Corporate Social Responsibility(CSR) can be broadly defined as a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis, something like working with employees, their families, the local community and society at large to improve their quality of life. Over the past few decades, the concept has received widespread attention in all walks of life, not only the industrial application, but also the academic community has done a lot of related research[3-5]. Even in the field of food, many researchers actively use the integration of multiple variables for the study of related issues[6-14].
As the food industry has an extremely close and direct relationship with the environment and consumers, it is second to none in the expansion of CSR, more and more food companies are turning their attention to CSR[15-17]. As consumers pay more attention to environmental protection and are more sensitiveto CSR, many food enterprises begin to invest in CSR in practice. For example, Nestl é, the world's leading beverage company, has been a leader in CSR, focusing on providing a healthier and happier living environment for individuals and families, helping to develop communities, and protecting and conserving limited natural resources for the next generation especially focus on protecting water resources. According to 2017 Consolidated Nestlé Environmental Performance Indicators, Nestlé saved 67.5% packaging source optimization during 2010-2017, 21.4% total on-site energy consumption from 2007 to 2017, 38.1% total water withdrawal during the period 2007-2017 etc. Research shows that consumers have more support to the food company which actively participate in CSR and have more purchase intentions. CSR initiatives have become a core part of business activities in the food sector.
CSR and CSR activities are important, but for food companies, in the course of business inevitably involved in some active or passive food safety scandals, a food safety scandal is also a crucial factor for a food enterprise and has an extremely important impact on the brand image of an enterprise. For example, Shanghai Husi Food Co. Ltd. supplied products containing expired meatto McDonald's, KFC, Pizza Hut, Starbucks, and Burger King. The products were sold in numerous countries including Japan. Since then, the company's brand image in China has tumbled and stalled. Again, for example, 2008 Chinese milk scandal has left dozens of children dead, tens of thousands suffering from kidney disease, and leading dairy companies have gone bankrupt etc. From the perspective of the enterprise, it is worth considering whether some actions(CSR ）that have a positive impact on the brand image of the enterprise may have a positive impact of minimize the damage of brand image after the food safety scandal. Therefore, this study examines the interaction between different level of CSR participation and the occurrence of food safety scandals on the impact of brand image and consumer loyalty. Among them, this study suggests that food safety scandals have negative moderating effects on the relationship between CSR participation and Brand Image of food enterprises. In other words, when a food safety scandal occurs, the impact of CSR engagement on Brand Image is diminished; when it is not, it is reinforced. In addition, the study will be conducted in the context of the Chinese food industry following a food safety scandal that shocked the world.
Ⅱ. Literature Review
2.1 CSR Engagement, Brand image, Customer loyalty
Since the emergence of the concept of CSR, the research on the relationship between CSR and Brand Image has been very deep and wide[20,21,22]. In many studies, researchers have suggested CSR is regarded as a good strategic marketing tool and it has significant influence on consumers’ behaviours[22-24], especially the brand image. This means that the choice of CSR Strategy and the degree of CSR participation gradually become a factor which can’t be ignored, different degree of CSR participation will closely affect consumer behavior. Among all kinds of industries, the food industry also bears the brunt. Research shows that in the food industry, for Food Enterprises, CSR strategy on the impact of the customer behavior is extremely important [24,25],even when product-harm crisis occur, CSR involvement can also have a positive impact on food companies to save the Enterprise.
Many papers on the relationship between CSR and brand image can be found, no matter what research methods are used, the results are unanimous that CSR has a positive effect on brand image[21,26]. According to the Redcardglobal, CSR engagement can be divided into four levels, namely Economic Responsibility, Legal Responsibility, Ethical Responsibility, Philanthropic Responsibility, get bigger in turn. As the first level of the lowest end, but also as the most basic social responsibility of the enterprise, to do a good job to obtain profits, pay wages to employees, produce the products consumers need or provide the services they need, this is the most basic and only way for a business to exist in society for a long time. There is no guarantee that companies will act legally when they are able to consistently generate significant economic benefits, and some are considered socially irresponsible if they violatet he law to generate significant economic benefits. Thus, complying with the law while reaping economic benefits is a second level of corporate social responsibility.
Companies create economic benefits by complying with the law, but if they can also comply with ethicsand ethics, they will move to a third level of social responsibility, such as being pro environmental and treating their suppliers and employees properly, even though it's not their job. The highest level of corporate social responsibility is giving back, not only at all three levels, but also consciously giving back, by donating money, volunteering to staff members from time to time, and so on. On brand image, until now, there are a lot of brand image models, many of these different types of brand image model is based on different types of products to distinguish[28-30]. Consumer packaged food brands have also been studied by  and . In this study, the CSR elements in the Brand Image model of packaged foods will be used. Previous studies that demonstrate the link between social initiatives and positive affective, cognitive, and behavioral responses by consumers[4,33-37]. Based on the conclusions of previous researchers, we propose the first hypothesis ：
Hypothesis1: CSR engagement has a positive ef ect on brand image.
Another very important factor is consumer loyalty. Customer loyalty exist as a result of the interaction between a customer’s relative attitude to a special brand, and also their repeat purchase behavior for that brand. Customer loyalty is a major strategic objective and focus in marketing. It has been suggested that brand image is a major driver of customer loyalty, and hence companies seek to increase the equity of their brands. Many researchers have studied the relationship between brand image, consumer satisfaction and Consumer Loyalty[39-41], Although the conclusion is that consumer loyalty is affected by both brand image and consumer satisfaction, and it is positive, brand image has a stronger influence on consumer satisfaction than consumer satisfaction. About the packaged food brand, some researchers found that in many factors of brand image, CSR Image has a considerable degree of positive impact on the customer’s purchase intention and customer loyalty. Therefore we propose the second hypothesis based on the above conclusion:
Hypothesis2: Brand image has a positive influence on Customer Loyalty.
2.2 Food safety scandal and the relationship between CSR engagement and brand image.
It is clear that consumer reaction to food safety scandals has been entirely negative, without exception. Another study found that when food safety scandals occur, consumers became very uncertain, anxious, and critical of their food. It is well known that food safety scandals have a negative effect on consumer behavior, but what if a company engages in active CSR activities and publicizes them before the scandal? Researchers have found that when consumers were exposed to information about CSR, their interest in the food industry increased dramatically. Still, consumers' perceptions of CSR are unlikely to fully offset the negative impact of the scandal on brand image. When it comes to food safety scandals or not, there has been little research on the impact of CSR involvement on consumer behavior. It is therefore bold to assume that, while varying degrees of CSR engagement can reduce the negative impact of food safety scandals on brand image, it can’t completely reduce it, and on the contrary, food safety scandals can dampen the positive effects of CSR engagement on Brand Image. Therefore we propose the third hypothesis:
Hypothesis3: Food safety scandal negatively moderates the relationship between CSR engagement and Brand image.
[Figure 1]Research Model
2.3 Food safety scandal issues in the Chinese food industry.
The current study focuses on the Chinese food market to examine the proposed hypotheses. The Chinese food market has grown considerably and arguably represents a major market in the world. Due to the emerging status of the Chinese food market as a major player and the increasing infl uence of Chinese people on both domestic and international markets, inevitably food company managers and researchers seek to understand the behavior of the Chinese food guest, especially in relation to worldwide attention to food safety and CSR.
Despite the significance of the matter, empirical evidence regarding environmental social responsibility is lagging from China’s food sector. The sporadic empirical evidence in the previous literature mainly focused on the managerial perspective to examine why CSR is a good strategic marketing tool and it has significant influence on consumers’ behaviours[22-24] and why CSR strategy on the impact of the customer behavior is extremely important in food industry[24,25]. But consumers’ perspectives regarding the company’s CSR engagement remain less explored especially when the food safety scandal occurs. Thus, the current study asks the following research question: in China, the world’s fastest-developing economy, do consumers exhibit higher levels of evaluation of food company toward businesses implementing different level of CSR engagement when food safety scandal occurs or not? In contributing to understanding this question, this research conducts its analysis concentrating on Chinese consumers in China’s food market.
3.1 Research design
This study adopted a 2×2 between-subjects experimental design using written scenarios describing each of four virtual situation. According to Ramirez, Mukherjee, Vezzoli, & Kramer, as a research method, scenarios methodology can help researchers to complete "interesting research". This experimental design can be targeted to set up a virtual situation, to create casual environment for the respondents. In addition, the virtual situations before writing the questionnaire could minimize the effects of the different subjects' related memories. The scenarios examine two levels of CSR engagement (high versus low) combined with a dairy company’s food safety scandal or none. In the scenarios, in order to make the most of the virtual events as real as possible, the scenarios’ description of CSR engagement will be based on the Starbucks, which has the most mature and sophisticated CSR system on the food and beverage industry. Meanwhile, the description of the food safety scandal will be drawn directly from the 2008 China milk scandal. After reading the assigned scenario, respondents went directly to the questionnaire, filling out the questionnaire according to their own feelings just now. Details of the scenarios were provided in the Appendix.
This study is based entirely on Chinese consumers, so the scenarios and questionnaire were translated into Chinese for testing.
The first part of the questionnaire is about checks of realism (two-items) and manipulation (one-item) for four scenarios (2×2: High and Low CSR engagement by Food safety scandal and None). This study used two items adopted from  to check the realism of the scenarios. A 7-point Likert-type scale for responses ranged from strongly disagree (1) to strongly agree (7) for the items of ‘There are social phenomena like this in real life.’ and ‘The corporate social responsibility activities mentioned above are believable.’ Respondents rated CSR engagement on the basis of random scenario content. In this study, an item in the questionnaire was set up to check the manipulation of CSR engagement, measured on a 7-point Likert-type scale ranging from strongly disagree (1) to strongly agree (7): ‘The corporate social responsibility activities mentioned above are positive.’
The second part of the questionnaire is about the questions of brand image and consumer loyalty. To measure the study variables, multi-item scales are used as the main method. The measures of brand image consist of four items measured on a 7-point Likert-type scale ranging from strongly disagree (1) to strongly agree (7), adopted from the  study, ‘I think the food products of JY are environmentally friendly’, ‘I think the food products of JY are healthy’, ‘I think the food products of JY are produced under good working conditions’, and ‘I think the food products of JY comes from suppliers who care about humen welfare’.
The study used four items to measure customer loyalty on a 7-point Likert-type scale, ranging from strongly disagree (1) to strongly agree (7): ‘I would still consider that JY was basically honest’, ‘I would still consider JY as my first choice to buy when I buy food product’, ‘I would still recommend JY to other people’ and ‘I would still have trust in JY’.
At the end of the questionnaire, the respondents' gender, age group, education level, region of residence and monthly household income were collected.
3.3 Sample and data collection
Data were collected from Chinese customers from different regions on Internet. Because China is such a large country and people’s perceptions regarding food safety scandal and CSR engagement may vary across different regions, we chose to collect data at the chat room with more than 500 people and included people from as many different regions as possible in order to reach people from a relatively large geographic area. In order to enhance the implementation of the experiment and improve the efficiency of data collection, in this study we use C language to develop an experimental software. Upload the program to the Network Server, we send links to open chat rooms of 500 or more people on Wechat (China's social networking software). The chat rooms are randomly generated, and the people in the chat room don't know each other, to improve the objectivity of the data.
Respondents who click on a link to access the site, read the instructions for the experiment, agree to participate in the experiment, and then consciously carry out the experiment. At the start of the experiment, the program will automatically show the respondent a scenario at random, respondents read the scenario carefully and then click next to proceed to the questionnaire and fill it. A total of 454 questionnaires, were collected in November 2018. 77 data files were deleted due to loss or invalidation (include foreigners), resulting in a total of 377 valid respondents. Among them, 205 people received the scenario of high CSR engagement, 172 people received the low one. About the food safety scandal, 194 respondents completed the survey, 183 respondents completed the none one.
4.1 Descriptive statistics of the sample
The study first performs a descriptive analysis of respondents’ profiles. According to the results [Table 1], Of 377 respondents, 185 are females, 181 are males and 11 are others. 31% respondents are 40-50 years old (N=117) followed by more than 50 years old (P=27.3%, N = 103), 20-30 years old (P= 23.3, N=88) and 30-40 years old (P=18.3%, N= 69).
[Table 1] Description of the respondents
Most respondents have a bachelor’s degree (P= 47.5%, N = 179) followed by an under high school degree (P=21%, N=79), associate degree (P=18.8%, N=71) and postgraduate degree (P=12.7%, N=48).
A majority of the participants are from Eastern China (P=67.6%, N=255). Most respondents’ household monthly incomes are under 11,000RMB (about 181,5000 KRW).
The study first performs a descriptive analysis of respondents’ profiles. According to the results [Table 1], Of 377 respondents, 185 are females, 181 are males and 11 are others. 31% respondents are 40-50 years old (N=117) followed by more than 50 years old (P=27.3%, N=103), 20-30 years old (P=23.3, N=88) and 30-40 years old (P=18.3%, N=69). Most respondents have a bachelor’s degree (P=47.5%, N=179) followed by an under high school degree (P=21%, N=79), associate degree (P=18.8%, N=71) and postgraduate degree (P=12.7%, N=48). A majority of the participants are from Eastern China (P=67.6%, N=255). Most respondents’ household monthly incomes are under 11,000RMB (about 181,5000 KRW).
4.2 Main Analysis
The study first checked the realism and manipulation. For scenario of the high level of CSR engagement, the mean values of the two questions are 5.61 and 5.21 with t-values of 66.23 and 53.68 when compared with a value of 4 (neutral). For the scenario of low level of CSR engagement, the mean values of the two questions are 5.36 and 5.11 with t-values of 47.15 and 48.25 when compared with a value of 4 (neutral). Therefore, the scenarios are realistic. Next, for the scenario of high level of CSR engagement, the mean value is 5.49 with a t-value of 56.36 when compared with a value of 4 (neutral), while for the scenario of low level of CSR engagement, the mean value is 4.92 with a t-value of 38.87. Statistically, this study successfully manipulates CSR engagement.
Then, the study performed exploratory factor analysis (EFA) using varimax rotation for two constructs of brand image and customer loyalty. Factor analysis confirms that all four items form brand image and all four items form customer loyalty. Meanwhile, the study estimates Cronbach’s Alpha to test the reliability of items for the two constructs; alpha values are 0.958 and 0.971 for brand image (BI) and customer loyalty (CL), respectively.
For the test of hypotheses, the study performed a multiple regression. To control for the control variables: region (REGION) and education level (EDU), the study included six dummy variables, REGION1, REGION2 and REGION3 that represent Northern, Eastern and Southern regions, respectively, with Central and Western region being a reference point; EDU1, EDU2 and EUD3 that represent Associate, Bachelor and Postgraduate degree, respectively, with High school degree being a reference point.
The results in step2 of Table2 support Hypothesis1: CSR engagement positively impacts brand image (t-value=10.758; p-value< 0.001). Although findings also show that FSS have an opposite signi ficant main effect on BI (t-value= -4.956; p-value< 0.001). Other results show that a significant coefficient of the interaction term, CSR×FSS, with a p-value<0.001. At the same time, t-value=-12.586<0 means that the effect of high level of CSR engagement on brand image is actually lower when the food safety scandal occurs than when it does not. Thus, the findings successfully to support Hypothesis 3. Finally, adopted from Baron & Kenny, the study performed a series of regression analyses to test Hypothesis 2, the mediating effect must satisfy three conditions: (1) The independent variable is significantly correlated with the dependent variable; (2) The independent variable is significantly correlated with the assumed mediating variable; (3) When the mediating variable is put into the regression equation, the correlation between independent and dependent variables was weaker or less significant. There was a significant effect of CSR engagement on Customer Loyalty (β=0.424), and a significant effect of CSR engagement on Brand image. The Regression Coefficient of the independent variable CSR engagement decreases (β= -0.040) after adding the mediating variable Brand image. It means the correlation between the independent variable CSR engagement and the dependent variable customer loyalty was weaker, brand image as a mediator between CSR engagement and customer loyalty.
[Table 2]Results of regression analysis for moderation by food safety scandal
[Table 3] Results of mediating effect test
Since the 2008 Chinese milk scandal, the brand image of dairy companies has plummeted, and the government has stepped up efforts to enact laws and improve the industry's systems. Companies have also been trying to salvage their brand image in various ways, but very little research has focused on improving it before the scandal. Different from the previous research, this study did not focus on whether the enterprises participated in CSR, but paid more attention to the influence of different degree of CSR engagement on brand image and customer loyalty. On this basis, the occurrence of food safety scandals as a moderating variable, in order to study whether the occurrence of food safety scandals in dairy enterprises have the impact on CSR engagement and brand image.
In this study, 2x2 between-subjects was designed as experimental material according to four virtual situations. In this study, the experimental procedures and questionnaires were posted online. The respondents first read the scenarios randomly, and then went through the questionnaire procedure. In this study, survey was used to collect data from Chinese consumers. Of these, 377 valid data were collected from a sample of average Chinese consumers over the age of 20.
Based on Chinese consumers, this study examines and confirms the major influence of CSR engagement on brand image. On this basis, this study also examined the moderating effect of food safety scandals on CSR engagement and brand image. The results are in line with the hypothesis. Finally, the mediating effect of brand image on the relationship between CSR engagement and consumer loyalty was examined. The results are also consistent with the hypothesis.
First, the validation of Hypothesis 1 suggests a simple complement to the results of previous studies (CSR has a positive effect on brand image)[21, 26]: This means that consumers not only have an obvious sense of CSR, furthermore, there were significant differences in the degree of CSR engagement. And, not surprisingly, the food safety scandals did have a negative moderating effect on CSR engagement and Brand Image. However, we note that the majority of consumers in this study were from eastern China. Eastern China is more urbanised than the rest of the country and as a result, residents in the east are more sensitive to information about CSR . That means people in eastern China are more sensitive to CSR activities and may even be extremely sensitive to four levels of CSR engagement.
This study contributes to the marketing literature by examining an industry-specific case (i.e. the dairy industry). Findings of this study also provide some managerial implications. But in fact, we find that some CSR studies on the Chinese market have found that consumers are not sensitive to CSR activities, and that they believe that Chinese society as a whole lacks a positive perception of CSR and green actions The government needs to set up more programs to spread green awareness. By comparing the results of this study, we believe that the results of previous studies may be more reliable than this study. Therefore, it is hoped that the scope of the study will be expanded in the future.
But going back to this study, according to the research results, food companies, especially dairy companies can learn from this study results, companies not involved in CSR activities as soon as possible to participate in CSR activities, companies that are already involved in CSR are trying to reinforce the scope of their CSR activities. China has grown economically into the world's second largest economy, and Chinese people's sense of social responsibility will grow along with the rapid improvement of their material life, managers must pay attention and attention to this point. But then again, these are not the most important, for a food business, the most important must be to adhere to a moral conscience, strictly monitor food safety, and engaging in CSR is only a small positive factor, because the most important and best is always a safe product.
Ⅵ. Limitations and recommendations
As a not mature enough master's degree thesis, there are some flaws and limitations in this study. First of all, data collection, this study used the online experiment + questionnaire model, in fact, through the network access to the data, the authenticity is very difficult to justify, I hope that future research can be carried out in a fixed experimental place. Second, based on the data collected, we find that the sources of study are too single. As China is the third largest country in the world in terms of land area, different regions have different levels of development and different people have different ideas, which will cause great interference. We hope that future studies will spread out the sources of research as much as possible. In addition, the fact that consumers have more or less the same memory for CSR engagement or food safety scandals, and that there may be other important variables at play. We hope that future studies will be better able to eliminate these confounders and find more and better variables for testing.