administrative region of China Reg arding the differences in history, politics, economics and culture between mainland from China and developed countries will be significant. With this understanding, a com parison between these two regions will also be explored. 4.2 Methodology A system of linear equations is used to analyze Chinese consumers' perception of pork quality , Y . (1) W here i denotes individual; j =1, 2,..., m , indicates the j th equation; is individual i 's independent variable vector of the j th equation; and is the error term. With the assumption that the error terms have zero mean and independen ce across individuals and homosk edasticity, we stack the system of equations into a seemingl y unrelated regression (SUR) model as follows: 36 (2) Based on the framework above, we specify the following empirical model given in equation (3). (3) As our interest is not only to reveal the quality perception of Chinese pork, but also to compare the difference in quality perceptions across Chinese and other imported pork, multiple equations with respect to individual countries are developed. The countries /regions un der consideration include the major players in the Chinese pork market: China (CN) , the United States (US) and European Union (EU). All three equations have the same independent or explanatory variables except for the E.U. equation , which does not incorpor ate the patriotism variable (denoted by P), as this is measured on a relative scale between China and the U.S. In this context Y are the country - perception of pork from different countries. This variable is measured by asking respondents to rate the pork quality score on a scale from 1 to 5, where 1 is the lowest rating and 5 is the highest. is a 3 x 1 vector of country - specific scores of pork regarding credence attribute claims including food safety, animal welfare and environment al issue s . Likert - scale questions 37 such as " how would you rate the food safety standard of pork produced in China/the United States/the European Union" were asked to obtain such information. (wh ich is not country specific) is a 10 x 1 vector of scores for search and experience attributes including color, fat cover, drip, marbling, freshness, packaging, price, taste, tenderness and origin . is a vector of basic demographic variables including gender, age, household size, education , children, senior, income and city dummies to control the city fixed effects 7 , 8 . P is a variable that capture s the relative level of patriotism and is defined as P = CN/US, where CN is the score o f consumer's love 9 for China (on a L iker t scale) and US is the score of love for the US (on the same Likert scale). We use P as a proxy to measure the relative level of patriotism of Chinese consumers, and use this as a control variable in our quality models as we hypothesize that Chinese consumer's level of patriotism will potentially influence the perception of domestic vs US pork. I is a dummy variable indicating consumers past experience on imported pork ( where a value of 1 denotes that respondent have purchased imported pork in the past) and F is a variable capturing the share of frozen pork purchased . Given the basic structure of models that have multiple equations and categorical dependent variables, it is appropriate to employ a multivariate ordered logit/pro bit to estimate the empirical model . The advantage of using a multivariate approach rather than a single equation approach is that we can conduct tests of coefficients across equations to test whether the coefficients on our v ariables of interest are significantly different among countries. Such 7 While we run the mainland regression, city dummy is included in this vector to control the fixed city effects. 8 A description of these v ariables can be found in Table 3 . 9 More precisely we used a Chinese character that represents a patriotic feeling when conveying this questio n. 38 hypothesis test will help us understand the difference in Chinese quality perception among pork from different countries. Nevertheless, given the complexity of implementing a multivaria te ordered logit/probit model, we use an alternative SUR method to achieve the same goal. As an estimation method dealing with a linear equation system, SUR requires dependent variables to be continuous. Although the dependent variables, in our case, are ordered scores generated from Likert - scale questions, we can employ SUR as a substitute or proxy for multivariate ordered regression given the compatibility between the two approaches. To illustrate this compatibility , we show that the results generated by SUR and multivariate ordered logit/probit are similar in terms of significant variables and average partial effects -- the main indicators of interest. Since the average partial effect is one of the main indicators we are using to test for compatibility, it is worth noting that the coefficient s from the result of the ordered logit regression are log odds ratios, rather average partial effect s . To illustrate this , suppose we have an ordered response model as: ( 4 ) Where Y is the ordered dependent variable whose value is ranging from (1) to (5) and X is a vector of independent variables. The coefficient s of the regression using the ordered logit are : 39 ( 5 ) Whose stand ard interpretation is that for one unit increase in the predictor, the response variable level is expected to change by its respective regression coefficient in the ordered log - odds scale ( while all other variables in the model are held constant). The average partial effect we desire can be computed as fol lows: (6 ) Therefore, we calculate the av erage partial effect based on (5 ) before comparing the se two estimation methods. Error! Reference source not found. show s the result of the comparison etween the OLS and ordered logit regressions. Taking the regression results of the Hong Kong data as an example, we can see that significance levels of variables in both estimation methods are the same and the average partial effects are very similar. We therefore find it appropriate to employ SUR instead of multivariate ordered regression to estimate the models in this case. One advantage of using such continuous linear estimation method is that it is computa tionally simple t o conduct hypothesis test s directly with the model coefficients, which inherently are the average partial effects we desire, and are more straightforward to interpret and compare. 40 4.3 Data summary Table 5 presents information about Chinese consumer's evaluation of pork attributes as well as the patriotism score and past experience regarding imported pork. Two observations are worth noting. First, for both mainland Chin a and Hong Kong consumers , their evaluations of Chinese pork are much lower than U.S. and E.U. pork in terms of food safety, animal welfare and environment issue. Thus, how these evaluations relate to consumer's perception of pork quality is of particular interest given our objective. Second, the patriotism score of consumers pork from both countries is of great interest given the scope of our study. 41 Table 5 : Perception statistics Variable s Description Mean or distributio n Mainland China Hong Kong Observations N=600 N=200 Experience and search attributes Color 5 if highest importance 1 if lowest importance 4.66 (0.61) 4.45 (0.70) Fat Cover 4.30 (0.87) 4.23 (0.78) Drip 4.23 (0.93) 3.81 (1.05) Marbling 3.92 (1.04) 3.70 (1.05) Fresh 4.37 (0.89) 4.68 (0.58) Packaging 3.45 (1.20) 3.19 (1.14) Price 4.16 (1.00) 4.26 (0.88) Taste 4.52 (0.79) 4.12 (0.95) Tenderness 4.52 (0.83) 4.09 (0.95) Origin 3.61 (1.28) 3.47 (1.09) Credence attributes Food safety China 5 if highest score 1 if lowest score 3.39 (1.10) 2.79 (1.01) U.S. 3.77 (0.96) 4.00 (0.69) E.U. 3.88 (0.91) 3.91 (0.77) Animal welfare China 5 if highest score 1 if lowest score 3.00 (1.23) 2.51 (1.02) U.S. 3.66 (0.92) 3.64 (0.82) E.U. 3.76 (0.92) 3.74 (0.77) Environment China 5 if highest score 1 if lowest score 2.91 (1.23) 2.32 (1.04) U.S. 3.70 (0.94) 3.71 (0.81) E.U. 3.82 (0.94) 3.78 (0.80) Qu a lity China 5 if highest score 1 if lowest score 3.57 (1.05) 3.19 (1.00) U.S. 3.68 (0.92) 4.01 (0.67) E.U. 3.82 (0.91) 3.85 (0.71) Other variables Patriotism China/U.S. 2.46 (1.49) 1.30 (0.97) Imported pork purchase (IPP) (%) 1 if purchased imported pork before 18.71 51.18 Frozen (%) The share of frozen pork 8.20 6.95 Note: Standard deviation in parentheses. 42 4.4 Results We apply the empirical model to both mainland China and Hong Kong data, in order to Hong Kong will help to understand preference heterogeneity among Chinese consumers. Estimation results for mainland China and Hong Kong are presented in Table 6 and Table 7 . 43 Table 6 : Mainland China regression Variables China U.S. E.U. Coefficient S E Coefficient S E Coefficient S E Male - 0.07 - 0.06 0.05 - 0.05 - 0.05 - 0.05 Age 0.00 0.00 0.00 0.00 0.00 0.00 Household Size - 0.01 - 0.02 0.00 - 0.02 0.02 - 0.02 Education - 0.06* - 0.03 0.03 - 0.03 0.08*** - 0.03 Children - 0.08 - 0.06 - 0.09* - 0.05 - 0.09* - 0.05 Senior 0.04 - 0.06 0.05 - 0.05 0.06 - 0.05 Income 0.01* 0.00 0.00 0.00 0.00 0.00 CityBJ 0.00 - 0.07 0.05 - 0.06 0.09 - 0.06 CitySH - 0.11* - 0.07 - 0.07 - 0.06 - 0.05 - 0.06 Patriotism - 0.01 - 0.02 - 0.05*** - 0.01 / / IPP - 0.08 - 0.07 0.17*** - 0.06 0.06 - 0.06 Frozen 0.00 0.00 0.00 0.00 0.00 0.00 Color - 0.05 - 0.05 0.02 - 0.04 - 0.01 - 0.04 FatCover - 0.04 - 0.03 - 0.03 - 0.03 - 0.03 - 0.03 Drip - 0.01 - 0.03 - 0.02 - 0.03 - 0.02 - 0.03 Marbling - 0.02 - 0.03 0.02 - 0.03 0.01 - 0.03 Fresh 0.11*** - 0.03 - 0.07** - 0.03 0.01 - 0.03 Packaging - 0.06** - 0.02 0.06*** - 0.02 0.02 - 0.02 Price 0.01 - 0.03 0.02 - 0.02 0.02 - 0.02 Taste 0.14*** - 0.05 0.02 - 0.04 0.06 - 0.04 Tenderness 0.00 - 0.04 0.08** - 0.04 0.07* - 0.04 Origin - 0.01 - 0.03 - 0.01 - 0.02 0.01 - 0.02 Food Safety 0.54*** - 0.04 0.47*** - 0.03 0.48*** - 0.03 Animal Welfare 0.05 - 0.04 0.19*** - 0.03 0.15*** - 0.04 Environment 0.18*** - 0.04 0.05 - 0.03 0.14*** - 0.04 Constant 1.00*** - 0.34 0.66** - 0.30 0.03 - 0.29 Observations 600 / 600 / 600 / R - squared 0.61 / 0.60 / 0.62 / Note: SE means Standard Errors. *** p<0.01, ** p<0.05, * p<0.1 44 Table 7 : Hong Kong regression Variables China U.S. E.U. Coefficient S E Coefficient S E Coefficient S E Male 0.10 0.11 0.11 0.10 0.14 0.09 Age - 0.00 0.00 0.00 0.00 0.00 0.00 Household Size 0.02 0.05 - 0.00 0.04 - 0.03 0.04 Education 0.10* 0.06 0.11** 0.05 0.04 0.05 Children - 0.08 0.10 - 0.04 0.09 0.03 0.09 Senior - 0.13 0.11 - 0.05 0.10 - 0.06 0.10 Income - 0.00 0.00 0.00 0.00 - 0.00 0.00 Patriotism 0.12** 0.05 - 0.00 0.04 / / IPP - 0.11 0.10 0.06 0.09 - 0.17** 0.08 Frozen - 0.00 0.00 - 0.00 0.00 0.01** 0.00 Color - 0.02 0.08 - 0.08 0.07 - 0.09 0.07 Fat Cover - 0.01 0.07 - 0.02 0.07 0.00 0.06 Drip 0.06 0.06 - 0.02 0.05 0.03 0.05 Marbling 0.08 0.06 0.02 0.06 - 0.01 0.05 Fresh 0.01 0.10 - 0.02 0.08 - 0.04 0.08 Packaging - 0.03 0.05 - 0.03 0.05 0.00 0.04 Price 0.07 0.06 0.08 0.05 0.00 0.05 Taste 0.02 0.08 0.16** 0.07 0.12* 0.07 Tenderness - 0.12 0.08 0.01 0.07 - 0.01 0.07 Origin 0.00 0.06 0.05 0.05 0.08* 0.05 Food Safety 0.55*** 0.07 0.24*** 0.07 0.34*** 0.06 Animal Welfare 0.12 0.09 0.07 0.06 - 0.10 0.07 Environment 0.06 0.08 0.08 0.06 0.32*** 0.06 Constant 0.73 0.56 1.38*** 0.53 1.29*** 0.50 Observations 200 / 200 / 200 / R - squared 0.58 / 0.25 / 0.42 / Note: SE means Standard Errors. *** p<0.01, ** p<0.05, * p<0.1 45 4.4.1 Pork Credence Attributes In terms of credence attributes (food safety, animal welfare, and environmental issues), food safety is the most important criterion for both mainland and Hong Kong consumers when evaluating pork quality. This result is not surprising given their exposure to numerous food incidents in China over the past decade. Through hypothesis testing we find that, for mainland consumers, the relationships between food safety and quality are not significantly different among pork produced in the three regions. This indi cates that mainland consumers have strong preference for food safety attributes in pork regardless of origin. For Hong Kong consumers, however, the role of food safety is more important when evaluating the quality of pork from China. This is evidenced by t he hypothesis test, where the relationship between food safety and quality of Chinese pork is significantly higher than for pork from developed countries. As Hong Kong consumers perceive that the safety of Chinese pork is worse than that of other developed countries pork (see Table 3), their preference for safe pork explains why Hong Kong consumers are not satisfied with the quality of pork imported from mainland China. The role of environmental issues is also observed in the mainland regression, where t he relationships between environment and quality are significant for pork from all three regions. The hypothesis test indicates that the relationship for Chinese pork is significantly stronger than for pork produced in the U.S. and the E.U. This finding in dicates that environmental issues play a more important role in determining the quality of domestic pork. Given that consumer rating of Chinese pork in terms of environment is much lower than that of U.S. and E.U. pork (Table 3), this result reveals that e nvironmental concern is another reason why 46 mainland consumers perceive domestic pork to be of lower quality. In contrast, Hong Kong consumers do not relate environmental issues to pork quality, and this could be attributed to the fact that pork from all th ese regions is imported, which has a lesser impact on the local environment. For mainland consumers, the relationship between animal welfare and quality are significant for U.S. and E.U. pork (which is not significant for Hong Kong consumers), indicating that mainland consumers recognize foreign animal - friendly production. For cons umers who support animal - friendly production, animal welfare is an important production process attribute used to evaluate pork quality. However it is worth noting that this result only applies to imported but not domestic pork. 4.4.2 Pork Search and Exp erience Attributes In terms of pork search and experience attributes, freshness and packaging are significant in the mainland regression, while there is no significant attributes (search and experience) in Hong Kong regression. The result can be interprete d as follows: the more mainland consumers prefer freshness, the more likely they are to perceive pork to be of higher quality. This result is plausible since for mainland consumers, Chinese pork means domestic production and fresh products , whereas pork fr om other countries means long - distance transportation and often frozen or chilled product. Moreover, the more consumers care about packaging, the more likely they are to attribute higher quality to pork from developed countries. This can be explained by th e fact that imported pork can only be purchased in high - end supermarket s in the mainland, which are often sold pre - packaged. 47 As there is little pork production in Hong Kong, pork from all these three regions is imported, which explains why freshness and pa ckaging are no longer significant in the Hong Kong model. Moreover, search/experience attributes are not significant, indicating that Hong Kong consumer preferences for these attributes do not affect their quality perception. 4.4.3 Patriotism Effect The effect of patriotism on Chinese consumer food preferences is revealed in the mainland low quality score to U.S. pork, even though their level of patriotism does not af fect their perception of domestic pork. This is due primarily to the fact that Chinese patriotism or nationalism is affected by political and economic issues; that is, the political affairs or the economic conflict between China and other countries will gi ve rise to a negative impression of foreign products for Chinese consumers (Sean 2013). This patriotism effect is also present for Hong Kong consumers , but in a slightly different form. Patriotism is found to have a positive effect on Chinese pork qualit y perception, but does not negatively affect the perception of US or E.U. pork products. This can be attributed to the fact that Hong Kong is often viewed as an international metropolis, where people are used to consuming products from all over the world. Also, as Hong Kong had a long history - welcome cultures as well as products from all over the world. Thus, consumer's patriotism is not likely to induce a negative impr ession on foreign product, but rather a positive impression on their own product. 48 4.4.4 Demographic Impacts In the mainland China regression, education has a significantly negative effect on the quality perception of Chinese pork and a positive effect on the quality perception of U.S. and E.U. pork, indicating that more educated consumers in mainland China are more likely to perceive pork produced in developed countries (U.S. and E.U.) to be of higher quality. This is due to the fact that more educated con sumers are more likely to know and understand that pork from developed countries is produced under strict and well - organized systems. Past experience consuming imported pork also significantly influences consumers' perception of pork quality, which has a n egative effect on Chinese pork quality and a positive effect on U.S. and E.U. pork. This indicates that for consumers who have purchased (and therefore consumed) imported pork, they are more familiar with pork produced in developed countries and place a hi gher quality value on them. In the Hong Kong regression, income has a significantly negative effect on the quality perception of Chinese pork and a significantly positive effect on U.S. pork. One potential explanation is that, for Hong Kong consumers, the y always make purchasing decisions among imported pork including both Chinese pork and U.S. pork. Consumers with higher income are more likely to purchase U.S. pork, which is often more expensive. As a result, they are more familiar with the product and th us perceive it to be of higher quality. However, for mainland China consumers, pork from developed countries is less accessible as it can only be found in limited international supermarket or high - end markets, and domestic pork is much more popular and dominant in mainland China. Therefore, even though some mainland 49 consumers with high income can afford imported pork, they are not as familiar with imported pork products. This can also help explain why the income effect in mainland China is i nsignificant. 4.5 Conclusions Consumers in mainland China and Hong Kong exhibit different preferences for pork, in terms of both tangible and intangible pork attributes. Specifically, among the three credence attributes evaluated, food safety is the most important criteria for consumers in mainland and Hong Kong to evaluate pork quality, no matter whether the pork is domestic or imported. Therefore, for foreign pork suppliers, their advantages on food safety control and quality management may help them exp lore sales in China. In addition, due to the rising concerns over the domestic environment, the environmental aspect of pork production is likely to In terms of tangible attributes (search and experience attributes), mainland consumers care more about the freshness and packaging of pork while they compare the quality of pork from not related to their quality perception. We attribute the insignificance of other physical pork characteristics to the fact that mainland consumers are not familiar with foreign pork. With this understanding, freshness and packag ing , which are the more tangible attributes unde r The role of patriotism in determining pork quality is also identified in this study; the more patriotic mainland consumers are, the more likely they are to perc eive foreign pork to be of 50 lower quality while patriotic Hong Kong consumers are more likely to perceive Chinese pork to be of higher quality. This suggests that, U.S. pork suppliers should promote U.S. pork carefully -- labeling U.S. pork could potentially help boost sales in China if marketing efforts promote the safety aspect of the product, however, these advantages may be neglected by consumers who are more patriotic. 51 5. Consumer WTP for quality attributes in domestic vs. imported pork 5.1 Introducti on In this chapter choice experiment approach, and estimate willingness to pay (WTP) for select attributes . The advantage of measuring WTP is that it provide s pricing information and help s compare the relative importance of specific pork characteristics . As we have shown the relative importance of credence attributes for Chinese consumers evaluating pork quality, measuring will help understand th eir current attitudes toward domestic and imported pork by revealing the monetary values of these pork characteristics . A classification of various WTP measurement methods is shown in Figure 4 . Breidert et al. (2006) divided WTP measurement approaches into the revealed preference approach and the stated preference approach, based on data collection methods. According to the definition given by Louviere et al. for the existing pr oduct, which only explains the existing situation. Stated preference refers can encompass hypothet ic al scenarios. Stated preference approaches allow researchers to test various scenarios or hypothes e s by adding shifts or shocks to existing markets in order to forecast market development and make policy suggestions. Given that our goal is to evaluate th e potential for US pork in China a market not widely accessible to Chinese consumers to this point this study focuses 52 Figure 4 : Classification of WTP measurement methods Source: Breidert et a l. 2006 53 There are two branches of the stated preference approach, direct and indirect surveys. The former directly asks respo presents commodity profiles with various prices and asks respondents to choose t he one they prefer to purchase (Marbeau 1987). As consumers have no incentive to reveal their true WTP in the direct surveys, and their valuation is hard to relate to the actual purchasing decision, a direct survey approach may result in extremely unreason able outcomes (Hanna and Dodge 1995). With this understanding, we will employ the indirect approach, specifically a choice experiment Choice experiment s have pr eference s regarding food products (Tonsor et al. 2009; Ubilava and Foster 2009; Loureiro and Umberger 2007; Lust et al. 2003 ). The main idea s according to their choice among various alternatives. Based on the Lancastrian approach to consumer theory, choice experiment s assume that the choice of the decision maker depend s on both the objective characteristics embodied in the product per se (e.g. , quality attributes) and the subjective characteristics of the decision maker (e.g. , education level, income level, etc.). This can be explained as follows: Lancaster (1966) indicated that the utility derived by the consumer does not arise from the good pe r se , but rather from the characteristics possessed by the good. In addition, he pointed out that goods can capture various characteristics that could be shared by other goods, but in aggregate the goods could encompass characteristics different from other s. With this understanding, the combination of multiple characteristics of the good 54 jointly determines the utility to the consumer after a choice is made . The basic premise of c onsumer theory is that individuals seek to maximize their utili ty subject to th eir resource allocations or budget constraints. Consumer s subjective characteristics determine their preferences, which in turn will determine the level of satisfaction (utility) obtained from purchasing goods. Therefore, the essence of choice experiment s is to simulate real purchasing situation s and ask consumers to make a choice over multiple alternatives. It is reveal ed. As the alternatives in the choice sets can be differentiated by tangibl e and intangible characteristics in the products, many studies have been conducted to investigate consumer preference s for various for pork appearance in 23 countries, Ngapo et al. (2007) differentiated 16 different pork chops in terms of color, fat cover, marbling, and drip and found that consumers from different countries have heterogeneous preferences for these tangible attributes. Specifically , color and fat cover played an important ro decision. There are also many studies investigating consumer preference s for intangible attributes. For attributes, Lusk e t al. (2003) designed a choice experiment in which consumers from France, Germany, the U.K., and the U.S. were asked to make choices upon various types of rib - eye steak. They found consumers from these countries represented heterogeneous preferences: Frenc h consumers perceived beef produced without hormones having higher value than U.S. 55 consumers did , and European consumers perceive beef produced without genetically modified corn as having higher value than U.S. consumers. In terms of animal welfare attributes, Tonsor et al. (2009) conducted a choice experiment with a focus on investigating the role of crate - in the U.S. state of Michigan. Pork chops varied in the choice experimen t in terms of price, farm size, production practice (gestation crate - free or typical), and country of origin. Through a random parameter logit model (RPL) and latent class models (LCM), they identified consumers heterogeneous preference s for pork chop att ributes. As there are two ways to promote crate - free production (voluntary versus a formal ban), the results indicated that the former received a significant premium from consumers, while the latter failed to . As the negative effects from meat production on environment ha ve been increasingly understood , the choice experiment approach has also been used to study meat consumer s preference s for environmental - friendly attributes. Zanoli et al. (2013) studied Italian beef consumers ly friendly attributes through a hypothetical choice experiment. They differentiated beef alternatives by production method (organic, conventional, or GM), production location (domestic or abroad), breed origin (local or no n - local), animal welfare, and appearance. Specifically, they introduced domestic production and breed origin as proxies for environmental attributes; the former was a proxy for food mileage, and the latter was a proxy for biodiversity conservation. As the results showed a significant WTP for organic beef and a relationship between the environmental issues and the 56 variation in WTP for organic, the authors concluded that environmental concern s played a role on Italian consumers demand for beef. In Chi na, choice experiment s have only recently been used to study consumer preferences for various food products (e.g., Ortega et al. 2011; Bai et al. 2013). For example, as Chinese as a result of various food saf ety scandals , Ortega et al. (2011) administered a choice experiment to estimate urban Chinese c onsumers WTP for multiple food safety attributes in pork. To take consumer heterogeneity into consideration, the authors constructed an RPL model and LCM models and found that Chinese consumers have a strong demand for food safety assurance. Specifically, among the attributes evaluated , government certification received the highest premium, followed by third - party certification, traceability , and a product - speci fic information label. Another study conducted by Ortega et al. (2011) investigate d Chinese consumers preference for ultra - high temperature (UHT) fluid milk with respect to food safety attributes. Five two - level at tributes including price, shelf life, government certification, third - party certification, and brand were incorporated in the choice experiment, and a RPL model was used to capture consumer preference heterogeneity. They found that government certification received the highest WTP from c onsumers, followed by the product brand and third - party certification. Longer - shelf - life UHT milk (longer than three months) was less valuable for consumers, receiving a negative WTP. I n order to understand the role of traceability on Chinese consumers purchasing decision s , Bai et al. (2013) designed a choice - based conjoint experiment to study consumer preferences 57 for traceable milk. The authors used a conditional logit model to analyze the choice experiment data and found that consumers have positive p reference for milk with traceability. Consumer WTP was related to the certificate issuer, where government certification received the highest WTP, followed by industrial association certification and third - party certification. As animal welfare and environ ment al issues have received increased attention in China, a few studies have been conducted to estimate consumers preference s for these credence attributes . For example, Zhao et al. (2011) interviewed visitors to gardens and zoos in China. The result of t he direct survey showed that Chinese citizens have a positive willingness to pay for animal welfare improvement. However, this work does not directly concern animals for food consumption, and thus more research is r equired to investigate consumer preferenc e and demand for animal - friendly food products . To the best of our knowledge, no existing research has addressed Chinese consumers In chapter 4, we have shown that Chinese consumers may relate animal welf are and environmentally friendly attributes to pork quality. In this study , a choice experiment will be employed to help understand whether these credence attributes can receive a positive price premium from Chinese consumers. In addition to pork credence attributes, country of origin, another important pork attribute, is also of great interest in this study. domestic pork versus pork imported from the U.S. and other countries. This will allow us to asses s the market potential for US pork in China. To achieve this goal, the potential effects of patriotism are also of importance. There is a substantial literature which has revealed 58 iotism or nationalism (Hong and Wyer 1989; Shimp and Sharma 1987; Zajonc 1980; Han 1988). Consumers who are patriotic may think that purchasing foreign products can be viewed as being dishonest to their country, and will hurt the domestic economy and resul t in job losses perception of product quality; a highly patriotic consumer is more likely to overestimate domestic product quality and underestimate that of imported pro ducts (Netemeyer et al. 1991; Umberger et al. 2003) . In our case, a comparison between mainland China and Hong Kong will help understand the effect of patriotism on Chinese consumer behavior. Given the extreme difference in socio - political nature between mainland China and western countries, patriotism may result in mainland Chinese consumers obtaining negative utility from consuming foreign pork . However, as Hong Kong has a long history of being a British colony and international city, Hong Kong consumers - just increase their favor of Chinese pork as opposed to receiving disutility from foreign pork. the effect on their perception of Chinese pork. In this chapter, we will study whether the relative level of patriotism gives rise to different purchasing decisions in mainland China and Hong Kong. 59 T his study has three objectives. The first objective is to examine the monetary importance of the pork credence attributes by estimating consumers WTP for them. The second objective is to estimate consumers WTP for imported pork, es pecially U.S. pork, in order to reveal the potential of U.S. pork in the Chinese market. The last objective is to understand how consumers patriotism affects their purchasing decision of domestic and imported pork, which will help understand the nature of the Chinese market. 5.2 Theoretical framework The theoretical framework of this study follows theory. The main highlights of Lancaster theory are as follows: 1) characteristics in the good give rise to u tility; 2) a good can incorporate multiple characteristics, which can be shared by other goods; and 3) goods in aggregate can embody different characteristics from those embodied in the goods separately. Applying this idea to our case, pork can be viewed a s a combination of multiple attributes, including search attributes (price, expiration date, appearance, etc.), experience attributes (tenderness, taste, etc.), and credence attributes (food safety, animal welfare, environmental issues). In concordance wit h our objective, we focus on credence attributes and examine how country of origin affects consumer preferences utilizing a choice experiment. As choice experiment s are designed to imitate the real purchasing decisions the consumer faces , it is worth ill ustrating consumers purchasing decision making process. Louviere et al. (2000) summarize this process stage - by - stage in Figure 5 . 60 Figure 5 : Source: Louviere et al. (2000) As shown in Figure 5 , the consumer must be motivated to purchase the product in stage 1 (e.g. , famil y hopes to eat pork for dinner). Then, in stage s 2 and 3, the consumer needs to evaluate and compare the attributes represented by the pork alternatives. Based on their needs and preferences, the appropriate type of pork will be chosen in stage 4 and 5 to maximiz e utility . After consumption , the credence attributes are re - evaluated. This re - evaluation in stage 6 help s determine future purchasing decisions. The basic idea behind this decision making process is that consumers select the product ( from mult iple alternatives ) that maximize s their utilit y . T he random utility model (RUM) is used in this study, which has been widely applied to studying consumers demand for 61 products with multiple attributes (Manski 1977; McFadden 1973) . Regarding the assumption discussed above, individual n maximizes his or her utility , U nit , from selecting a good i from a finite bundle of J alternatives in choice set C in scenario t (Train 2003). Therefore, individual n will choose product i from J alt ernatives if and only if: (7) U nit > U njt j i , Specifically, utility U nit is a combination of both a deterministic a nd stochastic term: (8) U nit = V nit nit Based on equation s (7) and (8 ), we obtain the probability of individual n choosing good i from J alternatives in choice set C: (9) And: (10) In order to identify the above probability, it is necessary to know the distribution of the error term . The typical way is to use the conditional logit model and assume that the error terms 62 are independently and identically distributed and follow an extreme value (Weibull) distribution. Then, the probability of choosing good i can be expressed as: (11) T he deterministic term V nj t , also called indirect utility, is defined as: (12) Where is the th attribute in product j , and respective coefficients. Using the conditional log it model to estimate the utility equation assumes that the coefficients of the product attributes are all constants. With this assumption, consumer preferences are treated as homogeneous, which could be very restrictive. The RPL model is introduced as a more robust tool to estimate random utility models, which allows heterogeneity in consumer preferences (McFadden and Train 2000). Under RPL, the allowing for h eterogeneous preferences among consumers. To specify the RPL model, researchers have typically assumed a normal or lognormal distribution for the attribute coefficients. One flaw of this assumption is that a normal or lognormal distribution could yield value s for the price coeffi cient that are either negative or close to zero , resulting in a counter - intuitive output for WTP (Scarpa et al. 2008). 63 A potential solution employed in this study is to assume that the coefficient for price is constant and the coefficients for other variables follow normal distributions . This assumption facilitates the interpretation of WTP as it follows a normal distribution. However , a fixed price assumption could also prove to ces for cost are assumed to be homogeneous. Therefore, researchers face a tradeoff between computational convenience and robustness. Following Train (2003), the probability given in equation ( 9 ) becomes: (13) Where the distribution of all the random parameters f (·) is normal, except for price, which is fixed in this study. 5.2.1 The choice experiment - to - face consumer interviews were conducted at point of purchase in Guangzhou and Hong Kong in the summer of 2014. In each city, the sample size is 200. Guangzhou is the third largest city in China and the capital of Guangdong province. As consumers in both cities have simil ar eating habits (Cantonese cuisine), choosing Guangzhou as the representative of mainland China could minimize unobserved differences in terms of tastes and other regional issues. In the choice experiment, respondents are facing three alternatives: two pork products with five different attributes ( price, country of origin, food safety claim, animal welfare claim, and an environment ally - friendly claim), and an opt - out option so that the respondent can choose 64 neither of the pork alternatives. A description of the attributes with their corresponding levels is shown in Error! Reference source not found. . Before the respondents took the choice xperiment, the description of each pork attribute was provided. The details are given in Table 9 . Table 8 : Attributes used in choice experiment Attribute Levels Description Price 20, 30, 40, 50 (30, 45, 60, 75) 10 Price expressed in RMB (HKD) per 500 g of pork Country of origin CN, US, IO Pork from China, the U.S., or international other countries Food safety claim Yes, No Yes if the pork has food safety claim; otherwise no . Animal welfare claim Yes, No Yes if the pork has animal welfare claim; otherwise no . Environmental - claim Yes, No Yes if the pork has an environmenta l l y - friendly claim; otherwise no . 10 The numbers in parentheses are the price levels for Hong Kong in HKD. 65 Table 9 : Choice experiment description Choice Experiment Next you will be provided ten choice situations, which contain the following attributes: Price Price is expressed in RMB (yuan) per jin of fresh chilled pork loin (500g). Country of Origin Origin where the pork was produced: mainland China, Imported -- US or Imported -- Other. Food Safety Claims Yes, indicates this product has any claims indicating the use of food safety practices that will reduce your likelihood of becoming ill. For example: certification, additive information and traceability. Animal Welfare Claims Yes, indicates this product has any claims regarding the use of animal welfare practices. For example: type of production, standards and procedures to ensure that pigs are treated without cruelty and are fed with food of a certain quality. Environmental Claims Yes, indicates this product has any claims regarding environmental standards in the pig farms. For example: water quality, soil quality, standards for limiting the carbon footprint and for maintaining a sustainable ecosystem. The choice sets were created using a D - optimal design generated from the full - factorial candidate set using a modified Federov search algorithm using the software Ngene 11 . The experimental design was based on a linear (in parameters) utility specification with null priors. Our design allowed for estimation of main and specific two - way interaction effects between country of origin and the credence attributes. A no purchase or opt - out option was 11 D - optimal designs minimize the D - error of the design, which is calculated as the weighted determinants of the variance - covariance matrix of the design, where the weight is an ex ponential weight equal to the reciprocal of the number of parameters to be estimated. 66 included to better simulate a real - world decision scenario ( Adamowicz et al ., 1998 ). To reduce the probability of respondent fatigue, the choice sets were blocked into four groups, with each survey participant evaluating ten choice scenarios. 5.2.2 Empirical model and willingness - to - pay estimation According to the design of the choice experiment, we construct two empirical models for indirect utility. Model I is given below: (14) Where P stands for price, US and IO are the country of origin, the U.S. and international other countries, respectively. FS, AW and EN are the dummy variables indicating whether the pork is provided with food safety, anim al welfare, and environmentally fri endly claims. As we incorporate the opt - out option in this experiment, the constant in this model refers to alternative specific constant which is the opt - out option , indicating consumers prefer domestic pork without any quality attributes . To el iminate potential confounding effects between the intercept and the attribute variables, the data was effect coded , by which , the constant is just the coefficient for opt - out. With effect coding, the estimated WTPs have to be multiplied by 2 to generate appropriate marginal WTP values . F ollowing the empirical model, the measure of the WTP for attributes is given by: (15) 67 Where WTP k 12 is the willingness to pay for k th attribute, k is the estimated coefficient for k th attribute, 1 is the estimated coefficient for price. The multiplication by 2 here only applies to the attributes , which are effect coded. The above WTP is generated as follows. Suppose we are estimating the WTP for attribute k . We can have two utility equations which are identical except that the levels of attribute k are different: ( 16 ) and ( 17 ) If we set these two utilities equal, then we will have: (18) With effect coding, the two levels of attribute k bec omes 1 and - 1, then equation (17 ) becomes : (19) 12 Technically, is the partial derivative of the utility function with respect to k th attribute while the other attributes are not present. This is due to the existence of the interaction terms in the utility function. 68 And the monetary difference (WTP) becomes: (20) We do not incorporate the patriotism variable in model I , so this model mainly focuses on WTP for selected pork attributes without considering the potential estimatin g the monetary effect of patriotism is of particular interest. Therefore, model II is set up below by add ing an interaction term between the patriotism score and the dummy variable US 13 . The patriotism score 14 is measured by comparing consum China and the U.S. (21) Where PA stands for the patriotism score. T he WTP for U.S. pork is measured by: (22) 13 measured between China and th e U.S only. 14 We used Likert - associated patriotism score is equal to the ratio China/US. Specific notation is provided in chapter 4. 69 As different patriotism scores indicate different levels of patriotism of consumer s , this corresponding patriotism level changes . A nonparametric test proposed by Poe et al. (2005) is employed to 5.3 Empirical results 5.3.1 RPL estimates and WTP for selected attribute s R esults for the estimated Model I and Model II are shown in Table 10 and Table 11 . We modified the WTPs in both cities by converting RMB and HKD into USD 15 . In Table 10 , we can see that almost all of the attributes have significant effect s on y . All the credence attributes in pork are perceived as valuable, but of country of origin for pork vary between Guangzhou and Hong Kong. Specifically, without any credence attribute claims, pork from the U.S. is less valuable than Chinese pork for consumers in Guangzhou , but more valuable for consumers in Hong Kong. Similarly, pork imported from other countries is less valuable than Chinese pork for consumers in Guangzhou but is not significantly different from Chinese pork for consumers in Hong Kong. Two interaction terms have significantly positive ef fects on consumer utility, indicating the complementary effects among the interacted attributes (Ubilava and Foster 2009). The significantly positive estimates of the interaction between the U.S. and food safety indicates that, although U.S. pork is less v aluable for consumers in Guangzhou, it can . 15 USD to CNY exchange rate in 2014 June was 6. 239, and USD to HKD exchange rate in 2014 June was 7.752. 70 Similarly, f or consumers in Hong Kong, US origin and food safety claim are complementary , as providing both attributes will f urther augment Moreover, providing a US origin and environmentally utility. The statistical significance of the standard deviation coefficient for each attribute asserts the hypot hesis of preference heterogeneity. 71 Table 10 : Parameter estimates of Model I from RPL Guangzhou Hong Kong Coefficient S.E. Coefficient S.E. PRICE - 0.05 (0.01) *** - 0.06 (0.01) *** US - 0.72 (0.15) *** 0.37 (0.12) *** IO - 0.81 (0.14) *** 0.02 (0.12) FS 0.85 (0.11) *** 0.67 (0.10) *** AW 0.40 (0.09) *** 0.33 (0.09) *** EN 0.55 (0.09) *** 0.48 (0.09) *** US*FS 0.20 (0.12) * 0.41 (0.13) *** IO*FS - 0.11 (0.12) 0.14 (0.11) US*AW - 0.11 (0.12) 0.06 (0.12) IO*AW - 0.05 (0.12) - 0.15 (0.12) US*EN - 0.09 (0.12) - 0.26 (0.11) ** IO*EN - 0.02 (0.12) - 0.05 (0.12) OPTOUT - 2.38 (0.34) *** - 5.59 (0.37) *** STDEV(US) 1.08 (0.11) *** 0.57 (0.09) *** STDEV(IO) 0.69 (0.10) *** 0.30 (0.11) *** STDEV(FS) 1.06 (0.10) *** 0.73 (0.10) *** STDEV(AW) 0.24 (0.13) * 0.32 (0.10) *** STDEV(EN) 0.29 (0.08) *** 0.26 (0.08) *** N 2000 2000 No. of parameters 25 25 Log - likelihood - 1712 - 1504 Pseudo R - squared 0.292 0.354 AIC/N 1.579 1.442 *** p<0.01, ** p<0.05, * p<0.1 Standard E rrors (S.E.) in parentheses. Note: The RPL model was estimated using NLOGIT 4.0. 72 Table 11 : Parameter estimates of Model II from RPL Guangzhou Hong Kong Coefficient S.E. Coefficient S.E. PRICE - 0.05 (0.01) *** - 0.06 (0.01) *** US - 0.34 (0.20) * 0.69 (0.16) *** IO - 0.77 (0.15) *** 0.03 (0.13) FS 0.96 (0.12) *** 0.68 (0.11) *** AW 0.39 (0.09) *** 0.33 (0.09) *** EN 0.62 (0.09) *** 0.51 (0.09) *** US*FS 0.22 (0.13) * 0.47 (0.15) *** IO*FS - 0.09 (0.13) 0.13 (0.12) US*AW - 0.12 (0.13) 0.08 (0.13) IO*AW - 0.02 (0.13) - 0.13 (0.13) US*EN - 0.08 (0.12) - 0.25 (0.12) *** IO*EN - 0.09 (0.13) - 0.07 (0.13) PA*US - 0.16 (0.06) *** - 0.30 (0.10) OPTOUT - 2.37 (0.34) *** - 6.01 (0.45) *** STDEV(US) 1.16 (0.12) *** 0.38 (0.21) * STDEV(IO) 0.77 (0.10) *** 0.38 (0.10) *** STDEV(FS) 1.07 (0.11) *** 0.68 (0.10) *** STDEV(AW) 0.23 (0.12) ** 0.35 (0.11) *** STDEV(EN) 0.27 (0.10) ** 0.25 (0.10) * N 2000 2000 No. of parameters 27 27 Log - likelihood - 1700 - 1496 Pseudo R - squared 0.297 0.357 AIC/N 1.571 1.438 *** p<0.01, ** p<0.05, * p<0.1 Standard Errors (S.E.) in parentheses. Note: The RPL model w as estimated using NLOGIT 4.0. Using the estimates of Model I WTP for ea ch quality attribute is calculated and shown in Table 12 . To calculate the 95% confidence interval, a parametric bootstrapping technique proposed by Krinsky and Robb (1986) is employed. Specifically, the Krinsky - Robb approach is used to simulate an asymptotic distribution of the WTP by randomly drawing from a multivariate normal distribution, constructed by the combination of the 73 coefficient estimates and the associated variance - covariance matrix from the RPL model. Relative to estimating a standard error using the delta method, Krinsky - Robb approach produces an analogous result without the assumption that WTP is symmetrically distributed (Hole, 2007). Table 12 : Willingness to pay, mean values [95% confidence interval] Attributes Guangzhou Hong Kong US $ - 4.93 [ - 7.37, - 2.80] $1.58 [0.57, 2.59] IO $ - 5.52 [ - 7.82, - 3.69] $0.09 [ - 0.98, 1.10] FS $5.86 [4.17, 7.88] $2.87 [2.02, 3.78] AW $2.78 [1.53, 4.18] $1.41 [0.62, 2.19] EN $3.75 [2.48, 5.13] $2.07 [1.32, 2.84] As shown in Table 12 , among all credence attributes, the food safety claim received the highest WTP from consumers (5.86 USD in Guangzhou and 2.87 USD in Hong Kong) , followed by the environmentally friendly claim (3.75 USD in Guangzhou and 2.07 USD in Hong Kong), and the animal welfare claim (2.78 USD in Guangzhou and 1.41 USD in Hong Kong). This result is consistent with our previous finding that Chinese consumers v iew food safety as the most important criteria when evaluat ing pork quality. As a result, their WTP for this attribute is the highest. Pork from foreign countries is perceived differently in each city; it is less valuable for Guangzhou consumers and more valuable for Hong Kong consumers. One potential explanation is that mainland China consumers are used to consuming domestic pork while Hong Kong consumers are more familiar with pork from other countries. Additionally, patriotism could be another factor in fluencing consumer preferences. As we have shown in 74 examine this idea by estima ting model II . Another thing worth noting here is that, even though mainland consumers have a negative WTP for U.S. pork, their WTP could be positive if U.S. pork is provided with a food safety claim (the coefficient for US*FS is significantly positive). 5.3.2 Guangzhou versus Hong Kong Consumers After obtain ing consumer WTPs for pork attributes in Guangzhou and Hong Kong, it is interesting to compare the difference of these WTPs in order to identify the potential different preferences between these two cities. To achieve this, we set up several one - sided test s using the nonparametric method 16 proposed by Poe et al. (2005) to examine the estimates from Model I . The results of the one - sided tests are shown in Table 13 . Table 13 : Comparison of WTPs for pork attributes in Guangzhou and Hong Kong Hypothesis Guangzhou Hong Kong p - value $ - 4.93 $1.58 0.00 $ - 5.52 $0.09 0.00 $5.72 $2.87 0.04 $2.78 $1.41 0.03 $3.75 $2.07 0.11 $2.26 $6.21 0.00 From Table 13 than that of Hong Kong consumers. This finding is consistent with our previous discussion that mainland consumers p refer domestic pork than pork from other countries, while Hong Kong consumers have a positive perception of foreign pork. One thing worth noting is that, 16 We refer to this method as the - 75 positive, but t consumers are mo re sensitive to food - welfare is significantly higher, t heir WTP for an environmentally friendly claim is not significantly higher than that of Hong Kong consumers. This could be explained by the close geogr aphic location between these two cities; consumers in both cities have similar surroundings and therefore similar pref erences for the environmentally friendly claim. 5.3.3 Patriotism effects on WTP for U.S. pork The distribution of the patriotism score in Guangzhou and Hong Kong is shown in Figure 6 . According to Figure 6 , we can see that consumers in Guangzhou are more patriotic more scattered, while the dominant group h as a score of 1. In order to understand how patriotism for consumers in each city based on their distributions (the most frequent groups) and calculate the WTP for U.S. po rk at the selected level of patriotism, using the estimates of Model II . For Guangzhou consumers, we select PA=1 (non - patriotic), PA=1.66 (slightly patriotic), and PA=5 (very patriotic). For Hong Kong consumers, we select PA=0.66 ( unpatrio tic ), PA=1 (non - patriotic), and PA=1.5 (slightly patriotic). The WTP for U.S. pork with different combinations of attributes at each level of patriotism , along with the 95% 76 confidence interval, are shown in Table 14 . Also, using Poe - test , we tested the hypotheses that in both cities, the WTP for U.S. pork under a higher level of patriotism is greater or equal to the WTP for U.S. pork with a lower level of patriotism. Figure 6 : Frequency of patriotism scores in each city 0 10 20 30 40 50 60 70 80 90 0.2 0.25 0.33 0.4 0.5 0.6 0.66 0.75 0.8 1 1.25 1.33 1.5 1.66 2 2.5 3 4 5 Guangzhou Hong Kong 77 Table 14 : WTP for different types of U.S. pork Attributes Guangzhou Hong Kong Patriotism Score=1 Patriotism Score=1 US $ - 3.72 [ - 5.89, - 1.63] $1.71 [0.69, 2.78] US*FS $4.29 [2.13, 6.65] $6.30 [5.13, 7.60] US*AW $ - 2.02 [ - 4.15, 0.01] $3.31 [2.16, 4.53] US*EN $ - 0.40 [ - 2.31, 1.50] $2.73 [1.72, 3.69] US*FS*AW $5.99 [3.48, 8.70] $7.90 [6.57, 9.39] US*FS*EN $7.61 [5.27, 10.20] $7.31 [6.10, 8.67] US*AW*EN $1.30 [ - 0.85, 3.41] $4.33 [3.17, 5.51] US*FS*AW*EN $9.31 [6.44, 12.23] $8.91 [7.51, 10.56] Patriotism Score=1.66 Patriotism Score=0.66 US $ - 4.44 [ - 6.61, - 2.54] $2.06 [1.04, 3.15] US*FS $3.54 [1.47, 5.81] $6.63 [5.34, 7.93] US*AW $ - 2.74 [ - 4.85, - 0.95] $3.65 [2.52, 4.87] US*EN $ - 1.17 [ - 3.01, 0.62] $3.06 [2.00, 4.12] US*FS*AW $5.24 [3.01, 7.83] $8.22 [6.80, 9.70] US*FS*EN $6.86 [4.75, 9.35] $7.64 [6.33, 8.97] US*AW*EN $0.58 [ - 1.21, 2.37] $4.65 [3.41, 5.94] US*FS*AW*EN $8.56 [6.07, 11.38] $9.22 [7.75, 10.80] Patriotism Score=5 Patriotism Score=1.5 US $ - 8.13 [ - 11.41, - 5.52] $1.25 [0.17, 2.31] US*FS $ - 0,14 [ - 3.04, 2.83] $5.82 [4.62, 7.02] US*AW $ - 6.42 [ - 9.54, - 3.64] $2.85 [1.62, 4.08] US*EN $ - 4.83 [ - 7.78, - 2.29] $2.23 [1.13, 3.29] US*FS*AW $1.57 [ - 1.37, 4.66] $7.41 [6.10, 8.90] US*FS*EN $0.37 [3.15, 6.01] $6.80 [5.61, 8.15] US*AW*EN $ - 3.12 [ - 5.92, - 0.42] $3.82 [2.59, 5.56] US*FS*AW*EN $4.86 [1.85, 8.07] $8.40 [7.01, 9.90] As shown in Table 14 , consumers in both cities present a decreased WTP for U.S. pork as their relative patriotism score increases. Three points are worth noting here. First, whil e all Hong Kong consumers in the three dominant groups present a positive WTP for U.S. pork without any other quality attributes, Guangzhou consumers in three dominant groups have a negative WTP . As most of respondents are included in the dominant groups, this result 78 reveals the fact that, in our survey, most of Guangzhou consumers perceive U.S. pork as less valuable than domestic pork if there is no other quality attributes embodied, but U.S. pork is more valuable for most of Hong Kong consumers compared t o Chinese pork. We find that while Guangzhou consumers become extremely patriotic, their WTP for U.S. pork is significantly decreased (the WTP at PA=5 is significantly lower than that at PA=1 or 1.66). This result confirms the previous discussion that patr iotism has a significantly negative effect their WTP is decreased as the relative patriotism score increases, the WTPs under all three levels of patriotism are not significa ntly different from each other. This result indicates that the patriotism effect on Hong Kong is not significant. One thing worth noting is that, while the patriotism score equals 1 (consumer is no t patrioti c) Guangzhou consu mer WTP for U.S. pork is still negative ( - 3.73 USD ). Therefore, there might be how patriotic the consumer is. Specifically, for Guangzhou consumers, who have a negative perception of U.S. pork and receive a significantly positive WTP 17 . Third, for those consumer s who are extremely patriotic (patriotism score of 5) in Guangzhou, providing food safety claims is not enough to improve their perception of U.S. pork. Only 17 Accurately, it is the case for consumers whose patriotism scores are not at the level of 5. For the extremely patriotic consum ers in Guangzhou, their WTP for U.S. pork is not significantly different from 0, which is also improved. 79 providing U.S. pork with food safety, animal welfare, and environmental - friendly claims together c an obtain a significantly positive WTP from extremely patriotic Guangzhou consumers. 5.3.4 level of patriotism the determinants of patrio tism is important for U.S. pork suppliers to promote sales in China. gender, age, education, household size, presence of children and senior in the household , and in come, to examine what kind of consumers are likely to be more patriotic. We employ Ordinary Least Square method (OLS) to estimate the regression model, and the regression results are given in Table 15 . Table 15 : Determinants of patriotism Guangzhou Hong Kong Coefficient S.E. Coefficient S.E. Male 0.38 (0.23) 0.03 (0.15) Age 0.01 (0.01) 0.01 (0.01) Education - 0.18 (0.12) - 0.08 (0.07) H.h. Size - 0.13 (0.11) - 0.17 (0.06) *** Children 0.43 (0.23) * 0.05 (0.13) Senior - 0.22 (0.22) - 0.04 (0.14) Income 0.01 (0.01) 0.00 (0.01) Constant 2.60 (0.69) *** 1.77 (0.51) *** Observations 200 200 R - squared 0.04 0.08 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 80 As shown in Table 15 , the children dummy is the only significant variable in the Guangzhou regression, indicating that consumers who are raising kids are likely to be more patriotic. In Hong Kong, household size is the only significant variable, showing that consumer s living in a larger household size is likely to be less patriotic. Thus, U.S. pork might be more valuable for this type of consumers. It is acknowledge that this inquiry on the determinants of consume area. 5.4 Conclusion and implication s pork attributes. The relative importance measured by monetary values in this study is consistent with our evaluation in chapter 4. Applying two empirical models to da ta from Guangzhou and Hong Kong, we compared the preferences of consumers in both cities. Also, 5.4.1 Summary of the pertinent results As all pork credence attributes a re valuable for consumers in both cities, food safety received the highest WTP from consumers, followed by environmental - friendly and animal welfare claims. In terms of country of origin, Guangzhou consumers perceive foreign pork to be less valuable, while Hong Kong consumers perceive foreign pork as more valuable. Nevertheless, due to the complementary effect between U.S. origin and the food safety attribute, U.S. pork is still valuable for mainland consumers if provided with the food safety claim and othe r quality attributes. 81 We used a Krinsky - Robb bootstrapping approach and complete combinatorial test to for food safety is significantly higher than that of Hong Kong cons umers, indicating that mainland consumers are more food - safety - sensitive. For pork from foreign countries, rception of foreign pork could be that, they are more used to consuming imported pork. Also, as Hong Kong consumers are relatively more world - minded, they may focus more on the quality attributes rather than whether the pork is domestic or imported. The e ffects of patriotism on consumer food preferences are also identified in this study. Specifically, we selected the top three patriotism levels in each city, and compared t while the effect on Hong Kong consumer is not significant. This is consistent with our lt in a preference for domestic product. Hong Kong consumers seem more world - minded: their level of patriotism may not affect their evaluation of pork quality attributes. 5.4.2 Implications As Chinese consumers become more sensitive to food quality, the rising demand for higher food safety standards creates opportunities for U.S. pork suppliers. Since U.S. suppliers have 82 the advantages of food safety control and quality management, making con sumers aware of these quality attributes could help U.S. pork receive a price premium in China. However, as important for U.S. suppliers to understand this issue b y segmenting and understanding consumers with different levels of patriotism prior to implementing marketing strategies. Highlighting the quality advantages and noting that US products have improved safety characteristics, are essential for U.S. pork suppl iers to promote sales on the Chinese market. 83 APPENDICES 84 A ppendi x A. Method Comparison Table A - 1 : Comparison between orde red logit regression and OLS Regression result Average partial effect Quality China Ordered Logit OLS Ordered Logit OLS Observations 2 0 0 20 0 Gender 0.25 0.10 0.08 0.10 - 0.32 - 0.11 Age 0.00 0.00 0 .00 0.00 - 0.01 0.00 Education 0.19 0.07 0.06 0.07 - 0.17 - 0.06 Children - 0.12 - 0.02 - 0.04 - 0.02 - 0.28 - 0.10 Senior - 0.33 - 0.10 - 0.11 - 0.10 - 0.32 - 0.11 Income - 0.01 * - 0.01 ** 0.00 0.00 0.00 0.00 Patriotism 0.33** 0.1 2 ** 0.11 0.12 - 0.16 - 0.05 Food Safety 1.91*** 0.5 8 *** 0.62 0.58 - 0.26 - 0.07 Animal Welfare 0.20 0.09 0.07 0.09 - 0.25 - 0.08 Environment 0.15 0.06 0.05 0.06 - 0.23 - 0.08 Other Variables Y Y Y Y *** p<0.01, ** p<0.05, * p<0.1 85 A ppendi x B. Original Survey ˚g˛\5F _______________________________________ F9 Block number # 1 a. b. c. d. OÉ$_ ______________________________________ _ / _______________________________________ _______________________________________ _______________________________________ B3 _______________________________________ 1. :______ a. +g b. 2. ______ ______ 3. ______ _______ a. b. c . d. e. 4 . ______ a. b. c. 5 . ( ): ______ 6. 18 ________ 7. 60 ________ 8 . ˝_ 9 . ˝_ 10 . 11 . 86 ____ ____ :___ __ a. C€D˜ b. PÁEŒ c. d. Jñ 12 . ____ ____ :__ __ _ a. C€D˜ b. PÁEŒ c. d. Jñ )Z6¹#¸ 13. _____ a. ˝_ b. c. 14 . ____ a. ˝_ b. c. 15 . ________ a. b. c. d. e. f. : ________________ 16. ____ a. b. 3 - 4 !Q c. 1 - 2 !Q d. 1 !Q e. f. 17. ___ a. 1 ˙Ô b. 1 - 2 ˙Ô c. 2 - 3 ˙Ô d.3 - 4 ˙Ô e. 4 - 5 ˙Ô f. 5 - 6 ˙Ô g. 6 - 7 ˙Ô h. 7 ________ ˙Ô 18. _________ 3 a . Gü7: ( FJ7: ) 6¹ b . 6»˛‚6¹ c . ˇÂPØ d . e . f . )Z6¹Oµ g . h . i . j . , #˘˝> __________ 19. ______ % ______ % _____ % 20. _______ a. b. 21. ˘… a. b. c. d. e. 22. a. b. c. d. e. 87 23. a. b. c. d. e. 24. a. b. c. d. e. 25. a. b. c. d. e. 26. a. b. c. d. e. 27. 1 - 5 1 5 a. 5 4 3 2 1 b. 5 4 3 2 1 c. 5 4 3 2 1 d. 5 4 3 2 1 e. 5 4 3 2 1 f. 5 4 3 2 1 g. 5 4 3 2 1 28. 5 3 1 a. 5 4 3 2 1 b . 5 4 3 2 1 c. 88 5 4 3 2 1 29. 5 3 1 a. 5 4 3 2 1 b. 5 4 3 2 1 c. 5 4 3 2 1 d. 5 4 3 2 1 e. 5 4 3 2 1 F9P¼ / !ÿ˙Ô !W#b (Ž/¿ 89 )ß ˇA F9˛0 )ZGü7:6¹ ,´ 30. : __________ / !ÿ˙Ô 31. ______ a. b. c. d. 32. _____ a. ˝_ b. 33. _____ a. ˝_ b. 34. _____ a. ˝_ b. F9NÈ 1 F9N© 1 F9N© 2 F9N© 3 - 1 /500 1 /500 F9N© 1 2 ˇˇ ˛9 ˛9 ˛9 ˛9 : 90 F9NÈ 2 F9N© 1 F9N© 2 F9N© 3 - 0.33 /500 0.33 /500 F9N© 1 2 ˇˇ ˛9 : F9NÈ 3 F9N© 1 F9N© 2 F9N© 3 - 0.33 /500 0.33 /500 F9N© 1 2 ˇˇ ˛9 ˛9 ˛9 : F9NÈ 4 F9N© 1 F9N© 2 F9N© 3 0.33 /500 0.33 /500 F9N© 1 2 ˇˇ ˛9 ˛9 ˛9 : F9NÈ 5 F9N© 1 F9N© 2 F9N© 3 1 /500 1 /500 F9N© 1 2 ˇˇ ˛9 ˛9 : 91 F9NÈ 6 F9N© 1 F9N© 2 F9N© 3 0.33 /500 - 0.33 /500 F9N© 1 2 ˇˇ ˛9 ˛9 ˛9 ˛9 ˛9 : F9NÈ 7 F9N© 1 F9N© 2 F9N© 3 1 /500 0.33 /500 F9N© 1 2 ˇˇ ˛9 ˛9 ˛9 : F9NÈ 8 F9N© 1 F9N© 2 F9N© 3 - 1 /500 1 /500 F9N© 1 2 ˇˇ ˛9 : F9NÈ 9 F9N© 1 F9N© 2 F9N© 3 - 1 /500 - 0.33 /500 F9N© 1 2 ˇˇ ˛9 ˛9 : F9NÈ 10 F9N© 1 F9N© 2 F9N© 3 1 /500 1 /500 F9N© 1 2 ˇˇ ˛9 ˛9 ˛9 ˛9 ˛9 : 92 35. ˘… a. b. c. ˛9˝& d. e. a. b. c. ˛9˝& d. e. a. b. c. ˛9˝& d. e. (Ž/¿ a. b. c. ˛9˝& d. e. )ß˘Ô a. b. c. ˛9˝& d. e. )Z6¹(© 36. CXGÿ ˛$ 37. 5 1 Gý?± 8¢#- 5 4 3 2 1 6Õ,V 5 4 3 2 1 5 4 3 2 1 4é*6 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 93 5 4 3 2 1 5 4 3 2 1 CX 5 4 3 2 1 5 4 3 2 1 38. CXGÿ 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 !W#b 5 4 3 2 1 5 4 3 2 1 39. 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 !W#b 5 4 3 2 1 5 4 3 2 1 40. (Ž/¿ +O)Z ,´ 94 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 !W#b 5 4 3 2 1 5 4 3 2 1 41. )Z6¹ ,´ )ß "d CX 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 !W#b 5 4 3 2 1 5 4 3 2 1 42. )» 6¹6<+O a. b. c. d. e. 43. ˛~ )Z6¹6<+O+õ a. b. c. d. e. 95 44. 6¹6<+O a. b. c. d. e. 45. )Z6¹6<+O+õ a. b. c. d. e. 46. ______ ____ a. < 2,000 g. 12,000 - 14,000 b. 2,000 - 4,000 h. 14,000 - 16,000 c. 4,000 - 6,000 i. 16,000 - 18,000 d. 6,000 - 8,000 j. 18,000 - 20,000 e. 8,000 - 10,000 k. 20,000 - 22,000 f. 10,000 - 12,0 00 l. >22,000 47. 5 1 "Ñ˛9 5 4 3 2 1 "Ñ˛9 L˚NÈ 48. _____ a. ˛9 b. "Ñ˛9 96 49. 5 1 '˛(a 5 4 3 2 1 50. 5 1 B'˘… '˛(a 5 4 3 2 1 M”BR˙_ˇ1 B3 _________ 97 REFERENCES 98 REFERENCE S Adamowicz, W., Boxall, P., Williams, M., & Louviere, J. (1998). Stated preference approaches for measuring passive use values: choice experiments and contingent valuation. American journal of agricu ltural economics, 80(1), 64 - 75. wil lingness to pay for milk traceability in China. Agricultur al Economics, 44(4 - 5), 537 - 544. Baines, R. N., & Davies, W. P. (1998). Quality assurance in international food supply. In Proceedings of the third international conference on chain management in agribusiness and the food industry. Wageningen, The Netherlands: Wageningen University. Balestrini, P., & Gamble, P. (2006). Country - of - origin effects on Chinese wine consumers. British Food Journal, 108(5), 396 - 412. Bech, M., & Gyrd Hansen, D. (2005). E ffects coding in discrete choice experiments. Health economics , 14 (10), 1079 - 1083. Becker, T. (2000). Consumer perception of fresh meat quality: a framework for analysis. British Food Journal, 102(3), 158 - 176. Breidert, C., Hahsler, M., & Reutterer, T. ( 2006). A review of methods for measuring willingness - to - pay. Innovative Marketing , 2 (4), 8 - 32. Bernués, A., Olaizola, A., & Corcoran, K. (2003). Extrinsic attributes of red meat as indicators of quality in Europe: an application for market segmentation. Food Quality and Preference, 14(4), 265 - 276. Brunsø, K., Bredahl, L., Grunert, K. G., & Scholderer, J. (2005). Consumer perception of the quality of beef resulting from various fattening regimes. Livestock Production Science, 94(1), 83 - 93. Calvin, L., Gale, F., Hu, D., & Lohmar, B. (2006). Food safety improvements underway in China. Amber Waves, 4(5), 16 - 21. Cang, L., Wang, Y. J., Zhou, D. M., & Dong, Y. H. (2004). Heavy metals pollution in poultry and livestock feeds and manures under intensive farmin g in Jiangsu Province, China. Journal of Environmental Sciences, 16(3), 371 - 374. Caswell, J. A., & Mojduszka, E. M. (1996). Using informational labeling to influence the market for quality in food products. American Journal of Agricultural Economics , 1248 - 1253. 99 Chen, J. (2007). Rapid urbanization in China: A real challenge to soil protection and food security. Catena, 69(1), 1 - 15. products to the Chinese domestic m arket. China Information, 24(1), 61 - 74. Cicia, G., & Colantuoni, F. (2010). Willingness to pay for traceable meat attributes: a meta - analysis. International Journal on Food System Dynamics , 1 (3), 252 - 263. Cooper, B. Aus frozen meat in smuggling bust. 201 2, June 21. Queensland Country Life. Retrieved from: http://www.queenslandcountrylife.com.au/news/agriculture/cattle/beef/au s - frozen - meat - in - smuggling - bust/2597788.aspx Retrieved from : http://dealbook.nytimes.com/2014/04/10/china - pork - companys - listing - could - raise - more - than - 5 - billion/?_r=1 Dalgaard, R. (2007). The enviro nmental impact of pork production from a life cycle perspective (Doctoral dissertation, Videnbasen for Aalborg UniversitetVBN, Aalborg UniversitetAalborg University, Det Teknisk - Naturvidenskabelige FakultetThe Faculty of Engineering and Science, Institut f or PlanlægningDepartment of Development and Planning). Den Ouden, M., Nijsing, J. T., Dijkhuizen, A. A., & Huirne, R. B. M. (1997). Economic optimization of pork production - marketing chains: I. Model input on animal welfare and costs. Livestock Production Science, 48(1), 23 - 37. Dransfield, E., Ngapo, T. M., Nielsen, N. A., Bredahl, L., Sjödén, P. O., Magnusson, M., & Nute, G. R. (2005). Consumer choice and suggested price for pork as influenced by its appearance, taste and information concerning country o f origin and organic pig production. Meat Science, 69(1), 61 - 70. Enneking, U. (2004). Willingness - to - pay for safety improvements in the German meat sector: the case of the Q&S label. European Review of Agricultural Economics , 31 (2), 205 - 223. Espejel, J., Fandos, C., & Flavián, C. (2007). The role of intrinsic and extrinsic quality attributes on consumer behaviour for traditional food products. Managing Service Quality, 17(6), 681 - 701. Food and Agriculture Organization of the United Nations, FAOSTAT datab ase (FAOSTAT, 2013). Available online at http://faostat3.fao.org/download/FB/FBS/E 100 Font i Furnols, M., Realini, C., Montossi, F., Sañudo, C., C ampo, M. M., Oliver, M. A., & Guerrero, L. (2011). country of origin, feeding system and meat price: A conjoint study in Spain, France and United Kingdom. Food Quality and Preference, 22(5), 443 - 451. Fu, C., & Yu, G. (2010). Estimation and spatiote mporal analysis of methane emissions from agriculture in China. Environmental management, 46(4), 618 - 632. A Report from the Economic Research Service. 2 Gao, Z., Ma, W., Zhu, G., & Roe lcke, M. (2013). Estimating farm - gate ammonia emissions from major animal production systems in China. Atmospheric Environment. Garvin, D. A. (1984). What does product quality really mean? Sloan management review, 26(1). Grunert, K. G., Bredahl, L., & Br unsø, K. (2004). Consumer perception of meat quality and implications for product development in the meat sector a review. Meat science, 66(2), 259 - 272. Grunert, K. G. (2005). Food quality and safety: consumer perception and demand. European Review of Ag ricultural Economics , 32 (3), 369 - 391. Grunert, K. G., & Valli, C. (2001). Designer - made meat and dairy products: consumer - led product development. Livestock Production Science, 72(1), 83 - 98. Gu, P., Shen, R. F., & Chen, Y. D. (2008). Diffusion pollution from livestock and poultry rearing in the Yangtze Delta, China. Environmental Science and Pollution Research, 15(3), 273 - 277. Harper, G., & Henson, S. (2001). Consumer concerns about animal welfare and the impact on food choice. EU FAIR CT98 - 3678, Centre for Food Economics Research, The University of Reading. Han, C. M. (1988). The role of consumer patriotism in the choice of domestic versus foreign products. Journal of Advertising Research , 28 (3), 25 - 32. Hanna, N., & Dodge, H. R. (1995). Pricing Policie s and Procedures . New York University Press. Hoffmann, R. (2000). Country of origin a consumer perception perspective of fresh meat. British Food Journal, 102(3), 211 - 229. Hole, A. R. (2007). A comparison of approaches to estimating confidence intervals for willingness to pay measures. Health economics , 16 (8), 827 - 840. 101 Hong, S. T., & Wyer Jr, R. S. (1989). Effects of country - of - origin and product - attribute information on pr oduct evaluation: An information processing perspective. Journal of Consumer Research , 175 - 187. Huang, J., & David, C. C. (1993). Demand for cereal grains in Asia: the effect of urbanization. Agricultural Economics, 8(2), 107 - 124. Hunt, N. Higher feed co sts to drive up meat prices in 2013. 2012, November 8. Reuters. Retrieved from http://www.reuters.com/article/2012/11/08/meat - prices - idUSL5E8M7F3O20121108 Issanchou, S. (1996). Consumer expectations and perceptions of meat and meat product quality. Meat Science, 43, 5 - 19. Janssen, M., & Hamm, U. (2012). Product labelling in the market for organic food: Consumer preferences and willingness - to - pay for differe nt organic certification logos. Food Quality and Preference, 25(1), 9 - 22. Jiang, P., Jin, S. Y., Hao, X. Z., Zhou, D. M., Li, L. Z., & Lv, J. L. (2010). Distribution characteristics of heavy metals in feeds, pig manures, soils and vegetables. J Agron Env iron Sci, 29, 942 - 947 Krinsky, I., & Robb, A. L. (1986). On approximating the statistical properties of elasticities. The Review of Economics and Statistics , 715 - 719. Krystallis, Athanasios, and Ioannis S. Arvanitoyannis. "Investigating the concept of me at 72.1 (2006): 164 - 176. Lancaster, K. J. (1966). A new approach to consumer theory. The journal of political economy, 132 - 157. Latouche, K., Rainelli, P., & Vermersch, D. (1998). Food safety issues and the BSE scare: some lessons from the French case. Food policy, 23(5), 347 - 356. Liljenstolpe, C. (2008). Evaluating animal welfare with choice experiments: an application to Swedish pig production. Agribusiness, 24(1), 67 - 84. Li, Y. X., & Chen, T. B. (2005). Concentrations of additive arsenic in Beijing pig feeds and the residues in pig manure. Resources, Conservation and Recycling,45(4), 356 - 367. Louviere, J. J., Hensher, D. A., & Swait, J. D. (2000). Stated choice methods: analysi s and applications . Cambridge University Press. 102 Loureiro, M. L., & Umberger, W. J. (2003). Estimating consumer willingness to pay for country - of - origin labeling. Journal of Agricultural and Resource Economics, 287 - 301. Loureiro, M. L., & Umberger, W. J. (2007). A choice experiment model for beef: What US consumer responses tell us about relative preferences for food safety, country - of - origin labeling and traceability. Food policy, 32(4), 496 - 514. Lusk, J. L., Nilsson, T., & Foster, K. (2007). Public pref erences and private choices: effect of altruism and free riding on demand for environmentally certified pork. Environmental and Resource Economics, 36(4), 499 - 521. Lusk, J. L., Roosen, J., & Fox, J. A. (2003). Demand for beef from cattle administered grow th hormones or fed genetically modified corn: a comparison of consumers in France, Lu Wang, C., & Xiong Chen, Z. (2004). Consumer ethnocentrism and willingness to buy domestic products in a developing country setting: testing moderating effects. Journal o f Consumer Marketing, 21(6), 391 - 400. Lukovitz , K. Survey: 78% Of U.S. Families Report Buying Organics. 2011, November 30. MarketingDaily. Retrieved from http://www.mediapost.com/publications/article/163210/survey - 78 - of - us - families - report - buying - organic.html Manski, C. F. (1977). The structure of random utility models. Theory and decision , 8 (3), 229 - 254. Marbeau, Y. (1987). What value pricing research today. Journal of the Market Research Society , 29 (2), 153 - 182. Mayfield, L. E., Bennett, R. M., Tranter, R. B., & Wooldridge, M. J. (2007). Consumption of welfare - friendly food products in Great Britain, Italy and Sweden, and how it may be influenced by consumer attitudes to, and behaviour towards, animal welfare attributes. International Journal of Sociology of Food and Agriculture , 15 (3), 59 - 73. Mccarthy, Mary B., and Spencer J. Henson. "Irish consumer perceptions of meat hazards and use of extrinsic information cues." Food Economics - Acta Agriculturae Scandinavica, Section C 1.2 (2004): 99 - 106. McFadden, D. (1973). Conditional logit analysis of qualitative choice behavior. McFadden, D., & Train, K. (2000). Mixed MNL models for discrete response. Journal of applied Econometrics , 15 (5), 447 - 470. 103 McOrist, S., Khampee, K., & Guo, A. (2011). Modern pig farming i n the People's Republic of China: growth and veterinary challenges. Revue scientifique et technique (International Office of Epizootics), 30(3), 961 - 968. Meuwissen, M. P., Van Der Lans, I. A., & Huirne, R. B. (2007). Consumer preferences for pork supply chain attributes. NJAS - Wageningen Journal of Life Sciences , 54 (3), 293 - 312. Montlake, S. Why Chinese Purchase Of Smithfield Could Be Game Changer For Pork Trade. 2013, May 30. Forbes. Retrieved f rom http://www.forbes.com/sites/simonmontlake/2013/05/30/why - chinese - buyout - of - smithfield - could - be - game - changer - for - pork - trade/ Morgan, L. A. (1985). The importance of quality. In Perceived Quality (pp. 61 - 64). Lexington Books Lexington, MA. Nelson, P. (1970). Information and consumer behavior. The Journal of Political Economy , 311 - 329. Netemeyer, R. G., Durvasula, S., & Lichtenstein, D. R. (1991). A cross - national assessment of the reliability and validity of the CETSCALE. Journal of Marketing Research, 320 - 327. Ngapo, T. M., Martin, J. F., & Dransfield, E. (2007). International preferences for pork appearance: I. Consumer choices. Food Quality and Preference , 18 (1), 26 - 36. Ngulu, F. Hong Kong strategic entry point into mainland China . 2014, June 13. The Namibia Economist. Retrieved from http://www.economist.com.na/2011 - 12 - 07 - 11 - 05 - 31/agriculture/5609 - hong - kong - entry - strategy - point - into - mainland - china Nielsen, B. L., & Zhao, R. (2012). Farm animal welfare across borders: A vision for the future. Animal Frontiers, 2(3), 46 - 50 Osborne, H. China: Eight Million Acres of Farmland too Polluted to Grow Food. 2013, December 30. International Business Times. Retrieved from: http://www.ibtimes.co.uk/china - eight - million - acres - farmland - too - polluted - grow - food - 1430641 Ortega, D. L., Wang, H. H., & Eales, J. S. (2009). Meat demand in China. China Agricultural Economic Review, 1(4), 410 - 419. Ortega, D. L., Wang, H. H., & Wu, L. (2009). Food Safety and Demand: Consumer Preferences for Imported Pork in Urban China. Journal of food distribution research, 40(3). 104 Ortega, D. L., Wang, H. H., Wu, L., & Olynk, N. J. (2011). Modeling heterogeneity in consumer preferences for select food safety attributes in China. Food Policy, 36(2), 318 - 324. Osborne, H. China: Eight Million Acres of Farmland too Po lluted to Grow Food. 2013, December 30. International Business Times. Retrieved from: http://www.ibtimes.co.uk/china - eight - million - acres - farmland - too - polluted - grow - food - 1430641 Rabobank Industry Note. Retrieved from: http://www.farminguk.com/content/knowledge/20120919_China%20Pork__Revised% 20FINAL%20Approved_WS.PDF Pan, C., & Kinsey, J. (2002). The supply chain of pork: US and China (pp. 02 - 01). Food Industry Center, University of Minnesota. Peng et al. Haze, air pollution, and health in China. 2013, December 21. The Lancet. Retrieved from : http://www.thelancet.com/journals/lancet/article/PIIS0140 - 6736%2813%2962693 - 8/fulltext Peter, S. Gray - Channel Economics: Drivers of Smuggling of Food and Agricultural Products into China. 2004. Columbia Business School. Retrieved from: https://www0.gsb.columbia.edu/mygsb/faculty/research/pubfile s/2091/Gray - Channel_Smuggling_v02.pdf MotherJones. Retrieved from: http://www.motherjones.com/tom - philpott/2013/09/chinas - antibiotic - meat - livestock Pig Progress. US: Pork p rice recovery threatened. 2013, March 14. Pig Progress. Retrieved from http://www.pigprogress.net/Pork - Processing/Markets/2013/3/US - Pork - p rice - recovery - threatened - 1199992W/ Poe, G. L., Giraud, K. L., & Loomis, J. B. (2005). Computational methods for measuring the difference of empirical distributions. American Journal of Agricultural Economics, 87(2), 353 - 365. Pouliot, S., & Sumner, D. A. (2008). Traceability, liability, and incentives for food safety and quality. American Journal of Agricultural Economics, 90(1), 15 - 27. Ramzy, A. Will China's New Food - Safety Laws Work? 2009, March 3. Time. Retrieved from : http://www.time.com/time/world/article/0,8599,1882711,00.html 105 Rawwas, M. Y., Rajendran, K. N., & Wuehrer, G. A. (1996). The influence of worldmindedness and nationalism on consume r evaluation of domestic and foreign products. International Marketing Review , 13 (2), 20 - 38. Scarpa, R., Thiene, M., & Train, K. (2008). Utility in willingness to pay space: a tool to address confounding random scale effects in destination choice to the A lps. American Journal of Agricultural Economics , 90 (4), 994 - 1010. Schumacher, T., Schroeder, T. C., & Tonsor, G. T. (2012). Willingness - to - Pay for Calf Health Programs and Certification Agents. Journal of Agricultural and Applied Economics , 44 (02). Sean. Nationalism in China. 2013, September 18. Learn Chinese Business. Retrieved from: http://learnchinesebusiness.com/2013/09/18/nationalism - in - china - %E7%88%B1%E5%9B%BD%E4%B8%BB%E4%B9%89/ Shimp, T. A., & Sharma, S. (1987). Consumer ethnocentrism: construction and validation of the CETSCALE. Journal of marketing research , 280 - 289. Snyder, Power than Before the 2008 Financial Crisis. 2014, September 22. Global Research. Retrieved from: http://www.globalresearch.ca/economic - recovery - for - the - u - s - middle - class - significantly - less - purchasing - power - than - before - the - 2008 - financial - crisis/5403937 Song, M., Liu, L. J., Wang, Z., Nanseki, T., & Nanshi, H . (2008). Consumers' attitudes to food traceability system in China: Evidences from the pork market in Beijing. Journal of the Faculty of Agriculture, Kyushu University, 53(2), 569 - 574. Steenkamp, J. B. E. (1 989). Product quality: An investigation into the concept and how it is perceived by consumers . Assen: Van Gorcum. Sun, X., & Collins, R. (2012 ). A preliminary study of chinese consumers' willingness - to - pay for fruit produced with sustainable attributes. Acta Horticulturae, 1006, 349 - 354. Supphellen, M., & Grønhaug, K. (2003). Building foreign brand personalities in Russia: the moderating effect of consumer ethnocentrism. International Journal of Advertising , 22 (2), 203 - 226. Sun, C., & Wu, H. (2012). Pol lution from animal husbandry in China: a case study of the Han River Basin. Water Science & Technology, 66(4). Ubilava, D., & Foster, K. (2009). Quality certification vs. product traceability: Consumer preferences for informational attributes of pork in G eorgia. Food Policy, 34(3), 305 - 310. 106 Tonsor, G. T., Olynk, N., & Wolf, C. (2009). Consumer preferences for animal welfare attributes: The case of gestation crates. Journal of Agricultural and Applied Economics , 41 (3), 713 - 730. Train, K. Discrete choice m ethods with simulation. 2003. Cambridge University Press , 24 , 25 - 26 Ubilava, D., & Foster, K. (2009). Quality certification vs. product traceability: Consumer preferences for informational attributes of pork in Georgia. Food Policy , 34 (3), 305 - 310. Umberger, W. J. (2003). Country - of - perceptions (Doctoral dissertation, University of Nebraska - Lincoln). U.S. Department of Agriculture, Foreign Agricultural Service. Global Agricultural Trade System Online. Available online at http://apps.fas.usda.gov/gats/default.aspx U.S. Department of Agriculture, Foreign Agricultural Service. Production, Supply and Distribution Online. Available online at http://apps.fas.usda.gov/psdonline/psdHome.aspx Verbeke, W., & Ward, R. W. (2006). Consumer interest in information cues denoting quality, traceability and origin: An application of ordered probit m odels to beef labels. Food Quality and Preference, 17(6), 453 - 467. Verlegh, P. W., & Steenkamp, J. B. E. (1999). A review and meta - analysis of country - of - origin research. Journal of economic psychology, 20(5), 521 - 546. Wang, F., Dou, Z., Ma, L., Ma, W., Sims, J. T., & Zhang, F. (2010). Nitrogen mass flow in China's animal production system and environmental implications. Journal of environmental quality, 39(5), 1537 - 1544. Wang, H., Dong, Y., Yang, Y., Toor, G. S., & Zhang, X. (2013). Changes in heavy metal contents in animal feeds and manures in an intensive animal production region of China. Journal of Environmental Sciences, 25(12), 2435 - 2442. Wandel, M., & Bugge, A. (1997). Environmental concern in consumer evaluation of food quality. Food quality and preference, 8(1), 19 - 26. Wang, Q., Fuller, F., Hayes, D., & Halbrendt, C. (1998). Chinese consumer demand for animal products and implications for US pork and poultry exports. Journal of Agricultural and Applied Economics, 30, 127 - 140. Wong, C. Y., Polonsky, M. J., & Garma, R. (2008). The impact of consumer ethnocentrism and country of origin sub - components for high involvement products on young 107 Logistics, 20(4 ), 455 - 478. Wu, L., Xu, L., & Gao, J. (2011). The acceptability of certified traceable food among Chinese consumers. British Food Journal, 113(4), 519 - 534 Xu, L., & Wu, L. (2010). Food safety and consumer willingness to pay for certified traceable food i n China. Journal of the Science of Food and Agriculture, 90(8), 1368 - 1373. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American psychologist , 35 (2), 151. Zanoli, R., Scarpa, R., Napolitano, F., Piasentier, E., Naspetti, S. , & Bruschi, V. (2013). Organic label as an identifier of environmentally related quality: A consumer choice experiment on beef in Italy. Renewable Agriculture and Food Systems , 28 (01), 70 - 79. Zhang, T., Sun, H. W., Wu, Q., Zhang, X. Z., Yun, S. H., & Kan nan, K. (2010). Perfluoro chemicals in meat, eggs and indoor dust in China: assessment of sources and pathways of human exposure to perfluoro chemicals. Environmental science & technology, 44(9), 3572 - 3579 Zhang, X. (2005). Chinese consumers' concerns abo ut food safety: case of Tianjin. Journal of international food & agribusiness marketing, 17(1), 57 - 69. Zhang, C., Bai, J., & Wahl, T. I. (2012). Consumers' willingness to pay for traceable pork, milk, and cooking oil in Nanjing, China. Food Control, 27(1) , 21 - 28. Zheng, S., Xu, P., Wang, Z., & Song, S. (2012). Willingness to pay for traceable pork: evidence from Beijing, China. China Agricultural Economic Review, 4(2), 200 - 215. Zhao, Y., & Wu, S. (2011). Willingness to Pay: Animal Welfare and Related Influencing Factors in China. Journal of Applied Animal Welfare Science, 14(2), 150 - 161. Zhu, Y. G., Johnson, T. A., Su, J. Q., Qiao, M. , Guo, G. X., Stedtfeld, R. D., & Tiedje, J. M. (2013). Diverse and abundant antibiotic resistance genes in Chinese swine farms. Proceedings of the National Academy of Sciences, 110(9), 3435 - 3440.