SOUTH-AFRICAN MNC FOREIGN EXPANSION INTO DEVELOPED AND DEVELOPING ECONOMIES

(1)

SOUTH-AFRICAN MNC FOREIGN EXPANSION INTO

DEVELOPED AND DEVELOPING ECONOMIES

An investigation on entry mode selection and entry success

By

FERDINAND JOHANNES LAURENS TERBERG

DDM: Advanced International Business Management and Marketing Student number: s1796313 (Groningen) & b130027074 (Newcastle)

(2)

2 ABSTRACT

In light of on-going globalization and the emergence of developing countries, this study investigates the differences in entry mode strategy and entry success for MNCs from a developing country (South-Africa) when expanding abroad into developed and developing markets. While prior research mainly focused on the expansion strategies of developed market MNCs, this research focusses on MNCs from a developing country. The analysis is based on 174 entries by 14 MNCs listed on the Johannesburg Stock Exchange (JSE) over the ten year period from 2001 until 2011. Logistic and linear regressions as well as cross tabulation chi square models were conducted in order to achieve the results.

The results indicate that there are differences in entry success and entry mode decision of South-African MNCs according to the level of development of the host country. Furthermore, the impact of determinants on entry mode decision and entry success differ when the full model was split into developing and developed host markets. Market size has a significant impact on the entry mode decision when expansion is done into developed countries; country openness and country risk have a significant impact on entry mode decision when expansion is conducted in developing markets. Additionally, the entry mode decision has a significant impact on entry success in developed markets. Entry timing and country risk have a significant impact on entry success in developing markets.

This research is the first to differentiate between expansions into developed and developing countries from a developing country perspective. Additionally, this research is the first to investigate factors that influence success while differentiating between the development level of the host country. Moreover, this study adds new insights to the current literature by investigating relationships that have not been investigated yet.

(3)

3 ACKNOWLEDGEMENTS

(4)

4

TABLE OF CONTENT PAGE

1. INTRODUCTION 6

2. LITERATURE REVIEW 11

2.1 Entry modes 11

2.2 Entry success 12

2.3 Differences between developed and developing market MNCs 13

3. HYPOTHESIS DEVELOPMENT 15

3.1 Determinants of entry mode decision 15

3.2 Determinants of entry success 20

3.3 Developing vs developed market (MNCs) 25

4. METHODOLOGY 28

4.1 Data 28

4.2 Dependent variables 30

4.3 Independent variables 33

4.4 Differentiating between developed and developing markets 35

5. RESULTS 36

5.1 Entry mode decision 36

6. DISCUSSION 44

6.1 Entry mode decision 44

7. IMPLICATIONS 51

8. LIMITATIONS & FURTHER RESEARCH 53

9. CONCLUSION 54

REFERENCES 56

(5)

5

LIST OF TABLES PAGE

TABLE 1: All independent variables and their expected impact 19 on entry mode decision

TABLE 2: All independent variables and their expected effect 24 on entry success

TABLE 3: Descriptive statistics of variables (entry mode decision) 36

TABLE 4: Correlation matrix (entry mode decision) 37

TABLE 5: Logistic regression results (beta coefficients and standard errors) 38 TABLE 6: Descriptive statistics of variables (entry success) 38 TABLE 7: Descriptive statistics of the variable entry mode 39

TABLE 8: Correlation matrix (entry success) 39

TABLE 9: Linear regression results (beta coefficients and standard errors) 40 TABLE 10: Cross-Tabulation table (entry mode and OECD membership) 41 TABLE 11: Cross-Tabulation table (entry success and OECD membership) 42

(6)

6 1. INTRODUCTION

The emergence of developing markets and their firms as important players in global markets has been a remarkable phenomenon of globalization in the twenty first century (Chittoor, 2009). In 2012, for the first time ever, developing economies absorbed more Foreign Direct Investment (FDI) than developed countries, accounting for 52% of global FDI flows (UNCTAD, 2013). Not only have the developing market economies been among the top destinations for inward FDI, the outward FDI flows from these economies have also been growing steadily. FDI from developing economies outflows reached $426 billion in 2012, a record 31% of the world total (UNCTAD, 2013). Additionally, data on internationalization indicators for the largest 100 Multinational Corporations (MNCs) headquartered in developing economies reveal a strong internationalization effort with steep increases in foreign assets and sales. The foreign assets of MNCs from these economies rose 19.7% in 2011, a rate faster than that of the largest 100 MNCs and almost double the remarkable 11% increase in domestic assets (UNCTAD, 2013). The rising international presence of developing markets and their MNCs require a greater attention to how they behave.

(7)

7 Martinez & Gastelaars, 2011) have recently highlighted the possibility of differences in international strategy according to the level of economic development in the host and home country.

In addition to differences in the characteristics of developed and developing countries, MNCs from both markets seem to differ as well. Despite the fact that MNCs from developing economies continue to expand into foreign markets, there is significantly less literature written about them compared to their developed markets counterparts. Many authors have observed that the characteristics of MNCs from developing economies differ from those of comparable MNCs from developed economies (Buckley, 2004; Mathews, 2002). More specifically, the belief that firms from developing economies approach internationalization differently has prompted Mathews (2002) to argue for alternative internationalization framework for MNCs from developing economies. Other authors have opted for extensions or adaptions of existing theoretical frameworks to describe the internationalization strategy of firms from developing countries (UNCTAD, 2006).

Given the recent trend that developing countries and their MNCs are getting a greater importance in the international economy and indications that developing country MNCs approach internalization differently than MNCs from developed countries, this study will focus on MNCs from a developing market. More specifically, this study will focus on two well established topics within the foreign expansion literature, namely: entry mode decision and entry success. The main four study outlines for this thesis are:

- Whether established determinants of entry mode decision and entry success will hold when examining the expansion of MNCs from a developing market.

- Whether the impact of these determinants will differ significantly when expansion of developing market MNCs is done into either developed markets or developing markets.

(8)

8 - Whether there will be differences in the degree of entry success when expansion is done by a developing market MNC into either a developed market or another developing market.

In order to find answers for the described study outlines, existing literature and theories on the determinants of entry modes and entry success will be examined at first. A lot of research has been conducted in the past on these determinants. However, a relatively low number of studies conducted their research based on developing market MNCs expanding into foreign markets. Hence it would be interesting to see if these determinants from established literature hold when examining MNCs from developing markets. Additionally, this study will check whether there will be differences in the impact of the determinants when expansion will be done into either developed or developing countries. As scholars have recently highlighted, there is a possibility of differences in international strategy according to the level of economic development of the host country (Williams et. Al, 2011). Williams et al. (2011) conducted a study checking the relative impact of established determinants on MNCs from a developed market (the Netherlands) expanding into developed and developing economies. This study achieved interesting results in the sense that differences in the relative impact of two well established determinants were found. However, this study has not been conducted based on MNCs from a developing economy and only two determinants were involved in this study. In addition to following hypotheses from existing studies, new hypotheses will also be developed based on the differences between MNCs from developed and developing markets. These hypotheses will be developed based on literature which will be collected from other topics in the international business theory.

(9)

9 to the firm in a foreign market (Alderson, 1957). Hence, it is important that managers of MNCs from developing countries are aware of the advantages and limitations of each entry mode. Managers of developing market MNCs should understand the theories that underlie the effect of appropriate entry mode selection and be able to use this knowledge in crafting an effective entry strategy.

For MNCs from developing economies it seems of great importance to enter a new market successfully. MNCs from developing markets have to deal with many challenges when entering foreign markets. To begin with, MNCs from developing markets have to overcome the so called late mover disadvantage (Chittoor, 2009). This is mainly due to the fact that the internationalization process of many developing countries started later than the process of countries located in the developed world. Next to that, they most of the time hail from less abundant resource environments and have to acquire resources and capabilities to successfully compete with established players from developed economies (Guillen, 2000). These arguments indicate that determinants and outcomes of entry success could be different for firms coming from developing markets compared to firms from developed markets. Hence it will be beneficial for research and developing market MNCs to gain greater knowledge on entry success.

(10)

10 South Africa as a country to conduct this study on. No study so far in the field of internationalisation strategy has focussed on South-African MNCs in particular. English is the main business language in South-Africa which makes the quest for information easier. Furthermore, South-Africa hosts many big corporations with international strategies. Companies like SABMiller, AngloAmerican, and MTN have a striking worldwide presence.

(11)

11 2. LITERATURE REVIEW

2.1 Entry Modes

This thesis focuses on three modes of entry: greenfield, acquisition, and joint venture/cooperative modes. All of these entry modes involve equity investments. They commit a firm to a strategic choice, as they cannot be changed easily and involve significant resource and time commitments (Elango & Sambharya, 2004). In a greenfield investment, a foreign firm starts operations on its own in a host country. The result will be the creation of new capacity/supply in a given country. One of the main implications of a greenfield investment is that the start-up firm must provide all the necessary resources and capabilities itself to overcome structural barriers, as well as risks due to liabilities of foreignness.

Donald (2008) describes an acquisition as the purchase of one business or company by another company or other business entity. In this thesis an acquisition would identify the takeover of 95% or more of all stocks from an established firm within the host country. One feature of an acquisition is that it could be conducted much faster compared with a joint venture or greenfield entry (Elango & Sambharya, 2004). One of the most important benefits of an acquisition is that the investing firm could be in full control of the established firm, thereby overcoming structural barriers in the host country and the liability of foreignness rather quickly. A difference with greenfield modes is the fact that acquisitions do not increase industry supply while greenfield modes do increase industry supply.

(12)

12 An important issue for a firm that has decided to expand into foreign countries is to make a decision between keeping and sharing control (Arregle, 2006). Control describes the amount of authority a firm has over the daily operations of its foreign affiliate. Normally, a firm's equity share, or ownership share, determines the amount of control a firm could exercise over the foreign affiliate (Anderson & Gatignon, 1986). The partner with the highest equity share in the foreign enterprise also has the most investment risks. Therefore the party with the largest ownership share often exercises the most authority over the daily operations. Douglas & Craig (1995) argue that each entry mode is associated with a certain amount of control. Generally, low control is associated with an entry mode that involves a small commitment of resources and full control is associated with an entry mode that involves a large commitment of resources. Within this thesis greenfield and acquisition modes hold the assumption of full control since a large commitment of resources is required. Joint ventures and other shared control modes identify cooperative modes of entry since a small commitment of resources is required.

2.2 Entry success

Every firm aims to achieve the most success in the market. However, it is hard to capture what universally is meant with success. Businesses tend to have different indicators of success according to the internal and external environment they operate in. Hence one of the most disputed issues in the field of studying success and failure of international market entry is to define and measure it (Johnson & Tellis, 2008). This is so because it is still not clear what the exact indicators of success are. Attempts to overcome this problem by the survey method have led to the well-known self-reporting bias (Golder & Tellis, 1993). Additionally, Luo (1998) argues that success is a time dependent phenomena, and at any given time, it may only be partial. To evade this problem, scholars have used multiple measures of success in the past, such as market share and profitability (Pan, Li & Tse, 1995), hazard rates (Li, 1995), and timing (Luo, 2001).

(13)

13 success, assessed on a five point scale. This graded measure of success enabled them to have a clearer overview of what success exactly is and measure the degrees of success. For this thesis the graded measure of success from Johnson & Tellis (2008) is used as an indicator for success since it gives a combined overview of the term based on preliminary research. The graded measure will further be outlined in the methodology chapter of this thesis.

2.3 Differences between developing and developed market MNCs

Developing markets constitute a tremendously diverse population of countries (Gammeltoft, Barnard & Madhok, 2010). Therefore it is hard to make generalizations across the group of countries. Nevertheless, a variety of features do set developing economies and their firms apart from their developed-world counterparts. A range of different specifications can be identified due to the authentic environment of the home country. First off all, in many developing markets, the (local) government plays a larger and more active role in the economy (Gammeltoft et al., 2010). This affects the national firms from these developing markets in the sense that they are likely to be more adapted to the preferences of the government. The support of the government could provide them with multiple privileges like: easier access to certain inputs, preferential financing, subsidies and other support. Due to their close relationship with the government, firms from developing markets tend to operate according to somewhat different logics when internationalizing. The government from the home country might specifically encourage their national firms to internationalize in order to strengthen international competitiveness. On the other hand, there are also other reasons for developing economy firms to internationalize. Especially volatility in the home market and efforts to escape bureaucratic restrictions apply as reasons to expand abroad (Peng, Wang & Jiang, 2008).

(14)

14 This has implications for the organization and governance of firms. For example, developing economy ﬁrms tend to be more horizontally and vertically integrated. In other words, developing market MNCs will hold and develop more capabilities in house and not buy some of these capabilities in the market like most developed and lean MNCs would. Furthermore, Khanna & Yafeh (2007) argue that MNCs from developing countries have a greater likelihood to organize themselves in diversiﬁed business groups in order to economize on scarcer technological, managerial, physical and other resources.

Third, firms from developing countries are often more reliant on social networks with ethnic, linguistic or cultural afﬁnities, in other studies also described as relational assets (Dunning & Narula, 2004). Due to the limited experience with having an international business many developing market MNCs tend to work with more closed networks and more personalized governance and control systems. As a consequence, expats play often a very important role in developing markets (Cho & Lee, 2003).

(15)

15 3. HYPOTHESES DEVELOPMENT

In this section hypotheses will be developed based on two dependent variables: entry mode decision and entry success. Both dependent variables are well known in the foreign expansion strategy literature. They have multiple characteristics in common since MNCs will most likely chose an entry mode which will predict the most success. However, within the literature the underlying determinants (independent variables) between the two tend to differ. Johnson & Tellis (2008) argue that there are multiple determinants which influence entry success and that entry mode selection is only one of them. Therefore this chapter takes both dependent variables and their underlying determinants into account.

3.1 Determinants of entry mode decision (1st dependent variable)

Cultural distance

(16)

16

Hypothesis 1a: Cultural distance is positively associated with South-African MNCs choosing cooperative entry modes rather than wholly owned subsidiaries.

On the other hand, the absorptive capacity of local partners is likely to be lower when there is high cultural distance. Local partners could have difficulties in implementing competitive advantages effectively since these advantages are likely to be harder to codify when cultural distance is large. Based on this, Kogut & Singh (1988) argue that firms might prefer to transfer its specific advantages in an internalized manner when cultures are very different. Through internalization firms are likely to be able to transfer other firm specific advantages more effectively which might help to compensate for the liability of foreignness that results from high cultural distance (Mahnke & Venzin, 2003). Furthermore, based on the transaction cost theory, cultural distance increases information asymmetry which consequently leads to increased monitoring costs. Hence, internalizing foreign activities would be more efficient and large cultural distance might lead to the desire to exert greater control over foreign activities (Tihanyi et al., 2005). Another argument advanced by Barkema, Bell & Pennings (1996) maintains that cooperation with local partners would double the complexity since the company expanding abroad would have to cope with the foreign culture of customers and with the different corporate culture of a cooperative partner (Brouthers & Hennart, 2007). It is suggested that wholly owned subsidiaries could avoid this problem. Therefore:

Hypothesis 1b: Cultural distance is positively associated with South-African MNCs choosing wholly owned subsidiaries rather than cooperative entry modes.

Market attractiveness

(17)

17 firm efficiency and hence firm profits. Brouthers (2002) argues that in these countries characterized by high market attractiveness, companies are expected to use vertical integration so that they could gain economies of scale and secure a long-term market presence. Additionally, firms tend to use wholly owned subsidiaries in attractive markets to gather all the financial rewards themselves without sharing profitability with any partners.

However, in the past different facets of market attractiveness have been described. Most authors argue that increasing market size leads to an enhanced resource commitment in the country. In other words: market size is positively related to internalization and wholly owned modes of entry (Erramilli, Agarwal & Kim, 1997). In large host countries firms could expect more risk since more resources have to be invested in these large markets. In line with the higher risks, Agarwal (1994) argues that firms could also expect higher returns. Higher returns are expected to come from the opportunity to gain economies of scale, based on the assumption that a high proportion of the costs of internalization is fixed (Buckley & Casson, 1996). Furthermore, according to the transaction cost theory, market size is also an indicator of transaction frequency, which also enhances the firm's tendency to internalize (Williamson, 1985). Thus:

Hypothesis 2: Market size is positively associated with South-African MNCs choosing wholly owned subsidiaries rather than cooperative entry modes.

Another proxy for market attractiveness is income level. In prior research, high income indicated stable economic conditions and therefore high market potential. Additionally, high income level corresponds with higher standards of living (Erramilli et al., 1997). A high standard of living is very attractive to most firms since it indicates that wealth is properly distributed and that a big part of the population is able to buy their products. This is likely to have a positive effect on sales. Consequently, with increasing per capita income, a higher likelihood of full ownership is assumed (Erramilli et al., 1997). Therefore:

(18)

18

Country openness

The term openness refers to the lack of regulatory and other obstacles to entry of foreign firms (Johnson & Tellis, 2008). A country's openness to foreign investment is likely to facilitate operations in the market better (Morschett et al. 2010). Hence it will improve the situation for entrants and make full ownership in an open country more attractive. In addition, an effect of the host country's openness to foreign investment on the entry mode choice is argued based on the follow the-client hypothesis. In countries whose economies are more open to international investment, it is likely that more foreign firms are active. In these international markets it is therefore also more likely that business already know each other from other countries of operation. Dunning & McQueen (1982) argue that it might be better to enter these open markets with a fully owned mode since most firms who are already active in the host country like to buy from a well-known company and from a long term supplier with who they already have a business relationship. When the firm would enter with a cooperative mode it will be harder for them to start these relationships due to the other party that is involved (Dunning & McQueen, 1982). This other party could have a negative influence on the image of the entering firm. Thus:

Hypothesis 4: Country openness is positively associated with South-African MNCs choosing wholly owned subsidiaries rather than cooperative entry modes.

Country risk

(19)

19 given firm (Agarwal, 1994). Transaction cost reasoning ignores the advantages of strategic flexibility (Brouthers, Brouthers & Werner, 2008). In a high risk country firms must create necessary flexibility in order to change to a different mode of entry when the initial mode becomes less efficient due to unpredicted changes in the environment (Anderson & Gatignon, 1986; Brouthers, 2002). Real option theory suggests that firms should keep the initial investment in a host country low when the value of an investment cannot be accurately predicted due to uncertainty (Brouthers et al., 2008). In this case cooperative modes are an attractive alternative. Furthermore, the resource based view (RBV) perspective argues that firms should avoid full ownership in high risk countries so that they could protect their resources better (Agarwal & Ramaswami, 1992). Therefore:

Hypothesis 5: Country risk is positively associated with South-African MNCs choosing cooperative entry modes rather than wholly owned subsidiaries.

The following table sums up the expected effect all independent variables have on the dependent variable entry mode decision:

Table 1: All independent variables and their expected impact on entry mode decision

Independent variable Expected effect on entry mode selection Cultural distance High cultural distance is expected to have a

positive impact on both cooperative and wholly owned modes of entry.

Market size Large market size is expected to have a positive impact on wholly owned subsidiaries.

Income level High income level is expected to have a positive impact on wholly owned subsidiaries.

Country openness High country openness is expected to have a positive impact on wholly owned subsidiaries. Country risk High country risk is expected to have a positive

(20)

20 3.2 Determinants of entry success (2nd dependent variable)

Entry mode

Anderson & Gatignon (1986) argue that the key difference between all modes of entry is the degree to which they give a firm control over its key marketing resources. The RBV and the transaction cost view predict alternative outcomes of success when choosing an entry mode. The RBV suggests that when degree of control increases, chances of success increase because the firm can deploy and protect key resources that are essential to success (Gatignon & Anderson, 1988). These key resources consist out of intangible and tangible properties. When a firm has control over these properties it has the freedom to use them flexibly which would enhance the chances of success. Overall, control provides two main benefits for firms. It safeguards key resources from leakage and it allows for internal operational control (Luo, 2001). Additionally, internalized firms are able to control key complementary resources better, such as access to local distribution channels. These complementary resources could be important to success in any given country. Therefore:

Hypothesis 6a: Wholly owned entry modes are positively associated with entry success of South-African MNCs.

The transaction cost view suggests that total cost increases with an increasing level of control. Additionally, transaction cost theory argues that the higher the resource commitment and desired control of an entry mode, the higher the cost. Wholly owned subsidiaries are categorized as expensive entry modes since a lot of resources are required to set up these operations (Pan & Chi, 1999). The entering firm is responsible for all of its expanses and is not able to share these expenses with partnering firms. These higher costs imply that higher levels of revenues are needed for the firm to break even and make a profit. Cooperative modes of entry entail less investment cost and therefore:

(21)

21

Entry timing

The role of market entry timing is critical in entering new markets (Pan & Chi, 1999). Prior literature suggests many reasons how early entry into international markets could positively enhance success. First of all, early entrants are able to lock up access to important resources such as distribution channels and suppliers. These key resources could host more value than other resources since early entrants could pick the best resources available. Second of all, early entrants have the opportunity to set consumer preferences (Mitchell, 1999). By shaping the market to their offering, firms could gain an additional advantage since consumers would be more inclined to buy their offering. Third of all, early entrants are able to benefit from governmental concessions and incentives. Some governments are keen to attract foreign entrants and offer valuable benefits to ‘’early movers’’ (Pan & Chi, 1999). Lastly, early entrants could gain valuable experience which could be an advantage towards later entrants. Therefore:

Hypothesis 7: Early entry is positively associated with entry success of South-African MNCs.

Firm size

There are several reasons that explain why larger firms have greater success than smaller firms. First of all, larger firms have the (financial) resources to exploit and attract other successful resources (Bonaccorsi, 1992). For example, large firms have the ability to acquire successful firms while most smaller firms do not have this ability. Second of all, larger firms are likely to possess more product and marketing specific knowledge than smaller firms. This knowledge could be crucial in concurring new markets. Third of all, larger firms are more capable to survive periods of negative performance than smaller firms. While smaller firms may not have additional resources to survive, big firms could survive a crisis and profit from reduced competition. Therefore it is arguable that:

(22)

22 Conversely, history shows that larger firms are not always more successful than smaller firms. The withdrawal of Wal-Mart from the Korean and the German market is remarkable since Wal-Mart is one of the largest firms in the world. There are a couple arguments which could explain these withdrawals and more generally, the main disadvantage of large corporations. Hitt & Cheng (2003) argue that large size diminishes organizational flexibility because of increasing bureaucracy. Additionally, Chandy & Tellis (2000) argue that this increasing bureaucracy has also a negative effect on the innovative capabilities of most firms. Hence:

Hypothesis 8b: Firm size is negatively associated with entry success of South-African MNCs.

Cultural distance

Cultural distance affects the effectiveness of an entry. It affects the underlying behaviour of customers in a host market and the execution and implementation of marketing and management strategies (Kogut & Singh, 1998). Additionally, Leclerc, Schmitt & Dube (1994) argue that cultural distance affects how customers derive meanings about the brand or product. In countries with high cultural distance, customers could derive opposite meaning as expected about the product due to the dissimilarity in cultures. This has a potential negative influence on success. Furthermore, the fact that most firms start their international activities in countries that are quite similar is an example of how culture influences market entry (Terpstra, Sarathy & Russow, 2006). Frankel & Rose (2002) argue that linguistic similarity is a far more powerful determinant of the volume of trade between countries than economic factors, such as exchange rates and trade barriers. Additionally, Barkema et al. (1996) show in their study that cultural barriers have a negative impact on organizational learning and hence they decrease a firm’s longevity in countries which are very cultural distant. Therefore:

(23)

23

Country openness

The term openness refers to the lack of regulatory and other obstacles to entry of foreign firms. Johnson & Tellis (2008) argue that a country’s openness positively influences entry success due to three reasons. First of all, country openness stimulates demand since it increases the variety of products offered for sale in the market. An increase in demand would most likely correspond to an increase in profits for a firm. Second of all, country openness increases competition on quality and hence it will improve the overall level of quality supplied. This improved level of quality would most likely have a positively influence on the level of sales. Third of all, as a country becomes more open, competition increases efficiency and lowers prices. As a result an increase in demand would likely occur. Therefore:

Hypothesis 10: Country openness is positively associated with entry success of South-African MNCs.

Country risk

(24)

24 hence the profits of a company (Erb et al, 1995). To sum up, the three sources of country risk host several disadvantages which negatively affect the success of an entrant. Hence it could be argued that:

Hypothesis 11: Country risk is negatively associated with entry success of South-African MNCs.

The following table sums up the expected effect all independent variables have on the dependent variable entry success:

Table 2: All independent variables and their expected effect on entry success

Independent variable Expected effect on entry success

Entry mode selection Wholly owned and cooperative modes of entry both are expected to have a positive impact on entry success.

Entry timing Early entry is expected to have a positive

impact on entry success.

Firm size Large firm size and small firm size are

expected to both have a positive impact on entry success.

Cultural distance Low cultural distance is expected to have a positive impact on entry success.

Country openness High (host) country openness is expected to have a positive impact on entry success. Country risk Low (host) country risk is expected to have a

(25)

25 3.3 Developing vs developed market (MNCs)

To improve their competitiveness, especially in terms of seeking to compete in global markets, firms based in developing economies are likely to want to enter developed economies to develop new resources and capabilities (Hoskisson et al., 2004). Firms coming from developing economies are likely to be at a disadvantage in global markets relative to firms coming from developed economies in terms of experience and assets they possess. Cantwell (1992) argues that developing market firms may be forced to enter developed markets in order to develop knowledge bases that are globally competitive. Acquiring new technological capabilities is likely to be one of the main goals. This international exploration allows them to catch up with developed market firms and enhance their long term performance (Frost, Birkinshaw & Ensign, 2002). In order to learn and develop new (technological) capabilities, the acquisition of developed market firms and their assets is likely to be the easiest and quickest way.

In addition, large distribution networks could pose a big obstacle to firms from developing economies that would like to enter markets in developed economies. Most developed markets require efficient and large distribution networks. These large distribution networks consist out of retailers, suppliers and other key stakeholders. Most developing market MNCs do not possess the capabilities to manage and set op these large distribution networks in developed markets. As a result, firms from developing economies are likely to use acquisitions to enter developed countries where a comparable firm from a developed economy would use greenfield operations. By acquiring a developed market MNC problems with the distribution network could be solved. For example, South African Breweries (SAB) established a strong distribution network and presence in the United States by acquiring Miller in 2002.

(26)

26 result of the acquisition, developing market firms could trade under the name of the acquired firm and avoid the conceptions. Hence:

Hypothesis 12: MNCs from South-Africa would favour acquisitions over greenfield and a joint venture mode of entry in developed countries.

Most developing markets have to deal with a weak institutional environment (Holtbrügge & Baron, 2013). The collaboration with a local partner is aimed to compensate difficulties arising from institutional pressures when entering a developing market. There are multiple arguments that describe the benefits of collaborating with a local partner in developing markets. First, foreign entities may obtain legitimacy easier with the help of a local partner (Xu & Shenkar, 2002). Local partners are likely to have better relationships with important stakeholders in the host country. These relationships could be essential in ‘’locking up’’ key resources. Second, local partners are able to take care of potential problems arising from institutional weaknesses such as insufficient intellectual property rights protection and corruption in developing markets (Holtbrügge, 2004). As a foreign firm with no experience how to deal with host country specific weaknesses, collaborating with a local partner could be crucial. Given these advantages of collaborating with a local partner it seems more likely that MNCs would enter developing markets through a cooperative entry mode and therefore:

Hypothesis 13: MNCs from South-Africa would favour a cooperative mode over a wholly owned mode of entry in other developing markets.

(27)

27 cost are lower than for firms coming from developed markets when they compete in developing economies (Geringer, Tallman & Olsen, 2000). On the other hand, the focus of developing market MNCs in developed countries mainly lies on exploration instead of exploitation (Wright et al, 2005). The goal of these firms in developed markets is not to enhance performance but to develop absorptive capacity and new capabilities (Zahra & George, 2002). This makes them on average less competitive in developed markets compared to developing markets. Hence, it could be argued that developing market MNCs will be more successful entering other developing markets instead of developed markets since their focus will lay more on exploitation instead of exploration and because of their distinctive capabilities they maintain while competing in a similar market. Therefore:

(28)

28 4. METHODOLOGY

4.1 Data

The methodology is an important way to structure actions of the research. A well-formulated problem indicates what kind of entities is expected to treat as sources of data (Thomas, 2004). Data collected for this thesis is of quantitative data type. For this study quantitative data is collected since it provides the most current information about the international expansion activities of South African MNCs.

The sampling frame for this study consisted out of 14 large MNCs from South Africa, according to their equity price, listed on the Johannesburg Stock Exchange (JSE) who conducted outward FDI during the ten year period during from 2001 until 2011. For this thesis not the 14 biggest MNCs listed on the JSE were investigated due to two reasons. First, some MNCs like British American Tobacco, Vodacom and Woolworths listed on the JSE are subsidiaries from a foreign (developed market) parent company. This study would like to discuss MNCs from South-Africa who started and grew up in the developing context not in a developed context. Therefore firms with a foreign (developed market) parent are not included in this research. Second, South-Africa is home to a lot of big mining and resources corporations. To control for industry characteristics of internationalisation some big MNCs in the resources industry were left out. In appendix A the list off all companies, their annual turnover and their industry is shown.

(29)

29 widespread international operations, including developed and developing countries. Second of all, stock listed companies publish audited annual reports from which foreign expansion moves are clearly reported. These annual reports were obtained from the corporate websites of these MNCs. Prominent researchers in the field of MNC international strategy and foreign expansion have previously used company reports as the principal data source (Gatignon & Anderson, 1988; Wiliams et al., 2011).

Evidence of new entries into foreign markets was found by a qualitative search on the English version of the company report using keywords such as: acquisition, greenfield, joint venture, start-up, investment and partnership. At the end of the report most companies had a section which showed their equity investments during the year. This was extremely helpful in cases where it was ambiguous how big the ownership share was in the new subsidiary. Company reports showed also to be a good help when researching the success of each entry. Evidence of entry success was gathered in company reports which were issued 1 and 2 years after the initial entry by searching through keywords such as the name of the new entity. The information given about the success in the report was not always very objective, it supplied however a good indication about the actual performance of the investment. The results were later always double checked through other sources of information.

In addition to company reports other sources were used to gather data on entry success. Electronic sources, such as LexisNexis and Orbis were used to receive data about market success.In these electronic sources data was gathered by searching for the name of the new entity and the name of the parent company. In some cases also helpful articles about the entry were found. Golder & Tellis (1993) argue that archival data must meet the following criteria to ensure validity:

•Competence: the capability of the informant to report correctly, •Neutrality: the lack of vested interest by the informant of the report, •Reliability: a long record for undisputed good reporting by the informant, •Corroboration: confirmatory evidence from a similar source, and

(30)

30 The competence criterion is met in this thesis because the data gathered comes from well-known sources and is from the time frame when the firms entered the host country. The neutrality criterion is satisfied because neutral authors composed the information. The reliability criterion is satisfied because all the sources that were used are reputable. Additionally, the corroboration criterion is met since for every entry at least two data sources were used. Furthermore, contemporaneity is satisfied because the data sources sorted the data with the oldest first to ensure that the reports closest to the event (the entry) are included in the sample.

After all, 174 entries were identified. In all cases the selected entry mode was clear. However, only in 95 cases the success of the entry was successfully identified. This was mainly due to the difficultness of receiving useful information about the success. Especially for smaller entries that did not have a great impact on the overall performance of the MNC, retrieving information about the success was hard. These smaller entities were too small to be found in the company reports and also too small to be found in any additional data sources. Therefore this thesis will work with two datasets which will entail a different size. A size of n=174 to measure entry mode selection and a size of n=95 to measure entry success. A size of n=95 might bring along some risks. One risk could be the possible bias on the findings since the smaller the sample size, the greater the chance of coincidences.

4.2 Dependent variables

Two dependent variables were included in this study: entry mode decision and entry success.

Entry mode

(31)

31 greenfield and acquisition modes as the same type of entry mode (wholly owned modes). To give more clearness in the outcomes of the study, this thesis makes a distinction between greenfield and acquisition modes since both have distinct features which sets them apart. However, a lot of the hypotheses in this thesis differentiate only between wholly owned and cooperative modes due to the fact that most of the research in the past also only differentiated between these two modes. Dummy variables will be appointed to differentiate between the three different types of entry modes. A 0 will account for a greenfield mode, a 1 will account for a cooperative mode and a 2 will account for an acquisition mode of entry.

As argued, a big part of the hypotheses are drawn based on wholly owned modes and cooperative modes. When differentiating between wholly owned (greenfield and acquisition) and cooperative (joint venture) modes, this study will follow the majority of the literature (Brouthers, 2002) to get to a decent cut-off point between a joint venture and a wholly owned subsidiary. An ownership share of 95% will be used as the cut-off point between a cooperative and a wholly owned subsidiary. In other words: an entry mode is identified as a wholly owned mode if the South African investing firm hold 95% or more equity ownership in the foreign venture and an entry mode is identified as a cooperative mode if the South-African investing firm hold between 5% and 94% equity ownership in the foreign venture. Dummy variables will be appointed to differentiate between wholly owned modes and cooperative modes. A 1 will account for a cooperative mode and a 0 will account for a wholly owned mode.

Entry success

(32)

32 Successful Entry: 5

• Market share leader

•Making more margins than their global margin

•Top three in market share •Well-functioning partnership •Above-average industry leadership •Exceeded investment criteria •Top three in industry profitability

Poor Entry: 2

•Fail in system integration and optimization

•No initial lead buyers •Underperformance •Conflicting expectations •Struggled to make headway •Priced out

•Stiff competition •Market restrictions

•Executives frustrated with entry Good Entry: 4

•Successfully selling •Met investment criteria •Growing shipments •Increasing investments

•Rapidly evolved into a major force in the industry

Failed Entry: 1

•Quit or withdrawal from market •Break up with cessation of venture

Acceptable Entry: 3 •Continuing operations

•Hope to recover investment in time •Entry awaiting removal of market restrictions

(33)

33 4.3 Independent variables

Cultural difference

This study employs the measure of cultural distance between the host and the home countries from Hofstede’s (1991) four cultural dimensions: power distance, individualism– collectivism, masculinity–femininity, and uncertainty avoidance. Following Kogut & Singh (1988), this thesis collapsed the individual scores into a single number by taking the distance of the four dimensions as follows:

CDsmt =

Where CDsmt is the country distance score between host country s and home country m in year t, Djst is the score on dimension j for host country s, and Djmt is the score on dimension j for home country m both measured in year t. Mitra & Golder (2002) argue that this measure of cultural distance has a long history of use in both the international marketing and the strategy literature.

Market size

The Gross Domestic Product (GDP) for the host country in the year prior to entry was used to measure market size. The information needed was obtained from the World Bank database. Log transformation was used to normalize the distribution of the variable.

Income level

(34)

34

Country risk

The measure of country risk needs to capture political, regulatory, and economic sources of risk (Erb, et al., 1996). Country risk indicators will be captured from the international country risk guide. Erb et al. (1997) argue that this guide possesses the greatest accuracy. This measure of country risk is based on a multidimensional measure for each component of country risk through political, financial, and economic risk.

A composite of country risk is produced by combining these three measures according to the following formula:

CPFER = 1/3 (PR + FR + ER)

Where CPFER is a composite of political, financial, and economic risk ratings; PR is the total political risk indicators; FR is the total financial risk indicators; and ER is the total economic risk indicators. The highest overall rating (theoretically, 100) indicates the lowest risk, and the lowest rating (theoretically, 0) indicates the highest risk.

Firm size

In the literature, firm size is typically measured by sales, number of employees, or assets. In this study, size of a firm based is measured based on its global sales volume of the year prior to the entry. Data was gathered from annual reports. Additionally, log transformation was used to normalize the data.

Entry timing

(35)

35

Country openness

The measure of openness is based on the fraction of foreign direct investment as a function of the host country’s GDP. FDI data will be gathered from the United Nations Conference on Trade and Development (UNCTAD) website. At the end the fraction was multiplied by a 100 to make the distribution more normal.

4.4 Differentiating between developing and developed countries

To differentiate between developed and developing countries, all entries will be split into two samples: entries into OECD countries and entries into non-OECD countries. The current members of the OECD represent the most economically powerful countries in the world. Their membership attests to the benefits of belonging to a free market system that fosters economic and industrial development (OECD, 2013). Along with the ties it maintains with some of the world’s emerging economies, all OECD countries together account for almost 80% of the world’s trade and investment (OECD, 2013). Therefore, OECD membership seems to be solid base to differentiate between developed and developing countries.

(36)

36 5. RESULTS

The results will be analysed according to the three building blocks of this thesis: entry mode decision, entry success and the differences between developed and developing countries.

5.1 Entry mode decision

In appendix B it is shown that the foreign expansion is conducted across a wide range of locations. Eventually, out of 174 entries, 89 wholly owned modes and 85 cooperative modes were conducted. Notable is the fact that a big part of the foreign expansion into developed markets was done into countries that have English as their primary (business) language and more specifically countries that like South-Africa were part of the former British Empire. Australia, the USA and the United Kingdom capture 44.4% of all entries into developed economies. Table 3 shows the descriptive information for the dataset. The 63 countries captured in the dataset represent a sufficiently wide range of cultures, country risk, market size, income level and country openness. All of the variables were normally distributed and were therefore perfectly suited for this study.

Table 3: Descriptive statistics of variables (entry mode decision)

N Minimum Maximum Mean Std

Cultural distance 174 19,1 73,9 47,2 12,18

Country risk 174 52,0 92,3 74,18 8,36

Market size 174 21,24 30,34 26,59 1,21

Income level 174 4,89 11,33 8,88 1,80

Country openness 174 3,4 97,8 33,92 20,19

(37)

37 Xu & Shenkar, 2002). However, all three variables still represent a different meaning and outcome. Therefore, this thesis continues to use all three variables independently.

Table 4: Correlation matrix (entry mode decision) Cultural

distance

Country risk Market size Income level Country

openness Country risk -0.308** Market size -0.357** 0.579** Income level -0.380** 0.803** 0.689** Country openness 0.055 0.315** -0.016 0.310** OECD -0.551** 0.728** 0.605** 0.841** 0.344** ** Correlation is significant

(38)

38 Table 5: Logistic regression results (beta coefficients and standard errors)

Model 1: All locations Model 2: OECD countries Model 3: Non-OECD countries Cultural Distance -0.032 (0.017) -0.007 (0.031) 0.004 (0.027) Market Size 0.194 (0.123) 0.693 (0.342)* -0.031 (0.145) Income level -0.027 (0.0194) -0.851 (0.86) -0.218 (0.254) Country openness 0.034 (0.011)** 0.029 (0.02) 0.035 (0.015)* Country Risk 0.082 (0.038)* 0.079 (0.083) 0.118 (0.05)* -2 Log likelihood 183.192 78.258 94.049 Chi-square 57.931*** 4.976 14.218*

Cox and Snell 0.283 0.06 0.142

N 174 81 93

*** p<0.001, ** p<0.01, * p<0.05

In table 6 and 7 all descriptive information of all variables is shown. As explained in the methodology chapter, the sample size of entry success consists out of n=95 entries. All scale variables represent a wide range and are all normally distributed. Additionally, the entry mode variable represented in table 7 shows no spectacular differences between the two different modes. With a difference of a bit more than 8% the two modes almost have the same total amount and are therefore perfectly suited for further valid analysis.

Table 6: Descriptive statistics of variables (entry success)

N Minimum Maximum Mean Std

(39)

39 Table7: Descriptive statistics of the variable entry mode

Frequency Percent

Cooperative mode 52 54.7%

Wholly owned mode 43 45.3%

Total 95 100%

The correlation matrix is shown in table 8. Like in the former section (entry mode decision) a relatively high correlation is found between OECD membership and country risk. This could be explained by the fact that most developing countries suffer from higher levels of risk compared to developed countries. In addition, the variable entry mode has a strong correlation with the variables country risk and OECD membership. The relationship between entry mode and country risk could be easily explained by the former section whereby country risk has a significant impact on entry mode decision. The correlation between entry mode decision and OECD membership could be explained by the aim of this paper since the author argues that there would be notable differences between entry modes into developed and developing markets. Hence, no variables were skipped and the decision was made to continue the study as planned.

Table 8: Correlation matrix (entry success) Cultural

distance

Country risk

Entry mode Timing Firm size Country

openness Country risk -0.202* Entry mode -0.221* 0.503** Timing -0.071 -0.170 -0.041 Firm size -0.090 0.013 -0.020 0.363** Country openness -0.089 0.199 0.372** 0.195 0.174 OECD -0.482** 0.650** 0.572** -0.181 -0.067 0.322**

** Correlation significant at the 0.05 level, * Correlation significant at the 0.01 level

(40)

40 No support is found for H7, H8 and H10. However, when the full model is split into developed (model 2) and developing (model 3) markets some interesting results appear which clarify it even more that there is a significant difference between the impact of most variables when expansion is either done into developed or developing countries. When expansion is done into developed markets the entry mode decision influences the entry success significantly. A more wholly owned mode seems to have a fair amount of impact on the entry success rate. Other variables in model 2 seem to have no severe impact on entry success. When expansion is done into developing markets both timing and country risk influence entry success significantly. An early entry and a low country risk cause a higher entry success in developing countries. However, this support is not huge since the variables suffer from relatively small beta’s. Other variables in model 3 seem to have no important impact.

(41)

41 5.3 Developing vs developed market (MNCS)

Table 10 represents the cross-tabulation table which is used to answer H12 and H13. In this table which represents OECD membership as developed countries, membership is linked to the entry mode decision. There appears to be an association between OECD membership and entry mode decision based on a significant pearson chi-square score (p=0.000<0.05). A cooperative mode is more likely than a greenfield or an acquisition mode of entry when expansion is done into developing markets by South-African MNCs. In 73.1% of the cases South African MNCs chose a cooperative entry mode into developing markets. This result indicates a strong support for H13. Additionally there is also support for H12. An acquisition is the most likely mode of entry when expansion is done into developed markets by South-African MNCs with a rate of 66.7%.

Table 10: Cross-Tabulation table (entry mode and OECD membership)

OECD Total 0 1 Entry mode 1 Count 12 10 22 % within OECD 12,9% 12,3% 12,6% 2 Count 68 17 85 % within OECD 73,1% 21,0% 48,9% 3 Count 13 54 67 % within OECD 14,0% 66,7% 38,5% Total Count 93 81 174 % within OECD 100,0% 100,0% 100,0%

(42)

42 65% of the entries in developed markets scored a 4 or a 5. These results clearly indicate that there is no support for H14.

Table 11: Cross-Tabulation table (entry success and OECD membership)

(43)

Table 12: Schematic overview of the results

FULL SAMPLE DEVELOPED COUNTRIES DEVELOPING COUNTRIES

H1a Cultural distance is positively associated with a cooperative entry mode

Coefficient -0.032 Coefficient -0.007 Coefficient 0.004

P-value N.S. P-value N.S. P-value N.S.

H1b Cultural distance is positively associated with wholly owned entry modes

H2 Market size is positively associated with wholly owned entry modes

Coefficient 0.194 Coefficient 0.693 Coefficient -0.031

P-value N.S. P-value * P-value N.S.

H3 Income level is positively associated with wholly owned entry modes

Coefficient -0.027 Coefficient -0.851 Coefficient -0.218

H4 Country openness is positively associated with wholly owned modes

Coefficient 0.034 Coefficient 0.029 Coefficient 0.035

P-value ** P-value N.S. P-value *

H5 Country risk is positively associated with a cooperative mode of entry

Coefficient 0.082 Coefficient 0.079 Coefficient 0.118

P-value * P-value N.S. P-value *

H6a Wholly owned modes are positively associated with entry success

P-value N.S. P-value * P-value N.S.

H6b Cooperative entry modes are positively associated with entry success

H7 Early entry is positively associated with entry success

Coefficient -0.042 Coefficient 0.028 Coefficient -0.086

P-value N.S. P-value N.S. P-value *

H8a Firm size is positively associated with entry success

H8b Firm size is negatively associated with entry success Coefficient -0.008 Coefficient -0.016 Coefficient 0.073

H9 Cultural distance is negatively associated with entry success

Coefficient -0.019 Coefficient -0.018 Coefficient -0.006

P-value ** P-value N.S. P-value N.S.

H10 Country openness is positively associated with entry success

Coefficient 0.003 Coefficient -0.002 Coefficient 0.002

P-value N.S P-value N.S. P-value N.S.

H11 Country risk is negatively associated with entry success

Coefficient 0.033 Coefficient -0.010 Coefficient 0.031

(44)

6. DISCUSSION

As with the results, the discussion chapter will be divided according to the three building blocks of this study: entry mode decision, entry success and differences between developed and developing country MNCs.

6.1 Entry mode decision

The regression model does not support the first hypothesis, since no significant relationship exists between cultural distance and the establishment chosen. Although two different directions were proposed, the empirical evidence is not conclusive. However, the beta coefficient is negative which points towards a positive relationship between cultural distance and cooperative entry modes. Other studies found the opposite relationship (Barkema et al., 1996; Barkema & Vermeulen, 1998) or did not report a significant effect (Brouthers & Brouthers, 2000). Furthermore, in their meta-analysis, Morschett et al. (2010) found no empirical evidence for the relationship between cultural distance and entry mode decision.

(45)

45 distant between both South-Africa and Uganda. Second, the model of Hofstede (1991) which was used to measure cultural distance might be outdated. One potential problem about this measurement lies in the fact that, for some countries, no indicators in the index are available. To overcome this difficulty, scores were assigned to these countries based on the score of others that were supposed to be culturally similar (Erramilli, 1991). Another potential problem is the fact that South-Africa itself is very culturally different. Due to its remarkable past South-Africa is a largely divided country with 11 official languages and 26 practiced religions. This makes it hard to determine a solid cultural index for South-Africa.

Also no support was found for the two variables which captured market attractiveness within the full sample regression analysis. This is contradictory with previous studies which suggested that the greater the market attractiveness, the greater the probability that the MNC will opt for a wholly owned entry (Brouthers & Brouthers, 2002; Taylor et al., 1998). A reason for the finding that market size does not have a great impact on the entry mode decision might be that the large number of competitors likely to be present in large markets might reduce the attractiveness of entry (Gomes-Casseres, 1990). In addition, the argument that the high amount of fix cost needed to invest and as an outcome the steep economies of scale to posit the use of wholly owned subsidiaries may be over evaluated (Morschett et al., 2010). Because firms need to invest more resources in larger markets, it is possible to argue that the necessary resources are easier to obtain through a partnership. A reason why no support is found for the influence of income level on entry mode decision might be the use of the wrong measuring instrument (Morschett et al., 2010). GDP per capita does not only explain a more attractive market but also a more stable institutional setting and market sophistication. In high income countries there usually is better protection of property rights and therefore the transaction costs are likely to be lower when enforcing rights in a partnership (Morschett et al., 2010). Therefore GDP per capita might be not an optimal indicator since the variable has effects that go beyond market attractiveness.

(46)

46 views the relevance of the host country’s GDP. An explanation for this could be based on the argument that high proportion of fixed costs and correspondingly steep economies of scale that is used in the literature to posit a positive influence on the propensity to use a wholly owned mode may be over-evaluated. Most developing markets have to deal with a weak institutional environment (Holtbrügge & Baron, 2013). This makes MNCs more cautious when they have to invest a high proportion of fixed costs in these weak institutional environments. On the other hand, most developed countries are better institutionally developed. This would probably make MNCs less cautious to invest a high proportion of fixed costs into a developed country which is needed to start up a wholly owned mode.

The fourth and fifth hypotheses in this study are supported since both country openness and country risk have a significant relationship with entry mode decision. Most remarkable is the confirmation of the fourth hypothesis which proposes a relationship between country openness and entry mode decision. In earlier studies results are rather inconclusive. One study displays negative significant results, and two studies result in coefficients of 0.00 and -0.000. Contractor & Kundu (1998) argue in their study that the relationship between country openness and entry mode decision should be studied with great caution since the theoretical arguments for a relationship are weak. Nevertheless, in this study a positive relationship is found which proposes that the greater a country’s openness, the greater the chance of a wholly owned entry mode. Less remarkable and more consistent with prior studies, country risk has a significant relationship with entry mode decision. In their meta-analysis Zhao et al. (2004) conclude that country risk is closely related to entry mode choice and that country risk leads to cooperative entry modes. This conclusion is supported in this study. Thereby the argument of resource protection and the enhancement of flexibility are obviously weighed more by South-African MNCs than the transaction cost argument, which supposes that companies desire full control in insecure environments in order to avoid opportunism.

(47)

47 instead of developed countries. Most developed countries have already got a fairly high level of openness and a low level of risk. Therefore variations within these variables will probably not have a considerate effect when expansion is conducted into developed markets. However, country risk and openness will be an important concern when expansion is planned into developing markets. The variations of these variables are higher and the mean of the two are lower which causes them to have a more severe impact.

The regression model does not support the sixth hypothesis since no relationship is found between entry mode decision and entry success. Previous articles in this field discovered a significant positive relationship with wholly owned modes and the success rate (Brouthers 2002; Johnson & Tellis, 2008). This study finds the same direction likely, whereby wholly owned modes causes higher entry success for South-African MNCs. However, stronger and clearer relationships were found when the model was split into developed and developing countries. A significant relationship was even found for developed countries whereby wholly owned modes positively influence entry success. Contrary, when expansion was conducted into developing countries the opposite relationship was discovered. In this case cooperative modes have a positive effect on entry success. This could be explained by the fact that each entry mode has its own distinctive capabilities which fit differently in the two different markets. Wholly owned modes fit better in the developed market where better institutions are in place and cooperative modes fit better in developing markets where institutions are largely underdeveloped.