The extension of animosity model of foreign product purchase: Evidence from a young vietnamese market segment

Nội dung text: The extension of animosity model of foreign product purchase: Evidence from a young vietnamese market segment

1. THE EXTENSION OF ANIMOSITY MODEL OF FOREIGN PRODUCT PURCHASE: EVIDENCE FROM A YOUNG VIETNAMESE MARKET SEGMENT PGS,TS Trang P.Tran - Department of Management, Marketing, and Information Systems State University of New York at Oneonta ThS Hồ Khánh Ngọc Bích – Trường Đại học Kinh Tế Huế ThS Lê Quang Trực - Trường Đại học Kinh Tế Huế TS Hoàng Trọng Hùng – Trường Đại học Kinh Tế Huế Abstract The primary purpose of this research focuses on how animosity from a young Vietnamese market segment affects their tendency to purchase foreign products from two different countries (China vs. USA). Data were collected from a student population in Vietnam with 300 respondents. Partial least squares (PLS) path modeling analysis was employed to test the proposed model. The findings show that the results are consistent with the existing literature. Out of six hypotheses capturing the main effects, only one relationship between consumer ethnocentrism on product judgment is not significant. Interestingly, we find that country of origin (China versus USA) did not moderate the effects of animosity, consumer ethnocentrism and cosmopolitanism on product judgment. The paper extends the model developed by Klien et al. (1998) by adding a new construct “cosmopolitanism”. The present research has brought new insights into the consumer animosity literature and developed new opportunities for better understanding the application of animosity into a particular developing country (Vietnam). Key words: animosity, consumer ethnocentrism, cosmopolitanism, product judgment INTRODUCTION More than ten years ago, a movement of boycotting American product spread out quickly in several countries in the world after American government sent their troop to Iraq, and officially joint the Iraq War. All “made-in-USA” products and businesses suffered. Windows of McDonald’s restaurants in Paris were smashed in Paris; American foods were removed from the menu at restaurants in Germany. The waves of anger moved beyond the boundary of European countries to Middle East. Starbucks, Big Mac, Microsoft, and Coca Cola which used to be beloved brands around the world became the target of boycotting movement (Kirschbaum, 2003). In America, a similar pattern developed, but in an opposite direction. Feeling angry because of French government’s decision of not joining the war, American customers rejected French wine, and French fries were changed to “Freedom Fries” (Ebenkamp, 2003; Amine, 2005). A similar confrontation emerged in Asia after so called the April 2005 event when a national history book changed the fact of Japanese past military acts after being approved by Japan government. The decision set an initial stage for protests and demonstrations taking place in China and South Korea (Liao et al., 2005). Angers toward Japanese government and their decisions were soon moved to Japanese products. A number of Japanese brands were cleared up from supermarkets in China (Liao et al., 2005). Most recently, ties between China and Vietnam become intense after China moved Haiyang Shiyou 981 oil rig into South China Sea where Vietnam claimed was part of their territory (BBC News, 2014). Demonstrations took place in several parts of Vietnam that then led to three days of anti-China unrest. Chinese businesses based in Vietnam were the target of angry demonstrators. Consequently, several factories were burned down and destroyed (BBC News, 2014). Vietnamese customers chose to stay away Chinese products as a way to show their frustration. 535
2. A growing body of literature has explored the phenomenon of animosity and associated its effects with marketing oriented constructs (Kalliny and LeMaster, 2005; Nijssen and Nijssen, 2004; Russell and Russell, 2006). A pioneering scholar in this trend was Klein et al.’s (1998) that examined the link between Chinese customer animosity toward Japanese products. That was followed by Klein (2002) where the animosity model was extended by including another construct: consumer ethnocentrism. Different from consumer animosity, consumer ethnocentrism refers to favorable attitudes toward domestic products and brands (Klein, 2002). Additionally, a number of other researchers have used this model to capture the influence of animosity on other activities, such as entry modes context, aversion toward economic superpowers, ties with countries with high level of international trade, and subsequent product choice (Kalliny and LeMaster, 2005; Nijssen and Nijssen, 2004; Russell and Russell, 2006). Despite the high number of research attempting to apply that model to a particular foreign product, research focusing on how consumer animosity affects their tendency to purchase products from two different foreign countries has been relatively scarce. This is the motivation for the current research. Moreover, to our best knowledge, the current research is the first research of this type designed to explore how the animosity model works in a developing country like Vietnam. Based on relevant literature, the current research extends the model developed by Klien et al. (1998) by adding a new construct “cosmopolitanism”. The objectives of the research are to answer the following research questions. 1. While the animosity model works in other country, does that work the same when applied in Vietnam? 2. Are there any differences in Vietnamese customers’ tendency to purchase products from China and the USA given the fact that Vietnam was in conflict with both countries at a different time? 3. In addition to animosity and consumer ethnocentrism, does cosmopolitanism have any impact on Vietnamese consumer tendency to purchase foreign products? The paper begins with theoretical perspectives to provide more background of key constructs and hypotheses. Methodology then follows in which measurement scales, Partial Least Square (PLS) approach, measurement model and structural model analysis are discussed. The paper concludes with general discussions and implications. LITERATURE REVIEW Consumer Ethnocentrism Ethnocentrism refers to one’s view of their own group in which they exaggerate their race or country as a center of the world and being better than others’ (Sumner, 1907). Each group believes that its own cultural values are greatest and scorn others’ folkways (Sumner, 1907). According to Sharma et al. (1995), ethnocentrism is a global scenario, which not only includes races, countries, groups, and social classes, but also develops into family pride, religious prejudice, racial discrimination and patriotism. Levine and Campbell (1972) emphasize ethnocentrism in terms of cultural protection, union and cooperation among community members and loyalty to the community. Consumer ethnocentrism refers to beliefs and conceptions about ethic judiciousness and appropriateness in purchasing foreign products (Shimp and Sharma, 1987). It is a result of patriotism, and the fear of losing control in economic benefits due to negative effects of import. In high-ethnocentric consumers’ opinions, purchasing foreign products is not only economically- oriented, but also ethically-oriented. They think that not buying overseas products means national 536
3. loyalty, and consuming them is blameworthy and contrary to national interests. All of those opinions, hence, lead to prepossessions against import (Shimp and Sharma, 1987). Consumer ethnocentrism, therefore, leads to partial evaluation: enhancing local products, lowering foreign products, or paying priority for home products and not willing to buy foreign products (Klein et al., 1998). The scale of consumer ethnocentrism, namely CETSCALE has been verified in several studies such as Nijssen, Douglas and Bressers (1999), and Kamaruddin, Mokhlis and Othman (2002). This scale is measured by 17 items in a unidimensional factor. Prior studies (Rosenbaum and Wong, 2009; Siemieniako, Kubacki, Gliska and Krot, 2011; Lantz, Nguyen and Tang, 2002) point out the negative influences of ethnocentrism on customers’ attitude towards foreign goods. In the survey for Drake University’s students in the USA, moreover, ethnocentrism negatively affects the recognition of International Commercial Agreements’ effectiveness (Luthy and Parsa, 1998). Douglas and Nijssen (2002), in their research across countries, indicate that ethnocentrism does not exist alone, but interacts with social and demographic components. More recent studies (Kamaruddin et al., 2002; Josiassen, Assaf and Karpen, 2011; Shah and Halim; 2011) assert that demographic characteristics impact at different levels on consumer ethnocentrism. Sharma et al. (1995) express four social psychology components of ethnocentrism in their study of factors influencing consumers’ attitude to imported products in the USA. These components include (1) openness for overseas cultures; (2) patriotism; (3) conservatism; and (4) cosmopolitanism. The result of their study shows that ethnocentrism has a negative correlation with the openness, but has positive correlations with the rest of factors. Sharma et al. (1995) also conclude that women are more ethnocentric than men, and the higher education and income, the less ethnocentrism. Animosity Animosity is defined as customers’ hate towards a foreign nation because of its past and present economic, political and martial events, which negatively influences shopping behaviours (Klein et al., 1998). Both ethnocentrism and animosity are emotional factors affecting foreign product evaluation and purchasing decisions (Rawwas et al., 1996; Klein et al., 1998). Animosity, however, differs from ethnocentrism in that the former expresses negative attitudes against specific nations while the latter refers to unfavorable attitudes against foreign countries as a whole (Shankamahesh, 2006). Consumer animosity is related to customers’ negative emotions toward their intention to buy and evaluation of products from a particular disliked group or country (Huang et al., 2009). Studies of animosity have been conducted in several developed countries such as the USA, Australia, New Zealand, Germany, Canada, France, but limited studies have been done in developing countries (Klein et al., 1998; Witkowski, 2000; Nijssen and Douglas, 2004; Amine et al., 2005; Bahaee and Pisani, 2009). Most previous studies disclose that consumer animosity can dramatically decrease sales of one country’s products in the victim country (Huang et al., 2009). For example, Klein et al. (1998) show that Australian and New Zealand customers refused to buy products from France and Germany because of the French nuclear tests in the South Pacific. Customers will probably express their animosity toward a nation when they believe that nation damages their country (Akdogan et al., 2012). Animosity is likely to be a protest of consumers in aggrieved country to international events. Ching (1996) and Onishi (2007) analyze Taiwanese consumers’ animosity toward Japanese goods reveals the negative effect of animosity on customers’ purchase decision. Similarly, in the study of the impact of animosity on Australians’ attitudes towards French products, Ettenson and Klein (2005) emphasize the inverse correlation between Australians’ animosity and their 537
4. willingness to purchase French products. Moreover, the results from the study of Bahaee and Pisani (2009) on the Iranian customers’ animosity toward the US products suggest that Iranians own different levels of animosity in terms of different demographic variables, including: age, education and abroad travelling experience. In more detail, Iranian students and females have higher animosity than non-students and males. Additionally, the researchers find out the negative relationship between Iranian customers’ animosity and willingness to purchase goods from the USA (Bahaee and Pisani, 2009). Cosmopolitanism In marketing, cosmopolitanism is considered as a concept to explain customer attitudes and decision-making. Riefler and Diamantopoulos (2009) define cosmopolitanism as a disposition of individuals to view themselves as citizens of the world rather than as citizens of a specific nation. It also shows open-mindedness to the world and other cultures (Skrbis et al., 2004). Cosmopolitanism has an impact on customers’ perceptions on product quality, purchase behaviour and willingness to buy foreign products versus home products (Bruning, 1997; Samiee, 2005). The higher level of cosmopolitanism leads to the more openness of trying foreign products (Riefler and Diamantopoulos, 2009). According to Riefler, Diamantopoulos and Siguaw (2012), a cosmopolitan consumer owns three characteristics: (1) an open-mindedness to engaging in different cultural and national origins, (2) an appreciation the globalization due to the availability of products from different countries and cultures, and (3) a willingness to purchase foreign products. Cosmopolitanism can encourage individuals to experience and discover new insights of other cultures (Thompson and Tambyah, 1999; Cannon and Yaprak, 2002), and induce a greater intention to consume foreign products (Rawwas et al., 1996). Further, there is a direct relationship between cosmopolitanism and product quality: customers owning high levels of international openness prefer products from foreign countries over products from their home country (Lee and Chen, 2008). Cosmopolitan consumers in developing countries favor products from developed countries (Strizahakova et al., 2008). The monopoly of government-owned companies in developing countries leads to the perception that domestic products will not be as good as those of developed countries (Batra, 1997). In recent years, however, that perception has been changed to increasing preference for domestic goods because of branding globalization, low-cost manufacturing and openness of import barriers in developing markets (Sharma et al., 2008). On the other hand, the results from the study of Papadopoulos, Heslop and Bamossy (1990) express that customers from developed countries are likely to underestimate products made from developing countries. Along with the growth of low-cost outsourcing in developing economies, they have changed their perceptions of such products (Chao, 1993). Developed nations’ cosmopolitans perceive such products as domestic products or favour foreign products because of carrying a domestic brand (Cannon and Yaprak, 2002; Steenkamp and de Jong, 2010). In Vietnam, there have been few studies of ethnocentrism and no studies of animosity and cosmopolitanism. The evidence proves that Vietnamese consumers have high level of ethnocentrism, give priority for consuming home products, and perceive that the purchase of domestic goods is moral and essential for the growth of their home land (Nguyen, 2004; Pham, 2012). Vietnamese are quite open-minded, not extremely conservative and highly globally- integrated (Pham, 2012). Although they support local products, they purchase and have positive view of products from other countries because of prior product quality and price (Nguyen, 2004). HYPOTHESES DEVELOPMENT 538
5. In this study, we investigate how consumer ethnocentrism, animosity and cosmopolitanism influence foreign product judgement and willingness to buy foreign products. Foreign product judgment is defined as customers’ perception towards foreign products in general (Klein et al., 1998). Riefler and Diamantopoulos (2007) view animosity as remnants of aversion, anger and hostility toward a despised foreign country. Several studies emphasize that there is an increasingly negative evaluation of products from a disdained country when animosity grows (Akdogan et al., 2012). For example, Japanese products are the last choice on the “proud to own” question in the United Kingdom and France due to animosity created since the World War II (Papadopoulos et al., 1990). Thus, it is hypothesized that: H1: Consumer animosity has a negative impact on product judgment. Patriotism, racism, and internationalism have been investigated as antecedents of consumer behaviour toward foreign products in previous studies. The results, however, are mixed. For example, Balabanis et al. (2001) find that internationalism or cosmopolitanism has an insignificant influence on Turk and Czech consumer view of products from another country. By contrast, from the study of Suh and Kwon (2002), global mindedness has a significantly positive correlation with consumers’ attitudes to non-home country products. Hence, this study aims to re-investigate the effect of cosmopolitanism on consumer evaluation towards the purchase of foreign products. Thus, the following hypothesis is proposed for further analysis: H2: Cosmopolitanism has a positive impact on product judgment. Consuming foreign products is considered to be an unpatriotic action (Sharma et al., 1995; Klein, 2002). Klein et al. (1998) suggest that ethnocentrism has a dramatic role in shaping consumers’ attitudes towards foreign products. Therefore, ethnocentric customers are more likely to denigrate the quality of products from other countries (Kasper, 1999; Klein, 2002). Previous studies have concluded that consumer ethnocentrism is negatively related to product judgment (Shimp and Sharma, 1987; Klein et al., 1998). As such, the following hypothesis is proposed: H3: Consumer ethnocentrism has a negative impact on product judgment. Animosity may lead to the boycott of products from a disliked country in order to revenge for its political or economic events and to protect their home economy from future harm from that country (Cui et al., 2012). For instance, in the USA, the level of consumers’ animosity toward Japan is measured as being low (Klein, 2002). That animosity, nevertheless, is still negatively related to their intention to purchase Japanese products (Klein, 2002). Consequently, regardless of the level of animosity, there may be a negative effect of animosity on customers’ willingness to buy products from the disdained country (Cui et al., 2012). Hence, the current study proposes the following hypothesis: H4: Animosity has a negative impact on willingness to buy. The results of many empirical studies have showed that consumer ethnocentrism and willingness to buy products from other countries have a reverse correlation (Sharma et al.,1995; Suh and Kwon, 2002). Park and his colleagues (2008) point out that consumer ethnocentrism aggresses the sales of global luxury brands among young Korean consumers. Zarkada-Fraser and Fraser (2002) also suggest a negative influence of consumer ethnocentrism on support for foreign retail outlets. Hence, we posit: H5: Consumer ethnocentrism has a negative impact on willingness to buy. Cheah and Phau (2006) suggest a relationship between consumers’ product judgment and 539
6. their purchase intention. There are several external factors affecting consumers’ perceptions of judging foreign products (Nguyen, Nguyen and Barrett, 2008; Shin, 2001). Moreover, Taewon and Ik-Whan (2002) show that product judgment has an important role in shaping purchase behavior to foreign products in certain cultural context as proved by consumers in Korea. Ahmed and his colleagues (2013) find out a significantly positive correlation between judgment of foreign products and purchase action. Thus, the following hypothesis is proposed: H6: Product judgment has a positive effect on willingness to buy There is a continuous question about whether different countries of origin (COOs) lead to different levels of impact of animosity on product judgment (Cui et al., 2012). Cui et al. (2012) conduct a survey to compare Chinese consumer’s perceptions of two different countries, including Japan and the USA. They reveal that Chinese consumers’ animosity towards the USA is lower than that towards Japan (Cui et al., 2012). As a result, their project judgment is more favourable for the US-made products than those from Japan (Cui et al., 2012). In Vietnam, Vietnamese people have animosity towards both the USA and China due to political and economic events, such as the Vietnam-USA war and the Vietnam-China war in the history. However, due to the recent event of HY981 oil rig between Vietnam and China, the influence of animosity on Chinese product judgment may be worse than that of US product judgment. Thus, we propose the following hypothesis: H7: Country of origin (USA/China) moderates the relationship between animosity and product judgment Jin et al. (2014) investigate the relationship between cosmopolitanism and COO. and indicate that although young consumers in developed countries possess a high level of world- mindedness, they positively evaluate products from their home and other developed countries, but disregard products from developing countries. Reversely, consumers from developing in general have good feelings toward imported goods from developed countries, but negative judgment towards imported goods from other developing countries, although they are still proved as cosmopolitans (Jin et al., 2014). Jin et al. (2014), nevertheless, focus on COOs of respondent groups, and the comparison between home product image and foreign product image. The effect of COOs of products towards the relationship between cosmopolitanism and product judgment is not apparent. Therefore, this study examines whether there is any correlation between COOs, cosmopolitanism and product judgment. The following hypothesis is proposed: H8: Country of origin (USA/China) moderates the relationship between cosmopolitanism and product judgment Nguyen’s (2004) study of Vietnamese consumers’ purchasing behaviors towards Japanese and Chinese motorbikes shows that Vietnamese ethnocentrism to Chinese products is negatively correlated with perceived quality and not correlated with perceived price. By contrast, concerning Japanese products, ethnocentrism of Vietnamese consumers has a positive correlation with perceived price and no correlation with perceived quality (Nguyen, 2004). Consequently, highly ethnocentric consumers are more likely to severely evaluate foreign products’ defects, but they still objectively judge dominance of foreign products (Nguyen, 2004). Duong (2011) also concludes that COO significantly impacts on consumers’ attitudes and buying intention. Thus, we propose the following hypothesis: H9: Country of origin (USA/China) moderates the relationship between consumer ethnocentrism and product judgment. METHODOLOGY 540
7. Sample Collection Students at a city in Vietnam were selected as respondents for our study for the following reasons. First, students own a greater view than non-students because of their good knowledge and education of historical, political and economic relationship between a foreign nation (e.g. the USA, China) and their aggrieved home nation (e.g. Vietnam) (Bahaee and Pisani, 2009; Cui et al., 2012). Second, these students are living and studying in Hue city, a market of enormous imported products, including US and Chinese products, and thus they are more likely to be exposed to these imported products and accurately evaluate these products (Cui et al., 2012). Since we have a total of 48 items in the questionnaire, the minimum number of respondents we should have is 240 (5 times of the number of items) (Hair et al., 2013). 310 self-administered questionnaires, which contained four pages with eight sections, were directly distributed to students in different classes. The participants were clearly explained all special terms and the purpose for doing the survey. Each participant spent about 15 minutes to complete the questionnaire in an independent, confident and faithful way. After completing the survey, the respondents were thanked for their cooperation. Due to incomplete answers, only 300 questionnaires were used for further analysis. In the sample, the average age was 21 years and 69% of the respondents being female. Measurements and Procedure All measurements were adapted from existing literature: cosmopolitanism from Cleveland et al. (2009), and the remaining scales including animosity, consumer ethnocentrism, product judgment, and willingness to buy from Klein et al. (1998). The measurements used 7-Likert scales with 1 indicated “Strongly disagree” and 7 “Strongly agree”. The original questionnaire was in English. Right translation was important in order to ensure the study work effectively and smoothly. First, the questionnaire was translated from English to Vietnamese for Vietnamese participants. The Vietnamese questionnaire version was then back-translated into English by three bilingual translators. After discovering differences between two English versions, the Vietnamese questionnaire was revised and applied in Vietnam. ANALYTICAL RESULTS PLS Approach Partial least squares (PLS) path modeling analysis was employed in the current research. PLS approach is a statistical technique designed to assess and evaluate estimated parameters in a complex, multivariate relationship between observed and latent (unobserved) variables (Ringle, Wende, and Will, 2005). This approach is more favorably selected in path analysis than covariance based approach (i.e., LISREL or AMOS) because it is not strictly bound by a large sample size or an assumption of normal data distribution. Additionally, PLS is preferred because of its ability to estimate the mediation effect in the model which is considered relatively complicated in covariance-based counterparts (Chin and Newsted, 1999; Haenlein and Kaplan, 2004). Two types of constructs are usually considered in PLS path analysis: reflective and formative. Graphically, a reflective construct is represented by arrows pointing from the latent variable to observed variables (or indicators) while a formative construct is portrayed by arrows pointing from observed variables or indicators to the latent variable (Chin and Newsted, 1999). Only reflective constructs were used in this study. The model was assessed through two sequential steps (1) the analysis of measurement model and (2) structural model analysis. In the first step, reliability and validity of latent constructs 541
8. were evaluated. After it was confirmed, the second step was performed where structural paths between exogenous and endogenous variables were estimated. Measurement Model The measurement model refers to the relationship between observed indicators and underlying (latent) constructs. High order component (HOC) of animosity was employed to capture the high order effect of three animosity constructs: War Animosity, Economic Animosity, and General Animosity (Hair et al., 2013). The measurement model consisting of reflective constructs is analyzed through assessment of two key criteria: reliability and validity (Chin, 1998; Hulland, 1999). Reliability is evaluated at two levels: at the indicator level and at the construct level. At the level of indicator, factor loadings were used for assessment. Thresholds of factor loadings are 0.4 or 0.5 are accepted depending on the research subject and discipline (Chin, 1998; Hulland, 1999). All items had acceptable factor loadings, greater than 0.5, except item 4 of Willingness to Buy (buy 4) and item 2 of Product Judgment (jud 2) were removed due to low loading, 0.519 and 0.231, respectively. Another criterion employed to assess reliability at the construct level is composite reliability (Nunnally and Bernstein, 1994). The value of 0.7 shows high reliability for the construct. Table 1 portrays high reliability in all latent constructs of the model as composite reliability values were well above the thresholds of 0.7. Table 1: Convergent and Discriminant Validity Composite AVE Animosity Consumer Cosmopolit Economic Product War Willingness Reliability Ethnocentri anism Animosity Judgment Animosity to Buy sm Animosity111 Consumer Ethnocentrism 0.812 0.421 0.135 0.649 Cosmopolitanism 0.918 0.652 ‐0.139 ‐0.021 0.808 Economic Animosity 0.903 0.652 0.197 0.148 0.188 0.807 Product Judgment 0.907 0.665 ‐0.144 0.058 0.117 ‐0.168 0.815 War Animosity 0.897 0.743 0.349 0.207 0.135 0.548 ‐0.297 0.862 Willingness to Buy 0.883 0.605 ‐0.339 ‐0.287 ‐0.091 ‐0.379 0.356 ‐0.423 0.778 Note: the numbers on diagonal are square root of AVE In PLS path analysis, both convergent and discriminant validity are used to evaluate validity of a construct. Average variance explained (AVE) is used to be indicative of convergent validity. AVE of a construct represents the amount of variance that indicators of that construct explain relative to the total amount of variance including the variance accounted for by measurement errors (Bagozzi, Yi and Phillips, 1991). A construct that has AVE of greater than 0.5 shows a high level of convergent validity. Accordingly, all latent constructs exceeded the minimum threshold of 0.5 except consumer ethnocentrism (0.421). In addition to convergent validity, discriminant validity is employed to measure validity of an underlying construct. If a constructs shows evidence of discriminant validity, the variance shared between that construct with the block of its indicators is greater than the variance shared by that construct with other constructs with each characterizing another block of different indicators (Hulland, 1999). That is obtained by comparing between the square root of the AVE and correlation coefficients between that construct with others. If the square root of the AVE is greater than the Pearson’s correlation coefficients, the evidence of discriminant validity of a construct is confirmed. Results, shown in Table 1, illustrated that all constructs had acceptable levels of convergent validity. As all criteria to evaluate reliability and validity were fulfilled, the results in the measurement model showed a good fit between the model and the data. 542
9. Structural Model The relationships between exogenous latent variables and endogenous latent variables are tested in the structural model. Unlike the covariance-based approach (i.e., LISREL), PLS path analysis is a technique that is preferred in the predictive modeling and is not restricted by statistical model fit indices as a result of the assumption of distribution free variance approach (Hulland, 1999). Instead, the quality of the structural model is assessed through a series of non-parametric tests. The results of these tests provided the following statistical figures. First, the range of values of the determination coefficients (R2) is from 0 to 1. The higher the R2 value is, the higher percentage of variance in an endogenous variable by the exogenous variables pointing to it. The R2 values, shown in Figure 1, were 0.11 and 0.334 for Product Judgment and Willingness to Buy, respectively. That means 11% and 33.4% of variance in Product Judgment and Willingness to Buy, respectively, were explained by three exogenous variables, Animosity (high order), Consumer Ethnocentrism, and Cosmopolitanism. Despite the shortage of explanations of what the adequate threshold values for R2 is (Stevens, 1999), implications from the results are the conceptual model fits the data reasonably. Figure 1: The Empirical Testing of the Model Second, direction and significance levels of path coefficients estimated in the PLS analysis are tested using a bootstrapping resampling procedure. Means of asymptotic t-statistics generated through this procedure are used to evaluate the goodness of path coefficients. The number of bootstrapping samples was set to 500 and standard error estimates were assessed through t-tests (Chin, 1998; Sellin and Keeves, 1997; (Hulland, 1999). The results pertaining to direction and significance levels of path coefficients were reported in the “Hypotheses Testing” section below. Finally, the effect size is a critical criterion to examine whether the impact of an independent latent variable on a dependent latent variable is substantive or not. The current research applied the method developed by relevant literature (Cohen, 1992; Schroer and Hertel, 2009) to test the effect size f2 in the PLS approach. As Cohen (1992) suggested, if f2 values are 0.02, 0.15, and 0.35, effect sizes are small, medium, and large, respectively. Table 2 shows that all the effect sizes were medium except the low effect of Consumer Ethnocentrism on Product Judgment (0.017). 543
10. Table 2: Summary of f2 Results Product Judgment Willingness to Buy Animosity Consumer Ethnocentrism 0.017 0.075 Cosmopolitanism 0.031 HOC 0.104 0.168 Product Judgment 0.105 Hypotheses Testing Main Effects This section captures the main effects of exogenous variables on endogenous variables that were illustrated in hypotheses from H1 through H6. Particularly, H1 suggests that animosity has a negative impact on product judgment. The results showed that this relationship was supported (b= -0.315, p PJ ‐0.315 ‐0.329 0.078 4.057 0.000 H1 supported CO ‐> PJ 0.169 0.168 0.096 1.770 0.077 H2 marginally supported CE‐> PJ 0.125 0.118 0.083 1.504 0.133 H3 not supported HOC ‐> WTB ‐0.356 ‐0.348 0.077 4.604 0.000 H4 supported CE‐> WTB ‐0.230 ‐0.239 0.040 5.815 0.000 H5 supported PJ ‐> WTB 0.276 0.284 0.060 4.570 0.000 H6 supported Panel B. COO Moderation Effect Coefficients Coefficients Difference t‐Value (China p‐Value Hypotheses Testing (China) (USA) (|China‐USA|) vs USA) (China vs USA) HOC ‐> PJ 0.118 0.114 0.003 0.013 0.990 H7 not supported CO ‐> PJ 0.161 0.316 0.155 0.601 0.548 H8 not supported CE‐> PJ 0.119 0.121 0.002 0.016 0.987 H9 not supported Note: HOC: high order component of Animosity, CO: Cosmopolitanism, CE: Consumer Ethnocentrism, WTB: Willingness to Buy, PJ: Product Judgment 544
11. Moderation Effects A moderating variable (or moderator) is one that changes the direction and/or significance of the effect of an independent variable on a dependent variable (Baron and Kenny, 1986). The moderator can be quantitative (i.e. age, or income) or qualitative or categorical (i.e., gender, education). In this study, the moderator (Country of Origin: China vs. USA) is qualitative. The method used to test the moderation effects was adapted from Hair et al. (2013), which was based on a comparison of path coefficients across two groups of data. To do so, there are three steps required to specify relevant parameters: (1) the number of observations in each group; (2) the path coefficients of each group; and (3) the standard errors of the parameter estimates of each group. The results of testing moderation effects were presented in Panel B of Table 3. The findings showed that the moderation effects of COO were not significant in all hypothesized relationships: animosity and product judgment (H7, p=0.990), cosmopolitanism and product judgment (H8, p=0.548), and consumer ethnocentrism and product judgment (H9, p=0.987). Therefore, H7, H8, H9 were not supported. The implications would be that the strength of the relationship between animosity, cosmopolitanism, consumer ethnocentrism and product judgment are not significantly different across two COOs. DISCUSSIONS AND CONCLUSIONS Discussions Built on the animosity model of foreign product purchase initiated by Klien et al. (1998), the current research seeks to answer three research questions: (1) while the animosity model works in other country, does that work the same when applied in Vietnam? (2) are there any differences in Vietnamese customers’ tendency to purchase products from China and the USA given the fact that Vietnam was in conflict with both countries at a different time? And (3) in addition to animosity and consumer ethnocentrism, does cosmopolitanism have any impact on Vietnamese consumer tendency to purchase foreign products? The findings, based on the student sample of 300, indicate that the model can be applicable in Vietnam with most of the relationship being significant. More specifically, consumer animosity has a significant negative impact on product judgment and willingness to buy; consumer ethnocentrism has a significant negative impact on willingness to buy; product judgment has a significant positive effect on willingness to buy. However, different from the existing literature, the current research shows that the impact of consumer ethnocentrism on product judgment is not significant. The results also show that the influence of cosmopolitanism on product judgment is marginal (b=0.169, p=0.077). Cosmopolitanism is defined as disposition of individuals to view themselves as citizens of the world rather than as citizens of a specific nation (Riefler and Diamantopoulos, 2009) and refers to the extent to which local people are open-minded to the world and other cultures (Skrbis et al., 2004). The evidence from this research somehow points out this relationship. In reality, it can be seen that the more open minded the customers, the more positive they feel about foreign products. The current research attempts to see the effect of country of origin (USA versus China) on customers’ perception about foreign products. Empirical findings show that Vietnamese customers do not see significant difference in terms of country of origin. In other words, the magnitudes of the relationship between animosity, cosmopolitanism, and consumer ethnocentrism and judgment of foreign products from USA and China are not significantly different. 545
12. Managerial and Theoretical Implications The current research has both managerial and theoretical implications. Managerially, the findings have provided international marketing managers with important insights. Consumer animosity happens as a result of national conflicts and affects negatively local consumer attitudes ranging from their belief about foreign products to their willingness to purchase products from the country they conflict with. As earlier discussed, the incident of China’s oil rig served as a catalyst for starting the waves of anger and frustration among Vietnamese consumers. Those waves moved quickly from negative attitudes toward Chinese government and their policies to the boycotting behaviors among local consumers toward Chinese products. No matter how much an international company spends, marketing strategy at that time simply does not work as effectively as it does in a peaceful environment. The lessons learned are international companies should temporarily stop major marketing plans and wait for these waves of frustration to disappear. The results of the study shed light on better insights about the effect of country of origin in the context of animosity. Global marketing marketers should keep in mind that both consumer ethnocentrism and animosity are key antecedents leading to negative attitudes about foreign products among local customers. In this research, a current national conflict, oil rig, is employed to illustrate the extent of animosity of customers toward China and a past national conflict, American war, is used for the American situation. However, the results point out that there are no differences among customers toward purchasing products made in those countries, no matter whether the war or conflict just take place recently or several years ago. With respect of global marketing development, it is important for the marketers to be aware that, within the current context, the strength of influences of key factors like animosity, consumer ethnocentrism, and cosmopolitanism on foreign project judgment is relatively the same, no matter where these products are made. Theoretically, the present research has brought new insights into the consumer animosity literature and developed new opportunities for better understanding the application of animosity into a particular developing country (Vietnam). Different from Klein et al. (1998), the findings confirm that consumer ethnocentrism does not affect product judgment significantly. A possible reason for that observation is that in Vietnam market, the negative effect of consumer ethnocentrism on foreign product judgment is not so strong. That means consumer judgment that can be affected by some factors is not impacted by the level of consumer ethnocentrism. This is in line with Douglas and Nijssen’s (2002) research that indicates that ethnocentrism does not exist alone, but interacts with social and demographic components. The finding is further supported by recent research that asserts demographic characteristics impact at different levels on consumer ethnocentrism (Kamaruddin et al., 2002; Josiassen, Assaf, and Karpen, 2011; Shah and Halim; 2011). To a certain extent, the findings provide evidence supporting the relationship between cosmopolitanism and product judgment. This would add to consumer animosity literature by sharing a balanced point of view between some antecedents that negatively affect local consumer attitudes about foreign products, and another antecedent that explains why local customers favor foreign products. By integrating two types of antecedents, the present research facilitates further development in both research domain including animosity and cosmopolitanism Limitations and Future Research This research suffers from a few limitations. Firstly, this research was conducted in one city in Vietnam, which may have some differences of cultures and tradition in comparison to other 546
13. cities. Moreover, our respondents were only students who are likely to own some distinct characteristics. As a result, our collected sample might partly represent Vietnamese consumers as a whole. Future studies may expand research domain in terms of geographical areas and social groups. Secondly, our sample total is only 300, which is not a great number although it is appropriate for SEM testing. Future research could test a significantly bigger number of respondents in order to enhance a higher level of sample’s representative. Thirdly, for the purpose of this research, a SEM approach was used to examine Vietnamese consumers’ purchase attitude and behaviour towards Chinese and USA products. The application of quantitative methods alone is considered valuable in identifying relationships among variables. However, it has limitations in explaining reasons causing those relationships. Adoption of qualitative methods along quantitative research in future studies may help to further discover the relationships between variables, especially the relationships that were not supported in this study. Finally, the study did not examine whether different demographic characteristics lead to various levels of correlations among variables. The results of this study show that all variables in the research model do not completely explain consumers’ attitude and behavior towards foreign products because other factors, such as economy status of home country and one country’s product availability in the market may have effects. These issues should be taken into account in future research. 547