of market effect”

Purpose: This article proposes and empirically tests the country of market (COMK) effect, which captures the consumer’s responses of home market to a country where the product is marketed. Design/methodology/approach: Study 1 applies a lab experiment about Chinese consumers’ purchase intention for printers marketed either in the US or China. Study 2 applies country level data to examine the impact of economic development of 22 host countries on the performance of 167 multinational retailers in their home country. Findings: Study 1 shows that the printers marketed in US attract a higher level of purchase intention than printers marketed in China. This COMK effect is more salient for printers manufactured in China than those manufactured in US. In addition, innovation and design factors corresponding to the host country’s image fully mediate the COMK effect. Results in Study 2 show that a retailer that markets its services in a host country with a higher (lower) level of economic development is likely to generate higher (lower) level of retailing performance in its home country. Furthermore, it is found that COMK effect is diminished as the level of economic development of a vendor’s home country increases. Research limitations/implications: In addition to the cognitive components of country image (e.g., design and innovation), consumers’ affective components may also influence the COMK effect. Future research could discuss the impact of consumer ethnocentrism and consumer animosity on consumers’ attitude towards the product marketed in other countries. Practical implications: Strategically, marketing products to a country with a favorable image could benefit vendors from an emerging economy. For manufacturers from developed countries, marketing a product within their own countries may enhance the associated innovation and design images while marketing the same product in an emerging market. Originality/value: This article proposes and tests a demand side country effect on consumers’ purchase intention for products marketed in other countries. It is in sharp contrast to the traditional country effect which focuses on the supply side effect (e.g., country of origin, country of manufacture, country of assembly etc.).


Introduction ©
Globalization has accelerated the outsourcing of production away from the West toward a few countries with significant manufacturing capabilities and/or cost advantages.With this diffusion of manufacturing technology, countries such as Korea and China have quickly developed powerful reputations for manufacturing.Many global destinations now possess the ability to deliver competitive manufacturing excellence that was the exclusive preserve of the West only a few decades ago.
It is well known that some countries mandate that manufacturers declare the country of origin (COO) information of their products.This rests on the belief that COO information is important for shaping beliefs and expectations about product performance.This study advocates a similar position with regard to information about country of market (COM), i.e., a mar-ket destination where a manufacturer's product has performed exceptionally.Although a rich literature exists on the COO effect, the COM effect remains relatively unexplored.It is not well known that firms may voluntarily disclose information about prestigious global markets where their product has performed especially well.For example, Katmerciler, a Turkish manufacturer of vehicle mounted equipment, advertizes its product locally by highlighting its export performance in dozens of other countries.In the world's largest C-to-C website (http://taobao.com),an affiliated e-commerce company of Alibaba, more than one million items carry "export" information to signal product quality.Historically, even strong brands such as Coca Cola have showcased export-oriented product versions in advertising messages (for an example from the 1920s, see http://www.adbranch.com/coca-colafor-export/).Finally, from a financial perspective, the stock performance of global firms such as Samsung, Nokia and Apple depends heavily on their respective products' performance in tough overseas markets.Such examples demonstrate that consumers may base their purchases on where a product is sold, in addition to the place where a product is made.These changes in production and consumption domains highlight the need to re-conceptualize and extend the well-known country of origin effect to accommodate consumers' perceptions of products marketed in other countries.The demand side effect based on where exported products are sold remains largely unexplored, with the lone exception of Yucetepe's (2003) unpublished dissertation on country of destination effect that focuses on the effect of export destination on consumers' product evaluations.
Our study re-examines product-related perceptions exemplified in the country of origin (COO) effect (Han, 1988;Han and Terpstra, 1989;Hsieh, 2004).It reaffirms a key assumption in the COO literature that consumers rely on perceptions of country image or equity to reach product purchase decisions.However, it differs from the COO literature in that such perceptions may not be exclusively tied to the country of origin.Instead, it highlights consumers' perception of equity of the country where a product is marketed, thereby allowing the product to also build a reputation based on that country's image.We term this as country of market (COMK) effect.
Two studies demonstrate the COMK effect, the boundary conditions and the mechanism of this effect.Study 1 applies a quasi-experiment approach.It focuses on the purchase intentions (PI) of Chinese participants toward printers that are marketed in either China or US, and that are manufactured in either China or US.The results show that printers marketed in US attract significantly higher level of purchase intention than those marketed in China.The COMK effect is more salient for printers manufactured in China.Besides, the innovation and design dimensions of country image fully mediate the COMK effect.To test the COMK effect on the retailing performance in the real world, study 2 applies country level data to examine the impact of the host country on the performance of a multinational retailer in its home country.We found that the higher the level of economic development of the host country, the higher the sales per unit area of the retailer in its home country.Findings also suggest that the COMK effect is weaker if the home country reflects a higher level of economic development.
Our research contributes to the literature of country effects in multiple respects: (1) it explores the demand-side country effect from goods domain to the services domain (i.e., retailing); (2) it explores the boundary condition of COMK effects; (3) it delineates the mechanisms that capture how COMK effects occur in a developing country; (4) it generates new findings and insights about consumer behavior that integrate and strengthen product marketing activities across international and domestic markets.In particular, it showcases the influence of a product's performance in overseas markets on its performance within a domestic market.
We organize the paper as follows.We, next, present a conceptual framework, along with related hypotheses.We, then, describe two empirical studies designed to test the COMK effects, and to illustrate their mechanisms and boundary conditions.Finally, we discuss the results and their limitations, and offer some directions for future research.

Country image
Country image represents the consumers' gestalt perception of a product within a particular country (Nagashimaya, 1970).It captures the strengths and weaknesses of an industry in a particular country at a general level (Pappu, Quesster and Cooksey, 2007;Wu, 2011).Country image typically embeds the equity of a country that is associated with the evaluation of a product (Shimp and Sharma, 1987;Lin and Chen, 2006).Therefore, consumers' ratings of products (and their purchase intentions) are likely to be closely associated with the equity of country where they are marketed.
The customer-based brand equity literature (e.g., Keller, 1993) offers useful implications on how consumers evaluate and purchase products based on the quality information provided by the country image.First, consumers use intangible extrinsic cues as a signal of overall quality.Examples of such extrinsic quality cues include price (e.g., Olson and Jacoby, 1972), warranties or guarantees, brand reputation (Jacoby et al., 1977) or seller reputation (Shapiro, 1982).In this case, consumers do not process the detailed product information.Particularly when consumers lack product information, extrinsic cues may serve as a signal of product quality (c.f.Erdem and Swait, 1998).
Similarly, a country's image may reflect consumers' country-specific beliefs about workmanship, durability and reliability (Erickson, Johansson and Chao, 1984).Consumers may infer the quality of a product from the image of the country where it was produced.This notion is exemplified in the COO literature that posits that products originating from a country with a favorable image tend to receive better evaluations (e.g., Han and Terpstra, 1988; Peterson and Jolibert, 1995; Verlegh and Steenkamp, 1999) relative to those originating from countries with unfavorable images.In a similar vein, the COMK effect may signal that a product possesses the design and innovation features that consumers in the host country expect.
Consumers may develop an overall product quality assessment based on country stereotypes in their minds.For instance, the literature suggests that consumers may react differently to the marketing mix of different competing brands.From their past experiences, consumers may develop impressions about a brand.Such an impression, or brand image, is associated with the product's functional and symbolic benefits.Similarly, a consumer's experience with a product that was marketed in a specific host country may shape or reinforce country-specific image(s).In other words, the COMK effect is driven by the equity of a prestigious host country that consumers' may rely on to evaluate a product, in contexts where a vendor markets its product within that host nation.Hence, COMK plays a similar role in shaping consumers' product quality judgments.For example, printers marketed in a more developed country are more likely to be preferred over those marketed in a less developed country.Based on the above argument, we propose that: H 1 : The vendor's equity for country of market (host country) positively influences consumers' purchases in its country of origin (home country).
COMK effect might be diminished when the vendor incorporates positive COO image.A vendor from a developed country might already possess positive features embedded in the country equity.For example, US-based vendors may already hold positive image with respect to design and innovation.For a US-based vendor, note that activities such as marketing a product within the US may convey similar quality information to consumers as in COO.Therefore, the product quality information conveyed by COMK might be redundant given the information gleaned from COO, if both COO and COMK effects are attributable to US.In contrast, the COMK effect is relatively more salient for a vendor from a less developed country, because it does not hold positive country equity by default.For these reasons, we expect a negative interaction effect of COMK and COO on consumer purchase as hypothesized next.
H 2 : The country of market effect would be reduced along with the vendor's equity for country of origin.
Country image is a multi-dimensional construct.It consists of three dimensions: design, innovation and pride (Pappu et al., 2007;Roth and Romeo, 1992).Design and innovation represent cognitive elements, while pride reflects the emotional and symbolic benefits perceived by consumers.

Design and innovation.
As the cognitive aspects of country image, design and innovation characteristics influence a consumer's purchase intention for a product that possesses favorable or unfavorable COO perceptions (Chuang and Yen, 2007;Hong and Wyer, 1989;Martin and Eroglu, 1993;Pappu et al., 2007).For example, US, as a country, is generally associated with high levels on the following desirable characteristics: rich history of innovation and product development, robust marketing expertise, demanding product design and per-formance standards, effective regulations to enforce those standards, and large, sophisticated and discerning consumer markets.To the extent that such country-specific characteristics serve as testimonials to the quality of a product that is marketed therein, a consumer might use the design and innovation information embedded in the country image for purchasing decisions.Based on the above argument, we propose: The design dimension of country image mediates the country of market effect in a developing country.
H 4 : The innovation dimension of country image mediates the country of market effect in a developing country.

Pride.
Consumers in a less-developed country may experience a feeling of pride when purchasing products marketed in developed countries.First, the respective country image(s) may provide symbolic and emotional benefits to consumers (Hong and Wyer, 1989;Zhou and Hui, 2003).Hence, consumers in an emerging economy often associate the consumption of products from western countries with social status enhancement and demonstration of wealth and power (Batra et al., 2000; Alden, Steenkamp and Batra, 1999; Tse, Belk and Zhou, 1989).Second, Escalas and Bettman (2003) found that consumers might imitate the consumption behavior of other consumers who belong to an aspirational group.For example, consumers in a developing country who contemplate the purchase of a product that is marketed in the US have the opportunity to imitate the consumption behavior of US consumers.They might feel proud of consuming a product marketed in the US, because the same product is also available to consumers in their aspirational group.As such, we propose: H 5 : The pride dimension of country image mediates the country of market effect in a developing country.

Design of study 1.
The objective of study 1 is to test the COMK effect, explore its mechanism and boundary conditions.Printer was chosen as the focal product because it was frequently used by, and was affordable for undergraduate students.The product description of the printer reflected the specifications of a mini black/white laser printer, including print speed, resolution, paper handing input and monthly duty cycle.The price of the printer was set as RMB 800 (equivalent to USD 128), which reflects the price set by a mainstream Chinese online retailer.We adopt 2 (COMK: China [coded as 0] vs. US [coded as 1]) x 2 (COO: China [coded as 0] vs. US [coded as 1]) between subject design.120 undergra-duates at a large southern university in China participated in this study.Respondents were paid RMB 10 for participation, and randomly assigned to one of the four experimental treatment cells.We did not identify a particular brand name for both products, because brand name incorporates country information (Pappu et al., 2007).
Respondents initially provided ratings for purchase intention (Dodds, Monroe and Grewal, 1991) for the product they were assigned to.They, then, rated the design and innovation measurement scales for the printer associated with corresponding COMK and COO experimental treatment assignments.We adapted Pappu et al. (2007) scales by adding COMK into the original COO scale.In other words, the items provided the name of COO (e.g., US), the name of COMK (e.g.China) and the specific nature of a product (e.g.innovative).Respondents, then, rated their subjective knowledge of the printer using a scale developed by Moorman et al. (2004).Finally, we collected the respondent's knowledge of COO and COMK using a 7 point scale for each factor.The measurement scales for purchase intention, design, innovation, knowledge of product, COO and COMK constructs used a 7 point Likert scale, with higher scores representing more positive ratings.Table 1 demonstrates the factor structure of the dimensions of design, innovation and pride.Appendix 1 reports the matrix of correlation between experimental conditions (COMK and COO), and underlying purchase intention, design, innovation, pride constructs.Study 1 found that COMK enhances consumers' purchase intention, especially for products manufactured in a developing country.In addition, country image of design and innovation were shown to fully mediate the COMK effect.

Study 2
To test the COMK effect on retailing performance in the real world, study 2 aims to test this in the retailing industry context.This study also seeks to investigate the boundary conditions of this effect.
Rapid internationalization in retailing took shape in the late 1990s when leading retailers like Walmart and Carrefour expanded to Asian and African countries (Wrigley et al., 2005).The retail industry offers sufficient international data for our analysis.

Sample.
The sample retailing firms we studied were from the Euromonitor database.This multicountry database records sales and operation area of retailers whose market share exceeds 0.5% of total retail sales of consumer goods in a country.We choose 167 multinational retailers which conduct business in 2 to 22 countries.Our sample covers the timeframe from year 2006 to 2011.Some retailers have a performance record of less than six years, and our final dataset yields 936 retailer-year observations and an unbalanced panel.

Variables and model specification of study 2.
We apply the sales per unit area of a retailer as a performance indicator in the home country, because land cost is the major investment in the retailing industry.We compute the impact of host countries to the business performance of a retailer in the home country by the weighted sum of GDP per capita of host countries.The weight is calculated by a retailer's sales in a host country over its total sales in the host countries.According to H 1 , the higher the weighted GDP per capita of the host countries, the higher the sales per unit area of a retailer in its home country would be.In addition, the GDP per capita of the retailer's home country signals the economic development level of that country.According to H 2 , we expect that the higher the GDP per capita of the home country, the lesser the impact of host countries would be on the performance of retailer in the home country.
At the macroeconomy level, we control for the GDP and population density of the home country.At the microretailer level, we control for the retailer's age and market share.The data for GDP and population density were obtained from the World Bank database.We relied on the websites of the retailers in our database to calculate the age of each retailer in each year.The market share for each retailer is calculated by dividing its sales over the total retail sales of consumer goods of a country.The mean and correlation of all the variables used in this study are reported in Appendix 2.
We apply standard panel linear model to analyze the data.The panel model controls for the individual retailer's unobserved heterogeneity (e.g., reputation) and allows us to examine the performance directly.
The Hausman (1978) test shows that there is a significant difference between the fixed effect model and random effect model (χ 2 = 194.123,df = 6, p < 0.001).Hence, we use the fixed effects model.shows the interaction of the impact of GDP of host countries, and the GDP per capita of the home country on retailer's performance in the home country is negative (β = -3.090,s.e = 1.499).The result implies that the impact of COMK effect from the host countries would be weaker in a home country with a high GDP per capita.Thus, H 2 is supported as well.The predictions for all other control variables with respect to retailer's sales per unit area are remarkably the same as in the main effect model.

Result of study 2.
Study 2 empirically supported the COMK effect in the retailing industry based on the performance data of multinational retailers in various countries.The result implies that the economic country image of the host countries might be a signal of retailing service quality.Yet, the strength of the signal of the host countries is weaker, when the home country equity independently serves as a signal of service quality.

Overall discussion
Across the two studies, our findings show that consumers rely on the image of the country where the product is marketed, and stimulate purchases in the home country.Therefore, the superiority of innovation and design images of favorable COMK (e.g., US as a market) may spill over to the vendors with unfavorable COO, if vendors from developing countries market their products in developed countries.Our research found that design and innovation images mediate the COMK effect on PI.Zhang et al. (2010) maintain that technology of foreign invested enterprises in China may spill over to Chinese manufacturers.Our research implies that spillover might occur more easily at the level of product markets.Hence, international trade competition might not be confined to the global marketing arena, but its influence may also spill over to competition within a country.
Strategically, marketing products to a country with a favorable image could benefit vendors from an emerging economy with respect to marketing efforts within their own country.Our study finds that vendors that market their products in a more developed country may enhance both consumers' PI and retailing performance.For vendors located in an emerging economy, marketing products in developed countries provide an alternative way to compete with manufacturers from developed countries in the local market.For example, Wahaha, a leading brand in the soft drink industry in China, promotes its product in China by leveraging its market exposure in US, Canada and Singapore.Studies in the COO literature also imply some spillover benefit for a product that is widely known.For instance, Winit et al. (2014) found that consumers favor global brands (i.e., brands that are sold worldwide) over non-global brands.
For manufacturers from developed countries, marketing a product within their own countries promotes the associated innovation and design images while marketing the same product in an emerging market.Hence, manufacturers from developed countries should market similarly designed and innovative products in both emerging economies and their own countries.For example, a prominent Japanese cosmetic manufacturer markets a sub-brand uniquely formulated for the skin characteristics of Chinese women.Unfortunately, this sub-brand is perceived to be at a lower end than its counterparts marketed in both Japan and China.Although customizing the product for Chinese consumers might be a good practice, branding the product locally might make things worse.

Conclusion
We empirically test a demand side COMK effect that differs from the supply side COO effect based on a quasi-experiment focusing on consumer purchase intention with COMK information and a retailing performance analysis.We found that vendor's equity of country of market positively influences purchase intention and retailing performance in the home country.The COMK effect is stronger for vendors from relatively less developed countries.Design and innovation factors of country image fully mediate the COMK effect in a developing country.

Limitation and future research
Compared with the large volume of research on the COO effect over the past 40 years, our research is an early study on the COMK effect.One limitation of our work is its focus on the cognitive components of COMK effects (e.g., design and innovation).We acknowledge that affective components may also influence COMK effects.For example, consumer ethnocentrism (Zolfagharian, Saldivar and Sun, 2014) and consumer animosity (Maher, Clark and Maher, 2010) may reduce consumers' purchase intention for products marketed in other countries.Future research could address these topics.In addi-tion, future research should investigate how product type and product involvement (Prendergast, Tsang and Chan, 2010) influence the COMK effect.Finally, the interaction of brand image and country image deserves research attention.

Table 3 .
Fixed effect model of retailer performance