Determinants of employment growth at MNEs: evidence from Egypt, India, South Africa and Vietnam.
Spillovers, in the form of technology transfer from a multinational enterprise (MNE) to local firms in the host country, are often proposed as an important rationale for establishing policy regimes conducive to foreign direct investment (FDI) in developing countries (Findlay, 1978; Borensztein et al., 1995). Technology transfer remains the cornerstone of the substantial literature on spillovers, though much of the evidence linking presence of MNEs to technological improvement of the local firms is indirect (Meyer, 2004). (1) Empirical studies have typically examined the link between the extent of foreign ownership in a domestic firm or foreign presence in a sector on the level or change in the multifactor productivity in that sector (eg, Aitken and Harrison, 1999). Evidence favouring a positive relationship have been explained in three ways: (a) demonstration effect of the MNEs' production processes on their local competitors, (b) improvement in the productivity of the firms that are in the supply chains of both the MNEs and the local firms, and (c) migration of labourers from MNEs to local firms either through employment or start-up entrepreneurial ventures.
Direct evidence is the strongest in favour of the impact coming from the migration of labourers from MNEs to their local competitors. For example, Hou and Gee (1993) and Tsiang and Wu (1985) have found evidence of technological spillovers from MNEs to local firms in Taiwan, South Korea, Hong Kong and Singapore. Similarly, Pack (2001) and Katz (1987) report evidence suggesting that local managers of MNEs often leave to start their own businesses, thereby facilitating technology transfer from the MNEs to local companies. This is also consistent with the recent experience of the Indian software industry (Commander, 2003). (2) The process of spillovers by way of recruitment of MNE personnel by local firms in emerging markets has been modelled theoretically (eg, Kaufman, 1997), and, not surprisingly, these models suggest that the extent of spillovers would depend on the (marginal) costs and benefits associated with the hiring of labourers by both MNEs and local firms.
The obvious questions, therefore, are why a MNE should transfer highquality technology from the parent firm to the developing country subsidiary, and why it should train its labour force efficiently to use this technology. These issues have been addressed in the theoretical literature on spillovers. Fosfuri et al. (2001) show that a MNE would transfer technology to a developing country subsidiary and train the local labourers only if the local firms do not directly compete with the MNE in the product market. Campbell and Vousden (2003) extend this line of argument and use a two-period model to argue that a MNE would limit the number of labourers that are trained in the use of the cutting edge technology, thereby increasing the skilled wage to the point where it is unprofitable for a local firm to employ these skilled labourers. Wang and Blomstrom (1992), however, posit that a MNE would seek to maintain its technological dominance over the local firms not by limiting technology transfers, but by progressively transferring better technology from the parent to the developing country subsidiary, thereby creating a virtuous circle of competition, spillovers and continual technological upgrade of both the MNE subsidiary and the local firms.
Systematic studies of the factors that impact on the growth of MNE subsidiaries and their decision to hire labourers, who are the conduit through which technology is transferred from the MNEs to their competitors, are very rare (Bloningen and Tomlin, 2001; Tan and Mahoney, 2005). To our knowledge no such studies exist as yet for emerging markets, so crucial issues for such economies, such as the impact of the policy environment and the institutional arrangements on employment creation, remain unexplored. Further, while there are a large number of studies that explore the determinants of firms' employment growth in general (see, eg, Hart, 2000; Fotopoulos and Louri, 2004), they do not take into consideration the strategic objectives of the firms that can be of paramount importance in the context of foreign direct investment by MNEs into emerging markets.
Our contribution is to examine the determinants of employment growth at MNEs, taking into consideration the strategic objectives of these firms, the impact of local conditions, and the companies' experience in those environments. Our analysis is based on a unique data set of 293 MNEs operating in four developing countries: Egypt, India, South Africa and Vietnam. These countries were chosen because they have relatively high levels of foreign direct investment among the cohort of non-Chinese emerging market economies, thereby facilitating the collection of a reasonable number of foreign subsidiaries through random sampling. Moreover, they differ significantly with respect to key independent variable in this study, namely, their policy and institutional environment, as well as their pool of skilled labourers,. We use this cross-country data set to estimate an employment function that links growth of employment at the MNEs to their characteristics, the nature of the parent MNEs control over the local subsidiary, the characteristics of product and factor markets in the local country, and local institutions.
Our results suggest that the extent of control that a MNE has over its affiliate, characteristics of the relevant industry in the host country, and institutional and regulatory environments of the host country are important determinants of its employment growth in emerging markets. This is an important finding for policy makers in emerging markets, who may be encouraging FDI to take forms, such as joint ventures, which fail to maximise employment growth and therefore potential spillover benefits.
The remainder of the paper is structured as follows: In the next section we briefly summarise the relevant aspects of the literature on the growth of firms, including employment growth. The data and variables are described in the forthcoming section and the regression specification and results are reported in the penultimate section. Finally concluding remarks are given.
EMPLOYMENT GROWTH AT MNES OPERATING IN EMERGING MARKETS
Firms would increase their workforce when the additional employees would add to the profitability of the operation. This may be the case if markets are growing and if complementary factors are available. The economics literature suggests that such growth opportunities would depend on the control over the host-country operation, the initial size of a firm and the availability of financial resources for expansion. In addition, more recent literature on MNEs, some of which was discussed in the previous section, indicates that factors like the quality of local institutions, familiarity of MNEs with the local business practices, as well as strategic objectives of these firms would also affect their expansion, and, thus, employment growth.
The literature therefore suggests that employment growth of MNEs operating in emerging markets is likely to be affected by the following factors:
Control over the host-country operation
A MNE can enter an emerging market either as a Greenfield operation, or by acquiring a local firm, or by setting up a .IV in partnership with a local firm. Although evidence about the impact of foreign ownership on firm growth is mixed (eg, Pfaffermayr and Bellak, 2000), more often than not, controlling equity share in the hands of foreign investors, namely, MNEs, has a positive influence on the profitability of firms in emerging market contexts (eg, Sabirianova et al., 2005). The literature concludes that a MNE would ideally want to enter an overseas market by way of a Greenfield project or acquisition of a local firm, both of which gives it full operational control over the local operation (see Estrin and Meyer, 2004).
However, it has been suggested that learning to operate in the foreign environment may be an important factor impacting growth of MNE operations in overseas markets (Blonigen and Tomlin, 2001). The literature on the choice of entry mode of MNEs in overseas markets argues that if the information regarding local conditions is imperfect and if the local institutions and factor markets are not well developed, MNEs choose to initially enter overseas markets in the form of JV with local firms (see, eg, Meyer, 2001; Sinha, 2001). But these foreign investors are usually concerned that transfer of technology will benefit the local partner, who may--in a worst-case scenario--emerge as a competitor (Buckley and Casson, 1998). In these models, therefore, MNEs choose to transfer cutting edge technology to their emerging market operations and grow only after they have acquired full operating control over the local operation, either through a buy-out of the local partner, or the acquisition of a local firm, or through the floatation of a new firm.
The extent of control a MNE has over its local operation is also likely to be influenced by the extent of reforms in the relevant host-country industry. For example, the ability of a MNE to open a new plant or acquire an existing plant to expand its production capacity will be influenced by factors like industrial licensing regimes and legislations aimed at limiting market power of individual firms.
Host-country institutions and MNE experience
The ability of a MNE to grow in an overseas market would also be dependent on the local institutional environment, and this is particularly true in the context of emerging markets (Khanna and Palepu, 2000; Khanna et al., 2005). If weak institutions lead to macroeconomic instability, it is likely to have a detrimental impact on employment growth at MNEs (Aizenman, 2003). Further, and in a more obvious way, a MNEs ability (and willingness) to expand its local operation would depend on factors like the flexibility of the local labour market, the ease with which contracts can he enforced in the host country, and the extent of corruption in host country. Ceteris paribus, a MNEs growth in an emerging market is likely to be positively influenced by the extent to which price setting and market behaviour has been liberalised, such that voluntary contracts made on the market can be enforced through the legal system, and negatively affected by corruption (Habib and Zurawicki, 2002). (3) Many of these factors can be offset by experience of operating in such institutional environments. Thus, relevant experience will ameliorate the increased transactions costs of operations in countries where market-supporting institutions are weak. Thus, local experience would enable a MNE to navigate the local business environment better, and thereby expand its operations, even if the formal institutions are somewhat underdeveloped.
Initial size of the firm
The benchmark for the relationship between initial size of a firm and its subsequent growth is Gibrat's law that argues that smaller firms are more likely to growth faster (see Sutton, 1997). However, to the extent that size offers firms 'pecuniary' and other economies of scale (Hart, 2000), larger firms may have an advantage over smaller firms in terms of expansion and this advantage may be reinforced if organic growth involves learning that is not costless (Dasgupta and Stiglitz, 1988). Finally, rapid growth of smaller firms may be hindered if the resultant increase of managerial cost more than outweighs the reduced average cost of production (Williamson, 1967; Penrose, 1980).
Strategic objectives of MNEs
We have taken into consideration the possibility that the behaviour of a MNE with respect to its size or level of employment might depend on its strategic objectives. MNEs usually set up operations in new countries either to cater to large local markets, or to gain access to local resources that are valuable in so far as the supply chains of the MNEs are concerned. (4) While, in general, MNEs tend to be more capital intensive than their domestic counterparts (Marsh et al., 1983), the strategic objective of a MNE may play a crucial role in determining its optimal labour-capital mix. For example, a MNE that has entered a developing country to take advantage of an abundantly available, and hence inexpensive, local resource--typically semi-skilled labour, is likely to adopt a labour-intensive technology in the host country, (5) whereas a market-seeking MNE might produce its product in the host country with an unchanged input mix because its strategy is to leverage on its brand to earn rents in the host-country market.
Availability of labour and complementary factors of production
In the context of emerging markets, we have to take into account possible paucity of resources that can constrain a MNEs ability to expand operations. There might be a shortage of (semi-) skilled personnel who can adapt quickly to the MNEs' technology (McDonald et al., 2002). This can either be a consequence of inadequate educational facilities, or because labourers with adequate skills may be organisationally embedded and/or geographically fragmented in emerging markets. For example, high-skilled managerial labour may be scarce because most skilled managers are owners of their own firms, and language and cultural barriers in large countries like India may prevent internal migration of skilled and semi-skilled technical labour. In addition, a MNE's ability to grow may be adversely affected by paucity of supporting factors of production like machinery and ICT services.
Local market structure
Elston (2002) has argued that firms that face fewer financial constraints are more likely to grow faster and this line of argument has found some empirical support (eg, Heshmati, 2001). To the extent that any expansion is financed by free cash flow accruing to a firm (Jensen, 1987), expansion of a firm's operation would depend on its profitability, and profitability, in turn, depends on the extent of competition faced by the firm in this market (Fosfuri et al., 2001). (6) In the context of MNE operations in emerging markets, a firm's expansion might also depend on the willingness of the parent MNE to finance growth of the emerging market operation (Ushijima, 2005). This willingness too is likely to depend on the profitability of the local operation that would determine the rate of return on the investment made by the parent MNE, and hence on competition.
Finally, we also control for local industry demand.
In other words, employment growth of MNEs in emerging markets is captured by the following function, which is made operational for empirical in the following section:
Employment Growth = f(Control, Institutions, Experience, Initial size, Strategic objectives, Factor availability, Local market structure, Local demand) (1)
In effect, we have built our empirical specification by drawing on a number of different yet related strands of theories of economics and international business studies. Our specification takes into consideration classical issues like the relationship between initial size and growth of firms, as also more recent debates like those concerning the impact of institutions on firm (especially MNE) behaviour in emerging markets. We draw on economic theories concerning the impact of informational asymmetry and uncertainty about contract enforcement on the entry mode choice, and subsequent behaviour of, MNEs in southern markets. And, at the same time, we take into account factors like strategic objectives of MNEs that are largely ignored in the economics literature, with a few exceptions (see, eg, Ethier, 1986), but are deemed important by international business theorists. This harmonisation of the relevant literature has allowed us to develop an empirical framework that can be used to examine the determinants of employment growth at MNEs, the importance of such growth being spillovers in ways that are discussed in the first section of this paper.
DATA, VARIABLE MEASUREMENT AND SPECIFICATION
The data set were collected from randomly selected MNE affiliates operating in Egypt, India, South Africa and Vietnam, using a survey instrument. The base population for the survey was defined as all registered FDI projects that were established in the four countries between 1990 and 2000 that had a minimum employment of 10 persons, and minimum of 10 per cent equity stake by the foreign investor. The time limit ensured that the information relevant to the decisions taken at the time of establishment of these firms was part of the organisational memory at the time of the survey. Similarly, the stipulations concerning size and equity stake of the foreign investor ensured that the firms included in the base population were not trading or sales offices, but rather were fully operational business operations. The questionnaire was structured to enable us to collect information about not only the characteristics of the local affiliates, but also about the perception of the affiliates about local conditions during the recent years of operation. After accounting for missing observations, we have usable information for 293 observations spread across the four countries. (7) The sampling methods and descriptive statistics for the dataset are described in detail in Chapter 2 of Estrin and Meyer (2004).
Our measure of employment growth is the growth rate of the labour force associated with an MNE affiliate from the inception of its operation in the host country to 2000 (EMPGROWTH). In order to account for the fact that an early entrant in the host-country market chalks up cumulative employment growth over a greater number of years, we control for the length of operation of a MNE in a host-country market using a linear time trend (FIRMAGE). The value of the time trend is 1 for a MNE affiliate that initiated operation in 1991, and 10 for a firm that initiated operation in 2000.
The measures of the other explanatory variables are as given below:
Our data set allows us to classify four different forms in which the local operations of a MNE came into existence, namely, Greenfield project, acquisition of a local firm giving the MNE unambiguously controlling equity stake, partial acquisition of a local firm that does not give the MNE an unambiguously controlling stake, and JV. The discussion in the earlier sections suggests that the MNEs are unlikely to expand local operations, especially in the form of employment, so long as they do not have operational control of the local operations. In other words, employment growth at a Greenfield project or an acquisition that gives a MNE outright and unambiguous control over the local operation is likely to be higher than the employment growth at a JV or in the case of a partial acquisition. At the same time, however, it is not obvious as to whether there is any systematic difference in the growth rates of Greenfield projects and outright or full acquisitions. Hence, we create three dummy variables, one each for Greenfield, full acquisition (ACQUISITION) and JV-partial acquisition (JV). In the regression model, reported later in the paper, Greenfield is the omitted category. On the basis of the above discussion we expect the coefficient of JV dummy variable to be negative, but we do not have any priors about the coefficient of the acquisition dummy variable.
We proxy institutional development by four measures, three of which are based on published data for the four host countries. The first is the corruption perception index published annually by Transparency International, which is based on an aggregation of multiple published indices that measure managerial perception of corruption (CPI). This rates countries in terms of the degree to which corruption is perceived to exist among public officials and politicians, and reflects the views of business people and analysts. We also use the economic freedom index (EFI), developed by the Heritage Foundation, which provides a wide range of information about institutional development. Variables contained in the aggregate index include capital market development, property rights enforcement, regulation, trade policy and government intervention in the economy. For each observation in our sample, the value of this index is its value for the year of the MNE's entry into the relevant emerging market. Since the index were available only for the 1995-2000 period, we had to leave out of our analysis the <5 per cent of the firms in the sample that entered the four countries between 1990 and 1995.
Given the empirical evidence about the impact of the nature of the legal system on the individual and collective behaviour of economic agents (Djankov et al., 2002, 2003), we control for the legal environment of the host countries using a dummy variable that takes the value 1 if a country's legal system is based on the Anglo-Saxon common law paradigm, and is zero otherwise (COMMONLAW). Our prior is that rate of expansion of emerging market operations, and hence employment growth at these operations, is likely to be affected positively by the extent of economic freedom and the existence of an Anglo-Saxon common law system, and negatively by the extent of corruption perception.
Finally, we use information collected in the survey about industry level reforms. The extent of industry-specific reforms was measured by the four country teams responsible for the survey, based on discussions with local firms as well as MNEs operating in those industries. Measures for four different aspects of reform, including privatisation of state-owned firms, were obtained on a Likert scale that runs from 1 (no policy changes) to 5 (major policy changes). Using Crombach's alpha, we ascertained that it is possible to club together the four measures of reform, and hence a composite measure that is a simple average of the four component measures was used for the analysis (IREFORM).
We use two different measures of a MNE's experience in the subsequent analysis, namely, a dummy variable that indicates whether or not a MNE had prior experience of operating in a host country (SPECIFICEXP), and a categorical variable that takes the value 1-4 depending on the number of clusters of emerging markets in which the MNE has operating experience (RELATEDEXP). After some experimentation, we chose four clusters; Africa, Asia (other than Japan), Central and Eastern Europe, and Latin America. We expect prior experience in emerging market conditions, especially within the relevant country itself, to have a positive impact on employment growth of a MNE.
In keeping with the focus of our paper, we measure initial size using the number of employees at a MNE during its first year of operation in the emerging market (INITSIZE). Gibrat's law suggests that, Ceteris paribus, the expected impact of initial size on the subsequent growth rate of employment at the firm would be negative.
We operationalise MNE strategy via the objectives of the firm, distinguishing between resource and market-seeking motives for FDI. We define a MNE as resource seeking if it sells < 50 per cent of its output in the host country's market, and capture the strategic motivation of the MNE using a dummy variable that takes the value 1 if a MNE is resource seeking, and zero if it is market seeking (RSEEKING). (8) To the extent that the main resource sought by MNEs in emerging markets is (semi-) skilled labour, resource-seeking MNEs might be expected to expand employment at their local operations faster than market seeking MNEs.
The availability of local resources is measured for three categories: qualified personnel itself (PERSONNEL), as well as related factors of production like machinery and equipment (MACHINERY), and IT and telecommunications services (ICT). The perceptions of the MNE affiliates that responded to our survey are measured on a Likert scale of 1 (never available) through 5 (readily available).
We measure the extent of competition faced by a MNE in the relevant host country using a categorical variable to measure the extent of local competition, where 1 stands for none and 5 stands for greater than 10 (COMPETITION).
We use as a measure of local demand for each 3-digit industry's product its average growth rate in the relevant emerging market during the 1990s (IGROWTH).
The specification used for our econometric analysis, derived from equation (1), therefore, is given by the following:
EMPGROWTH = [[beta].sub.0] + [[beta].sub.1] ACQUISITION + [[beta].sub.2] JV + [[beta].sub.3] IREFORM + [[beta].sub.4] EFI + [[beta].sub.5] CP + [[beta].sub.6] COMMONLAW + [[beta].sub.7] SPECIFICEXP + [[beta].sub.8] RELATEDEXP + [[beta].sub.9] INITSIZE + [[beta].sub.10] RSEEKING + [[beta].sub.11] PERSONNEL + [[beta].sub.12] MACHINERY + [[beta].sub.13] ICT + [[beta].sub.14] COMPETITION + [[beta].sub.15] IGROWTH + [[beta].sub.16] FIRMAGE + [epsilon] (2)
The descriptive statistics are reported in Table 1. About half of the firms in our sample of 293 MNE operations in emerging markets are in the form of JV, while another 36 per cent are in the form of Greenfield projects. Cross-border acquisitions account for only about 12 per cent of the sample, largely on account of South Africa.
The average growth rate of employment at the MNE subsidiaries included in the sample was 25.7 per cent between the time of their (post 1990) entry into the host country and the year 2000. An average MNE entered a host country in 1996, that is, this translates into around 6 per cent growth in employment per annum. This is a small number given the low initial level of employment of about 181, and the average output growth of 11.8 per cent in the host-country industries to which these MNEs belong. In other words, it is not obvious that high growth of the local industry necessarily translates into employment growth at MNEs. We shall revisit this issue later in the paper.
The four emerging markets represented in the sample ranked fairly high in terms of both economic freedom and corruption perception, with an average score of over 3.5 each on a 5-point scale. Of the four countries, two--India and South Africa--are believed to have legal systems that approximate the Anglo-Saxon system.
Nearly half (45 per cent) of the MNEs had commercial experience in the relevant host country prior to setting up manufacturing operations in it, and about 70 per cent of the MNEs had operational or commercial experience in similar emerging markets. Indeed, average MNEs in the sample had operational experience in about 25 countries. However, this figure was influenced by MNEs like Coca Cola and Pepsi that have operations in over 175 countries.
Most of the MNEs in the sample, about 71 per cent, are market seeking. However, to the extent that they require local resources like qualified personnel, ICT and machinery and equipment, they do not face much difficulty in acquiring them. The average score for the degree of difficulty in obtaining these resources varies between 3.6 and 4.2, where 1 indicates 'never available' and 5 indicates 'readily available.' This suggests that supply side constraints were unlikely to have been binding in the context of employment growth at these MNEs, perhaps because these firms pay efficiency wages.
The MNEs faced a modest degree of competition in their host-country industries. An average MNE's host-country industry included five to 10 competitors.
Our empirical analysis involves the econometric estimation of equation (2). In keeping with relevant literature, we also control for all unobserved characteristics of the host countries and the 2-digit industries of the MNEs in these host countries that may affect employment growth of their local operations (Bloningen and Tomlin, 2001; Tan and Mahoney, 2005). In addition, we take into consideration the possibility that the strategic objectives of the MNEs may have different impact on employment growth in different industries, and hence we control for interaction between the resource-seeking dummy variable and the dummy variables capturing industry-specific factors as well.
The regression results are reported in Table 2. All specifications reported in Table 2 have been estimated using ordinary least squares, with the appropriate correction for hetereskedasticity. The McFadden's adjusted [R.sup.2] estimates for the regressions are about 0.18, and the F-statistics for the specifications are significant at the 1 per cent level. These statistics are entirely consistent with goodness of fit measures of cross-sectional regressions involving less than 300 observations. An interesting aspect of the results is that as the indices measuring economic freedom and corruption perception, and the dummy variable capturing the impact of (common law) legal systems on employment growth are introduced into the specification, in columns 2 and 3, respectively, the goodness of fit of the regressions does not change, but the significance of the controls for host countries declines noticeably. In other words, in our regression specification, the host-country dummy variables were largely controlling for cross-country variations in institutions.
Not surprisingly, the cumulative (or overall) employment growth of a MNE affiliate increases with the age of the firm. The other results reported in Table 2 that are consistent with our priors are as follows:
* Relative to a Greenfield project, employment growth is likely to be significantly slower if the host-country operation of a MNE is in the form of a JV with a local firm. However, there is no statistically significant difference between employment growth at a Greenfield project and an operation initiated by way of cross-border acquisition that gives the MNE full control of the host-country operation.
* The existence of an Anglo-Saxon legal system has a positive impact on the rate of employment growth.
* The annual average employment growth of a MNE is inversely related to the initial size of the host-country operation of the firm. This is consistent with Gibrat's law.
* Employment growth at a MNE's emerging market operation is inversely affected by the extent of competition faced by the latter in the host country.
Our results, however, also have a couple of counter-intuitive elements. To begin with, availability of qualified personnel is seen to have had a negative impact on employment growth at the local operation of the MNE. This result possibly implies a plausible quality-quantity trade-off in employment. If the available labour force is qualified, that is, have the appropriate skills, it is possible for MNEs to expand their operations, when required, by eliciting greater work effort from the existing employees who are presumably paid efficiency wages, rather than by way of employment growth.
Further, the rate of expansion of a MNE affiliate's emerging market operations is inversely related to the growth rate of the host-country industry to which it belongs. There are two possible explanations for this outcome. First, the turnover of local industries is typically measured in local currencies, which are significantly influenced by exchange rate fluctuations. For example, the exchange rate for the Indian rupee declined from 35.68 rupees per US dollar in September 1996 (ie, the time of entry of an average MNE into one of the four host countries) to 46.76 rupees per US dollar by the end of 2000. However, MNEs are interested in the expansion of their business operations as measured in the currency of their home base. Hence the > 11 per cent growth rate of local industries may not be an accurate reflection of the growth in the business operations of the MNEs. Second, the growth of turnover in the local industries may reflect a change in the output mix of these industries, that is, higher value addition, which is usually accompanied by a capital-favouring input mix.
There has been a significant growth in the literature on the impact of initial size, financial constraints and ownership structures on the growth of firms, often measured in terms of rate of change of employment. There has been a parallel growth in the literature on the factors that are likely to affect employment growth at emerging market operations of MNEs. However, there have been few empirical investigations of employment growth at MNE operations, especially in the context of emerging markets. Yet, there is evidence to suggest that training provided by MNEs to their work force is a significant channel of technology spillover in a host country. Using primary firm-level data from four emerging markets, therefore, we address this lacuna in the literature.
A key result is that employment growth at emerging market operations of MNEs is adversely affected if the local operation is a JV with a host-country firm. Given the reasonable assumption about the positive relationship between training of MNEs' work force and technology spillover/transfer, which implies that the likelihood and/or magnitude of the spillover/transfer is likely to be positively related to the rate of employment growth at MNEs, this result has an important policy implication. It suggests that FDI policies of developing countries that aim to foster technology transfer from MNEs to the domestic firms by limiting the ability of the former to operate wholly owned subsidiaries in these countries, whether by way of Greenfield projects or crossborder acquisitions, may be counterproductive. In view of the fact that, on account of domestic political compulsions and other reasons, many emerging markets restrict the extent and nature of control a MNE can have over its local affiliate, this may help to explain why the empirical literature on FDI finds little evidence of FDI-driven technology enhancement in developing countries.
We are grateful to Simon Commander, Stephen Gelb, Danchi Tan, two anonymous referees and editor Jeffrey Miller for helpful comments, to the Egyptian, Indian, South African and Vietnamese country teams for the collection of data, and to Caitlin Frost, Maria Bytchkova, Gherardo Girardi for excellent research assistantship. We remain responsible for all remaining errors. The research was made possible by a research grant from the UK's Department for International Development.
Aitken, BJ and Harrison, AE. 1999: Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. American Economic Review 89(3): 605-618.
Aizenman, J. 2003: Volatility, employment and the patterns of FDI in emerging markets. Journal of Development Economics 72: 585-601.
Altenburg, T. 2000: Linkages and spillovers between transnational corporations and small and medium-sized enterprises in developing countries: Opportunities and policies. Proceedings of the UNCTAD X Special Round Table 'TNC-SME Linkages for Development,' Bangkok.
Bartlett, C and Ghoshal, S. 1989: Managing Across Borders: The Transnational Solution. Harvard Business School Press: Boston.
Blomstrom, M and Kokko, A. 1998: Multinational corporations and spillovers. Journal of Economic Surveys 12(2): 1-31.
Blonigen, BA and Tomlin, K. 2001: Size and growth of Japanese plants in the United States. International Journal of Industrial Organization 19: 931-952.
Borensztein, E, de Gregorio, J and Lee, J-W. 1995: How does foreign direct investment affect economic growth. Working paper no. 5057, National Bureau of Economic Research: Cambridge, MA.
Buckley, PJ and Casson, MC. 1998: Analyzing foreign market entry strategies: Extending the internalization approach. Journal of International Business Studies 29(3): 539-562.
Campbell, N and Vousden, N. 2003: Training and technology transfer. Australian Economic Papers 42: 35-49.
Commander, S. 2003: The origins and dynamics of the software industry in emerging markets. CNEM, London Business School: Mimeo.
Dasgupta, P and Stiglitz, JE. 1988: Learning-by-doing, market structure and industrial trade and policies. Oxford Economic Papers 40: 246-268.
Djankov, S, La Porta, R, Lopez-di-Silanes, F and Shleifer, A. 2002: The regulation of entry. Quarterly Journal of Economics 117(1): 1-37.
Djankov, S, Shleifer, A, Glaeser, E, La Porta, R and Lopez-di-Silanes, F. 2003: The new comparative economics, volume. Policy Research Working Paper No. WPS 3054, The World Bank: Washington D.C., http://wdsbeta.worldbank.org/external/default/WDSContentserver/ IW3P/IB/2003/07/08/000094946_03061904092112/Rendered/PDF/multi0page.pdf.
Dosi, G. 1988: Sources, procedures and microeconomic effects of innovation. Journal of Economic Literature 26(3): 1120-1171.
Elston, JA. 2002: An examination of the relationship between firm size, growth and liquidity in the neuer market. Discussion Paper No. 15/02, Economic Centre of the Deutsche Bundesbank, http://www.bundesbank.de/vo/download/dkp/2002/15/200215dkp.pdf.
Estrin, S and Meyer, KE. 2004: Investment Strategies in Emerging Markets. Cheltenham: Elgar.
Ethier, WJ. 1986: The multinational firm. Quarterly Journal of Economics 101(4): 805-834.
Findlay, R. 1978: Relative backwardness, direct foreign investment, and the transfer of technology: A simple dynamic model. Quarterly Journal of Economics 42(1): 1-16.
Fosfuri, A, Motta, M and Ronde, T. 2001: Foreign direct investment and spillovers through workers' mobility. Journal of International Economics 53: 205-222.
Fotopoulos, G and Louri, H. 2004: Firm growth and FDI: Are multinationals stimulating local industrial development? Journal of Industry, Competition and Trade 4(3): 163-189.
Gershenberg, I. 1987: The training and spread of managerial know-how: A comparative analysis of multinationals and other firms in Kenya. World Development 15: 931-939.
Habib, M and Zurawicki, L. 2002: Corruption and foreign direct investment. Journal of International Business Studies 32(2): 291-307.
Hart, PE. 2000: Theories of firms' growth and the generation of jobs. Review of Industrial Organization 17: 229-248.
Heshmati, A. 2001: On the growth of micro and small firms: Evidence from Sweden. Small Business Economics 17: 213-228.
Hill, H. 1982: Vertical inter-firm linkages in LDCs: A note on the Philippines. Oxford Bulletin of Economics and Statistics 44: 261-271.
Hou, C and Gee, S. 1993: National systems supporting technical advance in industry: The case of Taiwan. In: Nelson R (ed). National Innovation Systems: A Comparative Analysis. Oxford University Press: New York, pp. 384 413.
Jensen, MC. 1987: The free cash flow theory of takeovers: A financial perspective on mergers and acquisitions and the economy. In: Browne LE and Rosengren ES (eds). The Merger Boom. The Federal Reserve Bank of Boston, pp. 102-143.
Katz, JM. 1987: Technology Creation in Latin American Manufacturing Industries. St Martin's Press: New York.
Kaufman, L. 1997: A model of spillovers through labor recruitment. International Economic Journal 11(3): 13-33.
Khanna, T and Palepu, K. 2000: The future of business groups in emerging markets: Long-run evidence from Chile. Academy of Management Journal 43(3): 268-285.
Khanna, T, Palepu, K and Sinha, J. 2005: Strategies that fit emerging markets. Harvard Business Review 83: 6-20.
Kokko, A. 1994: Technology, market characteristics, and spillovers. Journal of Development Economics 43: 279-293.
Marsh, L, Newfarmer, R and Moreira, L. 1983: Employment and Technological Choice of Multinational Enterprises in Developing Countries. ILO working paper no. 23.
McDonald, F, Tuselmann, HJ and Heise, A. 2002: Foreign direct investment and employment in the host regions. European Business Review 14(1): 40-50.
Meyer, KE. 2001: Institutions, transaction costs and entry mode choice in Eastern Europe. Journal of International Business Studies 31(2): 357-367.
Meyer, KE. 2004: Perspectives on multinational enterprises in emerging economies. Journal of International Business Studies 34(4): 259-277.
Pack, H. 2001: Technological change and growth in East Asia: Macro versus micro perspectives. In: Stiglitz JE and Yusuf S (eds). Rethinking the East Asian Miracle. Oxford University Press for the World Bank: Oxford, pp. 95-142.
Penrose, ET. 1980: The Theory of the Growth of the Firm. Blackwell: Oxford.
Pfaffermayr, M and Bellak, C. 2000: Why foreign-owned firms are different? A conceptual framework and empirical evidence from Austria. Discussion Paper No. 115, Hamburg Institute of International Economics, http://www.hwwa.de/Publikationen/Discussion_Paper/2000/115.pdf.
Sabirianova, K, Svejnar, J and Terrell, K. 2005: Foreign investment, corporate ownership and development: Are firms in emerging markets catching up to the world standard? Working Paper No. 734, William Davidson Institute, University of Michigan: Ann Arbor.
Sinha, UB. 2001: International joint venture, licensing and buy-out under asymmetric information. Journal of Development Economics 66: 127-151.
Sutton, J. 1997: Gibrat's legacy. Journal of Economic Literature 35(1): 40-59.
Tan, D and Mahoney, JT. 2005: Examining the penrose effect in an international business context: The dynamics of Japanese firm growth in US industries. Managerial and Decision Economics 26(2): 113-127.
Tsiang, S and Wu, R. 1985: Foreign trade and investment as boosters for take-off: The experiences of the four Asian newly industrialized countries, In: Galenson W (ed). Foreign Trade and Investment: Development in the Newly Industrializing Asian Countries. The University of Wisconsin Press: Wisconsin, pp. 301-332.
Ushijima, T. 2005: Internal capital market and the growth and survival of Japanese plants in the United States. Journal of Japanese and International Economies 19(3): 366-385.
Wang, J-Y and Blomstrom, M. 1992: Foreign investment and technology transfer: A simple model. European Economic Review 36: 137-155.
Williamson, OE. 1967: Hierarchical control and optimum firm size. Journal of Political Economy 75: 123-138.
(1) The evidence linking MNE presence in a sector in a developing country to technological improvements among domestic firms in the same sector is mixed (see Blomstrom and Kokko, 1998). Specifically, it has been argued that the extent of the spillover would significantly depend on the ability of the local firms to assimilate the new technology in their own production processes (Kokko, 1994).
(2) There is also evidence from countries like Kenya that only a small proportion of labourers (especially managers) leave MNEs to join local firms in developing countries (Gershenberg, 1987). Further, Altenburg (2000) reports that spin-off electronics companies in Malaysia maintain close relations as suppliers and subcontractors with the MNE, while Hill (1982) makes similar observations in the Philippine appliance and motorcycle industry. But, by and large, inter-firm mobility of labour is widely accepted as a channel through which technology spillovers of FDI take place (see Dosi, 1988).
(3) Note that the overall extent of economic freedom may have an overlap with the regulations governing FDI in emerging markets.
(4) This distinction has important implications for how investors set up their operations (Buckley and Casson, 1998). Local production for local markets eliminates the cost of transporting the product from production locations in other countries, and are also able to eliminate tariffs from the retail price of the product, thereby making it competitive vis a vis the domestic and import competition in the new location. Local production facilities also endow MNEs with the flexibility in the production process that is required to appeal to local tastes and preferences (Bartlett and Ghoshal, 1989). Resource-seeking investment, on the other hand, permits MNEs to leverage the resources available in the new production location--be it petroleum in Egypt, precious metals in South Africa, skilled IT personnel in India or cheap low-skilled labour ill Vietnam--to give its global operation a competitive edge over its rivals.
(5) The Hecksher-Ohlin theory about international specialisation argues that most developing countries have a comparative advantage in labour-intensive products. Therefore, resource-seeking MNEs are even more likely to opt for labour-intensive production techniques if they use developing country as a location for downstream units of their supply chain, or as an export base to the rest of the world.
(6) As we have seen earlier in the paper, threat of competition may adversely affect employment growth (and employee training) for other reasons as well (Campbell and Vousden, 2003).
(7) Of these, 23 per cent are from Egypt, 22 from India, 30 per cent from South Africa and 25 per cent from Vietnam. In related research in which roughly the same sample was used, we used the Heckman two-step method to examine whether there is a selection bias with respect to the MNE affiliates that did not provide information that led to the generation of missing values. Our analysis suggested that the missing values are random in nature, and do not owe their origin to selection bias.
(8) Since the 50 per cent cutoff point is plausible yet ad hoc, we experimented with other cut off points ranging from 30 to 70 per cent, but the econometric results, reported later in the paper, remained unaltered.
SUMON KUMAR BHAUMIK (1), SAUL ESTRIN (2) & KLAUS E MEYER (3)
(1) Brunel Business School, Brunel University, Social Sciences Building, Uxbfidge, Middlesex UB10 9NW, UK. E-mail: Sumon.Bhaumik@brunel.ac.uk
(2) London Business School, UK.
(3) University of Reading, UK.
Table 1: Descriptive statistics Standard Mean deviation Growth rate of employment (percent) 25.69 38.57 Ownership and control Percentage of MNEs with Greenfield project 36.18 Percentage of MNEs with full acquisition 12.82 Percentage of MNEs with JV (and partial 50.98 acquisition) Institutional environment in host country Index of economic freedom [cross product] 3.84 0.65 Index of corruption [cross product] 3.59 0.56 Percentage of host countries with Anglo-Saxon 50.00 legal system Liberalisation and privatisation prior to 2.41 0.82 establishment of affiliate [cross product] Experience Percentage of MNEs with in country experience 45.00 50.00 Percentage of MNEs with experience in other 70.00 46.00 emerging markets Initial size Number of employees at start of operations 181.70 485.61 Strategic consideration of MNE Percentage of MNEs that are resource seeking 29.00 45.00 Resource availability in host country Availability of qualified personnel 3.75 0.90 [cross product] Reliability of IT and telecommunications 4.19 0.90 services [cross product] Availability of machinery and equipment 3.62 1.29 [cross product] Characteristics of industry in host country Growth of industry turnover 11.76 11.58 Number of competitor [cross product] 3.60 1.27 Control variable Firm age 5.85 2.44 Note: [cross product] indicates that the variable was measured using a 5-point scale. Table 2: Determinants of performance (OLS estimates with robust standard errors) Model 1 Model 2 Constant 43.87 *** (14.80) 0.36 (60.47) Ownership and control Acquisition -8.44 (9.36) -10.10 (9.96) Joint venture -13.05 *** (4.80) -14.20 *** (5.20) Institutional and regulatory environment in the host country Index of economic freedom 2.51 (9.58) Index of corruption perception 7.63 (6.38) English (common) law system Extent of liberalisation of the 0.82 (2.34) 1.67 (2.97) host country's industry Operating experience in emerging markets Prior experience in host -0.15 (4.47) 0.55 (4.82) country Prior experience in similar 0.48 (4.37) -0.83 (4.46) emerging markets Initial size Initial level of employment -0.01 *** (0.00) -0.01 *** (0.00) Strategic interest of MNE Resource seeking 4.70 (4.68) 4.37 (4.81) Resource availability in host country Availability of qualified -5.45 ** (2.23) -5.65 ** (2.36) labour Availability of ICT services -3.37 (2.22) -3.28 (2.29) Availability of machinery and 0.94 (1.66) 1.34 (1.80) equipment Characteristics of industry in the host country Average growth of host-country -0.41 ** (0.16) -0.41 ** (0.16) industry during the 1990s Extent of domestic competition -2.74 * (1.54) -3.12 * (1.61) Control variables Age of the firm 3.70 *** (0.84) 4.57 *** (1.30) Industry Yes Yes Resource seeking x industry Yes Yes Host country Yes *** Yes ** Adjusted [R.sup.2] 0.175 0.183 F-statistic (Prob>F-statistic) 3.47 (0.00) 3.26 (0.00) Nobs 304 293 Model 3 Constant 0.37 (60.47) Ownership and control Acquisition -10.10 (9.96) Joint venture -14.21 *** (5.20) Institutional and regulatory environment in the host country Index of economic freedom 2.51 (9.58) Index of corruption perception 7.63 (6.38) English (common) law system 29.90 *** (10.53) Extent of liberalisation of the 1.67 (2.97) host country's industry Operating experience in emerging markets Prior experience in host 0.55 (4.81) country Prior experience in similar -0.83 (4.46) emerging markets Initial size Initial level of employment -0.01 *** (0.00) Strategic interest of MNE Resource seeking 4.37 (4.80) Resource availability in host country Availability of qualified -5.65 ** (2.36) labour Availability of ICT services -3.28 (2.29) Availability of machinery and 1.34 (1.80) equipment Characteristics of industry in the host country Average growth of host-country -0.41 ** (0.16) industry during the 1990s Extent of domestic competition -3.12 * (1.61) Control variables Age of the firm 4.57 *** (1.30) Industry Yes Resource seeking x industry Yes Host country Yes * Adjusted [R.sup.2] 0.183 F-statistic (Prob>F-statistic) 3.26 (0.00) Nobs 293 Dependent variable: average growth rate of labour force per annum, from start of operations until 2000. Note: (1) The values within parentheses are standard errors. (2) *, ** and *** imply significance at 10, 5 and 1 per cent level of significance, respectively.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||multinational corporations|
|Author:||Bhaumik, Sumon Kumar; Estrin, Saul; Meyer, Klaus E.|
|Publication:||Comparative Economic Studies|
|Date:||Mar 1, 2007|
|Previous Article:||Shock therapy and the Washington Consensus: a comment.|
|Next Article:||Changing preschool enrolments in post-socialist Central Asia: causes and implications.|