INTRODUCTION

The growth of international branded hotel chains is a phenomenon of the late twentieth century. It is easy to forget that this growth is quite recent. Conrad Hilton opened his first overseas property in Puerto Rico in 1948 and Pan Am opened its first Inter-Continental in Brazil in 1949 (Gee 1994). In 1998, the world's top 40 chains operated 21,336 properties throughout the globe (Bailey 1998).

As hospitality firms reach maturity in their own markets they expand overseas into new markets. In doing so, such firms need to make choices as to which countries they select for market entry. There is a wide range of criteria for making such entry decisions. This paper explores the extent to which  there  are  differences  between  the attractiveness of a country for general market entry and the attractiveness of a country with regards hotel market entry.

The paper concludes that countries that are generally attractive with regards market entry are not necessarily attractive to hotel chains and vice versa.   Likewise, different firms or brands may select quite different criteria with regards to prioritising target country locations. Finally, entry mode choice is sometimes determined by company policy irrespective of country criteria.

INTERNATIONAL HOTEL CHAIN GROWTH

An international hotel chain can be defined as a firm which has direct investments and other major forms of contractual agreement (such as franchises or management contracts) in more than one country (Dunning and McQueen 1982). The number of such companies is growing considerably. Few however have reached global status - that is to say they have hotel properties on at least five of the six economically viable  continents  (Gannon  and Johnson 1995).

Gannon and Johnson (1995) suggest expansion strategies of such global chains falls into three main groups.  Asian hotel companies had relatively few properties when compared with European and American  chains,  but  a  high  degree  of internationalisation, as measured by the proportion of chains located outside their home country (Inter-Continental and Orient-Express owned by Japanese companies at the time of this study, and Ramada owned by a Hong Kong based company).  North American chains (Choice, Hyatt, Sheraton, Marriott and Radisson) were the least international, with a relatively low proportion of hotels outside the USA, although some groups are actively pursuing internationalisation  strategies.  For  instance, Marriott's corporate franchise deal with Whitbread Hotels in the U.K. Finally European chains, have been more international than American chains but less so than Asian ones.

Since 1995, this situation has not changed very much despite the significant amount of merger and acquisition activity within the industry.  In 1995 there  were  only  9  hotel  companies  with representation in more than 30 countries (Bass? Sheraton and Westin combined as Starwood, Marriott International, Accor, Granada, Hyatt, Hilton International, Carlson, Choice), and this was the same in 1998 (Bailey 1998). Of the world's ten largest hotel chains, in 1998 four chains have more than 90% of their rooms in the USA (Cendant, Promus, Patriot American, Hilton Corporation). So it is clear that internationalisation and size are not related. It is clear therefore that some hotel chains have decided to be represented outside their home country as a matter of deliberate policy and others have concentrated on their home market.

STRATEGIC CHOICE AND HOTEL CHAIN DEVELOPMENT

Once the decision to 'go international' has been made, a chain then has a choice with regards to which countries or destinations to prioritise and what 'entry mode' to adopt. Market entry can be achieved with minimal operational activity and no investment in a country through franchising, or with operational activity but no or limited investment through a management contract, or with operational control and direct investment through the development and ownership of a hotel property in the target country. The decision which of these entry modes to adopt is influenced by two main criteria, those external to the firm and firm specific. External criteria are those that derive from the attractiveness of a country or location with respect to hotel market entry; whereas firm specific criteria are those determined by the firm with regards market entry policy. Some firms are completely ruled by such policies, i.e. they adopt a single mode of entry. For instance Hyatt operate solely on a management contract basis.

It seems likely that these two decisions about entry and mode of entry are highly iterative, and may vary from chain to chain. However, we shall now consider each of these in turn in relation to Southeast Asia and Indo-China.

GENERAL MARKET POTENTIAL v. HOTEL DEVELOPMENT POTENTIAL

Hotel chains have to make a basic decision as to whether to have representation in a country or not. There are a number of ways for countries to be indexed in terms of their potential for market entry. The MSU-CIBER Index, illustrated in Table 1, is typical of such approaches. It uses eight composite criteria to arrive at an overall ranking based on weighted dimensions for a number of countries in Southeast Asia and Indo-China. The weightings are arrived at through an analysis (typically transaction cost analysis) of actual market entry by firms into countries and the extent to which these various factors have influenced such decisions.

However, there is evidence to suggest that hotel companies do not rely solely on these general criteria but take account of other specific factors. For instance, hotels do not solely rely on internally generated demand.  Occupancies may be greatly influenced by international travel behaviour, a factor not included in the general analysis of market potential, such as the MSU-CIBER Index. It may
also be the case that criteria used to determine general  market  potential,  such  as  market receptivity, are not deemed significant for hotel market entry. Dunning and Kundu (1995) identify a number of industry specific variables such as size and nature of host city, availability of appropriate inputs such as labour supply, and economies of supply.

It is clear, based on logical deduction and previous research, that market entry by hotel chains is not solely governed by the same criteria that may be applied to market entry in general, as illustrated in Table 1. A different set of criteria may be used. If this is the case, in Southeast Asia and Indo-China does the ranking of country attractiveness as shown in Table 1 change when hotel market entry is considered?

Table 2 illustrates the type of dimensions that might be used for hotel market entry analysis and the index that results from this. Some of these factors derive from studies of entry mode choice (discussed in the next section).  Clearly, the rankings that emerge are greatly influenced by the relative weighting of the various factors. The weighting presented here is based on the authors' opinion (one has been a hotel development executive with a major international chain for 6 years).

Quite clearly a new ranking emerges. China emerges as the most attractive country for hotel entry. Korea is an attractive country for general investment, it is less attractive to hotel chains; whereas Thailand is more attractive to hotel entry than general investors. Indonesia has the least potential in both cases.

However, it is the case that firms may decide to enter a market even if the country scores low on a market potential index. This is because there are various forms of entry that reduce the level of risk such as license agreements and joint ventures. Indeed there are some countries, such as China, which place legal restrictions on foreign investment and ownership. The hotel industry has a long history  of  developing  alternative  forms  of management or 'affiliations', which comprise two broad categories namely management contracts and franchises. It is generally agreed that direct ownership and management of a hotel property carries the highest risk, a management contract less risk, and franchising even less. Hence hotel chains may decide to enter markets that are deemed to be risky by using an entry mode that reduces this risk.

The only large-scale, empirical study of hotel market entry mode choice is that by Contractor and Kundu (1998).  They adopted a transaction cost analysis approach, which identified 12 potential influences on the entry mode choice made by hotel chains. These can be broken down into country-specific variables, firm structural factors, and firm strategy and control factors (as perceived by executives). These are shown in Figure 1. The factors that were found to most influence market entry mode were country risk, level of development, international experience (measured by number of years since first overseas development), extent of foreign business (% of hotels outside home country), strategic control over quality and reservation system and brand. Cultural distance, penetration by foreign businesses, size of the chain, need for size, economies of scale and investment in training were not significant.

Other studies have adopted a qualitative approach to entry mode choice by hotel chains. Saunders and Renaghan (1992) investigated hotel development  by  interviewing  28  hospitality executives working in Southeast Asia. They stress the potential importance of socio-cultural difference. In their view, it is this factor that will ultimately ensure the success or failure of an American or European chain's market entry strategy. Zhao and Olsen (1997) also identified socio-cultural concern as one of eleven factors influencing entry mode choice. Other external (i.e. country specific) factors were  political,  economic,  technological,  and ecological. They also found some factors, which they described as task environmental factors, namely competitors, customer, property location, partner characteristics and human factors.

HOTEL DEVELOPMENT IN SOUTHEAST ASIA AND INDO-CHINA

On the basis of the discussion so far we can make three propositions about hotel chain penetration in Southeast Asia and Indo-China.

• Proposition 1: Hotel chain penetration will be greatest by Asian chains, followed by European, and then American.
• Proposition 2: Hotel chain market entry is most likely in those countries that provide the highest potential return at the least perceived risk.
• Proposition 3: Hotel market entry mode will reflect the perceived level of return and risk.

Table 3 identifies the number of hotels operated by the ten largest chains in the region. The first proposition is not substantiated.   Asian hotel chains do not have a major presence across the region and tend to be concentrated in their home market, such as Oberoi. Two European chains do operate one-third of the hotels in the region, but Bass does so largely through its Holiday Inn brand, which was originally an American chain.  The history of chain ownership and recent merger and acquisition activity suggests a more detailed analysis is required before any assumptions can be made about exactly how the country of origin affects market entry in the industry.

Table 3 also allows some analysis of the relative attractiveness of countries. Again a complex picture emerges.  The least attractive country as far as market entry or hotel entry - Indonesia - actually has the largest number of chain hotels.  Whereas Hong Kong, Korea and Singapore, which were identified as the top three potential countries for hotel entry, have the least number of hotels in this sample.

There may be a number of reasons for this one of which is methodological. The sample of countries and chains may not be representative of the region as a whole. There are certainly some countries, such as Bangladesh, with considerably fewer international chain hotels. A larger, if not complete, sample might provide a more accurate assessment. Another reason may be that although the se countries are different, the range of difference is not so great that hotel chains discriminate between them to prioritise one country over another. For instance, if physical distance from the country of origin of a chain is important, all these countries are long-haul and relative to the US or Europe 'close' to each other. Table 2 also illustrates that country risk does not vary widely between the ten countries, whereas market size does.

However, another explanation for why market entry  does  not  appear  to  reflect  country attractiveness is entry mode. The range of market entry alternatives may mean that penetration occurs wherever the return may justify and the risk of entry is offset by the choice of mode. We cannot therefore simply look at hotel entry, as illustrated in Table 3, and expect to find most development in high return - low risk countries or cities.  But we might expect to find more direct ownership and leasing in these countries, and more franchising and contracts in low return/high risk environments. This is illustrated in Figure 1. Using criteria from Tables 1 and 2, countries are categorised as high return/low risk, high return/high risk, low return/high risk, or low return/low risk.   The nature  of  entry  mode  choice,  from  eight international hotel chains is then identified in each of these categories.

What we find is that management contracts are by far and away the most dominant form of entry. However, franchising is significant in those
countries that have the highest risk and lowest return,  which  may  reflect  this  mode's characteristics.

It  must  also  be  remembered  that  the environment is not static but highly dynamic. The so-called Asian currency meltdown demonstrated the volatility of this region and the implications this might have for entry mode decisions. Hence by looking at data concerning country risk, tourist arrivals and so on for any given year, along with that years level of hotel development, the linkage between the two is difficult to demonstrate. Hotel projects typically take three years from conception to opening, during which time country and market conditions may have changed dramatically.

It also assumes that hotel development is a planned,  strategic  process,  when  experience suggests that it may be far more opportunistic than that. For instance "ACCOR's position in the market [Asia Pacific] has been, to a large extent, opportunistic" (Teboul 1998). The chain specifically targeted Indonesia with its mid-market brands, Novotel and IBIS, in order to not compete with the other chains in the region and to take advantage of the Indonesian government's support of tourism development.  Likewise  "Kwek  (CDL  Hotels International Chairman and Managing Director) is among the rare hotel investors / owner-operators who has the ability and strategy that includes taking advantage of opportune purchases and growing his business worldwide through a marriage of acquisitions, equity positions, and management contracts" (Weinstein 1999).

CONCLUSION

The lack of obvious fit between some of the proposed aspects of hotel development does not necessarily mean that development is solely opportunistic. Indeed, it may be the case that qualitative studies of hotel development based on interviewing executives, which generally suggest the process is highly contigent, may mask patterns of development that development executives are themselves unaware of. What is needed is an approach that includes a larger sample of hotels and countries which reflects the dynamic nature of the development process. Transaction cost analysis is such an approach. It stresses the importance of firm-specific variables which may explain how firms bear the transaction costs of negotiating, monitoring and enforcing contracts with external parties (Williamson 1975). It is especially effective  in  explaining vertical integration decisions and entry mode choice of service firms (Erramilli and Rao 1993).

We have currently collected data on 80 countries for the period from 1978 to 1998, along with data on every hotel opening in each country. For 1,000 of these hotels we have identified its entry mode, but we expect the final sample to comprise more than 1,500 hotels.   Using Eviews, this data will be analysed to identifies those variables that appear to have influenced the entry mode.

We hypothesise that four or five key variables will emerge as significant in determining entry mode choice, but that these will only explain a proportion of all the decisions. The unexplained proportion will reflect the contingent, opportunistic or qualitative aspect of hotel development worldwide.

REFERENCES

Bailey, M. (1998) The International Hotel Industry, Travel & Tourism Intelligence: London

Contractor, F.J. and Kundu, S.K. (1998) Modal Choice in a World of Alliances. Analyzing Organizational Forms in the International Hotel Sector, Journal of International Business Studies, 29:2, 325-358

Dunning, J.H. and Kundu, S.K. (1995) The Internationalisation  of  the  Hotel  Industry, Management International Review, 35:2, 101 - 133 Dunning,  J.H.  and  McQueen,  M.  (1982) Transnational Corporations in International Tourism, UNTC: New York

Erramilli, M.K. and Rao, C.P. (1993) Service firms' international entry mode choice: a modified transaction-cost analysis approach, Journal of Marketing, 57 (July), 19-38

Gannon, J. and Johnson, K. (1995) The Global Hotel Industry: The Emergence of Continental Hotel Companies in Progress in Tourism and Hospitality Research, 1:1, 31-42

Saunders,  H.A. and Renaghan,  L.M.  (1992) Southeast Asia: A New Model for Hotel Development, Cornell , HRA Quarterly, October 16-23

Taboul, J. (1998) ACCOR in Indonesia, Insead Euro-Asia Centre: INSEAD, Fontainebleau

Weinstsein, J. (1999) Built to Last, Hotels' Investment Outlook, June 16-24

Zhao, J.L. and Olsen, M.D. (1997) The antecedent factors influencing entry mode choice of multi-national lodging firms, International Journal of Hospitality Management, 16:1, 79-98.