Global Leadership

Global Brand Architecture Position and Market-Based Performance: The Moderating Role of Culture M. Berk Talay, Janell D. Townsend, and Sengun Yeniyurt

ABSTRACT Companies expend vast resources to create product and brand portfolios in the global marketplace. Yet knowledge of the market-based performance implications of various positions in a firm’s portfolio architecture is lacking in the litera- ture. To further the understanding of managing brands in the global marketplace, the authors develop a conceptual framework based on the tenets of signaling theory, explore the relationship between global brand architecture and market-based performance, and consider how culture moderates this relationship. The results of the analyses, from a panel data set of 165 automotive brands operating in 65 countries from 2002 to 2008, reveal that global brands per- form better in the marketplace than their nonglobal counterparts. Cultural values indeed provide boundary conditions for this relationship, suggesting that alternative strategies for some markets may be advisable.

Keywords: global brands, global brand architecture, culture, signaling, panel data analysis

S trategically managing the identity of brands and products in a global environment is among the most challenging activities for executives of com-

panies of all sizes. Choosing the markets to serve and the means to serve them is a fundamental issue, and immense resources are expended to implement strategies to achieve performance objectives. The dynamics of globalization and competition have caused multi- national companies to evolve from parochial strategies focused on individual markets toward more complex portfolio management strategies that transcend national boundaries. Developing the best range of focus for brands and products goes beyond features and attrib- utes; in the contemporary environment, it also includes strategic geographic scope considerations.

The geographic range strategy that firms employ to man- age their brands is characterized by a hierarchical struc- ture of products and brands present in global markets; that is, firms often employ a variety of options in their

branding strategies, from local/domestic branding to branding with regional, multiregional, or global orienta- tions. This phenomenon is referred to as a global brand architecture (GBA) and specifically refers to the portfolio of brands a firm controls on a continuum of geographic scope and degree of consistency (Douglas, Craig, and Nijssen 2001; Townsend, Yeniyurt, and Talay 2009). Developing a rational GBA is a key element of a firm’s overall international marketing strategy because it offers a foundation to leverage its brands’ equity across mar- kets, integrate acquired brands, and rationalize global strategies (Douglas, Craig, and Nijssen 2001).

To develop and deploy a GBA effectively, it is important to understand the market-based performance outcomes derived from various options within the GBA. Although the GBA is not a new concept, actual market-based per- formance of brands at various positions in a GBA has yet to be grounded with empirical support; that is, exist- ing studies have not addressed whether global brands actually perform better than single-country, regional, or multiregional brands. This study expands the extant

M. Berk Talay is Associate Professor of Marketing, University of Massachusetts Lowell (e-mail: [email protected]). Janell D. Townsend is Associate Professor of Marketing, Oakland University (e-mail: [email protected]). Sengun Yeniyurt is Associate Pro- fessor of Marketing, Rutgers University (e-mail: yeniyurt@rci. rutgers.edu). Seigyoung Auh served as associate editor for this article.

Journal of International Marketing ©2015, American Marketing Association Vol. 23, No. 2, 2015, pp. 55–72 ISSN 1069-0031X (print) 1547-7215 (electronic)

Global Brand Performance 55

56 Journal of International Marketing

research on global brands by investigating how GBA affects a brand’s performance in the global marketplace.

It is likely that global brands perform better in different markets because of the influence of national culture. Different regions of the world seem to have varying mechanisms through which global brand perceptions and attitudes are formed, processed, and employed in purchase decisions. Akdeniz and Talay (2013) find mul- tifarious effects of culture as moderators of the relation- ship between product signals and performance. Research in the Chinese market has shown that global retail brands influence customers through different functional and psychological values (Swoboda, Penne- mann, and Taube 2012). In Eastern European markets, global brands have been found to be perceived as a pass- port to global citizenship (Strizhakova, Coulter, and Price 2008). Dimofte, Johansson, and Bagozzi (2010) find support for the notion that ethnic cultures within a country act as moderators of global brand quality per- ceptions. Therefore, it is well established that national culture plays a significant role in directly influencing consumer financial decision making and also moderates the impact of marketing efforts by the financial services firm (Petersen, Kushwaha, and Kumar 2015).

Although culture has been addressed as an important con- sideration from many perspectives, research on its role in moderating the relationship between branding strategies and market performance is lacking. To that end, we iden- tify extrinsic boundaries based on cultural factors, which can cause the performance of alternative strategic approaches to vary in global markets. This enables us to contribute to the global branding literature by extending the understanding of how culture affects managerial deci- sions, which will help improve opportunities for successful GBA management. We suggest, and empirically demon- strate, that the effects of culture are likely to change the relationship between the signal sent by the brand’s posi- tion in the GBA and the brand’s market-based perfor- mance. Research from a broad array of disciplines has posited culture as a measure of values and beliefs and has widely viewed it as a precursor to the acceptance of global brands in a country due to the advent of global consumer cultures (Alden, Steenkamp, and Batra 2006). When in doubt, consumers will choose products and brands that have synergies with their values and beliefs, which are cor- related with their cultural heritage.

Moreover, consumer attitudes toward global brands, perceptions of their quality, and purchase likelihood form the foundation of much of the global brand litera- ture (Özsomer and Altaras 2008; Steenkamp, Batra,

and Alden 2003). Recent international marketing stud- ies have focused on perceptions of global versus local brands. Consumer attitudes toward global and local products (Steenkamp and De Jong 2010), brand exten- sions of global or local origin (Iversen and Hem 2011), and perceived quality and global brand purchase likeli- hood (Özsomer 2012) all underlie attempts to under- stand the differences between global and local brands. Psychological mechanisms that create differences in attitudes toward global brands from developed versus developing countries have also been of interest to researchers (Alden et al. 2013; Guo 2013; Swoboda, Pennemann, and Taube 2012). However, market-based performance metrics are missing from the global brand- ing literature. Although the aforementioned studies related to global brand perceptions make an important contribution, they all employ perceptual data derived either from controlled experimental designs or from survey-based studies. The current study, in contrast, considers actual strategic brand deployment through the GBA, rather than consumer perceptions of deploy- ment. This is important because creating tangible mar- ket performance is among marketing’s most fundamen- tal contributions to the firm. Therefore, our research fills a gap in the literature, in that we present actual market-based data to assess the performance implica- tions of the various geographic range options, as mod- erated by national culture. Specifically, we provide mar- ket share by brand as a means to evaluate global brand performance because market share and its derivatives are the most salient measures of market performance in all marketing literature streams. In addition, as Steenkamp (2014) notes, market share is a valued out- come for global brands and is a topic Chabowski, Samiee, and Hult (2013) suggest as an important foun- dation for further research.

Overall, our study makes several contributions to the lit- erature. We expand the research on local versus non - local (Batra et al. 2000; Zhou, Yang, and Hui 2010) and local versus global (e.g., Steenkamp and De Jong 2010) dichotomies and demonstrate that brand multinational- ity is a hierarchical continuum. We demonstrate the actual market-based performance effects of the GBA and illustrate that they depend on, along with many other factors, a country’s cultural milieu.

The remainder of this article is structured as follows. First, we present a framework based on a conceptualiza- tion of global marketing strategies and actions (signals), which lead to market-based performance outcomes for brands (sales). Next, we develop hypotheses and present a longitudinal econometric model to test them. Our

Global Brand Performance 57

model is fit with a data set that has extensive temporal and geographical coverage and is among the most com- prehensive in the global brand strategy literature. Specifically, we develop a longitudinal model utilizing brand-level data for 165 automotive brands (e.g., Chevrolet, Mini) owned by 96 companies (e.g., General Motors, BMW) from 18 countries, operating in 65 mar- kets from 2002 to 2008. We present the results of our analysis and conclude with a discussion of managerial implications and limitations of the research.

LITERATURE REVIEW AND CONCEPTUAL FRAMEWORK

Global firms often own many products and brands that can vary in scope across and between national markets. To manage these brands and products as portfolios, firms are increasingly implementing unam- biguous international brand architectures in which ranges and geographic scope are employed as strategic means of facilitating both brand consistency and dif- ferentiation across international markets (Douglas, Craig, and Nijssen 2001). This phenomenon seems to be in response to the emergence of market segments that transcend national boundaries (Hofstede, Wedel, and Steenkamp 2002). That is, for a firm to be success- ful in all the markets it serves, it may choose to vary the mix of brands available in some countries and offer only a certain number of brands in each of its markets. For example, Toyota has managed its namesake brand globally but has offered brands such as Lexus and Scion as primarily domestic brands (ironically, in the United States rather than Japan, its country of origin) for a time before incrementally expanding into other markets.

In line with the notion of managing brands geographi- cally, GBA has four basic strategies: global, multi - regional, regional, and domestic (Townsend, Yeniyurt, and Talay 2009). This is reflected in the strategic orien- tations taken by companies operating in the variety of contexts in the global environment. While progressive in nature, there are differences in characteristics and tactics that enable firms to achieve these positions.

Global brands are present in all major market regions of the world and employ an integrated approach to stan- dardization across markets. Multiregional brands may be present in several markets, across several continents, but they do not have a centralized or standardized mar- keting program across geographic markets; they are not present in all triad markets (i.e., Asia, Europe, North

America). Regional brands are offered in multiple coun- tries in one geographic region (Rugman and Collinson 2004). Morrison, Ricks, and Roth (1991) suggest that a multiregional approach may actually provide the best strategic balance between global and domestic brands and optimize performance. In a study of global automo- tive manufacturers, Schlie and Yip (2000) show that most were following a regional strategy, with a few moving toward a global orientation. Douglas and Craig (2011) argue that firms should develop alternatives to global strategies, such as semiglobal marketing strate- gies, to maximize their performance.

Single-market brands are those sold in an individual national market. Even in the age of globalization, single- market brands still have a place in a firm’s GBA. Although a single-market brand serves only one national market, studies such as Kapferer (2002) suggest there are means for these brands (usually but not always domestic) to compete through local knowledge and flexibility. Appealing to patriotism and ethnocentrism are strategies these types of firms undertake (Klein 2002). Although there is reason to believe that brands with more narrow geographic approaches may provide attractive options for consumers, mounting evidence seems to support the notion that global brands act as a quality signal.

These positions in a GBA provide different signals in the markets in which they do business. For example, global brands have been established as signals of quality (Steenkamp, Batra, and Alden 2003). Cues such as these are used as a means to mitigate the effects of uncer- tainty. Consumers may not know about the quality of a product or brand, and they will use available informa- tion in their process of evaluation. Consumer product evaluation cues are either extrinsic (i.e., the cue is not physically part of the product) or intrinsic (i.e., the cue is a core product attribute) (Richardson, Dick, and Jain 1994). “Globalness” of a brand is therefore an extrinsic feature that consumers interpret in their choice process and is a reflection of the brand’s position in a GBA.

The GBA management process is conceptualized as the practices employed in the implementation and monitor- ing of global brand strategies for the brands in a firm’s portfolio. The scope of brands in the GBA is thus a mea- sure of the breadth of a brand’s global strategy. Alden, Steenkamp, and Batra (2006) show that attitudes toward consumption options are clustered along a con- tinuum of local–hybrid–global orientations, and they suggest that there are market-driven arguments to be made for the role of alternative geographically based

58 Journal of International Marketing

strategies. Research has indicated that global dispersion and geographic scope, coupled with local market knowledge, facilitate the launch of brands globally (Yeniyurt, Townsend, and Talay 2007). Furthermore, GBA is an important strategic consideration of a brand’s position and stage of internationalization (Townsend, Yeniyurt, and Talay 2009). Steenkamp and De Jong (2010) suggest that, because of the variety of attitudes toward local and global products, it may be best for firms to vary their branding and product portfolio strategies across markets.

Our framework (Figure 1) provides a means to under- stand how culture moderates the relationship between GBA and market-based performance. Fundamentally, a global company’s marketing program activities com-

prise the strategy, structure, and process undertaken by the organization. These activities occur in a dynamic environment, in which situational context can establish boundary conditions (e.g., culture) that determine a var- iation in relationships between brand strategies and market-based outcomes. From a distribution perspec- tive, a GBA is driven by the external environment, which is broad and varied across geographic and cul- tural boundaries (Douglas, Craig, and Nijssen 2001). The central tenet of the framework we employ is the contextual premise that market-based performance returns on global brand marketing strategies will vary on the basis of boundary conditions presented by extrin- sic cues in the market environment. Essentially, in what market contexts are brands at different positions in a GBA likely to perform better?

Several extrinsic brand cues potentially affect the rela- tionship between a brand’s position in GBA and its mar- ket performance. In this research, we investigate the role of cultural concepts and how they alter the effects of global brand marketing strategies on market-based per- formance. The most dominant cultural paradigm uti- lized when analyzing and assessing culture is that of Hofstede (1983), which forms the basis for a significant proportion of the cross-cultural studies undertaken in the literature. In light of the prominence of Hofstede’s cultural dimensions, we use them in our analyses of cul- ture’s role in the relationship between GBA and perfor- mance. We believe that these dimensions will also mod- erate the relationship between global brand marketing strategies and market-based performance. The develop- ment of the hypotheses that follow is based on the con- ceptualization of GBA as a progressive set of categoriza- tions ranging from single country to global.

Power Distance

Hofstede, Hofstede, and Minkov (2010, p. 61) define power distance as “the extent to which the less power- ful members of institutions and organizations within a country expect and accept that power is distributed unequally.” Low-power-distance cultures tend to be egalitarian and attribute less importance to differences in prestige, wealth, and status in their interpersonal relationships. In contrast, high-power-distance cultures emphasize prestige, wealth, and authority as crucial factors in forming social classes as well as shaping the relationships between them. Attaining and maintaining prestige in such societies are important sources of per- sonal satisfaction. People living in high-power-distance cultures are sensitive to social norms and tend to exhibit conformity to the norms of the classes with

Figure 1. Conceptual Framework

GBA

Global brands

Multiregional brands

Regional brands

Cultural Dimensions

Power distance

Individualism

Masculinity

Uncertainty avoidance

Market

Performance

Control Variables

Country of origin

Market commitment

Market size

Luxury

Local brand

GDP per capita

GDP growth rate

Population

Human development

Global Brand Performance 59

which they are affiliated as well as those to which they aspire. Therefore, relative to the other categories of geographic scope in a GBA, brands higher up in the GBA should be more important signals in high-power- distance cultures because they might have a stronger influence on increasing perceived quality and decreas- ing perceived risk. Those brands could more strongly symbolize power, prestige, wealth, and status (i.e., val- ues that are more emphasized in high-power-distance cultures), and such societies exhibit stronger motiva- tions to follow and imitate their aspirational social classes, which are more likely to consume global brands. Therefore,

H1: Brands with a higher position in a GBA (i.e., with a broader geographic scope) exhibit bet- ter (worse) market-based performance in countries with a higher (lower) level of power distance.

Individualism

The relative degree of individualism/collectivism exhib- ited by a national culture is believed to act as a moder- ator to the relationship between global brand marketing strategies and market-based performance. People from individualist cultures tend to view themselves as liber- ated and distinct from others in their society. Primary importance is given to the well-being of their immediate family and themselves rather than to society as a whole. In contrast, people from collectivist cultures feel that they belong to a group and will be more likely to allow the needs of the group to come before their own indi- vidual needs.

Two important distinctions between individualist and collectivist cultures make this dimension highly relevant for our study: patriotism and consumer ethnocentrism. Patriotism refers to love for and a sense of pride in one’s own country, a sacrificial devotion to it, respect and loy- alty to its people, and protection of it against outsiders (Balabanis et al. 2001). Patriotic consumers regard pro- tecting their country’s economic interests and support- ing domestic producers as their duty and, thus, show high intentions of buying domestic products and low intentions of buying foreign products.

Consumer ethnocentrism, in contrast, refers to “the beliefs held by consumers about the appropriateness, indeed morality, of purchasing foreign-made products” (Shimp and Sharma 1987, p. 280). Research has shown that consumer ethnocentrism predicts, albeit

with varying precision among product categories, con- sumers’ preferences to favor and purchase domestic products over their foreign counterparts (Balabanis et al. 2001). Ethnocentric tendencies have been found to be better predictors of purchase of domestic versus for- eign products than demographic and marketing-mix variables (Herche 1994). The literature has also sug- gested that collectivist cultures exhibit significantly higher levels of patriotism and consumer ethnocen- trism than individual ist cultures, and the people in such cultures are more likely to subordinate their per- sonal interests for the country’s welfare (Hofstede, Hofstede, and Minkov 2010). Therefore, consumers from individualist cultures would be more receptive to brands with higher levels of the GBA, whereas those from collectivist cultures would be more receptive to brands from their home countries because this would support the group to which they belong. We hypothe- size the following:

H2: Brands with a higher position in the GBA (i.e., with a broader geographic scope) exhibit bet- ter (worse) market-based performance in countries with a higher level of individualism (collectivism).

Masculinity

Masculinity is among the cultural dimensions that should make a difference as to the type of brand that would be successful in a market. Masculine cultural val- ues suggest assertiveness, achievement, and acquisition of wealth as more important in a society (Hofstede, Hofstede, and Minkov 2010). In masculine cultures, successes are more important than caring for others or improving the general quality of life for everyone in the society (Hofstede 1983). People may demonstrate achievement by having the latest and most prestigious possessions, which is essentially a proxy for success and reflects a given level of status. Thus, status purchases and conspicuous consumption are more prevalent in masculine cultures. Consumers in masculine cultures tend to buy more expensive watches and real jewelry, fly business class on leisure trips more frequently, and, most notably for this study, are more likely to favor foreign products and brands over their domestic counterparts (Hofstede, Hofstede, and Minkov 2010). Indeed, one rather ubiquitous way of demonstrating success and achievement in highly masculine cultures is to purchase global brands (De Mooij and Hofstede 2011) because they signal power, prestige, wealth, and status. There- fore, we expect the following:

60 Journal of International Marketing

H3: Brands with a higher position in a GBA (i.e., with a broader geographic scope) exhibit better (worse) market-based performance in countries with a higher level of masculinity (femininity).

Uncertainty Avoidance

Uncertainty avoidance captures “the extent to which people feel threatened by ambiguous situations and have created beliefs and institutions that try to avoid these” (Hofstede, Hofstede, and Minkov 2010, p. 418). This dimension refers to the endeavors of certain cultures to increase stability and predictability and to eschew ambi- guity. Because of our focus on the variance in the inter- pretation of different types of brands in different cul- tural milieus, it is the most relevant cultural dimension for this study. Consumers in high-uncertainty-avoidance cultures are more risk averse and less tolerant of ambi- guity. They tend to reduce aversion and ambiguity by seeking and favoring credible signals.

Moreover, consumers in these markets tend to utilize mar- keting information more frequently and intensely because they are less sensitive to search costs and more willing to collect and process information than consumers in low- uncertainty-avoidance cultures. In this study, we propose that consumers from high-uncertainty-avoidance cultures will place more emphasis on brands as cues of product quality and process information about them more intensely than consumers from low-uncertainty-avoidance cultures. This is because information search and process- ing in the purchasing process should positively correlate to a culture’s risk-aversion level (Dawar and Parker 1994). As a brand’s position increases in the GBA, it will signal superior quality, higher customer demand, and proven success in various parts of the world, all of which decrease the perceived risk. Stated differently, we posit that the effects of global brands will be stronger in high- uncertainty-avoidance cultures because these cultures are more sensitive to ambiguity.

H4: Brands with a higher position in a GBA (i.e., with a broader geographic scope) exhibit bet- ter (worse) market-based performance in countries with a higher (lower) level of uncer- tainty avoidance.

METHODS Sample Description

We investigate the relationship between a brand’s global marketing strategy and its market-based performance in

the context of the global automotive industry for the period 2002–2008 using a data set of 165 brands (e.g., Chevrolet, Mini) owned by 96 companies (e.g., General Motors, BMW) from 18 countries, operating in 65 countries. This data set not only enables us to cover approximately 90% of the global automotive industry but also represents the most extensive temporal and spa- tial coverage in the literature to date (Figure 2).

The automotive industry is relevant for this research because of its economic and strategic importance, along with its geographic scope. It has been experiencing robust growth due to increasing global demand, particularly from emerging markets; developed markets have been relatively stagnant and, therefore, more competitive. Vehicle production has more than doubled since 1975, from 33 million to 73 million in 2007. Moreover, the automotive industry is characterized by broad-based research and development and extensive supply chain management, with many marketing operations conducted on a global scale (Talay, Dalgic, and Dalgic 2010).

The automotive industry is also rife with highly dynamic market competition, with continual launches of new products into existing markets and new brands and products into new markets. The emergence of developing countries as major players in the industry—as consumers on the one hand and manufacturers and suppliers on the other—as well as a convergence in demand characteris- tics has created an even greater need to embrace the busi- ness concepts associated with globalization. Further- more, automobiles are highly image-conscious products that typically include a substantial level of purchase involvement and are relatively costly to dispose of after purchase. We believe, therefore, that this highly turbu- lent and global industry is a suitable context for analyz- ing the performance of global brands.

Dependent Variable

This study examines the impact of cultural milieu on the link between the GBA and market performance. Our dependent variable is the market share of brand i in country j in year t as proxy of market performance. This performance criterion is widely used in the international marketing literature (Guo 2013; Iversen and Hem 2011; Swoboda, Pennemann, and Taube 2012), thus allowing for comparisons between our study and other extant works. Furthermore, both scholars and practitioners consider this variable an important performance indica- tor (Farris et al. 2006). However, market share of a brand in a given country may not reflect performance in terms of profit, return on investment, or shareholder

Global Brand Performance 61

value. Market share figures for this study are extracted from a data set containing 165 brands (e.g., Chevrolet, Mini) from 18 countries, owned by 96 companies (e.g., General Motors, BMW), operating in 65 countries from 2002 to 2008. This proprietary data set was provided by CSM Worldwide, which specializes in providing fore- casting and market intelligence solutions to manufactur- ers, suppliers, and financial organizations in the global automotive industry. To control for the effects of skew- ness in distribution and outliers in the data, we used the natural logarithm of this variable in our analyses (e.g., Talay, Calantone, and Voorhees 2014).

Independent Variables

GBA. We use three dummy variables to denote global, multiregional, and regional brands. Following Rugman and Collinson (2004), who define a global brand as hav- ing at least 20% of its sales in each of the three regions of the broad “triad” of the European Union, North Amer- ica, and Asia, we operationalize the position of a brand in the GBA by its presence in three continents: Asia, Europe, and North America. To be more precise, we identify a brand as global if it has operations in all three of these continents (e.g., Ford, Honda, Mercedes-Benz). If a brand

is present in multiple countries and continents but is not a global brand, it is identified as multiregional (e.g., Seat, Pontiac, Volga). Regional brands refer to brands that have operations in only one continent, albeit in multiple countries within the same continent (e.g., Dacia, Holden, Morgan). Finally, if a brand is available in only one coun- try, it is identified as a single-country brand, which we use as the base case in the analyses.

Cultural Dimensions. To capture the effects of culture, we utilize Hofstede, Hofstede, and Minkov’s (2010) scores for the cultural dimensions of power distance, individualism, masculinity, and uncertainty avoidance. Using these scores individually rather than as a compos- ite enables us to gain deeper insights about the varying effects of each cultural dimension on the relationship between a brand’s locus in the GBA and its market per- formance.

Control Variables

While our conceptual and empirical models include a set of factors that are plausibly related to the market share of a brand i in country j in year t, we acknowledge that our data set still has limitations. Indeed, there may be other

Figure 2. Markets Included in This Study

62 Journal of International Marketing

factors (e.g., marketing spending, distribution, financing options) influencing the brand performances that we are not able to observe because of the geographical breadth and temporal depth of our data set. This raises the con- cern that there may be unobserved heterogeneity in our data because the factors that are not accounted for in our model are correlated with those that are accounted for, and this may lead to a bias in the estimated effects. There- fore, we include a series of control variables, which may affect the market performance of brand. Specifically, we account for the effects of a brand’s commitment to a country, the level of development of a market along with its actual and potential size, whether a brand is a luxury brand, and whether a brand is a local brand.

We operationalize market commitment as follows: the models of brand i sold in country j in year t as a percent- age of the entire set of models brand i offered in year t. For example, BMW’s global product-market commit- ment levels to the Malaysian and U.S. markets in 2005 was 81.8% and 90.9%, respectively, because it offered only 9 models in Malaysia and 10 models in the United States, of the 11 total BMW models available in 2005.

We also distinguish luxury brands and nonluxury brands because they may be subject to different demand char- acteristics (Dubois, Czellar, and Laurent 2005). The lit- erature has shown that being “local” in a market can change demand for a product both for better and for worse (Batra et al. 2000). Therefore, in an attempt to control for the effects of being a local brand, we also include a dummy variable indicating whether a brand is sold in its home country at year t.

We also include a brand’s national origin as a control variable. Nationality is an important driver of brand identity management strategies, and even brands with a global approach are guided by the principles derived from the national heritage of the brand. Therefore, we include a variable for nationality in the model to control for the effects of the variance in global orientations and strategies observed by brands from different countries (Chryssochoidis, Krystallis, and Perreas 2007). We oper- ationalize country of origin using a set of seven dummy variables for China, France, Germany, Italy, Japan, South Korea, and the United States. Each dummy variable equals 1 if the home country of the brand is the country that dummy variable denotes, and 0 otherwise. Brands from other countries establish the base condition. Brands from these countries capture 138 of the 165 brands in our data set and make up 89.19% of all sales during the 2002–2008 period.

We capture the level of development in a country with three variables: gross domestic product (GDP) growth rate, GDP per capita based on purchasing power parity (PPP), and the Human Development Index. We measure the potential size of the market in country i using the country population at time t, whereas the actual market size is operationalized as the total number of vehicles sold in country i at time t. Similar to our dependent variable, we used the natural logarithm of market size to control for the effects of skewness in distribution and outliers in the data.

Model Development

We have compiled a panel data set of country-level annual sales composed of repeated observations of brands. To test our hypotheses, we therefore employ ran- dom-effects cross-sectional time-series models corrected for serial correlation of errors (Wooldridge 2002). Random-effects models rest on the assumption that ai and xit are uncorrelated (Wooldridge 2000, p. 449). The transformation of the data is conducted in such a way that the variance–covariance matrix of the composite error has a block diagonal pattern that requires general- ized least squares estimation. This allows for an assess- ment of the variation both within and between each group of observations. The random-effects estimation is more efficient than the alternatives because it uses a weighted average of both types of estimators when we assume zero correlation between the explanatory variables and the composite error (Kennedy 2003, p. 307). The model we test has the following structure:

(1) yit = xitb + (ai + eit),

where ai ~ (a, s2a) and eit ~ (0, s2u). Because the random- effects estimation method assumes that ai and xit are uncorrelated, it is necessary to assess whether this assumption is true. This is accomplished through com- parison of the estimates obtained from the fixed-effects model and the random-effects model (Hausman 1978). The Hausman test assesses the null hypothesis that the coefficients estimated by the efficient random-effects estimator are equivalent to the ones estimated by the consistent fixed-effects estimator. Performing the Haus- man test yields a c2 statistic of 29.3 (d.f. = 19, p = .061). The null hypothesis cannot be rejected, and the assump- tion that ai and xit are uncorrelated can be made, indi- cating that the random-effects model will not produce biased coefficients. Thus, both theoretically and method- ologically, random-effects estimation is an appropriate specification for this data set.

Global Brand Performance 63

We also checked for serial autocorrelation in errors as suggested by Wooldridge (2002) using the xtserial routine in Stata 13.0. Although several tests for serial autocorre- lation in panel data models have been proposed, this rela- tively new test requires fewer assumptions (Wooldridge 2002). The results of the Wooldridge test indicate an F(1, 2,924) = 164.559 with Prob > F = .0001. Therefore, the null hypothesis is strongly rejected, indicating that there is serial correlation in the composite error and implying that (1) pooled ordinary least squares estimation will be inef- ficient, and generalized least squares will be a better esti- mation technique, and (2) an auto regressive (e.g., AR[1]) disturbance term should be included in the model. Fur- thermore, the results obtained through the random- effects estimation indicate a first-order serial correlation (r) of .579. The modified Durbin–Watson test statistic = .659 and the Baltagi–Wu locally best invariant test = 1.127 also confirm the need to correct for serial autocor- relation. Therefore, we used the xtregar procedure in Stata 13.0 to allow for first-order autoregressive correla- tion among the disturbance terms.

Tables 1 and 2 present the descriptive statistics and pair- wise Pearson correlations for the key variables, respec- tively. A relatively higher mean value of a dummy variable, which can range between 0 and 1, indicates that our data set contains more observations of that variable. Market size (i.e., the total annual sales of all brands) varies significantly from 5,722 vehicles sold in a country (Macedonia in 2002) to approximately 17 mil- lion vehicles (the United States in 2005).

We tested for multicollinearity and found that both the average and the maximum variance inflation factor val- ues were less than 10, a commonly used cutoff value (Koutsoyiannis 1977), which we attribute to the wide spatial and temporal coverage in our data set. The highest average and maximum variance inflation factor values in the estimated model are 2.77 and 7.87, respectively.

Table 3 presents the results of our estimations. We find that the market shares of global brands are 217% [exp(1.153) – 1 ¥ 2.168] higher than those of brands operating in single country (b = 1.153, p < .01), while multiregional brands (b = .796, p < .01) and regional brands (b = .663, p < .01) have approximately 121% and 94% higher market shares than their single-country (baseline condition) counterparts, respectively. Overall, we observe that the brands with higher positions in the GBA tend to sell more than brands with lower positions.1

We found all of the estimated coefficients for the inter- action effects between cultural dimensions and GBA lev-

els to be statistically significant and in the hypothesized direction (i.e., positive), indicating support for our hypotheses. That is, for any cultural dimension, the magnitudes of the coefficients of the GBA levels are ranked as global > multiregional > regional. This sug- gests that being higher up in the GBA not only improves the positive effects of individualism and masculinity on market performance but also decreases the negative impacts of power distance and uncertainty avoidance. To further elucidate the moderating effects of cultural milieus on the link between GBA position and market performance, we conducted a spotlight analysis. This technique uses basic statistics to analyze the simple effect of one variable at a particular level of another variable, continuous or categorical (Spiller et al. 2013). The results of the spotlight analyses, which are consis- tent with our findings, appear in Appendix B.

Table 1. Descriptive Statistics

Variable M SD

Market share 5.161 1.938

Global brand .690 .462

Multiregional brand .229 .420

Regional brand .038 .191

Power distance 56.283 2.886

Individualism 5.300 22.587

Masculinity 48.241 22.371

Uncertainty avoidance 68.234 21.512

China .026 .159

France .070 .255

Germany .254 .435

Italy .089 .285

Japan .194 .395

South Korea .045 .207

United Kingdom .031 .172

United States .239 .426

Market commitment 43.843 28.772

Market size 1.1 ¥ 106 2.3 ¥ 106

Luxury brand .296 .456

Local brand .044 .205

GDP per capita (PPP) 21,783.690 14,334.580

GDP growth rate 4.194 3.115

Population 8.9 ¥ 107 2.5 ¥ 108

Human development .805 .098

64 Journal of International Marketing

Ta bl

e 2.

B iv ar ia te C or re la tio n M at rix o f V ar ia bl es

V ar

ia b le

1

2 3

4 5

6 7

8

9

1

0 1

1 1 2 1 3 1

4 1 5 1 6

1 7 1

8 1 9 2 0 2

1 2 2 2 3

1 . M ar k et s h ar e 1

2 . G lo b al b ra n d . 1 0 1

3 . M u lt ir eg io n al b ra n d .0 4 –. 4 1 1

4 . R eg io n al b ra n d .0 5 –. 3 1 –. 1 0 1

5 . P o w er d is ta n ce .0 4 –. 0 1 .0 0 –. 0 5 1

6 . In d iv id u al is m – .0 6 .0 0 .0 1 .0 8 –. 3 8 1

7 . M as cu li n it y . 0 7 –. 0 2 –. 0 1 .0 0 .1 9 – .0 2 1

8 . U n ce rt ai n ty a vo id an ce – .1 2 .0 2 .0 3 .0 0 .3 8 – .3 5 .0 9 1

9 . C h in a –. 0 8 –. 2 3 .1 4 –. 0 1 .1 4 – .1 9 .0 6 –. 0 7 1

1 0 . F ra n ce . 1 7 –. 0 7 .0 5 .0 4 .0 2 – .0 3 –. 0 1 .0 4 –. 0 4 1

1 1 . G er m an y . 0 5 .1 7 –. 0 9 –. 1 1 –. 0 2 . 0 2 .0 1 .0 3 –. 0 9 –. 1 6 1

1 2 . It al y . 0 6 –. 1 4 .1 6 .0 5 –. 0 3 . 0 4 –. 0 1 .0 4 –. 0 5 –. 0 9 –. 1 9 1

1 3 . Ja p an . 0 9 .2 0 –. 1 6 –. 1 0 .0 2 – .0 2 .0 0 –. 0 1 –. 0 7 –. 1 3 –. 2 9 –. 1 5 1

1 4 . So u th K o re a .0 7 .1 4 –. 1 1 –. 0 4 .0 2 – .0 4 –. 0 1 .0 0 –. 0 3 –. 0 6 –. 1 3 –. 0 7 – .1 1 1

1 5 . U n it ed K in gd o m .0 0 –. 2 9 .1 0 .4 5 –. 0 8 . 1 0 .0 0 –. 0 2 –. 0 3 –. 0 5 –. 1 1 –. 0 6 – .0 9 –. 0 4 1

1 6 . U n it ed S ta te s – .0 2 .0 4 .0 3 –. 0 6 –. 0 4 . 0 4 –. 0 2 –. 0 2 –. 0 8 –. 1 6 –. 3 3 –. 1 8 – .2 7 –. 1 2 – .1 0 1

1 7 . M ar k et c o m m it m en t . 0 5 –. 2 4 .1 0 .1 5 –. 0 9 . 1 7 .0 6 –. 0 3 .0 6 .0 1 .2 4 .1 1 – .3 6 –. 1 4 . 0 8 – .0 1 1

1 8 . M ar k et s iz e . 0 2 –. 0 1 –. 0 3 .0 2 –. 0 4 . 2 2 .2 6 –. 1 8 .0 8 –. 0 5 .0 0 –. 0 3 – .0 1 –. 0 2 – .0 1 . 0 3 . 1 4 1

1 9 . L u x u ry b ra n d –. 1 8 .2 4 –. 2 2 .0 1 –. 0 9 . 1 0 –. 0 1 .0 0 –. 1 0 –. 1 8 .2 0 .0 3 – .1 8 –. 1 4 . 0 8 . 1 5 . 3 4 . 0 5 1

2 0 . L o ca l b ra n d . 1 1 –. 1 8 .0 1 .0 6 .0 5 – .0 1 .1 4 –. 1 6 .2 7 –. 0 3 –. 0 6 –. 0 2 – .0 7 –. 0 3 . 0 4 – .0 5 . 2 8 . 3 7 –. 0 5 1

2 1 . G D P p er c ap it a (P P P ) –. 0 6 .0 0 .0 1 .0 9 –. 3 6 . 4 3 –. 0 7 –. 1 7 –. 1 5 –. 0 4 .0 3 .0 4 – .0 3 –. 0 4 . 0 9 . 0 4 . 1 7 . 1 5 .1 2 – .0 3 1

2 2 . G D P g ro w th r at e .0 1 –. 0 3 .0 0 –. 0 6 .3 8 – .3 9 –. 0 2 –. 0 8 .1 9 .0 1 –. 0 3 –. 0 3 . 0 0 .0 2 – .0 6 – .0 3 – .0 7 – .0 5 –. 0 8 . 0 7 –. 4 5 1

2 3 . P o p u la ti o n . 0 3 –. 0 4 –. 0 2 –. 0 2 .3 8 – .2 0 .3 9 –. 0 7 .1 8 –. 0 3 –. 0 2 –. 0 5 . 0 0 .0 0 – .0 3 – .0 2 . 0 4 . 3 5 –. 0 2 . 3 2 –. 4 1 . 1 3 1

2 4 . H u m an d ev el o p m en t – .0 8 .0 0 .0 1 .0 9 –. 4 1 . 3 5 –. 1 3 –. 0 2 –. 1 8 –. 0 2 .0 4 .0 6 – .0 5 –. 0 4 . 0 9 . 0 5 . 2 1 . 1 3 .1 2 – .0 5 .4 0 – .4 6 – .4 5

N o te s: A ll c o rr el at io n s si gn if ic an t at p < . 0 5 e x ce p t th o se i n b o ld .

Global Brand Performance 65

Table 3. GBA, Culture, and Market Performance: Estimation Results

Variables Coefficient SE Relative Effect Size

Main Effects

Global marketing strategy

Global brand 1.153*** .155 .038

Multiregional brand .796*** .128 .035

Regional brand .663*** .097 .036

Culture

Power distance –.128*** .017 .038

Individualism .036** .016 .017

Masculinity .134*** .021 .035

Uncertainty avoidance –.070*** .012 .034

Country of Origin

China .680** .268 .018

France .405*** .072 .033

Germany .239*** .058 .027

Italy .159** .066 .018

Japan .257*** .056 .029

South Korea .352*** .079 .028

United Kingdom –.015 .269 .000

United States .207*** .059 .024

Control Variables

Market commitment .146*** .018 .039

Market size .176*** .046 .026

Luxury brand –.195*** .019 .042

Local brand .229*** .043 .032

GDP per capita (PPP) .092** .045 .015

GDP growth rate .014 .021 .005

Population .529*** .142 .025

Human development .109** .051 .016

Year Dummies

2003 –.020* .013 .012

2004 –.014 .019 .006

2005 –.029 .024 .010

2006 –.022 .029 .006

2007 .030 .034 .007

2008 .091** .038 .018

Interaction Effects

Power distance ¥ Global brand .208*** .053 .026

Power distance ¥ Multiregional brand .174*** .042 .027

Power distance ¥ Regional brand .137** .055 .018

Individualism ¥ Global brand .216*** .044 .030

66 Journal of International Marketing

We also tested, for each cultural dimension, whether the coefficient of an interaction effect is significantly different from the other two interactions. We found that for all four cultural dimensions, the coefficients of the interactions for the global brands are statistically higher than those for the regional brands, indicating that brands with a higher posi- tion in the GBA have higher market shares. However, the coefficients for the interactions of multiregional brands are not statistically different from those of global brands for individualism and uncertainty avoidance at the 95% level.2 Therefore, we conclude that whereas H1 and H3 are fully supported, H2 and H4 are only partially supported.

As for the control variables, our results show that a brand’s commitment (b = .146, p < .01), market size (b = .176, p < .01), localness (b = .229, p < .01), GDP per capita, population (b = .529, p < .01), and human development index (b = .109, p < .05) have significant and positive effects on market share. Our results show that, ceteris paribus, luxury brands have lower market shares than their nonluxury counterparts (b = –.195, p < .01), as we expected. However, GDP growth rate does not seem to have a significant impact on market share (b = .014, p > .10).

We also present the relative effect size of each variable in our estimations. This helps us go beyond the statisti- cal tests of significance, which indicate the likelihood that the impact of a covariate differs from random expectations and enable us to understand the relative magnitudes of the variables in the analyses (Gielens and Steenkamp 2007). An examination of the relative effect sizes of the interactions between the cultural dimensions and the GBA levels shows that the uncertainty avoid- ance has a strong moderating effect on the link between GBA levels and market share.

DISCUSSION

This study expands understanding of how brands per- form in global markets and has some important impli- cations for managing global brand and product portfo- lios. We find support for the notion that creating a global brand marketing strategy that includes incorpo- ration of global brands is an effective way to improve market-based performance. These findings have both academic and practical relevance and help improve global brand managers’ knowledge.

Individualism ¥ Multiregional brand .183*** .034 .032 Individualism ¥ Regional brand .125** .050 .018 Masculinity ¥ Global brand .177*** .034 .031 Masculinity ¥ Multiregional brand .133*** .048 .020 Masculinity ¥ Regional brand .123*** .032 .026 Uncertainty avoidance ¥ Global brand .209*** .032 .036 Uncertainty avoidance ¥ Multiregional brand .184*** .034 .032 Uncertainty avoidance ¥ Regional brand .147*** .022 .036

Number of observations 20,314 Number of groups 2,942 c2 3,642.38 Prob > c2 .000 R2 within .0950 R2 between .3991 R2 overall .3831 su 2.230 se .773

*p < .10. **p < .05. ***p < .01.

Table 3. Continued

Variables Coefficient SE Relative Effect Size

Global Brand Performance 67

Our findings suggest that a global brand marketing strategy affects a brand’s performance in the global mar- ketplace, in that brands that are higher in the GBA hierar- chy tend to perform better in individual country markets. The depth and breadth of the data set we employ gives cause to believe that this is an enduring phenomenon, and it supports the notion that there is a benefit to global branding programs. Although several studies have found support for consumer-based constructs such as brand credibility and purchase likelihood, this is the first study to employ actual market-based data to support the perfor- mance implications of the position of a brand in a GBA.

Theoretical Implications

Signaling is a sparsely understood concept in the inter- national marketing literature. However, it is known that signals are a powerful driver of consumer perceptions and responses, influencing market outcomes. Our study contributes to the understanding of the role of different signals sent by a brand’s position in the GBA. Using sig- naling theory, we developed four hypotheses regarding the boundary conditions established by cultural factors that cause performance variations in alternative strate- gic GBA approaches in global markets. This study also provides important empirical evidence that some cul- tural values alter how global brand market strategies perform in a given market. All of the cultural dimen- sions we tested exhibit significant effects for the direct effects of culture on market performance.

This study fills a lacuna at the nexus of three separate, but related, research streams. We contribute to the understanding of (1) the role of local versus global brands, (2) the moderating role of culture on the link between marketing signals and firm performance (e.g., Akdeniz and Talay 2013; Alden et al. 2013; Erdem, Swait, and Valenzuela 2006; Özsomer 2012), and (3) the evolution of global brands (e.g., Douglas, Craig, and Nijssen 2001; Özsomer and Altaras 2008).

Our results advance the literature on global brands by extending the findings of previous studies based on local versus nonlocal (e.g., Batra et al. 2000; Zhou, Yang, and Hui 2010) and local versus global (Steenkamp and De Jong 2010) dichotomies. Our study demonstrates that, rather than a dichotomy, brand multinationality can be better conceptualized as a hierarchical continuum, the performance effects of which depend on, along with many other factors, the cultural fabric of a country. Specifically, we find that the effects of power distance and masculinity on the link between multiregional brands and market share are different than their effects

on the performance implications of global and regional brands. This suggests that although cultures regard global brands differently from their nonglobal (e.g., regional, local) counterparts—thus corroborating the findings of previous studies—the regional and multire- gional brands might differ in terms of their performance outcomes from both local and global brands. This fur- ther indicates a multiple-category conceptualization (i.e., local, regional, multiregional, and global) might advance our knowledge in this field. In general, we find that all of the cultural dimensions amplify the impact of the brand multinationality (i.e., global, multiregional, and regional) signal on market performance. This implies that cultural values moderate the signals brands send.

In addition, we provide further support for the theoretical foundations of the evolution of global brands. Although there are advocates for alternative globalization processes, it is clear from the accumulated body of evidence that there is support for incremental movement from local to regional to global brands. Our supporting analysis related to temporal precedence (Table A1) enables us to incorpo- rate the evolution of a brand within the GBA. Our find- ings indicate progressive positions through the GBA that engender higher levels of market share, in support of pre- vious research in this area. This is an important contribu- tion to the literature because it helps researchers better understand the nature of this phenomenon and build foundations for further research. Moreover, whereas con- sumer attitudes and understanding of psychological mechanisms have provided the basis for the extant litera- ture on global brands thus far (e.g., Alden et al. 2013; Guo 2013; Iversen and Hem 2011; Özsomer 2012; Swoboda, Pennemann, and Taube 2012), we employ actual market- based performance measures in our study. Our strategic findings are largely consistent with the findings from psy- chology-based work, which indicate the robustness of the link between global branding strategies and their perfor- mance outcomes, both perceptual and market-based.

Managerial Implications

The results of this study, in which the position in a GBA is positively associated with improved market-based performance, support the argument that global brands are a viable strategic option for brand managers. We believe that the findings from this study can help man- agers. For example, if a brand is not participating inter- nationally, our research suggests that international par- ticipation is worth pursuing. Furthermore, managers of regional and multiregional brands may want to consider increasing their global positions, depending on the cul- tural dimensions of the potential markets for entry.

68 Journal of International Marketing

Indeed, the interactions of cultural dimensions with the GBA are potentially managerially useful for strategic planning purposes. Brands with a higher position in a GBA achieve better market-based performance in coun- tries with higher scores on Hofstede’s cultural dimen- sions. This suggests that brands might consider concen- trating their globalization efforts in countries that exhibit these values. One possible approach to doing so is to overlay the geographic maps developed by Hof- stede with the brand’s geographic footprint to identify potential markets to enter or expand operations. It is well established in the literature that cultures that rank higher in power distance place more emphasis on status symbols, whereas those with higher uncertainty avoid- ance scores tend to avoid ambiguity. In such cultures, companies can benefit from entering the market with their global brands and emphasizing their “globalness” in their marketing communications.

Findings from this study may also be useful for man- agers as they formulate their global portfolio strategies. Varying the position of a brand to best match its other attributes to the preferences of the cultural norms across geographic ranges may further enhance performance. Our findings provide a foundation for establishing a firm’s portfolio efficiently and in such a way as to maxi- mize the performance of all the brands systematically and collectively. That is, employing the tenets of a GBA in conjunction with culture may enable a firm to reap greater benefits than considering brands individually.

We believe that our control variables warrant some dis- cussion because they not only validate the results of the previous studies but also reveal some unexpected find- ings. For example, considering the amount of invest- ment required for the brand and the relative magnitude of the purchase for the consumer, it would not be rea- sonable to assume that automotive sales are independ- ent of factors such as GDP per capita and market size. We also find that local brands tend to have higher mar- ket shares than their foreign competitors. Although this may suggest a strong liability of foreignness effect, the source of this advantage may be due to regulatory or infrastructure issues more so than to global brand mar- keting strategies.

Luxury brands, by definition, are niche players or “spe- cialists” focusing on a very narrow segment of the mar- ket; by design, they are intended to get a smaller share of a market at higher prices. Compared with generalist brands, which produce for everyone in the market, lux- ury brands are intended to be exclusive. Overall, the

findings from this study extend our knowledge of how brands in a GBA perform and provide a useful set of premises on which to build further research.

LIMITATIONS AND DIRECTIONS FOR FURTHER RESEARCH

This study is limited in its generalizability because we conducted it in a single industry related to a specific, and relatively unique, durable consumer product. It may be useful to examine whether the findings of this study are applicable—and, if so, to what extent—to other types of products and services. For example, future stud- ies could replicate this study with low-involvement products such as fast-moving consumer goods. Alterna- tively, researchers might analyze GBA in the context of products of a more complex nature or credence prod- ucts, which would improve the external validity and generalizations of the results.

In addition, it would be worthwhile to add consumer- based perceptual measures. Attitudes toward global brands have been one of the most researched areas in the recent literature, and a logical next step would be to inte- grate the perceptual with the actual. This would allow for discernment of brand positioning in a global environ- ment and might provide mediation between brand strate- gies and market-based performance. The implications of our study suggest that image strategies might be a media- tor between a brand’s global architecture and market- based performance. Measuring the reception of signals that could convey notions such as actual or aspirational social status might also have value and would build on existing consumer behavior literature. Further research should incorporate consumers’ perceptions of a brand’s image into a model of actual market-based performance to test for these mediation effects.

A possible extension of this study could include the analysis of the effects of the interactions between cul- tural and macroeconomic conditions on the market per- formances of global brands. For example, the argument has been made in the literature that people in contempo- rary high-context cultures, especially in less affluent countries, are influenced by “dominant” consumers from low-context cultures (which are typically devel- oped countries). Drawn to the “good life” promised by the dominant hegemony (Üstüner and Holt 2007), high- context people are more likely to embrace low-context cultural values than the other way around, possibly cov- eting low-context symbols (e.g., [car] brands).

Global Brand Performance 69

In this study, we employ the most traditional market- ing measure of market-based performance. Although sales and derivatives such as market share have been widely explored in the literature, they have not been addressed with respect to global brand performance. Operational measures such as the costs associated with entering and managing country markets could be incorporated into future studies as alternative meas- ures of market performance.

Despite these limitations, we believe this study provides a noteworthy and relevant explanation of how a variety of strategic positions relate to market-based perfor- mance in a global environment. We hope that our find- ings will stimulate further research in this important area of study as managers continue to look for ways to develop and execute global brand and product strategies and manage portfolios in the global marketplace.

APPENDIX A: TEMPORAL PRECEDENCE OF GBA LEVELS OVER MARKET PERFORMANCE

The results presented in Table 3 already satisfy two of the three requirements of causality: concomitant varia- tion and elimination of other plausible alternatives. Concomitant variation is established by the statistically significant results for the hypothesized relationships between the GBA, cultural milieu, and performance. We

minimize the effects of other plausible causes using a set of 22 additional variables (i.e., country of origin, con- trol variables, and year dummies).

Estimating the effects-lagged independent variables on the dependent variable is a simple way to establish tem- poral precedence (i.e., the third requirement of cau- sality) because we have a panel data set. Therefore, we estimated the lagged effect of GBA (i.e., global, multire- gional, and regional) on market share and the lagged effect of market share on GBA using the same covariates in our main analyses (except for the interaction terms).

Specifically, we created a new variable, GBA level, which takes the values of 1, 2, 3, and 4 for local, regional, multiregional, and global brands, respectively. This variable helps us capture the evolution of a brand within the GBA. Then, we ran the following six models: (1) Models A1, A2, and A3, in which we estimated the one-, two-, and three-period lagged effects of the GBA level on market performance to test the GBA level’s tem- poral precedence on market performance, and (2) Mod- els B1, B2, and B3, in which we estimated the one-, two-, and three- period lagged effects of the GBA level’s mar- ket performance (i.e., the natural logarithm of sales of brand i in country j in year t) to test market perfor- mance’s temporal precedence on GBA level. The results suggest that advancement within the GBA drives sales, not the other way around (Table A1).

Table A1. Establishing Temporal Precedence of GBA Levels over Market Performance

A: Lagged Effects of GBA Level on Market Performance

Dependent Variable: Market Performance Coefficient SE

Model A1: One-period lag in GBA level .1701* .005

Model A2: Two-period lag in GBA level .1952* .006

Model A3: Three-period lag in GBA level .2526* .006

B: Lagged Effects of Market Performance on GBA Level

Dependent Variable: GBA Level Coefficient SE

Model B1: One-period market performance .0026 .021

Model B2: Two-period market performance .0028 .003

Model B3: Three-period market performance .0035 .003

*p < .01. Notes: We included the same culture, country-of-origin, and control variables in the main model in Table 3, along with the year dummies in our estimations, but we do not report them here for brevity.

70 Journal of International Marketing

Appendix B. Spotlight Analyses of the Differences Between the Effects of GBA Levels in Market Performance for a Systematic Selection of the Values of the Cultural Dimensions

Global Brands Multiregional Brands Regional Brands Local Brands

Dimension: Power Distance 10 7.702 7.008 6.896 5.156 40 7.740 7.046 6.933 5.194 70 7.778 7.084 6.971 5.231 100 7.815 7.121 7.009 5.269 Dimension: Individualism 10 6.574 6.462 6.268 4.722 40 6.923 6.810 6.617 5.071 70 7.272 7.159 6.966 5.419 100 7.620 7.508 7.314 5.768 Dimension: Masculinity 10 6.844 6.731 6.538 4.991 40 7.015 6.902 6.709 5.163 70 7.186 7.073 6.880 5.334 100 7.357 7.245 7.051 5.505 Dimension: Uncertainty Avoidance 10 7.001 6.888 6.695 5.149 40 7.034 6.921 6.728 5.181 70 7.066 6.954 6.760 5.214 100 7.099 6.987 6.793 5.247

Notes: All coefficients are significant at the .01 level.

Appendix C. Tests of Significance of the Differences Between the Interaction Coefficients

1 2 3

Dimension: Power Distance 1. Power distance ¥ Global brand N.A. 6.91** 7.23** 2. Power distance ¥ Multiregional brand N.A. 5.88** 3. Power distance ¥ Regional brand N.A.

Dimension: Individualism 1. Individualism ¥ Global brand N.A. 1.44 13.79** 2. Individualism ¥ Multiregional brand N.A. 11.64** 3. Individualism ¥ Regional brand N.A.

Dimension: Masculinity 1. Masculinity ¥ Global brand N.A. 8.14** 1.25** 2. Masculinity ¥ Multiregional brand N.A. 4.12* 3. Masculinity ¥ Regional brand N.A.

Dimension: Uncertainty Avoidance 1. Uncertainty avoidance ¥ Global brand N.A. .25 6.31** 2. Uncertainty avoidance ¥ Multiregional brand N.A. 5.47** 3. Uncertainty avoidance ¥ Regional brand N.A.

*p < 05. **p < .01. Notes: Significance is based on c2 tests. N.A. = not applicable

Global Brand Performance 71

NOTES

1. We empirically validated the temporal precedence of the GBA over sales. Appendix A presents the details of our analyses.

2. Appendix C reports the results of these tests.

REFERENCES Akdeniz, M. Billur and M. Berk Talay (2013), “Cultural Varia- tions in the Use of Marketing Signals: A Multilevel Analysis of the Motion Picture Industry,” Journal of the Academy of Mar- keting Science, 41 (5), 601–624.

Alden, Dana L., James B. Kelley, Petra Riefler, Julie A. Lee, and Geoffrey N. Soutar (2013), “The Effect of Global Company Animosity on Global Brand Attitudes in Emerging and Devel- oped Markets: Does Perceived Value Matter?” Journal of International Marketing, 21 (2), 17–38.

———, Jan-Benedict E.M. Steenkamp, and Rajeev Batra (2006), “Consumer Attitudes Toward Marketplace Globaliza- tion: Structure, Antecedents, and Consequences,” Inter- national Journal of Research in Marketing, 23 (3), 227–39.

Balabanis, George, Adamantios Diamantopoulos, Rene Den- tiste Mueller, and T.C. Melewar (2001), “The Impact of Nationalism, Patriotism and Internationalism on Consumer Ethnocentric Tendencies,” Journal of International Business Studies, 32 (1), 15–75.

Batra, Rajeev, Venkatram Ramaswamy, Dana L. Alden, Jan- Benedict E.M. Steenkamp, and S. Ramachander (2000), “Effects of Brand Local and Nonlocal Origin on Consumer Attitudes in Developing Countries,” Journal of Consumer Psychology, 9 (2), 83–95.

Chabowski, Brian R., Saeed Samiee, and G. Tomas M. Hult (2013), “A Bibliometric Analysis of the Global Branding Lit- erature and a Research Agenda,” Journal of International Business Studies, 44 (6), 622–34.

Chryssochoidis, George, Athanassios Krystallis, and Panagiotis Perreas (2007), “Ethnocentric Beliefs and Country-of-Origin (COO) Effect: Impact of Country, Product and Product Attributes on Greek Consumers’ Evaluation of Food Prod- ucts,” European Journal of Marketing, 41 (11/12), 1518–44.

Dawar, Niraj and Philip Parker (1994), “Marketing Universals: Consumers’ Use of Brand Name, Price, Physical Appearance, and Retailer Reputation as Signals of Product Quality,” Jour- nal of Marketing, 58 (April), 81–95.

De Mooij, Marieke and Geert Hofstede (2011), “Cross-Cultural Consumer Behavior: A Review of Research Findings,” Journal of International Consumer Marketing, 23, 181–92.

Dimofte, Claudiu V., Johny K. Johansson, and Richard P. Bagozzi (2010), “Global Brands in the United States: How

Consumer Ethnicity Mediates the Global Brand Effect,” Jour- nal of International Marketing, 18 (3), 81–106.

Douglas, Susan P. and C. Samuel Craig (2011), “Convergence and Divergence: Developing a Semiglobal Marketing Strategy,” Journal of International Marketing, 19 (1), 82–101.

———, ———, and Edwin J. Nijssen (2001), “Integrating Branding Strategy Across Markets: Building International Brand Architecture,” Journal of International Marketing, 9 (2), 97–114.

Dubois, Bernard, Sandor Czellar, and Gilles Laurent (2005), “Consumer Segments Based on Attitudes Toward Luxury: Empirical Evidence from Twenty Countries,” Marketing Let- ters, 16 (2), 115–28.

Erdem, Tülin, Joffre Swait, and Ana Valenzuela (2006), “Brands as Signals: A Cross-Country Validation Study,” Journal of Marketing, 70 (January), 34–49.

Farris, Paul W., Neil T. Bendle, Phillip E. Pfeifer, and David J. Reibstein (2006), Marketing Metrics: 50+ Metrics Every Exec- utive Should Master. Philadelphia: Wharton Publishing.

Gielens, Katrijn and Jan-Benedict E.M. Steenkamp (2007), “Drivers of Consumer Acceptance of New Packaged Goods: An Investigation Across Products and Countries,” Inter- national Journal of Research in Marketing, 24 (2), 97–111.

Guo, Xiaoling (2013), “Living in a Global World: Influence of Consumer Global Orientation on Attitudes Toward Global Brands from Developed Versus Emerging Countries,” Journal of International Marketing, 21 (1), 1–22.

Hausman, Jerry A. (1978), “Specification Tests in Econo - metrics,” Econometrica, 46 (6), 1251–71.

Herche, Joel (1994), “Ethnocentric Tendencies, Marketing Strategy and Import Purchase Behaviour,” International Mar- keting Review, 11 (3), 4–16.

Hofstede, Frenkel Ter, Michel Wedel, and Jan-Benedict E.M. Steenkamp (2002), “Identifying Spatial Segments in Inter- national Markets,” Marketing Science, 21 (2), 160–78.

Hofstede, Geert (1983), “The Cultural Relativity of Organiza- tional Practices and Theories,” Journal of International Busi- ness Studies, 14 (2), 75–89.

———, Gert Jan Hofstede, and Michael Minkov (2010), Cul- tures and Organizations: Software for the Mind. New York: McGraw-Hill.

Iversen, Nina M. and Leif E. Hem (2011), “Reciprocal Transfer Effects for Brand Extensions of Global or Local Origin: Evi- dence from Norway,” International Marketing Review, 28 (4), 365–411.

Kapferer, Jean-Noel (2002), “Is There Really No Hope for Local Brands?” Brand Management, 9 (3), 163–70.

72 Journal of International Marketing

Kennedy, Peter (2003), A Guide to Econometrics. Cambridge, MA: MIT Press.

Klein, Jill Gabrielle (2002), “Us Versus Them, or Us Versus Everyone? Delineating Consumer Aversion to Foreign Goods,” Journal of International Business Studies, 33 (2), 345–64.

Koutsoyiannis, Anna (1977), Theory of Econometrics. London: Macmillan.

Morrison, Allen J., David A. Ricks, and Kendall Roth (1991), “Globalization Versus Regionalization: Which Way for the Multinational?” Organizational Dynamics, 19 (3), 17–29.

Özsomer, Ayşegül (2012), “The Interplay Between Global and Local Brands: A Closer Look at Perceived Brand Globalness and Local Iconness,” Journal of International Marketing, 20 (2), 72–95.

——— and Selin Altaras (2008), “Global Brand Purchase Likeli- hood: A Critical Synthesis and an Integrated Conceptual Frame- work,” Journal of International Marketing, 16 (4), 1–28.

Petersen, J. Andrew, Tarun Kushwaha, and V. Kumar (2015), “Marketing Communications Strategies and Consumer Financial Decision Making: Role of National Culture,” Jour- nal of Marketing, 79 (January), 44–63.

Richardson, Paul S., Alan S. Dick, and Arun K. Jain (1994), “Extrinsic and Intrinsic Cue Effects on Perception of Store Brand Quality,” Journal of Marketing, 58 (October), 28–36.

Rugman, Alan M. and Simon Collinson (2004), “The Regional Nature of the World’s Automotive Sector,” European Man- agement Journal, 22 (5), 471–82.

Schlie, Erik and George Yip (2000), “Regional Follows Global: Strategy Mixes in the World Automotive Industry,” European Management Journal, 18 (4), 343–54.

Shimp, Terence A. and Subhash Sharma (1987), “Consumer Eth- nocentrism: Construction and Validation of the CETSCALE,” Journal of Marketing Research, 24 (August), 280–89.

Spiller, Stephen A., Gavan J. Fitzsimons, John G. Lynch Jr., and Gary H. McClelland (2013), “Spotlights, Floodlights, and the Magic Number Zero: Simple Effects Tests in Mod- erated Regression,” Journal of Marketing Research, 50 (April), 277–88.

Steenkamp, Jan-Benedict E.M. (2014), “How Global Brands Create Firm Value: The 4V Model,” International Marketing Review, 31 (1), 5–29.

———, Rajeev Batra, and Dana L. Alden (2003), “How Per- ceived Brand Globalness Creates Brand Value,” Journal of International Business Studies, 34 (1), 53–65.

——— and Martijn G. de Jong (2010), “A Global Investigation into the Constellation of Consumer Attitudes Toward Global and Local Products,” Journal of Marketing, 74 (November), 18–40.

Strizhakova, Yuliya, Robin A. Coulter, and Linda L. Price (2008), “Branded Products as a Passport to Global Citizen- ship: Perspectives from Developed and Developing Coun- tries,” Journal of International Marketing, 16 (4), 57–85.

Swoboda, Bernhard, Karin Pennemann, and Markus Taube (2012), “The Effects of Perceived Brand Globalness and Per- ceived Brand Localness in China: Empirical Evidence on West- ern, Asian, and Domestic Retailers,” Journal of International Marketing, 20 (4), 72–95.

Talay, M. Berk, Roger J. Calantone, and Clay M. Voorhees (2014), “Coevolutionary Dynamics of Automotive Compe- tition: Product Innovation, Change, and Marketplace Sur- vival,” Journal of Product Innovation Management, 31 (1), 61–78.

———, Orkunt Dalgic, and Tevfik Dalgic (2010), “Shareholder Value Creation in International Joint Ventures: The Case of the U.S. Automotive Industry,” Strategic Management Review, 4 (1), 1–17.

Townsend, Janell D., Sengun Yeniyurt, and M. Berk Talay (2009), “Getting to Global: An Evolutionary Perspective of Brand Expansion in International Markets,” Journal of Inter- national Business Studies, 40 (2), 301–320.

Üstüner, Tuba and Douglas B. Holt (2007), “Dominated Con- sumer Acculturation: The Social Construction of Poor Migrant Women’s Consumer Identity Projects in a Turkish Squatter,” Journal of Consumer Research, 34 (1), 41–56.

Wooldridge, Jeffrey M. (2000), Introductory Econometrics: A Modern Approach. Independence, KY: South-Western College Publishing.

——— (2002), Econometric Analysis for Cross Section and Panel Data. Cambridge, MA: MIT Press.

Yeniyurt, Sengun, Janell D. Townsend, and M. Berk Talay (2007), “Factors Influencing Brand Launch in a Global Mar- ketplace,” Journal of Product Innovation Management, 24 (5), 471–85.

Zhou, Liang, Zhiyong Yang, and Michael K. Hui (2010), “Non- Local or Local Brands? A Multi-Level Investigation into Con- fidence in Brand Origin Identification and Its Strategic Impli- cations,” Journal of the Academy of Marketing Science, 38 (2), 202–218.

Copyright of Journal of International Marketing is the property of American Marketing Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.