How EMNEs and DMNEs can attract applicants in emerging and developed countries – a cross-national conjoint analysis on the role of country-of-origin and CSR

Abstract

Multinational enterprises (MNEs) have increasingly expanded worldwide, so they need to recruit talent abroad. Nevertheless, the extant recruitment literature lacks an international perspective, as most research has been conducted in a single country context and in developed countries, creating a gap regarding how MNEs can develop optimal international recruiting strategies, considering not only developed but also emerging countries. Based on a cross-national conjoint analysis in an emerging country (Vietnam) and a developed country (the US), we calibrate the relative importance of organizational/job attributes, especially those with symbolic value such as MNEs’ country-of-origin (emerging vs. developed) and the three dimensions of CSR (economic, social and environmental responsibilities), together with other instrumental factors (e.g. pay, career opportunities), to young applicants in the two countries. Our results reveal some differences. Applicants from an emerging market attach more value to the economic dimension of CSR; still, they also value to the social and environmental CSR dimensions. Unexpectedly, emerging market MNEs (EMNEs) suffer more from the liability of emergingness in other emerging countries than in developed countries. Despite the persistence of certain cross-national differences, the overall influential structure on job choice remains largely similar across countries, opening up the possibility for a global employer branding strategy. Our results, moreover, suggest that symbolic attributes are as important as instrumental attributes to applicants. Most interestingly, the economic dimension of CSR and the country of origin are considered even more important than salary in Vietnam.

Keywords:

• Job choice
• international recruitment
• emerging market multinational enterprises
• CSR
• country-of-origin
• conjoint analysis

Introduction

By their very nature, multinational enterprises (MNEs) must attract an adequately skilled workforce in different host countries. The scarcity of skilled workers has intensified the global competition for talent, prompting MNEs to strategically position themselves in the international labor market (Held & Bader, 2018).

Drawing on the idea that job choice serves a social identity function conveying information about a person’s social identity (Tajfel, 1974), Lievens and Highhouse (2003) propose that when making their job choice, applicants are concerned with not only basic instrumental or tangible benefits, such as salary and career opportunities, but also the symbolic images that they ascribe to the organization in their job choice to maintain their self-identity, enhance their self-image or express themselves.

MNEs in foreign host countries have to compete with domestic players to attract local applicants. One notable distinguishing feature is their country-of-origin (COO), making it crucial to comprehend the effect of COO on job choice. Symbolic organizational image inferred from COO has been shown to have an effect on employer attractiveness in developed host countries in previous research (Held & Bader, 2018; Zhang et al., 2020), but not for emerging market labor markets, which are becoming increasingly relevant as host countries for MNEs.

Another symbolic organizational image has emerged as a dominant attribute in the public discussion – corporate social responsibility (CSR). A recent survey by Deloitte (2019) reported that 38% of millennials and Gen Z have backed away from companies that negatively impact the environment and society. Already in 2011, a study by Montgomery and Ramus showed that CSR was among the top five factors considered by MBAs from North America and Europe during their job search. Recently, Le and Morschett (2023) demonstrated that CSR positively affects employer attractiveness, and Dauth et al. (2023) highlighted the positive impact of top management team diversity which is one element of CSR. Overall, CSR acts as a reputational asset that sets firms apart from their rivals and provides an advantage in the ‘war for talent’ (Jones et al., 2019).

Although research has shed light on the role of COO (Alkire, 2014; Held & Bader, 2018; Holtbrügge & Kreppel, 2015; Zhang et al., 2020) and CSR (Hong & Kim, 2017; Montgomery & Ramus, 2011) in recruitment, four central premises pertaining to these attributes merit investigation. First, little is known about the extent to which applicants value these symbolic attributes relative to instrumental attributes like salary. Most prior studies applied (1) direct estimation, i.e. directly asking respondents to rate the importance of individual attributes, or (2) experimental policy-capturing, i.e. asking respondents to evaluate scenarios describing various levels of a set of attributes and then regressing their responses on the attributes to capture how they use the available information to make the overall evaluation. The former approach gives little insight into how the ratings of individual attributes are used in actual decisions and it is strongly criticized for eliciting socially responsible responses (Karren & Barringer, 2002). The latter improves such limitations, but most policy-capturing studies examined only two or three attributes and failed to consider their interdependency and relative importance in realistic trade-off decisions (Ronda et al., 2021). Indeed, realistic job offers hardly combine the best of everything. Hence, the common approach used in extant literature has hindered capturing the trade-off decisions faced by applicants and has not effectively addressed the important, yet unanswered question of how much MNE’s COO and CSR matter in job choice relative to other important factors (Jones & Willness, 2013). Obtaining answers to this question offers a more stringent assessment of the effects of MNE’s COO and CSR, allowing firms to focus on what really matters given their limited resources.

Second, while CSR is multifaceted, extant literature primarily focuses on its overall effects and overlooks the differential influences of its facets on candidates (Zhao et al., 2022). Le and Morschett (2023), while showing that CSR in general has a positive effect on employer attractiveness, call for a de-composition of its dimensions to gain deeper insights, enabling firms to make informed decisions and allocate resources effectively.

Third, the interplay between COO and CSR effects remains unclear. Such an international perspective on CSR is missing from the extant literature (Cooke et al., 2019), as the most frequently studied moderators so far are individual-level variables. Jones et al. (2019), nevertheless, underlined that although the vast majority of extant individual-focused CSR research is micro in nature, ‘the scientific study of these phenomena is by no means restricted to the individual level’ (p. 302).

Finally, it still remains largely unknown how different facets of employer image drive applicants’ job choices in a cross-national context. Research has focused narrowly on developed North American and European host countries (except China), often with developed market MNEs (DMNEs) as employers, as exemplified by a recent review of Cooke et al. (2019) that only 6% of the studies include emerging market MNEs (EMNEs). These findings have limited utility in an increasingly globalized work context, with the center of economic gravity slowly shifting to emerging and developing economies, not only restricted to China. MNEs have rapidly expanded their operations into emerging markets, hence the increasing need to recruit talent in these markets. This invokes questions of whether MNEs can adopt a global employer branding or should customize to the emerging countries. Investigating to which degree job preference is culture-free or culture-bound assists MNEs in managing the tension between global integration and local responsiveness, one of the key challenges they face (Rosenzweig, 2006). Further, Cooke et al. (2019) indicate that despite the fact that CSR is at the forefront on the global corporate agenda in today’s socially conscious market environment, its role in international HRM is neglected.

Therefore, the study aims to address these gaps by exploring the following questions. First, we extend the international recruitment literature by applying conjoint analysis to investigate: What is the relative importance of (1) MNE’s COO (developed vs. emerging) and (2) three CSR dimensions, encompassing the economic, social and environmental responsibilities—the three defining pillars of the triple bottom line concept (Elkington, 1998), as compared to more traditional instrumental attributes in job choice?

Second, we advance the understanding of how job and organizational attributes influence applicants’ job choices in a cross-national context by examining whether and to what extent the relative importance of MNE’s COO and CSR in job choice varies across countries. In this study, we collect data in an emerging country (Vietnam) and a developed country (the US), with potential applicants’ from these countries evaluating job offers from EMNEs and DMNEs. We include emerging markets, as labor markets and as MNEs’ origin, to contribute to the under-researched ‘human side’ of emerging markets as well as EMNEs (Meyer & Xin, 2018) and to simultaneously address questions of increasingly practical relevance, such as: Do EMNEs suffer from liabilities of emerging in emerging countries? Do emerging market graduates care for CSR? Can EMNEs overcome their liabilities via CSR? And if yes, via which CSR dimensions in which labor market? While recruitment research has found that EMNEs suffer from double disadvantages of liabilities of foreignness plus liabilities of emergingness when recruiting talent in developed markets (Held & Bader, 2018), the existence and the magnitude of the liabilities of emergingness in emerging markets have still been neglected in the literature. Given the continuing growth of EMNEs in both emerging and developed markets, it is imperative to address this gap (Zhang et al., 2020).

Theoretical background

The influence of instrumental and symbolic attributes on job choice

Research on applicant attraction to organizations has identified several outcome variables, including organizational attractiveness (e.g. Lievens & Highhouse, 2003), job pursuit intentions (e.g. Hong & Kim, 2017), and job choice (e.g. Montgomery & Ramus, 2011). This study focuses specifically on job choice because scholars tend to overlook applicant decisions and instead focus on organizational attractiveness as the focal dependent variable. Organizational attractiveness is a positive affective attitude toward considering an organization a desirable place to work (Highhouse et al., 2007). Rynes (1991, p. 436), nevertheless, has stressed that ‘we have virtually no information about how preferences and intentions are converted into actual job choices’. Chapman et al. (2005)’s meta-analysis substantiates that while organizational attractiveness is a significant antecedent of job choice, it should not be treated as a direct proxy for job choice itself. Boswell et al. (2003) define job search and eventual job choice as a dynamic decision-making process where applicants evaluate organizational/job attributes to choose jobs in organizations that best fit their needs. Lievens and Highhouse (2003) categorized these attributes into two broad categories—instrumental and symbolic attributes. The former refers to tangible attributes with utilitarian value, such as pay, benefits, career opportunities, whereas the latter denotes intangible attributes with symbolic images, such as prestige or credibility derived from an organizational affiliation. Table 1 presents a compilation of studies related to organizational/job attributes and their impacts on job search process and job choice in an international recruitment context.

Table 1. Overview of organizational and job attributes influencing applicant attraction and job choice in international recruitment contexts.

Author(s) (year)	Sample	Attributes of job or organization^b	Methodology/Contribution	Limitations
Lievens et al. (2001)	359 final-year engineering and business students in Belgium	Size Pay mix Level of internationalization Level of centralization	Using policy-capturing and experimental methods to investigate the effect of four organizational characteristics on applicant attraction	Limitation in using hypothetical scenarios (in which respondents are not facing a job choice situation) to measure attribute importance
Moy and Lee (2002)	200 final-year business students in Hong Kong	Career advancement opportunities Salary Job security Managerial relationships Fringe benefits Working conditions Involvement in decision making Responsibility given Marketability	Using direct estimation: respondents were asked to rank the importance of nine job attributes in their job choice and their perceptions of these attributes offered by SMEs and MNEs	Direct estimation poses a high risk of social desirability
Newburry et al. (2006)	4,605 US citizens	Level of internationalization Foreign ownership	Using policy-capturing method with online interview to investigate the effect of two foreignness dimensions on applicant attraction	Only two attributes and no test of relative important of the attributes
Montgomery and Ramus (2011)^1	759 MBAs in North America and Europe	Task attractiveness Geographic area Financial package Ethical reputation	Using conjoint analysis to measure the relative importance weights for a set of 14 attributes, including CSR, on applicant’s job choices	Findings may not be valid for applicants in emerging markets.
Baum and Kabst (2013b)	1,626 engineering upper-level undergraduate students in Germany, Hungary, China, and India	Work climate Career opportunities Work-life balance Task attractiveness Payment attractiveness	A cross-national study using policy-capturing method to investigate whether the impact of five employer image facets on application intentions varies across countries	Some potentially relevant attributes, such as CSR and COO, are not investigated.
Alnıaçık et al. (2014)^a	300 university students in Latvia and Turkey	Work climate Salary Recognition/appreciation from management	A cross-national study using direct estimation method to explore the differences in the importance levels of 25 items corresponding to the functional, economic and psychological aspects of employer branding between two countries	Direct estimation poses a high risk of social desirability
Alkire (2014)	626 German, French and American final-year students	COO Industry	Using policy-capturing and experimental methods to explore the influence of COO and country-of-origin images on applicant attraction of EMNEs in developed countries	Studies did not test whether EMNEs could compensate for their lower COO attractiveness by offering higher salaries
Holtbrügge and Kreppel (2015)^1	726 German citizens	COO Industry Salary Job security Work-life balance Self-dependent and creative work climate
Held and Bader (2018)	287 German business students	COO image Corporate character image
Zhang et al. (2020)	374 Dutch citizens	Cognitive COO image Affective COO image
Hong and Kim (2017)	189 Korean students	COO CSR	Using policy-capturing and experimental methods to investigate whether good CSR weakens the negative relationship between emerging COO of an MNE and its applicant attraction	No test of relative importance of COO and CSR attributes vs. other instrumental attributes such as salary

^a There are more than 10 attributes used in these studies. Space does not allow the listing of all attributes, but the top-rated attributes are listed in the attribute column.

^b The most important attributes (where applicable) are in bold.

Previous studies in Table 1 predominantly focused on the effects of few instrumental attributes (e.g. pays, benefits). However, there is evidence in the literature emphasizing the importance of incorporating both instrumental and symbolic attributes to comprehend applicant attraction (Kumari & Saini, 2018; Van Hoye et al., 2013). Van Hoye et al. (2013) found that organizations were better differentiated based on symbolic traits like sincerity and innovativeness, rather than instrumental dimensions like task demands. However, their study only focused on the symbolic dimensions of employer image, without explicitly examining the signals that provoke these symbolic inferences, limiting our understanding of how organizations can shape their image. Addressing this limitation, Kumari and Saini (2018) specifically investigated the role of CSR as a symbolic signal, comparing it with two instrumental attributes, namely career opportunities and work-life benefits. However, applicants typically consider a broader range of factors when making job choice decisions.

According to social-identity theory (Tajfel, 1974), identifying with groups is central to a person’s self-concept. Ashforth and Mael (1989) presented organizational identification as a specific case of social identity in recruitment context, where employees identify with employing organizations to enhance self-esteem and self-consistency needs. Highhouse et al. (2007) further explain an elaborated mechanism of symbolic attraction, suggesting that applicants select jobs in organizations with symbolic image in accordance with their social-identity concerns. To better understand this mechanism, it is important to examine a firm’s image for something specific, rather than its image per se (Barich & Kotler, 1991). Specifically, symbolic image inferences are translated into general impressions of impressiveness and respectability, which map on the two primary social-identity concerns for applicants in their job choice: (1) social-adjustment concerns relate to the need for approval from significant others; and (2) value-expression concerns relate to the need to express one’s ideals and values. To make impressiveness or respectable inferences, applicants must receive signals or cues about the organizations. Relying on signaling theory, Rynes (1991) noted that applicants, as external stakeholders, often lack complete information about potential employers; consequently, they must rely on the available information as signals to make inferences about unknown organizational characteristics and working conditions to reduce the information asymmetry. As a result, firms invest in sending signals to bridge the information gap and influence desired outcomes. For instance, a firm’s diversity management policy can serve as a signal of its working conditions, values and norms, ultimately impacting its attractiveness as an employer. Extant literature has revealed that applicants consider various factors as signals, including a company’s COO (Held & Bader, 2018), CSR (Jones et al., 2014) or top management team nationality diversity (Dauth et al., 2023).

Relevance of COO and of the three dimensions of CSR as signals for symbolic inferences

Applicants are nowadays also able to work, in their home countries, for MNEs from different origins. COO characteristics, particularly development levels, imprint on a country’s companies (Newburry, 2012). For example, Held and Bader (2018) and Holtbrügge and Kreppel (2015) observed a ‘hierarchy of biases’ phenomenon, in which firms from developed markets were systematically more attractive than firms from emerging markets. Specifically, developed COO can render symbolic value of prestige or high status and serve as a signal of impressiveness, thereby addressing applicants’ social adjustment concerns.

Furthermore, CSR, broadly defined as ‘a company’s commitment to minimizing or eliminating any harmful effects and maximizing its long-run beneficial impact on society’ (Mohr et al., 2001), is associated with valued characteristics such as empathy, honor and integrity and generates a respectable impression (Highhouse et al., 2007; Kim & Park, 2011), thereby fulfilling applicants’ value-expression concerns and exerting a positive influence on job choice. This is particularly relevant to growing social awareness and willingness to contribute to the society and environment. The G250 CSR reporting rate rocketed from only 35% in 1999 to 96% in 2020 (KPMG, 2020), indicating organizations’ efforts to signal their CSR and attract applicants. When examining MNEs, it is worth noting the similarities between MNEs and new ventures, particularly in terms of challenges they face. Both MNEs and new ventures operate under high levels of uncertainty and suffer from limited legitimacy as they expand across multiple countries or venture into uncharted territory (Held & Bader, 2018; Moser et al., 2017). Thus, we can draw connections to the literature on new venture employer attractiveness¹. Moser et al. (2017) have extended legitimacy theory, initially explored by Navis and Glynn (2011) in the context of new venture’s appeal to investor, to the recruitment context. New ventures and MNEs alike need to establish, besides heterogeneous identity claims, homogenous identity claims that signal attributes like trustworthiness, credibility, and reliability, thereby enhancing their perceived legitimacy. CSR can serve as a potential homogenous claim for MNEs, signaling their legitimacy and contributing to their overall attractiveness.

At the core, the three dimensions of the triple bottom line concept, aligning the economic, social and environmental responsibilities of business, are the defining pillars of twenty first century CSR (Elkington, 1998). First, the economic dimension of CSR is not limited to the firm’s financial profit but includes its contribution to the community’s economy. This includes creating wealth and employment, generating innovation, delivering competitively priced products/services that satisfy consumer needs, paying taxes and avoiding actions that might damage the firm’s license to operate in the long run, such as questionable payments and bribery—an ethical challenge confronting MNEs operating in emerging countries (Księżak & Fischbach, 2018). In this sense, the economic dimension directly benefits (prospective) employees as the better the economic sustainability of the firm, the lower the risk of going bankrupt or having to lay-off people (Duarte et al., 2014). Second, the social dimension of CSR is concerned with the direct and indirect business impacts on people, including employees, supply chain workers, customers and local communities (Capelle-Blancard & Petit, 2017). This dimension is attracting great attention due to increased public distrust toward MNEs’ business practices, as exemplified in scandals, especially regarding human rights violations under ‘sweatshop’ conditions in developing countries (White et al., 2017). Third, the environmental dimension of CSR has been at the center of the agenda for sustainability since the 1980s. There is ever-growing concern for the environment due to unprecedented environmental challenges and the widespread environmental movement (Scientists4Future, 2021). According to Fridays for Future (FFF, 2021), from August 2018 to June 2021 more than 100,000 climate strikes involving 14 million participants took place worldwide. Prior studies have predominantly focused on the social and environmental dimensions when investigating organizational attractiveness, overlooking the economic dimension (Hong & Kim, 2017; Jones et al., 2014; Kim & Park, 2011). However, the economic dimension is desirable for applicants as it signifies better long-term employment prospects. Additionally, facing the enduring global economic recession from 2008 and the recent COVID-19 outbreak, people are increasingly concerned about economic sustainability, with a prevalent fear of mass layoffs, poverty and financial risk to governments and public programs (Laborde et al., 2020).

Relevance of national contexts

MNEs operating across national borders need to attract a broader applicant pool from diverse host national contexts (Held & Bader, 2018). The term national context is rather broad and denotes an indefinite list of a country’s contextual factors (e.g. economic, political, institutional, educational, social, and cultural environment) that potentially can shape the preference structures of applicants (Lievens & Slaughter, 2016). In this study, we focus on the country development level (i.e. developed versus emerging). The contingency perspective suggests that applicants in emerging countries are embedded in different economic and institutional environments from those in developed countries, so they may attend to different signals (Dögl & Holtbrügge, 2014). Indeed, Connelly et al. (2011) called for more research on the collective beliefs about a signal. Responding to this call, we add the role of national context (developed versus emerging economies) to the receiving end of signaling theory.

Hypotheses

Country-of-origin of the company

The object of interest in our study are MNEs that recruit new employees in foreign markets, i.e. outside of their home market. Researchers have found that an MNE’s COO influences local applicants’ perceptions about its attractiveness as an employer and, consequently, their job choices (Held & Bader, 2018; Hong & Kim, 2017). Due to limited information about foreign MNEs and bounded rationality in host countries, COO acts as a heuristic base for judgment and as a simplifier encompassing other organizational aspects (Newburry, 2012). Specifically, COO development levels may dominate the impact of a specific COO. When knowledge is limited, applicants from developed countries may incline to lump all emerging countries together, and applicants from emerging countries, on the other hand, may incline to lump all developed countries together. Since EMNEs suffer from double disadvantages of liabilities of foreignness plus liabilities of emergingness, they are at a disadvantage compared to DMNEs (Zhang et al., 2020). In the context of recruitment, the signal sent by EMNEs from emerging countries, which are ranked lower in the country hierarchy and characterized by prevalent institutional voids, lower levels of industrialization, less advanced technology, and late liberalization, evokes a less impressive, lower-status, and less competent image during applicants’ screening process (Held & Bader, 2018; Holtbrügge & Kreppel, 2015), thereby failing to address applicants’ social adjustment concerns. Thus, applicants are less inclined to take a job at EMNEs, giving rise to the following hypothesis:

Hypothesis 1: COO influences job choice, with EMNEs being less preferred than DMNEs.

CSR

Since CSR is fundamentally a multidimensional construct, we disentangle the effects of economic, social, and environmental CSR. We expect that, as interrelated but distinct CSR components (Elkington, 1998), these three dimensions will exert unique influence on applicants. Unlike the other two dimensions, the economic dimension is not purely symbolic. Scholars have acknowledged the challenge of distinguishing between instrumental and symbolic attributes as they can intertwine and impact each other (Dittmar, 1992; Schuitema et al., 2013). While the economic dimension can indeed have symbolic elements when it signifies the firm’s commitment to ethical conduct, it can also have instrumental aspects by contributing to long-term employment prospects (Guerci et al., 2015). It signals the firm’s competence, financial stability, and is associated with an impressive image. This dimension addresses applicants’ social adjustment concerns and appeals to those who value status.

While the differences between social and environment dimensions seem trivial, Jones et al. (2014) studied the signaling mechanisms of two CSR practices and suggested that community involvement, as relatively less helpful for business, actually exert a stronger pride signal than pro-environmental practices. The social dimension signals a firm’s engagement in social causes, reflecting its benevolence and creating respectable impression. This dimension addresses applicants’ need to express their pro-social values.

On the other hand, the environmental dimension signals a company’s dedication to environmental stewardship. It generates a respectable impression and resonates with environmentally conscious applicants. Moreover, we anticipate that compared to social CSR, environmental CSR, fueled by advancements in green technologies and products (Dögl & Holtbrügge, 2014), can demonstrate a firm’s innovation and forward-thinking mindset, thereby holding potential to go beyond creating a respectable impression and instead leave an impressive image.

Overall, each dimension sends different signals and tap into different applicants’ social-identity concerns. While we expect and intend to investigate the differing effects, in the absence of both a consensus and a theoretical argument to posit which of them is more important than the others (Capelle-Blancard & Petit, 2017; Elkington, 1998), instead of one general hypothesis, three sub-hypotheses are postulated as follows:

Hypothesis 2: The (a) economic, (b) social and (c) environmental dimensions of CSR each have a positive impact on job choice.

The moderating role of national context (developed vs. emerging economies)

The signaling process is incompletely examined without consideration of the receiver of the signal—local applicants in different countries and the contingency perspective indicates the effectiveness of HR practices depends on the contextual factors, e.g. economics, institutions, and culture (Cooke et al., 2019). While MNEs strive for global standardization to capture the benefits from economies of scale, local adaptation are necessary due to cultural, economic, and institutional differences (Rosenzweig, 2006). It is inconclusive whether MNEs should use a standardized or an adapted approach to attract global talent. The question of whether to use a standardized or adapted approach to attract talent globally remains inconclusive (Baum & Kabst, 2013b).

Regarding the developed/emerging host country taxonomy, emerging countries differ from developed countries in economic and institutional conditions (Dögl & Holtbrügge, 2014; Zhang et al., 2020). Deresky and Christopher (2015) indicated that a country’s living conditions have a significant effect on the incentive preferences of individuals. In developed markets, where post-materialistic values hold significant relevance, symbolic attributes like COO and social and environmental CSR may exert stronger signaling effects. On the other hand, in emerging markets characterized by lower income and living standards, instrumental attributes such as financial incentives and career prospects have been found to be important (Caligiuri et al., 2010). Although real wages have risen rapidly in emerging countries, an immense gap between developed and emerging economies remains (ILO, 2020). This substantial economic gap prioritizes economic concerns and immediate benefits, overshadowing the value of MNE’s COO and social and environmental CSR signals in emerging markets.

Institutional conditions also contribute to the differences in social and environmental CSR valuation. While in developed countries, regulations and laws in favor of CSR have long been established, this subject has only recently gained attention in emerging economies (Dögl & Holtbrügge, 2014). Furthermore, emerging countries are characterized by institutional voids that necessitate companies to establish their own commitments and credibility (Khanna & Palepu, 2010).

Notably, the economic CSR stands apart from the others as it encompasses instrumental aspects, contributing to long-term employment prospects. In emerging markets, with a high prevalence of temporary and informal workers, who often face low job security and limited legal protection, the economic CSR becomes a powerful signal (Ohnsorge & Yu, 2022). In contrast, developed markets typically have more stringent employment protection legislation (Khanna & Palepu, 2010), reducing the need for the economic dimension as a prominent signal. Overall, the differences in economic and institutional conditions between emerging and developed markets shape the signaling dynamics and the perceived relevance of these attributes in each respective market.

Moreover, signaling theory postulates that individuals will rely less on the available signals, particularly from COO, in their decision-making process when they are familiar with a foreign company (Han, 1989). Since applicants from emerging markets are usually more familiar with EMNEs than those from developed markets, reducing their reliance on COO signals. The mere-exposure effect literature further shows that familiarity triggers positive reactions (Zajonc, 1968). Turban et al. (2001) also documented the positive effect of familiarity on a firm’s attractiveness as an employer. This logic, in line with signaling theory, implies that applicants from emerging markets know EMNEs better and are more open to evaluating them, which can help lessen the magnitude of liabilities of emergingness. Thus, we hypothesize that (Figure 1):

Figure 1. Research model.

Hypothesis 3: The relative importance of COO in job choice is moderated by the national context, such that COO is less important in emerging than in developed countries.

Hypothesis 4: The relative importance of the (a) economic, (b) social and (c) environmental dimensions of CSR in job choice is moderated by the national context. In emerging countries, the economic dimension is more important, while the social and environmental dimensions are less important than in developed countries.

Methodology

Developing an attribute list

To identify important attributes besides COO and CSR to include in our conjoint analysis, we first conducted a search of the relevant literature. We systematically combined keywords denoting concepts related to employer branding (e.g. employer attractiveness, organizational attractiveness, applicant attraction, recruitment, job choice) with keywords denoting international recruitment (e.g. international companies, foreign companies, cross-cultural, cross-national). We extracted the top attributes from previous studies (Table 1). Two studies revealed work climate to be the most important, and one study each showed salary, level of centralization, career advancement opportunities, and task attractiveness to be the most important.

Next, we conducted two pilot studies to validate our attribute list. First, we carried out 20 semi-structured interviews among potential US and Vietnamese job-seekers. The interviews followed a flexible structure, allowing for in-depth exploration of participants’ preference structures. Questions focused on their expectations and preferences regarding factors that are particular important in their job choice. The results revealed that salary, task attractiveness, work climate, career advancement opportunities and person-organization value fit were mentioned the most frequently, which partly confirmed our previous attribute selection. Second, we examined whether actual job postings contained information on the selected attributes. We analyzed the content of 50 randomly selected job postings of MNEs from major job boards, namely LinkedIn and CareerBuilder. Statements about work climate were present in 66% of the job postings, followed by information about COO (58%) and about at least one CSR dimension (50%). The postings provided less information about advancement opportunities (44%), task attractiveness (34%) and salary (32%). Information about level of centralization was mentioned in only 12% of the job postings. These validation steps resulted in a final list of eight salient attributes that applicants are likely to get informed about in real job ads: salary, work climate, career advancement opportunities, task attractiveness, COO and the three dimensions of CSR.

Conjoint analysis

Having identified the most relevant job attributes, we applied conjoint analysis to investigate their value and relative importance in employer choice decisions. The conjoint analysis, with its decompositional approach, where the attribute values are decomposed from the overall evaluations of different attributes in combination, can truly calibrate the relative importance of a set of attributes (Baum & Kabst, 2013a; Ronda et al., 2021). Conjoint analysis has been intensively used in marketing to study consumer preference and trade-offs among multiple attributes of a product or service. Recently, it has also been applied to analyze job preferences among applicants (Baum & Kabst, 2013a; Montgomery & Ramus, 2011; Ronda et al., 2021).

For the specificity of our study, we chose choice-based conjoint analysis (CBC), using stated choices (Rao, 2014). First, choice tasks mimic the real behavior more closely than profile ratings, leading to a greater external validity. Second, respondents can make each choice independently without worrying about rating scale consistency over profiles. Third, choice-based methods predict choice probability directly with easily reported estimation results (Elrod et al., 1992; Moore, 2004). Fourth, adaptive conjoin analysis (ACA) which is often used for larger sets of attributes, has been shown to have a limitation in pricing research because it tends to understate the importance of price (Orme, 2013). In the recruitment context, the price (or, salary) is an important attribute and the trade-off between salary and COO as well as CSR is relevant for this study, which further implies the use of CBC.

Sample

We collect data in Vietnam and the US, which allows us to compare applicants’ perceptions of emerging and developed countries. We distinguish between developed and emerging labor markets because previous studies find evidence that individual incentive preferences may differ between countries with different economic and institutional conditions (Deresky & Christopher, 2015; Dögl & Holtbrügge, 2014). Therefore, the country choice was based on the fact that the USA is the largest economy in the world with strong institutions. Vietnam, on the other hand, is one of the fastest-growing emerging economies with severe institutional voids and challenges (Pham & Dang, 2022). These countries are also considered attractive destinations for foreign investment (Jelili, 2020) and hold divergent ideological and cultural backgrounds, which represents an appropriate context for studying and comparing work value systems across countries.

Vietnamese and US undergraduate students are the main focus of the study because they constitute the majority of the future qualified workforce for MNEs operating in these two countries. Also, the range of potential starting salaries can be identified, allowing us to derive willingness to forego financial benefits for CSR from the conjoint analysis from public data. Reviewing former studies, we find that convenience sampling among students is common in research on applicant attraction and job preference (Alkire, 2014; Baum & Kabst, 2013a; Held & Bader, 2018). The survey language was English, being the lingua franca for MNEs, so describing company/job profiles of MNEs in English enhances the validity of our experiment. English is, in addition, compulsory at Vietnamese schools from the secondary level onward (Phuong & Nhu, 2015). Moreover, the use of choice instead of scales make the study more resistant to potential language effects occurring in cross-national research (Harzing et al., 2009, 2021).

The survey was conducted in July–September 2021 at seven universities and via a crowdsourcing platform. The crowdsourcing platform is suitable for our study because a considerable proportion of crowdworkers are young and currently studying at university (ILO, 2018, p. 36). To control for the quality of responses, we implemented filter and validity-check questions and time limit options in the survey, following the best practices suggested by Cobanoglu et al. (2021). The resulting sample comprised n = 1,023 students (Vietnam n = 513, US n = 510), 47.5% of whom were male (Vietnam 48.1%; US 46.9%). The average age was 24.59 (Vietnam 23.36; US 25.8), 72.3% stated that they would finish their studies within one or two years (Vietnam 78.6%; US 66.1%) and 82.6% were seeking a job (Vietnam 76.8%; US 88.4%). To check the representativeness of our samples (particularly for gender distribution), we drew on national statistics. The results (% of male in the national statistics of Vietnam and the US is 53 and 43%, respectively) indicate that the demographic characteristics of our sample are comparable with those of the general population of students in the observed countries.

Furthermore, to test for internal consistency, we included a dominated option within fixed choice tasks that has at least one attribute at a worse level and no attribute at a better level compared to the alternative, in line with Bryan and Parry (2002). If a participant selected this dominated option, it is reasonable to infer that the participant showed a lack of engagement and, consequently, made random choices. We excluded all responses that failed the test by selecting the non-dominant alternative, resulting in a final sample of n = 907.

Measurement

To develop measures for the attributes, we consulted previous studies (Baum & Kabst, 2013a, pp. 1399–1400; Montgomery & Ramus, 2011, pp. 23–24), followed practical guides on conjoint analysis (Ofek & Toubia, 2014) and interviewed HR experts working for MNEs in the respective countries. We used two levels for each attribute to avoid too many profile cards and, consequently, participant overloading, except for salary and the three CSR dimensions, for which we used three levels each.

We operationalized COO as the location of corporate headquarters to avoid confusion when determining COO due to the increasing globalization of the world economy. According to UNCTAD (2016), the UK, Germany, China and India are the most promising investors from developed and emerging economies, thus, these four countries represent a relevant selection of MNEs’ home countries. A pretest was conducted to ensure that these two developed and emerging countries are, in the context of recruitment, homogeneous within and heterogeneous between. Therefore, by randomly varying the country within one group, we reduce idiosyncratic country influence beyond the distinction in developed vs. emerging.

As one of our primary focuses is the CSR attribute, we highlighted three levels for each CSR dimension: good reputation, neutral reputation and poor reputation. We adapted the descriptions of Evans and Davis (2011, pp. 475–476) to clearly depict the relevant CSR challenges faced by MNEs.

The descriptions of the attribute levels seen by participants from both countries were identical, except for salary, which was adapted to the income levels in the USA and Vietnam, respectively. Four HR experts thoroughly reviewed and confirmed the realism and appropriateness of all the level descriptions, ensuring that they adequately portrayed each attribute’s different levels. We also pre-tested our list of attribute levels on 180 participants from Vietnam and the US to further refine the descriptions and ensure there were no ambiguous level descriptors. Specifically, we asked participants to rate all the attribute levels in a 7-point Likert scale ranging from ‘not attractive at all’ to ‘very attractive’. One-way ANOVA and t-tests showed that there are significant differences in the perceived attractiveness between attribute levels (p < 0.05). The attributes and their respective levels in the study are presented in Table 2.

Table 2. Attributes and descriptions of attribute levels.

	Attributes	Attribute levels
1	COO of the company	Developed market (UK, Germany) Emerging market (China, India)
2	Salary^a (fixed starting salary in the first year of employment)	VN	US
		10,300,000 VND pm 12,000,000 VND pm 13,700,000 VND pm	$49,000 p.a. $57,000 p.a. $65,000 p.a.
3	Career advancement opportunities – opportunities for promotion, development, qualification and further education	Fast – accelerated career track to senior positions Slow – no promotion for the first 3 years
4	Work climate – work attitude of boss, colleagues, and other members of the organization	Colleagues are friendly and supportive Colleagues are distant and competitive
5	Task attractiveness – how interesting the work is, and amount of routine work involved in your work	Very interesting, diversified work Mostly unvarying routine work
6	Economic sustainability – organization’s growing potential in terms of revenue, earnings, and the number of employees; good track record of paying taxes regularly; and not engaging in any unethical practices such as questionable payments and bribery	Profitable and growing business and good reputation with no record of lawsuits, investigations or fines Profitable but stagnant business and no particular reputation for business practices Business is losing money. The company faced several lawsuits and recently paid a large fine concerning a major bribery scandal.
7	Social sustainability – organization’s reputation for fair treatment and pay of suppliers and of workers throughout the organization’s supply chain	Good reputation for treatment of suppliers and of company’s workers in all production countries No particular reputation for treatment of suppliers and of company’s workers Poor reputation for treatment of suppliers and of company’s workers in many production countries
8	Environmental sustainability – organization’s reputation for commitment to environmentally sustainable development by minimizing the impacts of its products and operations.	Good reputation concerning environmentally sustainable development No particular reputation concerning environmentally sustainable development Poor reputation concerning environmentally sustainable development

^a Following the approach of Baum and Kabst (2013a), the salary figures represent the lower quartile, median and upper quartile of starting salaries among business graduates in 2020 (e.g. National Association of Colleges and Employers (NACE) salary survey; www.adecco.com.vn). The reported figures were inflated to 2021 and adjusted slightly after we interviewed the HR experts and conducted the pre-test experiment.

Procedures

We developed an online survey using Sawtooth Software, a company specializing in conjoint survey tools. The number of choice tasks per participant was set to 12; each task required a choice between three randomly generated profiles which were randomly selected from 1,296 possible combinations of attributes and factor levels (2⁴ × 3⁴), based on nearly-orthogonal and level-balanced designs. A sample choice task is presented in the Appendix A. The questionnaire was generated automatically for each participant, in this case out of a pool of 300 different versions with 12 choice tasks each. Based on our attribute list and sample size, this setting was chosen to achieve optimal balance between not overloading participants and achieving a manageable amount of attribute level overlap for robust estimations at the individual level (Orme, 2019). We also tested the design by using the test design facility integrated into the software. Of the 12 choice sets, two were fixed holdout tasks used solely for cross-validation, following Orme (2019). Each attribute was carefully described at the beginning of the choice task to avoid ambiguity. We instructed the participants that any non-described characteristic, such as working schedule, type of work and location, should be considered to be identical across the three profiles to eliminate potential effects of extraneous factors on their job choices. In addition, attribute order was varied across subjects, allowing us to control order effects while maintaining consistency in attribute display within subjects to minimize confusion. Afterward, additional information about individuals’ personal characteristics (e.g. age, gender, education level, international experience, familiarity with COO) was gathered.

After the survey was conducted, choice models were developed to estimate the utility values for each attribute that produced the maximum likelihood fit to the actual choices. Choice models were traditionally estimated only at the aggregate respondent level by pooling across all individuals. However, with the introduction of the Hierarchical Bayes (HB) estimation, heterogeneity can also be accommodated at the individual respondent level (Hein et al., 2020). HB’s strength is its ability to estimate individual-level utilities given only a few choices from each individual by borrowing information from the sample population and determining the optimal degree to which the upper- and lower-level models influence individual estimates. Thus, we used HB for our final model. In our final model, we computed the hit rate and the mean absolute value of the difference between the actual and simulated shares for our two holdout tasks that were not used to develop the model. The average hit rate across respondents is 65.05%, close to the hit rates for previous CBC studies (SawtoothSoftware, 2021). Table 3 also shows that the simulated choice shares are very similar to the actual choice shares (mean absolute error = 3.52), supporting the validity of our model.

Table 3. Validity check.

	Profiles	Actual choice shares (%)	Simulated choice shares (%)	Absolute error
Holdout task 1	1	55.5	55.4	0.1
	2	44.4	40.4	4
	3	0	4.2	4.2
Holdout task 2	1	50.3	49.5	0.8
	2	49.7	44.1	5.6
	3	0	6.4	6.4

Results

Hypotheses tests

We relied on the relative importance weights for each attribute to test our hypotheses. The overall results are first presented to test Hypotheses 1 and 2. Then, the results for each country are shown for Hypotheses 3 and 4.

Table 4 reports the importance weights for each attribute across the 907 respondents. For perspective, the last column shows the importance of each attribute relative to salary. Further, attributes are combined into different groups, in which the importance weights between attributes in different groups are statistically significantly different from each other, but only insignificantly different from other attributes within a group (p < 0.05).

Table 4. Average and relative importance weights.

	Attribute and rank of attribute^a		Importance weight	Confidence interval 95%	Relative importance
Total (n = 907)	Group 1	1. Salary	17.03	[16.33; 17.74]	1.000
		2. COO	16.95	[16.08; 17.83]	0.995
		3. Economic dimension of CSR	16.24	[15.62; 16.86]	0.953
	Group 2	4. Environmental dimension of CSR	12.03	[11.60; 12.47]	0.706
	Group 3	5. Social dimension of CSR	11.06	[10.67; 11.45]	0.649
		6. Advancement opportunities	10.72	[10.26; 11.18]	0.629
	Group 4	7. Work climate	8.92	[8.50; 9.35]	0.524
	Group 5	8. Task attractiveness	7.04	[6.63; 7.45]	0.413
US (n = 445)	Group 1	1. Salary	21.98	[20.51; 23.45]	1.000
	Group 2	2. COO	15.48	[14.34; 16.62]	0.704
	Group 3	3. Environmental dimension of CSR	13.90	[13.14; 14.25]	0.632
		4. Economic dimension of CSR	13.29	[12.55; 14.04]	0.605
		5. Social dimension of CSR	12.41	[11.75; 13.06]	0.565
	Group 4	6. Advancement opportunities	8.98	[8.33; 9.62]	0.408
	Group 5	7. Task attractiveness	7.09	[6.51; 7.66]	0.323
		8. Work climate	6.88	[6.37; 7.40]	0.313
Vietnam (n = 462)	Group 1	1. Economic dimension of CSR	17.95	[16.97; 18.92]	1.000
		2. COO	16.69	[15.44; 17.94]	0.930
	Group 2	3. Salary	14.58	[13.48; 15.39]	0.813
	Group 3	4. Advancement opportunities	11.52	[10.84; 12.21]	0.642
		5. Social dimension of CSR	10.74	[10.18; 11.31]	0.599
		6. Environmental dimension of CSR	10.63	[10.10; 11.16]	0.592
		7. Work climate	10.48	[9.79; 11.18]	0.584
	Group 4	8. Task attractiveness	7.40	[6.81; 7.98]	0.412

^a Attributes are ordered in descending order of importance weights.

Attributes are combined into different groups, in which the importance weights between attributes in different groups are statistically significantly different from each other, but insignificantly different from other attributes within a group (p < 0.05).

COO is ranked second with a relative importance of 0.995 for salary, with DMNEs being preferred over EMNEs (U_developed = 56.16; U_emerging = −56.16). Thus, Hypothesis 1 is supported. The economic dimension of CSR is ranked third and has 95% of the importance of salary, followed by the environmental and social dimensions. As expected, for all dimensions, good CSR is preferred over neutral and poor CSR (Economic: U_good = 35.17; U_neutral = 7.12; U_poor = −42.28; Environmental: U_good = 24.89; U_neutral = 2.38; U_poor = −27.28; Social: U_good = 28.31; U_neutral = −2.91; U_poor = −25.39). Therefore, Hypotheses 2a–c are supported. Salary, COO, and the economic dimension of CSR are essentially equal in importance and they are statistically and substantively more important than environmental and social dimensions of CSR.

Table 4 also reports the importance weights for each country. It becomes apparent that the national context moderates the relative attribute importance. To statistically test the moderating effects, we considered the importance weights for each respondent as the dependent variable and the country (Vietnam = 1) as the independent variable in a set of regression analyses (Table 5). We found no significant effect from the national context on the perceived value of COO. Consequently, Hypothesis 3 is not supported. Nevertheless, despite the insignificant difference between the countries for this variable, it is surprising to observe that COO is ranked higher and significantly more important than salary in Vietnam. There are, however, significant effects from the national context on the perceived value of the three CSR dimensions (economic: β = 0.102, p < 0.01; social: β = −0.066, p < 0.05; environmental: β = −0.107, p < 0.01). This shows that applicants from emerging markets, here Vietnam, attach more value to the economic dimension of CSR but less to the social and environmental dimensions. Thus, Hypotheses 4a, 4b and 4c are supported. As a robustness check, we conducted t-tests to compare the importance weights between the respondents from Vietnam and the US (Table 6). The t-test results converged with the regression analysis.

Table 5. Regression analysis.

Independent variable	National context (Vietnam = 1)
Dependent variables	Value of COO	Value of economic CSR		Value of social CSR		Value of environ­mental CSR		Value of salary		Value of advance­ment opportu­nities	Value of work climate		Value of task attractive­ness
B	−0.024	0.102**		−0.066*		−0.107**		−0.121***		0.054	0.074*		0.008
F	0.510	9.465		3.906		10.551		13.536		2.654	5.001		0.058
Equation R-square	0.001	0.010		0.004		0.012		0.015		0.003	0.005		0.000

Note: n = 907. National context was coded as 0 = US, 1 = Vietnam. Bold is to signify the significant values.

*p<.05; **p<.01; ***p < 0.001.

Table 6. T-tests.

	Value of COO		Value of economic CSR		Value of social CSR		Value of environ­mental CSR		Value of salary		Value of advance­ment opportunities		Value of work climate		Value of task attractive­ness
Mean differences (Vietnam vs. US)	−0.016		0.019**		−0.008*		−0.014**		−0.026***		0.029^+		0.029*		0.005

⁺ p<.10; *p<.05.; **p<.01; ***p < 0.001. Bold is to signify significant values.

Supplementary analysis

Table 7 presents the correlation matrix between the attributes. There are significant negative correlations between salary and other attributes, namely COO and the CSR dimensions. It is reasonable to infer that for the potential candidates, salary serves as a compensatory rather than complementary factor for COO and CSR. On this basis, equity theory (Adams, 1965), which addresses how individuals evaluate social exchange relationships, seems to apply when applicants choose among job offers (Baum & Kabst, 2013a). Our results imply that applicants use salary to resolve perceived inequity between job offers. In other words, applicants may be willing to accept a lower salary to work for a firm from a developed (versus emerging) country or with good (versus neutral) CSR.

Table 7. Correlation matrix between the conjoint analysis attributes.

	COO	Salary	Advancement opportunities	Work climate	Task attractive­ness	Economic dimension of CSR	Social dimension of CSR	Environ­mental dimension of CSR
COO	1
Salary	−0.075*	1
Advancement opportunities	−0.041	0.076*	1
Work climate	−0.050	−0.015	−0.009	1
Task attractiveness	0.017	−0.064^+	0.011	−0.032	1
Economic dimension of CSR	−0.149***	−0.363***	−0.036	0.010	−0.029	1
Social dimension of CSR	−0.082*	−0.223***	0.026	−0.025	−0.014	0.032	1
Environmental dimension of CSR	−0.048	−0.231***	−0.051	0.007	−0.031	0.001	0.046	1

⁺ p<.10; *p<.05; ***p < 0.001.

Using the conjoint market simulator simulating a thousand competitive job offerings via a sampling of scenario approach, we can attach monetary value to these factors. The market simulator is useful for converting raw conjoint (utility) data into managerially useful data, such as simulated choices or shares of preference/choice. Specifically, competitive profiles can be introduced within a simulated market scenario; then the simulator is run for each respondent to predict their choice according to their utilities with the assumption that the profile with the highest total utility is chosen and finally computes the percentage of respondents projected to choose each profile (also referred to as a share of preference).

Basically, the method to derive willingness to pay involves first simulating a job choice for the less favorable version of a job offer (e.g. emerging COO, neutral CSR) against a set of other competitive offerings. Next, the job offer is enhanced (e.g. developed COO, good CSR) and the algorithm tries to decrease the salary to drive the share of preference back down for the enhanced job offer. The reduction in salary that leads to equality in preference shares between enhanced and base versions of job offers is the willingness to pay. We repeated the simulation multiple times using 300 bootstrap samples with 30 competitive scenarios per bootstrap sample to estimate confidence intervals for willingness to pay for CSR and COO (Orme, 2021).

The results are displayed in Figure 2 for the US sample and in Figure 3 for the Vietnamese sample. In particular, American applicants would be willing to forgo 15.6% of their salary to work for DMNEs (versus EMNEs) (31.8% for Vietnamese applicants). This is also illustrated by our first fixed holdout task, in which the profiles are identical, except for the differences in terms of COO and salary: developed COO is combined with the medium salary in profile 1, while emerging COO is combined with a higher salary level in profile 2. Despite the 14% difference in salary, DMNEs (55.5% probability of being chosen) were still evaluated more favorably than EMNEs (44.4% probability).

Figure 2. Willingness to pay in terms of a reduced salary ($) for US sample.

Figure 3. Willingness to pay in terms of a reduced salary (thousand VND) for Vietnamese sample. Note: The figures denote the mean willingness to pay along with the lower and upper bound of the 95% confidence interval, using bootstrap sampling with 300 samples and 30 competitive sets per sample. Mean willingness to pay as % of expected salary in parentheses.

Concerning CSR, American applicants would be willing to forgo about 3.6, 3.3 and 5% (10.7, 10.1 and 3.8% for Vietnam) of their salary in return for good (versus neutral) economic, social and environmental CSR, respectively.

Furthermore, there are significant negative correlations between COO and the economic and social CSR dimensions. It can be inferred that good CSR may help EMNEs to close the gap with DMNEs regarding applicant attraction. Thus, we also used a market simulator to explore this interaction. First, we simulated job choice with two identical profiles, excluding COO. Then, we changed the profile of EMNEs to include good CSR and ran the simulation again. Table 8 shows that good CSR helps EMNEs to strongly reduce the gap in the share of preference with DMNEs. For US respondents, the preference for DMNEs over EMNEs is reduced from 30.2 to 9.4%; for Vietnamese respondents, it is reduced from 46 to 8%.

Table 8. Market simulation for share of preference.

US
	Simulation 1 (neutral CSR) (%)	Simulation 2 (good CSR) (%)
EMNE	34.9	45.3
DMNE	65.1	54.7
Difference	30.2	9.4
Vietnam
EMNE	27.0	46.0
DMNE	73.0	54.0
Difference	46.0	8.0

Discussion

We find that COO and CSR are important factors to attract applicants. Specifically, our findings indicate that EMNEs are confronted with a significant HR challenge when attracting talent not only in developed markets but also in emerging countries. This is consistent with lists of the ‘100 Best Companies to Work for’ in Vietnam and in the US (Fortune, 2020; VnExpress, 2020), which are dominated by DMNEs. In Vietnam, only 9 EMNEs appear in the top 100; in the US, not a single EMNE appears on the list. Previous research has documented the recruiting challenge facing EMNEs in developed markets (Held & Bader, 2018; Holtbrügge & Kreppel, 2015), but less is known about EMNEs in emerging markets. Pertinently, our simulations and holdout task indicate that offering a much higher salary (at least 15% in the US and 31% in Vietnam) is needed to keep EMNEs competitive with their counterparts from developed markets in the war for talent. This effect was also demonstrated in our holdout task.

Concerning CSR, our conjoint results show that each of the three dimensions is important. Particularly, both American and Vietnamese applicants would be willing to forgo a part of their salary in return for good (versus neutral) economic, social and environmental CSR. Moreover, we find a significant interaction between COO and CSR; good CSR indeed helps EMNEs close the gap to DMNEs.

We reveal cross-national differences in attribute valuation, in line with Baum and Kabst (2013b). In particular, the national context influences the importance of the three CSR dimensions, but not the importance of COO. However, despite insignificant differences in COO importance, it is interesting to observe that in Vietnam, COO (ranked 2nd) is more important than salary (ranked 3rd). Unexpectedly, EMNEs suffer more from the liability of emergingness in emerging countries. The mere-exposure literature claims that familiarity can lead to greater liking. On the other hand is the argument that familiarity can breed contempt, i.e. the more experience individuals have with an object, the more likely they will encounter evidence of dislike, which might result in less liking (Norton et al., 2007). This confrontation signifies that the familiarity-perception relationship may be contingent upon context (Zhang et al., 2020) and that cross-national differences in attribute valuation cannot be explained solely by economic and institutional condition differences. National context can also encompass aspects of culture (Cooke et al., 2019). Self-concept theory could be integrated to explain this finding. Markus and Kitayama (1991) distinguish the independent and interdependent construal of self. While the former is dominant in Western cultures, with the belief that distinct individuals are inherently separate, the latter is more common in Asian cultures with the idea that identity also lies in one’s social, familial and professional relationships. For those with interdependent selves, public perceptions and social roles are central to their identity, leading to the Asian focus on the ‘face’, which was conceptualized by Goffman (1955, p. 213) as ‘the positive social value a person effectively claims for himself by the line others assume he has taken during a particular contact. Face is an image of self-delineated in terms of approved social attributes’. Though it is a universal concept, face is particularly important in collectivist cultures of Asia and is claimed to significantly regulate individual behavior (Kim & Nam, 1998). Researchers have developed the concept of face and face consumption to explain Asia being one of the biggest markets for luxury goods from the West despite the population’s relatively low income (Biondi, 2020; Li & Su, 2007). Face consumption could be transferred to the recruiting context. Due to the importance of ‘face’, Vietnamese applicants place more emphasis on publicly visible factors, especially MNEs’ COO. Indeed, Woodard et al. (2016) find evidence that employer prestige linked to an MNE’s COO is a more visible indicator of success than salary and position. In other words, because of the prestige and impressive image that applicants ascribe to Western companies, working for DMNEs might make applicants feel they have ‘face’ that they would lack if working for EMNEs.

On the one hand, as expected, applicants from Vietnam (an emerging market) attach less value to the social and environmental dimensions, but on the other hand, due to high prevalence of informal and temporary workers in emerging markets, they put more emphasis on the economic dimension of CSR compared to applicants from the US (a developed market). This is because the economic dimension is not purely a socially responsible dimension but contains an egoistic dimension linked to long-term job security (Guerci et al., 2015). For instance, in Vietnam, non-standard and temporary workers make up nearly 70% of employment, whereas in the US, the figure is 3.5% (ILO, 2016).

Implications and limitations

Theoretical implications

We applied conjoint analysis in a cross-national comparison to examine how different employer image attributes impact job choices and whether these impacts vary across countries. Subsequently, this study contributes to the literature in the following ways. First, we contribute to signaling theory by highlighting the significance of MNEs’ COO and CSR as important signals in job choice decisions. Specifically, our findings demonstrate that these signals play a role in shaping applicants’ job choice, alongside other salient workplace aspects. Although a growing number of studies examine the attractiveness of COO and CSR for applicants, many rely solely on perception-based evaluation of employer attractiveness in quasi-experimental designs with compositional approach and, consequently, fail to capture the actual job choice event and calibrate the relative importance of these symbolic attributes when evaluated with other aspects. This study enriches the understanding of applicants’ job-choice process by applying a novel decompositional approach, which more accurately reflects the process in which applicants compare job offers with one another holistically and make trade-off decisions in real life. While we confirm the orthodoxy that salary, the most basic instrumental attribute, always plays an important role, other attributes with symbolic images, particularly MNEs’ COO and CSR, are gaining importance. Moreover, our study advances the understanding of MNE’s COO and CSR signaling by demonstrating that these symbolic attributes not only influence applicant attraction, but also matter in later phase of job choice. The existing evidence highlights the temporal aspect of MNEs’ COO and CSR signaling and their enduring impact on applicants’ evaluations and choices throughout the job choice process. This implies that MNEs’ COO and CSR signals should not be considered as standalone aspects but rather as integral components that shape applicant perceptions at different stages. Applicants may engage in ongoing information search and evaluation, continuously reassessing the attractiveness of organizations, in line with literature on sequential information processing (Hogarth & Einhorn, 1992). Therefore, to inform theory in the signaling effects of both instrumental and symbolic factors, it is necessary for researchers to examine recruiting as a multi-stage, sequential process.

Second, we enhance the understanding of signaling theory by demonstrating that different CSR dimensions have different signaling effects. While prior research has predominantly focused on examining CSR’s overall effects or the social and environmental dimensions, our research highlights the underexplored economic dimension. While the three dimensions are interrelated in the conceptualization, we find empirical evidence that each dimension indeed exerts a distinctive influence on job choice as their important weights are not significantly correlated (Table 7). This raises an intriguing possibility: the three CSR aspects influence job choice through heterogeneous underlying signaling mechanisms, beyond merely respectable impression. Additionally, we shed light on the role of social identity, suggesting that it may not be as crucial in shaping job choice based on the environmental and social dimensions as it is for the economic dimension. Applicants, especially those from emerging markets, are more likely to prioritize economically sustainable companies because it aligns with their identity as successful individuals. Thus, integrating different CSR dimensions into one overall scale or relying upon any one dimension as a surrogate for the others may be flawed, which lends support to Zhao et al. (2022) call for a more fine-grained conceptualization of CSR to better understand its multifaceted effects on recruitment-related outcomes.

Third, the study expands the horizons of individual-focused CSR research and contributes to the international HRM literature by providing preliminary evidence suggesting an interplay between MNEs’ COO as a signal for impressiveness and CSR as a signal for respectability on job choice. Our results show that while good social and environmental CSR (ranked 5^th and 6^th on the importance scale) alone may not strongly drive job choice of applicants from emerging markets, they have a notable effect for EMNEs, in helping them to close the gap with DMNEs. This implies that impressions of impressiveness and respectability are interrelated. These impressions also map on two fundamental dimensions underlying inferences in social cognition, namely competence and warmth (Fiske et al., 2007). There is, moreover, evidence that warmth impression has halo effects, spilling over into competence (Fiske et al., 2007), explaining why CSR can diminish the magnitude of liabilities of emergingness. Furthermore, the relevance of CSR for EMNEs and its potential to bridge the gap with DMNEs can also be interpreted in the light of the literature on new venture employer attractiveness, as both operate in highly uncertain environments and encounter difficulties in establishing legitimacy (Moser et al., 2017; Navis & Glynn, 2011). New ventures and EMNEs alike need to effectively signal their legitimacy as trustworthy, credible, and reliable employers to attract applicants. Future research can contribute the literature on the liability of origin, especially the liability of emergingness by aligning it with literature on new venture employer attractiveness. This integration of findings from the new venture context can enhance our understanding of the mechanisms through which EMNEs can enhance their legitimacy and overcome their country-of-origin constraints.

Finally, by shifting the focus from individual values to the national differences, we expand Highhouse et al. (2007)’s symbolic attraction mechanism and show that preferences for job/organizational attributes are contingent upon national context encompassing aspects of socio-economics, institutional environments, and culture. This also contributes to the international standardization versus adaptation discussion. Our study reveals that national differences lead to different valuations for some employer image attributes. This finding lends support to Baum and Kabst (2013b) middle-ground contingency approach: some instrumental facets (e.g. work-life comfort, salary) require cross-national adaptation while others (e.g. advancement opportunities) allow for a higher degree of standardization. We advance this statement by demonstrating that this approach is true not only for instrumental but also symbolic attributes. Therefore, aligning cross-national studies with a nuanced view of different instrumental and symbolic attributes could be pertinent for future international recruitment.

Practical implications

Our findings provide several managerial implications. To attract highly qualified applicants from diverse cultures, managers of MNEs need to understand the antecedents of applicants’ choice and to coordinate their employer branding activities on an international scale to optimize the cost-benefit ratio. Despite the persistence of certain cross-national differences, the overall influential structure on job choice remains largely similar across countries. Thus, MNEs can adopt a universal positioning to strengthen the brand image and capture the benefits of global economies of scale and scope. In particular, a combination of COO, salary, CSR and advancement opportunities can attract talent.

MNEs cannot change their COO, but they can develop suitable recruitment strategies acknowledging its effect. For instance, DMNEs should emphasize their positive country-of-origin image in international recruitment because COO is important in both developed and emerging markets and applicants are willing to forego a substantial portion of their salary to work for DMNEs. In contrast, for EMNEs to overcome the constraints associated with their COO and attract top talent, they should prioritize offering attractive advancement opportunities and actively engage in CSR initiatives, which serves as signals of their legitimacy as trustworthy and reliable employers.

Salary and advancement opportunities are, unsurprisingly, important attributes. However, symbolic factors, such as CSR, which stand a better chance of distinguishing a firm within the same industry, are as important. Developing a positive CSR reputation is a strategic choice as CSR has been found to enhance person-organization fit, which significantly improves job performance and commitment (Jones et al., 2014; Kim & Park, 2011). Although our findings indicate that a high salary can compensate for poor CSR, companies may run the risk of hiring ill-fitting applicants, leading to high turnover in the long term.

Furthermore, it is crucial for managers to recognize that CSR encompasses distinct dimensions appealing to different individuals, necessitating separate communication of each dimension while considering resource limitations. By gaining insights into the specific impacts of each dimension, managers can make more informed decisions on resource allocation and effectively address the specific concerns of target applicants. While academics and HR consultants emphasize social and environmental aspects, firms should not neglect communicating the economic dimension of CSR. Applicants are likely to be more concerned with practices contributing to the firm’s long-term survival and profitability, especially when experiencing economic difficulties such as the global recession and the COVID-19 pandemic. In particular, for MNEs operating in emerging markets where much of the labor force is temporarily employed, emphasizing job security via economic CSR is highly recommended. While economic CSR should be the main focus of communication, the other two dimensions should not be neglected but can be communicated with less intensity.

Last but not least, on the evidence of lasting impact of COO and CSR on applicants’ decision-making, MNEs need to carefully manage and strategically leverage their COO and CSR signals to convey a consistent image throughout the entire recruitment process.

Limitations and future research

The paper has certain limitations. First, although the choice of conjoint analysis method is one of the study’s strengths, the design faces challenges pertaining to external validity. Specifically, the scenarios presented to participants are designed in tabular form, potentially resulting in a low level of immersion. Future studies should aim to foster a high level of immersion by closely aligning the experimental and real-world settings, thereby enhancing external validity. Notably, technological advances have offered avenues for achieving this aim by changing the way scenarios are presented. For instance, with Virtual Reality technology, researchers can replicate the job-seeking process, enabling participants to navigate through a virtual job market (Aguinis & Bradley, 2014). Exposing respondents to realistic recruitment websites (e.g. Le & Morschett, 2023) would also increase external validity. In the present study, this method was not applied because the tabular form was necessary to conjointly analyze and compare a rather large number of job choice criteria, but the method could be used to re-assess a limited set of criteria and the results of the present study.

Second, job choice has been shown to be influenced by further factors, such as work-life balance or person-organization fit. However, due to the chosen methodology, we were unable to include a larger number of attributes. The number of required full-profile evaluations in the conjoint analysis grows exponentially with the number of attributes. Future studies could examine more or other attributes. Moreover, while our approach allowed a more accurate measures of applicants’ trade-offs, we have not identified the noncompensatory attributes of job choice, based on a non-compensatory violation of fit perspective (Held & Bader, 2018). Future research could explore this issue and enrich our understanding of the job-choice process by distinguishing between the attributes that serve as knock-out criteria and those that can be compensated.

Third, we only covered four MNE COOs. Although these countries are relevant and interesting for scholars and practitioners, future research should cover a larger number for more representative results. The Next-Eleven countries, for example, may hold interest because their companies are also expanding into developed and other emerging countries. While prior research has focused primarily on the major emerging economies, such as BRIC, as the home countries of EMNEs, attention should now be shifted to the Next-Eleven countries (Cooke et al., 2019; Held & Bader, 2018).

Fourth, we only collected data on applicants’ choices in one emerging country (Vietnam) and one developed country (US), which imposes certain limitations on our capacity to distinguish the effects of culture and institutional differences from each other and from other contextual factors (e.g. educational, social, political situations, etc.). It would be fruitful to replicate the study across a broader range of national contexts. When data on more countries with divergent national institutions, cultures and economic development are available, it would be possible to explore which economic/institutional factors and cultural dimensions underlie the differences in potential applicants’ valuation of employer image facets. Another limitation regarding the sampling is that we used a convenience sample of students, which may bring the generalizability of our results into question. Nevertheless, it does not diminish the importance of our contribution as students constitute the majority of the qualified workforce for companies and companies are targeting their recruitment efforts at making themselves appealing to young students (Jones et al., 2014).

Finally, our survey was distributed during the global COVID-19 crisis. The current economic downturn, coupled with large-scale social restrictions and mass layoffs (Laborde et al., 2020), might have influenced the importance of economic sustainability, particularly in emerging markets. While the relative importance of some traditional attributes is likely to remain stable over time, new attributes may prevail in the future. Longitudinal studies, moreover, could lend insights into how the role of a MNE’s COO changes in the future, especially when EMNEs establish themselves in the global markets and become better known by applicants worldwide.

Acknowledgement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data of this study are available from the corresponding author, upon request.

Notes

1 The authors thank an anonymous reviewer for the advice to integrate this literature stream.

Appendix A.

Sample choice task