Leaving home : Relocation choices of South African multinational enterprise ( MNEs )

As multinational enterprise (MNEs) from emerging markets grow into maturity they face significant challenges surrounding their identities and home country bases and, as a result, we see them making different choices. National champions choose to retain their corporate headquarters and identity from their emerging market homes while corporate emigrants relocate their head-offices abroad. Targets of mergers and acquisitions (M and A) may move to the acquirer's home or stay where they were born. Through an examination of South African-born firms that are cross-listed on both the Johannesburg securities exchange and one other exchange, we identify the differences between these evolutionary paths and discuss the drivers and constraints on relocation decisions.


INTRODUCTION
A shifting balance of economic power towards emerging markets has been evident for some time as the world's economic centre of gravity has moved East and South, away from organization for economic cooperation and development (OECD) countries and towards emerging economies (OECD, 2010).More recently, we have begun to witness the growing economic power of businesses coming from these emerging markets.As such, the location advantage of developed markets and the firms within their borders is changing.In 2005, 34 of the fortune global 500 companies were from emerging markets.Only four years later, this number had risen to 73 (Fortune, 2009).Even this increase, however, understates the true nature of the transition that is underway.Mature MNEs from developed markets have been able to retain their dominant positions through the acquisition of emerging market firms and this is a partial explanation for the limited representation of emerging market multinational enterprises (EMNEs) in the global 500 ranking.A more important reason may be the relocation of EMNEs themselves, from their emerging market homes to developed countries.This latter phenomenon has yet to receive much attention from researchers.
EMNEs have become key actors in foreign direct investment and cross-border acquisitions too (Guillén and García-Canal, 2009).However, there is no scheme or taxonomy for describing the strategy of 'infant MNEs' as they embark on internationalisation; this case falls between Corresponding author E-mail: sklein@uvic.caTel: 250-721-6420 Author agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License the cracks (Ramamurti, 2009).Extant research has focused on the location choices of MNEs in aggregate and has related this to individual factors, such as taxation (Barrios et al., 2008;M. Desai and Hines, 2002;Devereux and Maffini, 2006;Voget, 2008).While the phenomenon of EMNEs moving their headquarters to developed countries has been observed, current research has not considered the combination of factors that motivate or constrain emerging market firms in their headquarters location choices (Bel and Fageda, 2008;Birkinshaw et al., 2006;Braunerhjelm, 2004;Brouwer et al., 2004).EMNEs do not bear the costs of relocation without reason, and the underlying logic of location advantage for headquarters in developed markets requires verification.Such an examination is especially timely given the vigorous and continuing debate surrounding the reasons for EMNE relocation (McNulty, 2010).
Although emerging markets are not homogeneous, they share some locational disadvantages, such as weak institutional environments, property rights regimes, legal systems, and others (Guillén and García-Canal, 2009).Differences between EMNEs themselves are also notable.Their origins, industries, competitive advantages, markets and internationalisation paths vary widely (Ramamurti, 2009).While many EMNEs develop firm specific abilities to compensate for their home-country location disadvantages (Guillén and García-Canal, 2009), others relocate their headquarters to developed countries, presumably in the pursuit of location-based advantages.
As the significance of cross-border trade grows, and firms internationalise, EMNEs may relocate to Industrialised countries for asset-seeking and market-seeking reasons; for example, Anglo American moved from South Africa to Britain and Mittal Steel moved from India to the Netherlands.Relocation may be a strategic necessity in order to allow greater access to capital and other resources; as typified by South African Breweries relocating to London "seeking access to capital markets better endowed than those at home" (Hoover's, 2010).
Countries, or cities, compete to attract MNEs and extract rents from headquarters located within their borders, creating "a developing market for international headquarters" UNCTAD (2003a).Some firms, however, retain their indigenous headquarters, grow very large and continue to manage global operations from outside "the triad" (Rugman, 2008), despite the supposed disadvantages of their locations, firms such as Petronas of Malaysia or document type conference paper authors Vladimir Alvarado (PDVSA) of Venezuela have stayed at home.These EMNEs are often supported by their home country governments and are often natural resource specialists.
The purpose of this paper is to better understand the phenomenon of firm relocation as it applies to emerging market firms.We describe the relocation decisions made by South African firms in light of prior research and assess the applicability of such work in an emerging market context.We outline the different expansion paths taken by EMNEs in an era of globalization and review the locational advantages of emerging versus developed markets.Based on these advantages, we discuss and categorise the predictors of MNE behaviour and highlight the likely implications for firms from emerging markets.We then examine these implications in the context of South African firms, and draw conclusions from the analysis.

Internationalization paths
As an MNE progresses from "infancy" to "adolescence" to "maturity", the consequences of its birthplace change.The importance of home-country advantage declines as an MNE evolves, regardless of its nationality (Ramamurti,2009).As such, the country of origin may have obsolescing relevance for an MNE over time, in terms of value chain elements, supply of senior management, capital supply or relative revenue.What may become more important is the country in which the MNE chooses to live.
In figure 1 below, three possible expansion paths for an EMNE are identified as globalization takes hold in its home country: relocate, remain at home or be acquired.The firms that pursue such options are labelled as corporate emigrants, national champions and targets, respectively.We speculate that different paths, as depicted, will have different growth trajectories 1 .Becoming a target may be associated with slower growth, while becoming a corporate emigrant may be a cause (based on managerial ambition) or a consequence (due to access to a larger global market) of more accelerated growth than would be the case for national champions.The "corporate emigrant" is conceptualised as a firm that relocates its headquarters to obtain location specific advantages while customer facing business units remain in place.The "national champion" is a firm that does not relocate and bears the costs, benefits, of this decision.The "target" is acquired by another MNE, which results in an effective transfer of headquarters functions.The target could then be seen as an "outside-in" firm that locates its headquarters and operations separately -managing from the "outside", with operations "in" the country.
The EMNE may choose to relocate its head office to a developed country in order to take advantage of the economies of agglomeration, such as access to physical and human capital (Dunning, 1998) as well due to increasing accountability to international stakeholders (Birkinshaw et al., 2006), while the customer facing business units remain elsewhere.
Three other investment types may explain the "outsidein" phenomenon (Dunning and Lundan, 2008): 1 This question of growth trajectories, and its empirical verification, is deserving of further study, but is outside the scope of this paper.The preliminary question, that is discussed here, is how these different options describe firm behaviour in practice.1. Escape Investment seeks to avoid disadvantageous conditions in the home country.These conditions may be heavy taxation, lack of economic dynamism or the unacceptability of the business type in question.Thus, EMNE relocation to developed countries may be motivated by the desire to disengage from disadvantageous conditions at home. 2. Support Investment seeks to augment the capabilities or activities of the firm.Thus, EMNE relocation to developed countries may be motivated by the need to substantiate previous investment.3. Passive Investment is akin to portfolio investing.
Here, a minority stake may be purchased in an existing firm or asset and the emphasis is not necessarily on the management of the investee.This form of investment does not add to the understanding of EMNE relocation.
In the case of EMNEs, relocation may also be driven by managers pursuing their own, personal objectives, even if they are not well aligned with those of the organization.Such an "agency" problem may reflect individual desires to escape an unpleasant environment or to enjoy a better quality of life which may be available at home.The agency problem, however, could also work the other way and limit the organization's mobility if key executives do not wish to leave the comforts of their home.

Location drivers and constraints
EMNEs have disrupted and heightened competitive markets, forcing incumbent, largely Western, MNEs to respond with product innovation, consolidation and reconfiguration of value chains (Guillén and García-Canal, 2009;Ramamurti, 2009).In doing so, EMNEs have become key actors in foreign direct investment and cross-border acquisitions.It is critical that their decisions obtain greater scrutiny.A comparison between traditional and MNEs and new EMNEs, is shown in Table 1 (Guillén and García-Canal, 2009).Four general motivations for the foreign investment of MNEs are natural resource seeking, market seeking, efficiency seeking and strategic asset seeking (Dunning and Lundan, 2008).Natural resource seekers look for resources abundant to a region.These resources may be physical such as mineral deposits and are typically location-bound.Alternatively, there may be abundant human resources ina location, such as inexpensive labour or skills -technical, managerial or marketing.Corporate emigrants look for skilled human resources, for example in management or marketing, in their relocation decisions.
Market seekers invest to supply goods or services to markets that have been serviced previously by exports from the investing country (Dunning and Lundan, 2008).These firms may be following the relocation of production by suppliers or customers; which may need local adaption of their products; may be taking advantage of reduced transportation costs; or may be following a defensive or aggressive competitive strategy.Corporate emigrants may be motivated by the need to market more intensively to customers in those countries.Efficiency seekers aim to benefit from economies of scale and scope and of risk diversification (Dunning and Lundan, 2008).Corporate Emigrants' relocation to developed countries may be motivated by the need to rationalise previous investments in those countries, or to allocate corporate, rather than business unit level, activity to a Low, because of their ingrained structure and culture more suitable location.Strategic asset seekers invest in line with a long term strategy, typically to secure long-term competitiveness.These investments seek to augment previous commitments and existing asset bases, or to exclude ownership advantages to other firms.These investments may not be strictly profitable in the sense required in the other investment motives described above (Dunning and Lundan, 2008).Corporate emigrants may be motivated by a belief that the new region will be increasingly significant in the future.As EMNEs' competitive advantages are "weak" and they need to upgrade their resources, EMNEs are motivated to acquire some of the same firm and country specific capabilities held by their more traditional competitors.If these advantages are concentrated in geographic areas and are freely available to all firms in that area, the firm will be motivated to relocate to a region rich in these assets.
If emerging markets are unsupportive of headquarters, and MNEs located in developed markets have location advantages, EMNEs will be motivated to relocate themselves.Such motivation assumes that any loss of country specific advantage from the home country will be recovered in the move, and that firms are sufficiently mature such that firm specific advantages, example, an "adversity advantage" (Ramamurti, 2009), are fully transportable.Firms that do not relocate, for whatever reason, carry the costs of their disadvantaged locations.If a firm does not relocate, and remains globally competitive, this implies that it possesses country or firm specific advantages of greater magnitude than the disadvantages that accrue from its headquarters location.The more disadvantaged a particular location, the more firms would be expected to relocate from that country.With economic growth, the rate of relocation should slow as the disparity reduces, assuming a global status quo in terms of capital and personal transportability.
Predating recent advances in communication and service technologies, location research until the 1990's was based on the presumption that management and production functions were co-located (Deschryvere, 2009).As such, no distinction was made between headquarters and production relocations.More recently, the corporate headquarters has been visualised as a collection of thee divisible functions: Financial, Legal and Managerial (Desai, 2009).Here, the processes and products are distinct for individual business units and production centres.Each of these functions has specific motivations for its location choice, as illustrated below.Due to reductions in communications and travel costs, "firms are redefining their homes by unbundling their headquarters functions and reallocating them opportunistically across nations.…and, consequently, the idea of firms as national actors rooted in their home countries is rapidly becoming outdated" (Desai, 2009: 409), as depicted in Figure 2 (below).
Given the role and needs of each of the three homes, different drivers for competitiveness push and pull EMNEs in their headquarters relocation decisions.Similarly, there are constraints on them that resist the drive to relocate.

Resource and efficiency seeking drivers
Given that the most basic input and product of the headquarters is information, the optimisation of information transfer is a key for competitiveness.Researching European and American headquarters relocations respectively, Bel and Fageda (2008) and Strauss-Kahn and Vives (2009) agree that transport infrastructure and the costs of tacit information exchanges are important for the location of headquarters of large companies.Headquarters also move in order to optimise value chain elements, particularly business services.Ono (2003) demonstrated the link between location and the inexpensive procurement of services such as advertising, accounting and legal services.In the United States, headquarters location decisions are largely driven by the presence of a large and varied local supply of business services rather than the presence of a large number of headquarters (Pennings and Sleuwaegen, 2000;Davis and Henderson, 2008; In reality, neither communication and transport connectedness nor the presence of a large and varied supply of business services are geographically universal.To remain competitive, the MNE must move to the location that offers the best advantages for both connectedness and value chain optimisation; the greater the relative disadvantage, the greater the motivation to relocate.Since part of the definition of an emerging market is a weakness in this support environment (FTSE Group, 2009), EMNEs would generally be more motivated to relocate abroad than firms from the developed world.If emerging markets have smaller pools of the skilled labour required for MNE corporate management, an impetus exists for EMNEs to relocate to the developed world in order to easily procure this resource.

Product and capital market seeking drivers
In Europe, Mucchielli and Saucier (1997) concluded that new product introduction is a cause of headquarters relocation.Many studies have shown that proximity to customers and the size of the product market in a host country are significant predictors of location choice (Head and Mayer, 2004;Pennings and Sleuwaegen, 2000;Strauss-Kahn and Vives, 2009).Birkinshaw,et al.(2006: 682) recognise that "it is now accepted that proximity to specialised labour, complementary suppliers and customers, and access to knowledge spillovers are all important benefits to the firm." Apart from proximity benefits, there may be an added marketing benefit or country-of-origin effect; that is, customers perceive the quality of a product as stemming from the company's location.Swiss watches are an obvious example.As emerging markets mature, they must overcome negative perceptions of their country of origin.For example, Toyota struggled in the 1960's to establish the creditability of Japan as a centre of automobile production; a situation paralleled by Korean automobile manufacturers in the 1990's.MNEs based in developed economies would not feel this motivation as strongly.
A similar "legitimacy effect" has been noted in the case of investors.MNEs improve their visibility and relationships with shareholders and financial institutions in a progressive pattern (Birkinshaw et al., 2006).They may start with the issuance of foreign depositary receipts, continue through to overseas listings, and finally end in a relocation of the corporate office to a global financial centre.This progression has been promoted by EMNEs as a demonstration of their commitment to global capital markets.The reward may be in terms of borrowing costs, stock liquidity and the value of corporate governance.Other benefits include analyst coverage, price discovery, disclosure regulations and investor protection (Desai, Financial  ).The act of moving, as a signal to markets, may be more important than the benefits of the relocation itself (Birkinshaw et al., 2006).
MNEs list their securities in the developed world to access investor capital.For example, news corporation relocated from Australia to the United States in 2004 to access more readily American investors that might, better appreciate media companies (Desai, 2009).Corporate headquarters are moved to get closer to important external influencers, primarily shareholders and financial market players (Birkinshaw et al., 2006).For EMNEs, generally from less well-endowed capital markets, this offers an even more powerful draw.
In situations of merger or acquisition, where there is a major change of shareholding, relocation is necessary to regain proximity to influencers.It has been shown that headquarters are often relocated following an increase of overseas share ownership or following a merger or takeover (Brouwer et al., 2004;Strauss-Kahn and Vives, 2009).In most cases, the location of the acquirer is chosen as the location for the united firm (Baaij et al. 2004).
Financial markets with a better reputation and a reduced risk perception, can demand a greater premium for their shares.Thus, the relocation of primary listing to a capital market in a lower risk country will increase the perceived value of the company (Mohamed and Finnoff, 2005).

Institutional resource seeking drivers
Key institutional drivers of location are tax incentives and labour institutions (Brouwer et al., 2004;Birkinshaw, et al., 2006;Strauss-Kahn and Vives (2009).Mooij and Ederveen (2001) found that 1% increase in host-country tax rate decreases foreign direct investment (FDI) in that country by 3.3%.Further, firm taxation has a marked impact on the choice of corporate location, in terms of both "push" and "pull" (Devereux and Maffini, 2006;Egger, 2009;Voget, 2008).In addition, the taxation of individual employees influences location choice, and does so increasingly as a firm internationalises (Braunerhjelm, 2004).Separate from the location of their corporate headquarters, firms have been found to relocate their nominal legal location away from the United States to escape taxation (Desai and Hines, 2002).Parent-country taxation is a predictor of the pattern of MNE expansion (Barrios et al., 2008).
The strength, and rigidity of the available labour pool is another important location factor (Birkinshaw et al., 2006).Labour market rigidity puts "a brake" on a host country's attractiveness, even within OECD countries (Delbecque et al., 2008).A less frequently discussed location factor is the lobbying and institutional power of the home country government in the international arena.Governments negotiate advantageous terms for their indigenous firms and these become country specific advantages."Tier-l bargaining between the governments of host and home countries occurs bilaterally or through multilateral institutions" (Ramamurti, 2001: 23).State bargaining power on behalf of national business, is labelled as tier-2 bargaining.Multilateral institutions may include such bodies as the World Bank, the international monetary fund (IMF) and the world trade organisation (WTO).These institutions write the macro-level rules on FDI that frame micro negotiations between the MNE and potential subsidiary host countries.Given that emerging market governments are seen to have less "voice" in such institutions (U.S.Department of State, 2009), there exists a driver to secure this advantage through relocation.
MNEs may also be motivated to relocate based on greater protection under law, including intellectual property, which may be offered by the host country (Desai, 2009), since emerging markets have a mixed record in legal enforcement and transparency (Transparency International, 2009).

Agency constraints
Agency concerns regarding the actions of corporate managers have been widely discussed during the recent global financial crisis.The effects of personal, rather than corporate, requirements may determine location choice (Braunerhjelm, 2004).Dominant shareholders, for example, may affect the decision making process to their parochial ends.Birkinshaw et al. (2006), in a developed country context, found that "the more concentrated the ownership of the MNE (in terms of the percent shareholding of the largest shareholder), the lower the likelihood of corporate headquarters or business unit headquarters moving overseas"(p.689).
In many emerging markets, fast-rising EMNEs have a large percentage of state ownership.Their governments may be motivated to preserve local jobs and resist relocation.It follows that the higher a firm's state ownership, where the state is an important external influencer, the lower the likelihood of relocation of the headquarters abroad.Alternatively, a dominant private shareholder may push for relocation to a nation seen to be more desirable.Here, capital flight or "escape investment" may apply (Dunning and Lundan, 2008).In a merger or acquisition, relocation may result because of either the removal or the introduction of different shareholder interests.Similarly, where managers seek personal relocation to a country with a higher quality of life, a concentration in private shareholding may motivate relocation from emerging markets.

Firm constraints
Firms and industries vary in many ways, such as their level of technology, their stage in the industry life cycle, the extent of global product and process standardisation, human capital requirements and capital demand.As such, some firms and activities appear better suited to emerging market production (Ramamurti, 2009).It does not follow, however, that companies are better suited to having their headquarters located in emerging markets by virtue of their industry.Trans nationality will be associated with relocation, either as a predictor, or as a motivation, for relocation.A firm's revenues, assets or employment may be concentrated geographically, even if reach a global scale.In such cases, it makes little sense to relocate the headquarters outside of that region since doing so would result in an increase in distance related transaction costs.If production or sales are predominantly in one area, that may be the best place to be settled.For example, the Tata group's corporate headquarters is likely to remain in Mumbai for the foreseeable future.This may be due to the mature industries which still dominant the production of the diverse group, primarily steel and automobiles.Alternatively, this may be because of the dominance of the Indian market, or both reasons together (Tata Group, 2010).

Summary of drivers and constraints
Driving and constraining factors for MNE headquarters location selection are summarized in Table 2 below.The generalised strength of each factor in emerging versus developed markets is shown.
As shown, emerging markets are less advantageous locations for corporate headquarters.EMNEs are motivated to relocate to developed markets in order to seek out location advantages in a new country of residence.Location choice will be driven particularly by financial market considerations, that is, to the countries housing financial markets with the greatest capital endowment and best reputations.EMNEs not only have more motivation to relocate, they also have a high rate of movement.While 6% of all sampled MNEs in one study relocated across national boundaries between 1997 and 2007, those from emerging markets had a relocation rate of 50% (Voget, 2008).Firm relocation within the United Sates was only 5% a year between 1996and 2001(Strauss-Kahnand Vives, 2009).If firms do not relocate, their decision may be explained by constraining agency effects, such as concentrated or state shareholding, or by a lack of relevance of international operations.Other agency problems, centred around management's desire for quality of life and low personal taxation, are national factors that may also constrain headquarter location choice (Desai, 2009).Based on the above, the following propositions may be advanced: P1: Emerging markets offer less locational advantage than developed markets for EMNE headquarters.

METHODOLOGY
To investigate the relationships identified above, a study was conducted on firms from South Africa.Using a single emerging market has advantages, which it controls for home country differences.South Africa is a good context to examine relocation effects as the international expansion of firms was artificially suppressed until the early 1990s, due to international sanctions against the Apartheid regime.Since the transition in 1994, South African forms have moved aggressively into global markets, with many relocating their headquarters.In 1994, no South African firm was among the 50 largest transnational corporations from developing economies, ranked by foreign assets (UNCTAD, 1996).
In 1997, there were three, and by 2001, there were five (UNCTAD, 1999;UNCTAD, 2003b).South Africa has experienced dramatic political adjustments in the last decades as well as discord between government and business (Klein and Wöcke, 2009) with a flight of human and financial capital recorded (Mohamed and Finnoff, 2005).As a result, South African MNEs have been motivated to relocate their headquarters to more advantageous locations.There has also been notable internal debate in South Africa over the motivations of corporate relocation McNulty, 2010).Finally, reporting standards in South Africa are high, even against global standards, allowing for transparency of shareholding (World economic forum, 2010).

Data sources
Data obtained from the Johannesburg securities exchange (JSE) were supplemented by other databases: International institute for management development (IMD) world competitiveness yearbooks, international finance corporation (2010), IMF (2009), the Zephyr database of Bureau van Dijk, (2010), McGregor BFA (2010), JSE (2010) and FTSE Group (2010).For differences in locational advantage, the latest available data in the IMD database were used.All available scores were selected between 2000 and 2010 and aggregates were used for analysis.The variables used as proxies for the concepts to be investigated, as well as the data sources, are tabulated below.(Table 3.) To investigate location choice, the latest available data in the McGregor and Zephyr databases were extracted.The variables used as proxies for the concepts to be investigated, as well as the data sources are tabulated below.(Table 4.) The oldest historical data available were chosen to mitigate any changes that have occurred since relocation.255 transactions were found to match the set criteria.Of these, only 109 transactions occurred between firms based in developed and emerging markets.For location prediction, the additional variables used as proxies for the concepts to be investigated, as well as the data sources are tabulated below.(Table 5.).The study frame included all 74 EMNEs cross-listed on the JSE and any other exchange over the period 2004 to 2010.Of these, only those registered before 2004 were selected for further analysis.The logic behind focusing only on cross-listed companies is that they have already moved to access a larger or different pool of capital and that relocation of their

Direct state ownership
The average percentage share ownership held directly by the home government for all years in which data was available, weighted 60% to 2006.
McGregor database

Total state ownership
The average total percentage share ownership held directly or indirectly by the home government for all years in which data was available, weighted 60% to 2006.

Foreign interest
Assets Transnationality: ratio of foreign assets to total assets a) Sales Transnationality: ratio of foreign sales to total sales Averaged for all years in which data was available headquarters is unlikely to be attributable to purely capital concerns.Data on private shareholder concentration was not available due to deficiencies in the data sources.Shareholding data were available from 2006 onwards.Where shareholder or transnationality data were missing, the company was excluded from the analysis.A total of 61 firms were available for further examination, of which 10 are corporate emigrants, 20 are outside-in firms and 31 are national champions.

Locational advantage (P1)
As a first step in examining relocation, it is necessary to verify that the locational factors for South Africa actually motivate firms to leave home.Mean scores for the various factors of locational advantage were compared across developed countries and emerging markets as a group as well as to South Africa in particular.An analysis of variance (ANOVA) was carried out; the Wilcoxon/ Kruskal-Wallis signed-rank, non-parametric test was used as the data were non-normally distributed.The results are summarised in Table 6 below.
As can be seen, "tier-2 bargaining" state power is 61% greater in developed markets, but the difference is statistically significant only at the p <0.10 level.Real personal taxes are very similar for both groups of countries.All of the other variables show significantly higher scores for developed markets.The scores for South Africa's "banking and financial service" and "real personal taxes" were 7.04 and 5.13 respectively, in both cases above the developed market mean score.For all other variables, the South African score was lower than the mean for developed markets, indicating that relocational pressures exist for South African EMNEs.An exploratory multivariate analysis of variance was conducted to compare the overall vector of means difference between developed and emerging markets.The results of this analysis are tabulated in table 7 below.
As can be seen, the two country types are significantly different, with the mean score for developed countries being higher, providing empirical support for the proposition that emerging markets offer less locational advantage than developed markets for MNE headquarters.

Location choices (P2)
The sample was then reviewed for known location choices.Those EMNEs that had made alternative location choices were catalogued as either corporate emigrants or outside-in firms.That is, those who had chosen to relocate corporate headquarters from South Africa or those who had chosen to place their headquarters in a country other than that of their major operations.Details of firms described as corporate emigrants and their destination of relocation are shown in the appendix (Table A2).All of the South African corporate emigrants chose to move to developed markets in either Europe or North America as would be expected from their locational advantages over other emerging markets.Details of firms described as outside-in, with the locations of their primary listing and primary operations are shown in the appendix (Table A3).Following the methodology of the London stock exchange, country of operation was taken to be the most significant geographical location for revenues or assets.For firms not listed in London, country of operation was derived from publically available data such as annual reports and firm websites.All but one of the outside-in firms chose to locate their headquarters in a developed market, despite their primary operations being in an emerging market.The curious nature of these firms is demonstrated well in the example of inter-phase-slip algorithm (IPSA).Despite having almost all sales and assets vested in one plant in South Africa, as well as future prospects being in South Africa, their head office is located in London.

Location predictors (P3)
Having shown that locational advantages of developed markets exist, we now turn to the question of why some firms do not relocate.Details of firms described as national champions are shown in the appendix (Table A4).National champions are compared to corporate emigrants in terms of their state ownership and transnationality of assets and sales in Table 8.
As can be seen, state ownership measures were 41% and 51% higher for national champions versus corporate emigrants, but these are not statistically significant at a p < 0.10 level.All transnationality variables show significantly higher values for corporate emigrants.2Overall, our results provide empirical support for the proposition that EMNE headquarters location are constrained by ownership factors and encouraged by levels of foreign business interest.

Discussion
Based on the factors chosen, there is significant evidence that emerging markets offer less locational advantage than developed markets for EMNE headquarters.EMNE headquarters in developed markets: 1. Are better able to communicate across internal and external boundaries, improving performance; 2. Are better supported by the supplier network desired by a headquarters office; 3. Have greater access to competent senior managers, required for business growth; 4. Have access to greater pools of equity capital, required to fuel business growth; 5. Benefit from lower country risk perception with investors; 6. Have access to cheaper pools of capital, required to fuel business growth; 7. Have access to better educational support, required to inform business growth; 8. Have economical legal and regulatory support, required for competitiveness; 9. Have greater protection of intellectual property, required for innovation; and, 10.Support access to a better personal quality of life for managers; and.11.Enjoy the benefits of greater weight in multi-lateral institutions.
All of the ten corporate emigrants in this study chose to move to developed markets in either Europe or North America.All but one of the twenty outside-in firms chose to relocate their headquarters to developed countries, choosing to bear higher distance-related transaction costs arising from their primary country of operation being in an emerging market.Both of these facts support an argument that developed markets offer greater locational advantages than emerging markets to EMNE headquarters.South African MNEs that did not choose to relocate were constrained by low levels of foreign interest.There is some evidence, but not very clear, that they may also have been constrained by higher levels of state ownership.There was evidence that relocating firms had businesses that were significantly more transnational than those that did not.Transnationality was not correlated with state ownership.
Our finding that transnationality is associated with headquarters location choice appears to be at variance with that of Birkinshaw et al. (2006) who suggest that business unit, not corporate; headquarters relocate when there is a large portion of sales and manufacturing activities overseas.This apparent contradiction may be present if increasing transnationality follows relocation (McNulty, 2010); transnationality here was measured after relocation and thus is more likely to be a result than a cause of relocation.In addition, while having a somewhat transnational business is necessary before a firm considers relocation, it may not be the reason for relocation.An ambition for transnationality may be a motivation for the move in the first place.
As data for the concentration of private ownership were not available, it was not possible to compare these results with those of Birkinshaw et al. (2006) who found that concentrated private ownership was likely to constrain relocation.We suggest that their finding would not hold true in South Africa or other emerging markets given the agency effect implied in the lifestyle quality offered in developed markets.Rugman andVerbeke (1992, 2001) show how firm specific advantages may emerge from multiple sources.These may be from the home country office, a foreign subsidiary or across a MNE's network.The relocation of the headquarters to a global financial centre may be one way in which advantages are developed.Those that accuse emerging market MNEs of a lack of patriotism or label relocation as capital flight should acknowledge the competitive necessity of firms seeking the same country specific advantages as their developed market rivals.

Conclusion
A model describing the internationalisation of EMNEs was developed, describing three possible trajectories.The "Corporate Emigrant" is conceptualised as a firm that relocates its headquarters to obtain location specific advantages, while business unit offices remain elsewhere.
The "national champion" is seen as a firm that does not relocate and bears the costs, or benefits, of this decision.The "target" is acquired by another MNE, which results in an effective transfer of headquarters functions.A further variation is the foreign "outside-in" firm.Here, a foreign firm locates its headquarters and operations separatelymanaging from "outside", with operations "in" country.
There is significant evidence that emerging markets offer less locational advantage than developed markets (Appendix 1) to EMNEs.It is clear that South African multinational enterprises relocated to developed markets, not to other emerging markets.EMNEs that did not choose to relocate were constrained by low levels of foreign interest, but the state did not necessarily restrict or impose transnationality.
The results imply that if emerging market nations seek to remain attractive to EMNE headquarters and the high value-added employment that they offer, they have significant ground to cover to improve their attractiveness on the factors measured.Specifically, the factors that remain under the control of the host state, the institutional and agency factors, need to be focused upon.In 2010, the South African treasury recognised that "The current regulatory framework has tax and exchange control aspects which are inhibitive to international headquarter companies seeking to leverage South Africa's infrastructure and skills base as a means of investing in the rest of the continent.... headquarters companies will be allowed to raise and deploy capital offshore without exchange control approval" (National Treasury -South Africa, 2010).Measures like this are typical kind required to retain EMNE headquarters.
Given the complexity of the location decision, shareholders must be watchful that managers will not motivate a particular choice for personal reasons.It has been demonstrated that developed markets offer better "quality of life" but the lifestyle of managers should not be allowed to compromise firm competiveness.

Limitations and directions for future research
While presenting intriguing findings, this study has some limitations.The sample size was very small with only ten corporate emigrants, and substantiation of the consequences of the three growth paths is required.We assumed location is equivalent to the nation of primary stock market listing, which may be an over-simplification (Desai, 2009).Further, the analysis does not compare the factor strengths of the nations chosen or prioritise them.As it is often the case, the variables chosen may also not faithfully characterise the factors as experienced by all EMNEs, implying a possible representation error.
Regarding state power, it could be argued that the influence of each European country was underrepresented, as, in terms of trade, the European Union (EU) is a single large block.For example, Luxembourg's total trade is small, but through its alignment with the EU, its overall influence may be seen as large.We also recognize that it is often regions, not nations, which attract headquarters.It is the square mile in the City of London that attracts financial firms as much as it is the country of Great Britain.This question of geographical boundary needs to be addressed more directly in future research.
It is recommended that future research should be broadened to include more countries, especially given the reality that the rate of relocation from South Africa has been high in recent decades.Emerging markets are diverse, and the case of South Africa may not be representative.Corporate headquarters relocations may be predicted by levels of state ownership and levels of foreign business interest elsewhere.Verification of this would add greatly to the understanding of MNE internationalisation and the impact of an emerging market origin.If country level variation is so great that this pattern is not replicated, further nuance could be brought to the understanding of the rise of EMNEs.Finally, clarity is required on the direction of causality between transnationality and relocation.

Table 1 .
A Comparison of "New" and "Traditional"  multinational enterprise (MNEs)

Table 2 .
Strength of location factors for corporate headquarters

Table 3 .
Concepts, variables and data sources for locational advantage

Table 4 .
Concepts, variables and data sources for location choice

Table 5 .
Concepts, variables and data sources for location predictors

Table 6 .
Summary of results for locational Advantage (P1)

Table 7 .
Result of means test for one-way MANOVA

Table 8 .
Summary of results for location predictors ** p< 0.01

Table A4 .
Contd.Following an acquisition, Highveld Steel and Vanadium is now known as Evraz Highveld Steel and Vanadium. Note.