Marketing Letters

, Volume 23, Issue 3, pp 683–700

What characterizes Chinese consumer behavior? A cross-industry analysis of the Chinese diaspora in Japan

Authors

    • Department of Industrial Engineering and ManagementTokyo Institute of Technology
  • Gulimire Abulaiti
    • Department of Industrial Engineering and ManagementTokyo Institute of Technology
  • Takao Enkawa
    • Department of Industrial Engineering and ManagementTokyo Institute of Technology
Article

DOI: 10.1007/s11002-012-9171-8

Cite this article as:
Frank, B., Abulaiti, G. & Enkawa, T. Mark Lett (2012) 23: 683. doi:10.1007/s11002-012-9171-8

Abstract

In order to profit from China’s enormous business opportunities, international firms need to know Chinese consumer preferences. To learn more about intrinsic Chinese consumer preferences and their distinction from other Asian consumer preferences, this study analyzes differences in the formation of customer satisfaction, repurchase intent, and word-of-mouth intent between Chinese-born and locally born consumers in Japan. Verifying culture-based hypotheses, cross-industry analyses show that Chinese-born consumers pay less attention to the public brand image and risk-related switching costs, but more attention to quality expectations, perceived value, experienced usefulness, and financial switching costs than Japanese consumers. Marketing strategies should account for these preference structures.

Keywords

ChinaCustomer satisfactionCustomer loyaltyJapanRepurchase intentQuality expectationWord-of-mouth

1 Introduction

As Western economies are stagnating and Asia exhibits strong economic growth, the world’s economic center of gravity is moving towards Asia (Grether and Mathys 2008). To profit from new market opportunities, it is thus imperative for Western marketing managers to gain a deeper understanding of consumer behavior in Asia. With a steady record of 10 % annual growth and a population 50 % larger than the entire developed world, China is expected to become the world’s largest economy sometime this century (CIA 2011). Moreover, millions of Chinese are moving abroad and forming major communities in other countries. While the Chinese diaspora already constitutes a major share of the population in numerous Asian countries, globalization may well result in a similar scenario in other developed countries (New World Encyclopedia 2010). Therefore, one of the major business challenges of this century will be to effectively sell to Chinese consumers, whether in China or the rest of the world (Garner 2006; Grether and Mathys 2008).

Despite this immense need for knowledge, the marketing literature has not yet sufficiently clarified the characteristics of Chinese consumer behavior. As pointed out by Heine (2010), a drawback of extant studies comparing consumer behavior in China and other countries is that they confound differences in consumer behavior with differences in both national regulations and the quality of locally offered products and services (Maccoby 1994; Meng and Nasco 2009; Sharma 2010; Zhou et al. 2002). Another drawback is that these studies mainly discuss differences between Western and Eastern consumer behavior, which is imprecise given the large cultural variance within Asia (Heine 2010; Hofstede and Hofstede 2005). Besides, the literature lacks ethnic majority–minority comparisons of consumer behavior within an Eastern context, although globalization increases the degree of such ethnic diversity (Heine 2010). To address these issues and thus contribute to the international marketing literature, this study will compare the preferences of Chinese-born and locally born consumers in a single Asian country, Japan. Research-oriented benefits of such a comparison are that it removes the confounding effects of different offers, quality standards, and national regulations, which are present in two-country comparisons of consumer behavior, and offers a more precise look at what is genuinely Chinese within Eastern consumer behavior. We chose Japan as the country of analysis because it is Asia’s most developed country and has seen a surge in Chinese immigrants over the past years (Japanese Ministry of Justice 2011).

Such a Japanese–Chinese comparison of consumer behavior is beneficial to marketing managers and policy makers in Western countries, Japan, and China. First, many Western managers underestimate intrinsic differences between Japanese and Chinese consumer needs (Meng and Nasco 2009). A better knowledge of these differences should help them improve marketing strategy and enhance customer satisfaction and loyalty in two of the world’s three largest economies (CIA 2011). Second, as the largest export destination of Japan, China is very important to Japan’s economy (Japanese Ministry of Finance 2011), but China’s anger at Japan’s wartime aggression still limits the sales of Japanese firms today (Japan Today 2010). Understanding Chinese consumers and satisfying their needs should be instrumental in overcoming war-related prejudices against Japanese firms. Third, Japan is the number two export destination of China (Chinese Customs 2011). It is crucial for Chinese firms to better understand Japanese consumer needs to improve their long-term export profitability.

To these ends, we will build on extant cultural knowledge (Heine 2010; Hofstede and Hofstede 2005) to develop hypotheses on Japanese–Chinese differences in the formation of customer satisfaction, repurchase intent, and word-of-mouth intent. We will test these hypotheses with consumer data from 13 industries in Japan.

2 Conceptual framework

Our conceptual framework examines the effects of Japanese–Chinese differences on the relationships among consumer attitudes and loyalty intentions. It adopts the structure of the American (ACSI) and European (ECSI) Customer Satisfaction Index models (Fornell et al. 1996; Johnson et al. 2001) and complements it with more recent knowledge of the role of switching costs (Aydin and Özer 2005; Burnham et al. 2003). These standard models assume that customer loyalty intentions are the consequence of customer satisfaction, which is an overall attitude towards products and services, and several attribute-specific attitudes. Customer satisfaction, in turn, is assumed to result from these attribute-specific attitudes.

In terms of customer loyalty intentions, our conceptual framework includes repurchase intent and word-of-mouth intent, which are important to firms as promoters of customer retention and acquisition (Johnson et al. 2001; Reichheld and Sasser 1990). As attribute-specific attitudes, it includes perceived value, the public brand image, current quality expectations, and experienced usefulness. Perceived value is defined as perceived quality compared to the price paid for a product or service. It influences customer satisfaction and intentions because consumers seek to make choices optimizing the return on their investments (Fornell et al. 1996; Johnson et al. 2001). The public brand image is the perceived overall public opinion of a brand. Consumers use the public brand image as a guide to bridge knowledge deficits, reduce risk, and receive recognition by other consumers for popular choices (Johnson et al. 2001). It thus influences customer satisfaction, repurchase intent, and word-of-mouth intent (Ball et al. 2004; Johnson et al. 2001; Türkyılmaz and Özkan 2007).

Standard marketing models also include pre-purchase quality expectations, that is, past expectations of the quality of previously purchased products and services, to explain current customer attitudes and intentions. However, empirical results indicate that such expectations have very weak or no effects (Fornell 1992; Fornell et al. 1996; Johnson et al. 2001). Their role is also strongly criticized from a conceptual perspective (Johnson et al. 2001). As an improvement of standard research models, we suggest that consumer behavior is driven not by past pre-purchase expectations of the quality of previously purchased products and services but by current expectations of the quality of currently offered products and services. Especially, intended future actions such as repurchase and word-of-mouth intent should draw on such current information (Dale 2003; Yi and La 2004), not just on past experiences as in the ACSI model (Fornell et al. 1996). Hence, our conceptual framework replaces the past quality expectations construct (difficult to measure anyway) by current quality expectations.

Another slight modification of standard research models is that we replaced their perceived quality construct by experienced usefulness. This follows the rationale that consumers appreciate less the nature of purchased products than their experienced contribution to consumers’ lives (Thompson et al. 2005). This replacement also reduces the criticized conceptual overlap of perceived quality and value (quality compared to price) (Johnson et al. 2001). Experienced usefulness impacts customer satisfaction (Chen et al. 2007) and may also affect intentions (Bloemer et al. 1999; de Ruyter et al. 1998).

As measures of the perceived market structure, our model includes time-related, risk-related, financial, and relational switching costs (Burnham et al. 2003). These are economic and psychological costs involved in switching from one brand to another and strongly influence customer loyalty intentions (Aydin and Özer 2005; Burnham et al. 2003; de Ruyter et al. 1998).

3 Development of hypotheses

Our hypotheses will deal with how Japanese–Chinese cultural differences should influence customer loyalty intentions (H1a) and moderate the relationships among attribute-specific attitudes, overall attitudes, and loyalty intentions (H1b–H4). They will mainly draw upon Hofstede and Hofstede’s (2005) cultural findings that uncertainty avoidance is lower and long-term orientation and collectivism are higher in China than Japan. We chose Hofstede’s cultural framework because it is the most established and validated one (Taras et al. 2010). Based on our extensive experience in Asia, we consider its cultural scores for Japan and China to be more realistic than those of prominent alternative frameworks.

Despite limited evidence from contexts other than comparisons between Western and Eastern countries, conceptual frameworks in cultural psychology tend to assume that cultural differences in collectivism (vs. individualism) may be the root of all other cultural differences (Heine 2010). For instance, they assume that uncertainty avoidance and long-term orientation both arise from collectivism (vs. individualism) (Heine 2010). However, Japanese culture is characterized by both higher uncertainty avoidance and lower collectivism than Chinese culture (Hofstede and Hofstede 2005), which contradicts this assumption and suggests that uncertainty avoidance is independent from collectivism. Our Japanese–Chinese comparison thus extends frameworks of cultural psychology and treats uncertainty avoidance as separate from collectivism.

Our first set of hypotheses results from these Japanese–Chinese differences in uncertainty avoidance (Hofstede and Hofstede 2005). The decision not to repurchase from a previous provider of (i.e., firm having provided) goods and services but switch to a competitor involves risk and uncertainty because incomplete information on competitors’ offers makes it difficult for consumers to predict whether they will be satisfied with their new choice (Burnham et al. 2003). Consumers who dislike risk and uncertainty should thus be more inclined to repurchase from a previous provider (Liu et al. 2001).

According to Hofstede and Hofstede (2005), the most defining cultural difference between Japanese and Chinese is Japan’s extremely high and China’s low level of uncertainty avoidance. As Japanese have a much greater tendency to avoid uncertainty than Chinese, they should also have a greater tendency to specifically avoid the risk and uncertainty involved in switching providers. Hence, Japanese consumers should react more strongly than Chinese to risk-related switching costs which are the level of risk involved in the process of switching providers (Aydin and Özer 2005; Burnham et al. 2003). In addition, the average level of repurchase intent should be higher for Japanese than Chinese because switching providers always involves a certain level of risk (Burnham et al. 2003). We thus argue that the Japanese–Chinese difference in repurchase intent should be analogous to the observed difference in worker loyalty. Japanese tend to work for the same firm throughout their lives, whereas Chinese tend to quickly switch firms if a seemingly better opportunity comes up (Lee 2010).

Fischer et al. (2010) report that risk aversion is the primary reason why consumers pay attention to the public brand image. If the average social opinion of a brand is more positive, it is less risky to choose that brand or recommend it to others. Hence, Japanese should value the public brand image more than Chinese-born consumers to satisfy their particular need for uncertainty avoidance (Hofstede and Hofstede 2005). This discussion translates to the following hypotheses.
  1. H1a:

    The level of repurchase intent is higher among Japanese than Chinese-born consumers.

     
  2. H1b:

    Risk-related switching costs more strongly affect Japanese than Chinese consumer behavior.

     
  3. H1c:

    The public brand image more strongly affects Japanese than Chinese consumer behavior.

     

Consumers may also express their desire for uncertainty avoidance by a preference for products which they associate with higher quality and thus fewer failures (Fornell et al. 1996). As Chinese culture is characterized by lower uncertainty avoidance than Japanese culture (Hofstede and Hofstede 2005), this perspective may suggest that quality expectations play a smaller role in the formation of Chinese than Japanese customers’ loyalty intentions (H2a).

By contrast, the perspective of long-term vs. short-term orientation, which the field of cultural psychology associates with collectivism vs. individualism (Heine 2010), may indicate the opposite. Hofstede and Hofstede (2005) report stronger long-term (vs. short-term) orientation for China than Japan. Long-term orientation refers to stressing future advancement, whereas short-term orientation refers to stressing past values. According to Hofstede and Hofstede (2005), China’s highly pronounced long-term orientation derives from Confucianism which strongly influenced China but not Japan (Shiba 2006).

In our conceptual framework, quality expectations designate consumer expectations of the quality of currently offered goods and services. Unlike other attribute-specific attitudes of our conceptual framework, they thus relate to products and services not purchased in the past but currently offered. Long-term-oriented consumers should base their decisions more strongly on these quality expectations to improve their situation in the future. By contrast, short-term-oriented consumers should rely more strongly on indicators of past experiences as they stress past values (Hofstede and Hofstede 2005). Since Chinese have a stronger long-term (vs. short-term) orientation than Japanese (Hofstede and Hofstede 2005), this rationale implies that Chinese-born consumers pay more attention to quality expectations than Japanese (H2b, opposite to H2a).
  1. H2a/b:

    Quality expectations less (a)/more (b) strongly affect Chinese than Japanese consumer behavior.

     
Cultural research also shows that China is a much more collectivist (vs. individualistic) society than Japan (Hofstede and Hofstede 2005; Shiba 2006). Since meta-analytical research reports that Japan is far less collectivist (vs. individualistic) than commonly perceived in the Western world (Money et al. 2006; Takano and Osaka 1999), Hofstede and Hofstede’s (2005) result can be considered a conservative indicator of the collectivism gap between Japan and China. As broadly shown in cultural psychology, this cultural difference suggests that maintaining personal relationships is more important to Chinese than Japanese (Heine 2010). Relational switching costs denote the psychological barriers to switching providers caused by the fear of damaging personal relationships with staff and peer consumers. The effect of relational switching costs on customer loyalty intentions is stronger for consumers who care more about relationships (Aydin and Özer 2005; Burnham et al. 2003). This effect should thus also be stronger for Chinese-born than Japanese consumers because Chinese are more collectivist and care more about relationships (Hofstede and Hofstede 2005).
  1. H3:

    Relational switching costs more strongly affect Chinese than Japanese consumer behavior.

     

Our last set of hypotheses is an implication of Japanese–Chinese differences in economic development. While China is much poorer and less developed than Japan, it was even more so a few decades ago, before China implemented the economic reforms that put it on the path to economic growth (CIA 2011). Hence, Chinese-born consumers spent their early childhood, which is crucial to the absorption of cultural values (Hofstede and Hofstede 2005), in a much poorer environment than Japanese. Average Chinese-born consumers have thus learnt to attach greater importance than Japanese consumers to both optimizing the benefits and avoiding waste of financial resources (Meng and Nasco 2009). This should lead to a number of Japanese–Chinese differences in consumer behavior.

Financial switching costs are financial costs caused by the process of switching from one provider of goods and services to another. In a decision on whether to switch providers, these costs represent a waste of money which is weighed against potential benefits of other providers (Aydin and Özer 2005; Burnham et al. 2003). As Chinese-born consumers attach greater importance to avoiding waste of financial resources than Japanese, their loyalty intentions should be more strongly affected by financial switching costs. Likewise, perceived value should be more important to Chinese-born than Japanese consumers. Perceived value is the ratio of perceived quality over price and represents the return on consumption investment (Fornell 1992). To avoid wasting financial resources, Chinese-born consumers should pay more attention to optimizing the return on their financial investments and thus to perceived value.

Similar to this argumentation, Chinese-born consumers should focus on products proven useful and cut back on products proven problematic or unneeded in order to minimize waste. Risk affinity (Hofstede and Hofstede 2005) and the importance of waste minimization cause Chinese-born consumers to buy low-priced products (Meng and Nasco 2009) whose operational benefits may not be entirely clear up front. As such a high-risk strategy inevitably leads to a number of negative product experiences, it is rational within this consumption strategy to value products proven useful and discard products proven problematic and unneeded in daily use. Chinese-born consumers should thus strongly value the experienced usefulness of previously purchased goods and services. By contrast, Japanese consumers may not need to pay as much attention to post-purchase experienced usefulness as Chinese if they purchase expensive and reputable products in the first place to avoid uncertainty (Hofstede and Hofstede 2005; Meng and Nasco 2009; see hypothesis H1c).
  1. H4a:

    Financial switching costs more strongly affect Chinese than Japanese consumer behavior.

     
  2. H4b:

    Perceived value more strongly affects Chinese than Japanese consumer behavior.

     
  3. H4c:

    Experienced usefulness more strongly affects Chinese than Japanese consumer behavior.

     

4 Methodology

In order to test our research hypotheses in a cross-industry context, we designed a questionnaire survey measuring Japanese and Chinese-born consumers’ attitudes towards their primarily used brand (as in the ACSI and ECSI methodologies: Fornell et al. 1996; Johnson et al. 2001) in 13 diverse (in terms of switching costs, products/services, repurchase cycles, prices, tangibility) industries relevant to both consumer groups in Japan: automobiles, game consoles, mobile phones, personal computers, shampoo (five products), airlines, banks, convenience stores, gas stations, hairdressers, hospitals, mobile carriers, and supermarkets (eight services).

The questionnaire design followed the procedures of Rossiter (2002), Drolet and Morrison (2001), and Bergkvist and Rossiter (2007, 2009) which revised traditional concepts of scale design. Specifically, these articles proved the traditional notion wrong that reflective multi-item scales are always preferable to single-item scales. Bergkvist and Rossiter (2007, 2009) proved that constructs with concrete singular objects and concrete attributes should be measured with single-item scales because additional items do not add any predictive validity, run the risk of tapping into other predictive attributes not covered by the construct, and increase common method variance. As examples of constructs with concrete singular objects and concrete attributes, they cited product-specific perceived quality and overall brand attitude. As our survey was to ask questions on consumers’ primarily used brand, which is a concrete singular object, and measure constructs falling into Bergkvist and Rossiter’s (2007, 2009) definition of concrete attributes, we thus used single-item measurement. We used 10-point scales to achieve sufficient discrimination between single-item response categories (Bergkvist and Rossiter 2007).

To provide additional scale validation, we collected data from 871 consumers in December 2010 and January 2011 on a single industry (mobile phones) using the multi-item scales from which our single-item scales were extracted. In the dataset these scales fulfill the requirements of convergent and discriminant validity (α > .7; average variance extracted [AVE] > .5; AVE > squared construct correlations; confirmatory factor analysis: CFI > .95; χ2/df < 3). Following the procedures of Bergkvist and Rossiter (2007) for assessing scale-specific predictive validity, we regressed single-item customer satisfaction/repurchase intent/word-of-mouth intent upon its antecedents modeled by multi-item scales (R2 = .50/.35/.40) and, alternatively, reduced single-item scales (R2 = .51/.34/.39). As the predictive validity (R2) of the multi- and single-item scales is virtually equivalent for our doubly concrete constructs, single-item scales are more efficient and should be used (Bergkvist and Rossiter 2007). In our study this efficiency allows for collecting data from multiple industries.

For the example of shampoo, the questionnaire of our main (not the validation) study includes the following questions (English translation; sources in Table 1): (repurchase intent) “How likely are you to buy your next shampoo from the same brand?”; (word-of-mouth intent) “How likely are you to recommend the brand of your shampoo to others?”; (customer satisfaction) “What is your overall satisfaction with this shampoo?”; (perceived value) “How is the quality of your shampoo in relation to the price you initially paid for it?”; (experienced usefulness) “Has this shampoo been useful in making your life more convenient and fulfilled?”; (quality expectations) “What are your expectations regarding the quality of the shampoo products currently offered by this brand?”; and (public brand image) “What is your perception of the overall public image of this brand?” To assess switching costs, respondents indicated the degree to which they agreed with the following statements if faced with the decision to repurchase from the same brand or switch to another: (time-related switching costs) “Switching to another brand involves additional time and effort”; (risk-related switching costs) “Switching to another brand involves quality-related risk”; (financial switching costs) “Switching to another brand incurs additional financial costs”; and (relational switching costs) “I am afraid to lose personal relationships (with friends, staff, other users, brand community) by switching to another brand.”
Table 1

Correlations and descriptive statistics

Variable

1

2

3

4

5

6

7

8

9

10

11

1

Repurchase intent

           

2

Word-of-mouth intent

.68

          

3

Customer satisfaction

.59

.57

         

4

Perceived value

.52

.56

.69

        

5

Experienced usefulness

.48

.49

.52

.53

       

6

Quality expectations

.57

.59

.62

.59

.58

      

7

Public brand image

.62

.63

.61

.59

.51

.67

     

8

Time-related switching costs

.18

.16

.07

.08

.16

.11

.10

    

9

Risk-related switching costs

.26

.29

.17

.17

.17

.25

.22

.58

   

10

Financial switching costs

.10

.16

.04

.08

.10

.10

.09

.53

.51

  

11

Relational switching costs

.10

.19

.08

.09

.07

.12

.09

.26

.34

.34

 

Mean

6.44

5.69

6.41

6.21

6.58

6.41

6.56

5.63

5.30

5.10

3.05

Standard deviation

1.98

2.15

1.67

1.65

1.84

1.77

1.61

2.65

2.55

2.73

2.49

Sample number

3,168

3,166

3,171

3,168

3,169

3,169

3,168

3,168

3,168

3,167

3,165

All correlations significant at p < .05. Scales for variables 1, 3, 4, 5 (adapted), and 6 (adapted) from Fornell et al. (1996); 2 and 7 from Johnson et al. (2001); and 8 to 11 from Burnham et al. (2003)

After a pre-test and refinement of our questionnaire, we collected data from 227 Japanese-born and 92 Chinese-born consumers in Japan from March to June 2008 (response rate: 41 %). In the Tokyo metropolitan area, which absorbed most of the recent surge in Chinese immigration (Japanese Ministry of Justice 2011), we personally distributed questionnaires in malls, public places, collaborating firms, and universities with the request to fill them out immediately or mail them back to us. We randomized the order of industries in the questionnaire across our sample to minimize order effects. To allow for a sound comparison of the effects of Chinese vs. Japanese culture on consumer preferences, we sought to assure that our Chinese-born respondents have not yet deviated from Chinese culture. Hence, we filtered potential Chinese-born respondents based on the criteria that they immigrated within the past 2 to 5 years and are proficient in Japanese. For additional verification of differences between our ethnic sub-samples, our survey included the following questions on individualism (IDV) and uncertainty avoidance (UA) with five-point semantic-differential scales from Hofstede and Hofstede (2005): (IDV) “Think in terms of ‘we’” (−2) vs. “Think in terms of ‘I’” (2); (UA1) “There should be no more rules than strictly necessary” (−2) vs. “There is an emotional need for rules, even if these were implicit” (2); (UA2) “Competition among colleagues usually does more good than harm” (−2) vs. “Competition among colleagues usually does more harm than good” (2). These questions on personal attitudes will serve for verification rather than testing our culture-based hypotheses because culture is an ideal standard of behavior which humans acquire through learning (especially during childhood) and take into account even if personal attitudes deviate from this standard (Hofstede and Hofstede 2005). Culture is thus independent from personality and cannot simply be measured at the personal level (Hofstede and Hofstede 2005).

For other purposes, we collected further data on IDV, UA1, and UA2 in Japan (greater Tokyo) from October 2008 to February 2009 and October to November 2009 (sample size: 1229) and in China (Beijing, Shanghai, Wuhan, Ürümqi) from September 2008 to March 2009 (sample size: 1012). Based on these samples, our variables have the following means for Japanese-born consumers in Japan/Chinese-born consumers in Japan/Chinese-born consumers in China: (IDV) .44/.21/−.08; (UA1) .30/−.08/−.02; (UA2) −.34/−.46/−.43. The differences between Japanese- and Chinese-born consumers in Japan are significant (p < .05) with regards to IDV and UA1 and marginally significant (p < .1) with regards to UA2. This corresponds with Hofstede and Hofstede (2005), although the differences are weaker and these personal scores should not be understood as cultural scores. Chinese-born consumers in Japan and China differ significantly with regards to IDV but not UA1 and UA2. Compared with Chinese in China, Chinese immigrants to Japan might thus differ slightly less from Japanese in terms of personal individualism. In the worst case, this might lead to weaker support for H3 than would be justified for a completely non-biased Chinese-born sample. However, our conceptual framework does not rely on personality but on culture which Chinese have learnt during childhood, independently of whether they moved to Japan or not (Hofstede and Hofstede 2005). Hence, negative consequences of this sample bias for our analyses should be limited.

The sample of our main survey on Japanese- and Chinese-born consumers in Japan is uniformly distributed across gender but relatively young compared with Japan’s population. The young age does not need to be a problem but may actually help compare Japanese and Chinese because Chinese immigrants tend to be relatively young as political obstacles previously made it difficult for Chinese to move to Japan (Schaefer 2007). To enhance this comparability further, our analysis will correct for Japanese–Chinese sample differences in age, gender, and industry selection. Across industries, our sample totals 4,282 (less in our analysis due to missing values) industry-specific responses. A comparison of early (immediate) and late (closest to non-respondents: late mail) respondents does not indicate any non-response bias in our data (Armstrong and Overton 1977). Table 1 summarizes the construct correlations and descriptive statistics.

To test for common method variance, our survey also contains a question on stress (Hofstede and Hofstede 2005), which is theoretically unrelated to the other constructs. As the correlations between this and the other questionnaire items are slightly negative, our study does not seem to suffer from problems of common method variance (Lindell and Whitney 2001). Harman’s single factor test does not indicate problems of common method variance, either (Podsakoff et al. 2003).

5 Results

In order to test our hypotheses, we conducted three hierarchical linear modeling (HLM; Kreft and de Leeuw 1998) analyses of our pooled cross-industry data with customer satisfaction, repurchase intent, and word-of-mouth intent as dependent variables. These cross-classified HLM models consist of two hierarchical dimensions. Along the spatial dimension, industry-specific responses for up to 13 industries per respondent (level 1) are nested in respondents (level 2). Along the industrial dimension, industry-specific responses (level 1) are nested in an industrial context (level 2). Accounting for this structure corrects for potential lack of data independence within hierarchical clusters, which might violate statistical assumptions of regression (Kreft and de Leeuw 1998). Moreover, this procedure corrects for respondent-specific net acquiescence response styles, that is, respondent-specific tendencies to provide generally higher or lower responses (Baumgartner and Steenkamp 2001).

As independent variables, the HLM models contain gender (0, male; 1, female), age, and a product (0) vs. service (1) dummy as control variables; customer satisfaction, perceived value, experienced usefulness, quality expectations, and the public brand image as provider-related attitudes; time-related, risk-related, financial, and relational switching costs (not included in the model for customer satisfaction); and the country of birth (1, China; 0, Japan) and its interactions with the other predictors as effect variables. All models include an intercept, level-specific random terms, and random terms capturing cross-level effect variations. We chose standard linear models because non-linear, polynomial models did not significantly enhance the model fit. The parameters were estimated by full maximum likelihood. Restricted maximum likelihood yielded virtually identical results. Multi-collinearity was not a problem in this study (Mason and Perreault 1991).

The independent variables explain 57.6 (R2), 53.5, and 52.0 % of the variance in customer satisfaction, repurchase intent, and word-of-mouth intent, respectively (see Table 2). These values are relatively good compared with the literature (Fornell et al. 1996; Johnson et al. 2001). In all analyses, the model fit increased significantly after adding interaction terms. Japanese-born customers’ satisfaction is most strongly influenced by perceived value, followed by quality expectations, the public brand image, and experienced usefulness. In terms of direct effects, repurchase intent is most strongly influenced by the public brand image, followed by customer satisfaction, quality expectations, experienced usefulness, risk-related switching costs, perceived value, and time-related switching costs. Word-of-mouth intent is most strongly affected by the public brand image, followed by quality expectations, perceived value, customer satisfaction, perceived usefulness, and risk-related switching costs. While the formation processes of repurchase and word-of-mouth intent are thus similar, a few differences emerge. Consumers pay more attention to customer satisfaction and time-related switching costs when deciding on repurchases, whereas they pay more attention to perceived value and quality expectations when recommending products and services. Switching costs exert weaker effects than described by the literature (Burnham et al. 2003).
Table 2

Japanese–Chinese differences in the formation of customer attitudes and intentions

 

Dependent variable

 

Customer satisfaction

Repurchase intent

Word-of-mouth intent

Hypothesis

Independent variable

γ

γ

γ

 

Intercept

.951***

−1.035***

−1.617***

 

Control variables

   Female (1; vs. male: 0)

.003

−.090

.072

 

   Age

−.029

.016

−.059

 

   Service (1; vs. product: 0)

.018

.831***

.349**

 

Provider-related attitudes

   Customer satisfaction [CS]

 

.259***

.167***

 

   Perceived value [PV]

.405***

.057*

.187***

 

   Experienced usefulness [ExpU]

.083***

.144***

.121***

 

   Quality expectations [QExp]

.211***

.148***

.205***

 

   Public brand image [image]

.182***

.382***

.349***

 

Switching costs [SC]

   Time-related SC

 

.055***

−.001

 

   Risk-related SC

 

.061***

.086***

 

   Financial SC

 

−.012

.011

 

   Relational SC

 

.003

.020

 

Country of birth

   China (1; vs. Japan: 0)

−.067

−.199

−.539

H1a: − (only rep. intent)

Moderating effects of the country of birth

   China × CS

 

.036

.176***

 

   China × PV

.018

.097*

.041

H4b: +

   China × ExpU

.121***

.075*

.046

H4c: +

   China × QExp

−.012

.100*

.125**

H2a/b: −/+

   China × image

−.156***

−.288***

−.285***

H1c: −

   China × time-related SC

 

−.025

.026

 

   China × risk-related SC

 

−.069*

−.055

H1b: −

   China × financial SC

 

.052*

−.005

H4a: +

   China × relational SC

 

.024

.061*

H3: +

Model fit

   HLM pseudo R2

.576

.535

.520

 

   –2 Log likelihood

9,484

10,827

11,041

 

   Sample number

3,166

3,160

3,159

 

The analysis corrects for personal differences in response styles

Hierarchical linear modeling (HLM) analysis: maximum likelihood estimation

HLM Pseudo R2 from Kreft and de Leeuw (1998)

*p < .05; **p < .01; ***p < .001

When accounting for indirect effects of attribute-specific attitudes on customer intentions via customer satisfaction, their total (direct + indirect) effects on repurchase intent are as follows (in descending order of strength): public brand image (.429), quality expectations (.202), experienced usefulness (.166), and perceived value (.161). Attribute-specific attitudes exert the following total effects on word-of-mouth intent: public brand image (.380), perceived value (.255), quality expectations (.240), and experienced usefulness (.134).

Table 2 shows that the country of birth does not exert any significant main effect on repurchase intent. Hence, hypothesis H1a is not supported, although the HLM coefficient shows the predicted tendency. Regarding the hypothesized moderating effects of the country of birth, our results indicate that risk-related switching costs have stronger effects on repurchase intent for Japanese than Chinese-born consumers (supports H1b). By contrast, their effects on word-of-mouth intent do not show Japanese–Chinese differences. The public brand image has much stronger effects on customer satisfaction and intentions for Japanese than Chinese-born consumers (supports H1c).

Quality expectations exert stronger effects on repurchase and word-of-mouth intent (but not customer satisfaction) for Chinese-born than Japanese consumers (supports H2b as opposed to H2a). Relational switching costs have a stronger influence on word-of-mouth (but not repurchase) intent for Chinese-born than Japanese consumers (supports H3). Moreover, financial switching costs and perceived value have stronger effects on repurchase intent for Chinese-born than Japanese consumers (supports H4a–b). This difference does not occur in the formation of customer satisfaction and word-of-mouth intent. The effects of perceived usefulness on customer satisfaction and repurchase intent (but not word-of-mouth intent) are stronger for Chinese-born than Japanese consumers (supports H4c).

An additional analysis including interaction terms of the product vs. service dummy and the effect variables does not indicate that the distinction between products and services significantly moderates our focal effects. Besides, our dataset includes industry-specific variables on consumer experience for several but not all industries. Analyses including these variables and their interactions with the effect variables do not indicate any deviation from our conclusions.

In summary, our analyses support all hypotheses except H1a and H2a, but the predicted Japanese–Chinese differences in consumer behavior occur at different locations within our conceptual framework. Beyond our hypothesized effects, we found that customer satisfaction is a stronger predictor of word-of-mouth intent for Chinese-born than Japanese consumers.

Our conceptual framework predicts these effects by cultural differences between Japan and China, that is, by country differences in behavioral rules acquired through learning (especially during childhood). As it is impossible to statistically validate the mechanisms behind these effects with a sample size of only two cultures, we drew on cultural research in the literature to support our hypotheses. However, it might be possible to obtain at least secondary evidence of the mechanisms behind these effects by measuring personality instead of culture and assuming similarity between the effects of personality and culture. While culture and personality exert independent influences on behavior, these influences are often (but not always) similar, even though cultural influences tend to be much larger (Hofstede and Hofstede 2005; Taras et al. 2010). At best, influences of personality (as a proxy measure of culture) may thus underestimate the magnitude of cultural influences.

To provide such secondary evidence, we used our personality traits (sample size: 319 respondents at level 2) of individualism (IDV) and uncertainty avoidance (UA1, UA2), which relate to our supported hypotheses H1b–c (mechanism: uncertainty avoidance) and H3 (mechanism: individualism). Based on our HLM models and the mediated moderation tests proposed by Muller et al. (2005), we tested whether IDV, UA1, and UA2 mediate (p < .1) the hypothesized moderating effects of the country of birth on the formation processes of customer satisfaction, repurchase intent, and word-of-mouth intent. While UA1 mediates the moderating effects of the country of birth on the effects of the public brand image (H1c) on customer satisfaction, repurchase intent, and word-of-mouth intent, UA2 mediates the moderating effects of the country of birth on the effects of risk-related switching costs (H1b) on repurchase and word-of-mouth intent. IDV mediates the moderating effects of the country of birth on the effects of relational switching costs (H3) on repurchase and word-of-mouth intent (all coefficient signs as predicted). These mechanisms of personality may point to analogous mechanisms of culture.

6 Discussion

Over the past decades, Asian countries, and China in particular, have demonstrated tremendous economic growth (CIA 2011). In order to profit from the enormous business opportunities arising from Chinese consumers’ increasing wealth and migration (Garner 2006), international firms need to know Chinese consumer preferences. Past international comparisons of consumer behavior mostly discussed broad differences between the East and West and confounded environmental with cultural differences in consumer behavior (Maccoby 1994; Meng and Nasco 2009; Sharma 2010; Zhou et al. 2002). To overcome these limitations (Heine 2010), our study discussed differences within Eastern consumer behavior and isolated cultural from environmental differences by analyzing the behavioral distinction between Chinese-born and locally born consumers in a single Asian country, Japan. Based on extant cultural knowledge of Japanese–Chinese differences (Hofstede and Hofstede 2005) in uncertainty avoidance (H1a–c, H2a), long-term orientation (H2b), collectivism (H3), and financial attitudes (H4a–c), we hypothesized that Chinese consumer attitudes and intentions are more strongly influenced than Japanese by relational switching costs (H3), financial switching costs (H4a), perceived value (H4b), and experienced usefulness (H4c) and less strongly by risk-related switching costs (H1b) and the public brand image (H1c). Further, we found that theoretical perspectives based on uncertainty avoidance (H2a) and long-term orientation (H2b) differ in their prediction of whether quality expectations have smaller (H2a) or greater (H2b) effects on Chinese than Japanese consumers. We also hypothesized that the average level of repurchase intent is higher for Japanese than Chinese-born consumers (H1a).

Our results support all hypotheses except H1a and H2a. Moreover, they indicate that Japanese–Chinese differences do not always occur uniformly across the formation processes of repurchase and word-of-mouth intent. Hence, it may make a difference for the hypothesized Japanese–Chinese behavioral distinction whether consumers decide on their own purchases (repurchase intent) or contribute to other consumers’ purchases (word-of-mouth intent). Specifically, we found that the Japanese–Chinese differences hypothesized as a consequence of cultural differences in uncertainty avoidance (H1b–c) and financial attitudes (H4a–c) are more pronounced in the formation of repurchase than word-of-mouth intent. Fear of uncertainty and financial strictness thus seem to affect consumers’ own purchase decisions, but not their contributions to other consumers’ purchase decisions. By contrast, the Japanese–Chinese difference in the effect of relational switching costs, hypothesized as a consequence of higher Chinese than Japanese collectivism, is more pronounced in the formation of word-of-mouth than repurchase intent. Hence, collectivism seems to exert stronger influences on communication than purchase behavior. Our results further reveal that customer satisfaction is a better predictor of word-of-mouth intent for Chinese-born than Japanese consumers, which may also relate to higher Chinese than Japanese collectivism (Hofstede and Hofstede 2005).

Based on our results, we recommend managers to allocate their limited marketing resources in a way that primarily optimizes the public brand image and secondarily enhances customer satisfaction, generates high consumer expectations towards currently offered products and services, delivers beneficial customer experiences, and offers great value in order to make customers engage in repurchase behavior and word-of-mouth referrals. This non-intuitive finding is valuable because Japanese managers traditionally consider customer satisfaction to be the primary driver of repurchase and word-of-mouth intent and underestimate the role of the public brand image (Abulaiti et al. 2010). In contrast to results of Western studies (e.g., Burnham et al. 2003), setting up hurdles to switching seems to be a less successful strategy in Japan. When marketing to Chinese-born (vs. Japanese) consumers, fostering a great public brand image and setting up risk-related switching barriers is less important, whereas the other success factors are more important. Chinese-born consumers are more responsive to high quality expectations and good value, whereas Japanese react more strongly to a good public brand image.

Hence, we advise managers serving Japanese customers to invest their limited marketing resources into signals of low risk such as the public brand image which conveys security as it reflects perceived experiences of many other consumers (Fischer et al. 2010). In order to influence Japanese customers’ behavior, it also seems to be particularly effective to set up high risk-related switching costs. For instance, firms might communicate that using competitors’ brands is risky and dangerous or that not purchasing certain options leads to risky situations (Burnham et al. 2003). By contrast, managers targeting Chinese consumers should focus on communicating good value and thus low prices. Moreover, they should effectively use contractual design and loyalty programs to set up financial switching costs (Burnham et al. 2003) to which Chinese customers are particularly responsive. In order to influence Chinese customer intentions, firms should further place special importance on proactive communication of new product benefits to leverage the particular importance of current expectations to Chinese customers. Besides, firms should build personal bonds with customers and thus relational switching costs in China, whereas Japanese customers seem to prefer more anonymous relationships, at least in B2C markets.

As the Japanese–Chinese differences in the formation processes of repurchase and word-of-mouth intent are not fully identical, marketing managers need to adapt their strategy based on whether their primary target is to retain existing customers by stimulating repurchase behavior or acquire new customers by stimulating word-of-mouth referrals. As China is a growing market (CIA 2011; Garner 2006), where acquiring new customers is more important than retaining existing customers (Reichheld and Sasser 1990), the benefits of effectively stimulating word-of-mouth referrals should be particularly high. Hence, marketing managers targeting China should pay particular attention to our findings on the success factors driving Chinese word-of-mouth intentions (see Table 2).

We recommend researchers modeling consumer behavior to consider the effect of the public brand image which is very strong but mostly overlooked, even by famous scholars (Fornell 1992; Fornell et al. 1996). As an innovative choice of our study, our models included two variables which the literature has virtually not taken into account: experienced usefulness and expectations regarding the quality of currently offered goods and services. Our analysis showed that their effects are of the same magnitude as other established effects and should thus receive more attention by researchers. In particular, we recommend researchers to replace the past pre-purchase quality expectations construct of many standard research models, whose effects are almost never significant (Fornell et al. 1996; Johnson et al. 2001), by our current quality expectations construct which exerts substantial influences on consumer behavior. As the effects of switching costs were much weaker than claimed by the literature (Burnham et al. 2003), they might not be as crucial as previously thought. Our finding of Japanese–Chinese behavioral differences in Japan also implies that studies using data from multi-cultural communities may not validly claim that their effects exist equally across all consumer groups. We thus recommend these researchers to include the country of birth or other cultural variables as moderators in their models. Moreover, our Japanese–Chinese comparison showed that uncertainty avoidance may not arise from collectivism (vs. individualism) as assumed by studies in cultural psychology whose choice of contexts masks the cultural variance within Asia (Heine 2010).

A limitation of our research is that our analyses included data from only 13 industries. Although this scale is more extensive than in most academic marketing studies, data from more industries might be necessary to entirely remove potential industry-specific biases. We thus encourage future research with greater budgets to retest our hypotheses with data from more industries. Another limitation is that we measured subjective consumer intentions which may not always fully concur with real consumer action. We thus recommend future research with access to objective measures to verify our results. Researchers might also analyze differences between Chinese consumers’ behavioral patterns in China and various other countries. Such research should be able to quantify the effects of country conditions on Chinese consumer behavior. While panel data is relatively difficult to obtain in Japan, where consumers have strong privacy concerns (Shiba 2006), future research with access to such data should retest our hypotheses with longitudinal data. In particular, this approach would become necessary when using objective measures of future consumer actions instead of consumer intentions.

Besides, we encourage scholars to conduct more cross-industry research because most current marketing studies suffer from a single-industry bias and tend to falsely generalize results beyond the specific context of their focal industry. Moreover, we advise international researchers to conduct more empirical studies outside the Western cultural hemisphere and be more tolerant in evaluating international research describing psychological effects distinct from what might be expected in Western cultures (also see Heine 2010). This might help enhance the still limited knowledge of international differences in consumer behavior and enable firms to better compete in the increasingly global marketplace.

Acknowledgments

We would like to thank the Co-Editor, Frank R. Kardes, and two anonymous reviewers for their helpful and constructive comments on earlier drafts.

Copyright information

© Springer Science+Business Media, LLC 2012