انتخاب سیستم توزیع در یک صنعت خدمات : تجزیه و تحلیل شرکت های بیمه بین المللی در ایالات متحده آمریکا

Service firms play an increasingly important role in the global economy. However, the internationalization strategies of such firms, and especially their distribution system choices, have been underexplored in the international management literature. One specific service industry that has internationalized rapidly in recent years is the insurance industry. This paper examines the determinants and performance implications of the choice by international insurance firms between two rival distribution systems: direct writing and independent sales agents. Drawing on the transaction cost theory-based literature on resource commitment, control, and risk, we develop hypotheses on the determinants of the choice between these two distribution systems and on the performance implications of this choice for insurance firms. Analyzing a sample of 168 distribution entries into the United States by insurance firms from six foreign countries over the 1992–2000 period, we find that cultural distance has a U-shaped effect on the probability of direct writing, and that an insurer's intangible assets have a positive effect on this probability. We also find that the direct writing system performs better in terms of profitability, but that the independent agency system performs better in terms of market share growth.

The choice of foreign entry mode is one of the most important strategic decisions that internationalizing firms have to make, and is therefore one of the core research topics in international management research (Werner, 2002). Many studies have examined the entry mode choices of manufacturing firms (for reviews, see Datta et al., 2002 and Slangen & Hennart, 2007). Yet, services play an increasingly important role in the US and global economy, and have in fact become more important than goods. For instance, “services produced by private industry accounted for 67.8% of US gross domestic product (GDP) in 2006, with real estate and financial services such as banking, insurance, and investment on top” (US Department of State, 2007: 7). Production of goods, on the other hand, accounted for only 19.8% of US GDP in 2006. Moreover, services differ from goods in several important ways. For instance, while goods are tangible and can be stored, services are often intangible and must be produced and consumed simultaneously (Capar & Kotabe, 2003 and Contractor et al., 2003). Despite the economic importance and unique nature of services, relatively few studies have examined the entry mode choices of service firms. These studies have often examined services in general, lumping expansions by firms operating in a variety of service industries (e.g., Agarwal & Ramaswami, 1992, Erramilli, 1991, Domke-Damonte, 2000, Erramilli & Rao, 1990, Erramilli & Rao, 1993 and Fladmoe-Lindquist & Jacque, 1995). Only a few studies examined the entry mode choices by firms operating in one and the same service industry. Agarwal and Ramaswami (1992) examined the determinants of the choice between licensing, joint ventures, and wholly-owned subsidiaries by US equipment leasing firms, while Erramilli (1996) examined the determinants of the choice by US and European advertising agencies between majority and minority ownership of foreign subsidiaries. Brouthers, 1995 and Brouthers et al., 1996 examined the determinants of the choice by US computer software firms between servicing foreign markets through wholly-owned subsidiaries, joint ventures, or contractual arrangements (i.e., licensing or franchising). Contractor & Kundu, 1998 and Erramilli et al., 2002, finally, examined the determinants of the choice between wholly-owned subsidiaries, joint ventures, management contracts, and franchising in the international hotel sector. To the best of our knowledge, studies of the entry mode choices made by firms operating in one specific type of service industry–financial services–are so far lacking. In this paper we fill this gap in the entry mode literature by examining the entry mode choices of firms operating in a financial service industry, i.e. the insurance industry. Specifically, we examine the choice between two distribution systems that insurance firms may use to sell insurances abroad: the direct writing system and the independent agency system. The independent agency system represents a low-control entry mode under which insurance firms close non-exclusive contracts with independent business owners who sell the products of several insurance firms. These business owners control the customer list, meaning that the insurance firm cannot replace the agent or contact clients directly without the agent's permission. The direct writing system, on the other hand, represents a high-control entry mode under which insurance firms sell their products through employed sales agents and/or through exclusive agents. Under this arrangement, the insurance firm retains the ownership of the customer list (Kim et al., 1996, Berger et al., 1997, Regan, 1997, Seog, 1999, Regan & Tennyson, 2000 and Cummins & Doherty, 2006). An important challenge that international insurance firms face is to maximize their performance in their overseas markets by serving these markets through the right distribution system. This study aims to gain insight into the way in which insurance firms choose between independent agent and direct writing distribution systems, and into how this choice affects their performance in the foreign market entered. Specifically, this study sets out to explore the main determinants of the choice by international insurance firms between selling their products in the US through independent agents or through direct writing. Moreover, it examines the ex post performance consequences of this choice for the performance of the insurance firms in the US. By doing so, we aim to improve our understanding of the strategic decision making of managers of internationally-operating insurance firms, and of the comparative performance of rival distribution systems commonly employed in the insurance industry. Our decision to study the insurance industry is motivated by several factors. First, the insurance industry has experienced a large volume of foreign direct investment (FDI) in recent years, mainly due to new technological advancements and governmental liberalization policies, which have created many opportunities for insurance firms to become active in foreign markets. Moreover, the insurance industry is one of the largest service industries in the US (US Department of State, 2007). Second, an insurance policy involves the payment of a premium over a long period of time in order to generate a specific type of benefit for the policyholder in the future. As a result, policyholders are likely to have long-term relationships with their insurance firms, and are likely to have a special interest in the performance of their insurer. Together with the lack of international management research on insurance firms and their entry mode choices, these specific characteristics make the insurance industry an interesting service industry to study. To examine the determinants and performance implications of the choice by insurance firms between direct writing and independent agents, we draw on two distinct bodies of research: the transaction cost theory-based literature on the level of resource commitment, control, and risk associated with foreign entry modes (Anderson & Gatignon, 1986 and Hill et al., 1990) and the literature on insurance distribution systems. We combine these two streams of literature to develop and test a set of hypotheses on (i) the determinants of the choice between independent agent and direct writing distribution systems, and (ii) the implications of this choice for the performance of the insurance firms in the US. Our study makes at least two contributions to the international management literature. First, while several studies in the risk management and insurance literature have examined the distribution systems used by insurance firms (Kim et al., 1996, Regan, 1997, Regan & Tzeng, 1999 and Seog, 1999), we are the first to study distribution system choices made by such firms from an international management perspective. Second, we are the first to identify the main strategic determinants of the choice between using independent agents and direct writing, and to simultaneously explore the performance implications of this choice for the focal insurance firm. Hence, our study is both descriptive, identifying what drives insurance firms to choose a specific distribution system, and normative, identifying whether specific distribution systems contribute more to firm performance than others.

Several studies in the risk management and insurance literature have examined the distribution systems available to insurance firms (e.g., Berger et al., 1997, Regan, 1997, Regan & Tzeng, 1999, Seog, 1999, Regan & Tennyson, 2000 and Cummins & Doherty, 2006). However, none of these studies has examined the choice between rival distribution systems from an international management perspective. We have attempted to fill this gap in the international management literature by exploring the antecedents and performance implications of distribution system choices by foreign insurance firms in the US market. We found support for a U-shaped effect of cultural distance on the choice of distribution system. Our results indicate that up to a certain point cultural distance has a negative effect on the likelihood that an insurer chooses a direct writing system. After this point, a greater the cultural distance to the US is more likely to result in the use of direct writing systems. This finding is in line with our theoretical arguments and adds to the debate in the entry mode literature on the effect of cultural distance on the choice between low and high-control entry modes. This literature has obtained conflicting results regarding the impact of cultural distance on entry mode choice. Some studies found that cultural distance stimulates the use of low-commitment entry modes, while others found that it stimulates the use of high-commitment modes (for reviews, see Brouthers & Brouthers, 2001 and Harzing, 2003). Our findings suggest that these conflicting findings may be caused by the existence of a curvilinear effect of cultural distance on the propensity of firms to choose high over low-commitment entry modes. Cultural distance up to a certain point stimulates the use of low-commitment entry modes (in our case independent agents), but after that point stimulates the use of high-commitment entry modes (in our case direct writing). This is a novel finding that reconciles the conflicting findings of prior entry mode studies. Our study also showed that insurance firms with substantial intangible assets are more likely to sell their products in the US through a direct writer distribution system. This finding is in line with the widely held view in the international management literature that firms with many valuable assets prefer to keep close control over these assets by exploiting them through higher-commitment entry modes, so as to avoid these assets being misused by opportunistic local exchange partners (e.g., Hennart, 1991). Our findings indicate that the protection of intangible assets is also an important strategic consideration for firms operating in the international insurance industry. Hypothesis 3 predicted that insurance firms with less concentrated (i.e., widely-dispersed) business activities would prefer independent agents over direct writing, so as to limit coordination costs (Barkema and Vermeulen, 1998) and reduce the risks associated with selling less standardized insurances (Regan & Tennyson, 1996, Regan, 1997 and Regan & Tzeng, 1999). We did not find support for this hypothesis, possibly because diversified insurance firms may perceive the risks associated with selling less standardized insurances to be limited, since they are at the same time active in many other lines of business. The sales volumes in these other lines of business can be used to compensate potential losses in more risky lines of business. This may limit the perceived need of diversified insurance firms to rely on independent agents. We also did not find support for our hypothesis that insurance firms with higher shares of complex insurances in their business lines would be more likely to use the direct writing system. The reason for this lack of support may be that while the sales efforts needed to sell complex insurances are easier to monitor in case of a direct sales force, such a sales force generally understands complex insurances less well than independent agents that are specialized in the sale of such insurances. As a result, the attractiveness of direct writing may not substantially increase with the share of complex insurances in an insurer's lines of business. An important question in the insurance literature is whether one distribution system performs better than the other. In our model we examined the impact of the choice of distribution system on a firm's performance along two performance dimensions: market share growth and profitability. We found that the independent agency system tends to perform better than the direct writing system in terms of market share growth, but that the direct writing system performs better than independent agents in terms of profitability. These findings support Berger et al.'s (1997) claim that independent agents tend to provide more support activities to insurance firms, allowing these firms to increase their market share more rapidly. Furthermore, our results also confirm the idea that in order to motivate independent agents to offer support activities, they need to be paid high commissions (Regan & Tennyson, 1996, Kim et al., 1996, Berger et al., 1997 and Cummins & Doherty, 2006). This reduces the profitability of the independent agency system vis-à-vis the direct writing system. Taken together, these findings suggest that there is no single optimal distribution system. An important managerial implication is that managers of internationalizing insurance firms should base their choice of distribution system on their main strategic aim of international expansion: Do they want to expand their activities abroad to boost their market share or to increase their profitability? If managers primarily want to gain market share, for instance to increase their firm's market power vis-à-vis its competitors, they are advised to rely on independent sales agents. If, on the other hand, managers primarily wish to increase their firm's profitability, they are likely to be better off using a direct writing system. Overall, our study offers valuable insights to managers who must make complex decisions in the international insurance market. For international insurance managers and domestic managers alike, this study can shed some light on the complexities associated with internationalization, the choice of distribution system, and the performance implications of this distribution system choice. Our findings should contribute to a better understanding of how international insurers decide between rival distribution systems and how distribution system decisions translate into firm performance. A limitation of our study is that we measured an insurance firm's intangible assets through its rank in the Top 100 Global Reinsurers (cf. Qian and Delios, 2008). Admittedly, this operationalization is suboptimal, but limited data availability prevented us from using alternative measures. Future research should therefore try to use alternative operationalizations of this variable. Another potential limitation is that we have studied one specific industry (the insurance industry) in one specific host country (the US), limiting the generalizability of our results. Nevertheless, given the unique features of the insurance industry and the US market, we feel that our single industry, single host country approach is justified and is in fact a strength of our study because it enabled us to keep industry and host country conditions constant. Still, we invite scholars to replicate or extend our analysis in one or multiple other (financial) service industries in other host markets, so as to increase the generalizability of our findings and further increase our understanding of the antecedents and performance implications of distribution system choices in service industries. Finally, future studies could explore the existence of interaction effects between our independent variables. For reasons of focus, we limited our attention to main effects, but it might be that some of our independents moderate the effects of others. This may particularly hold true for cultural distance, which has often been found to be a moderating variable in the entry mode literature (e.g., Brouthers & Brouthers, 2001 and Slangen & Hennart, 2008). Future studies could shed more light on this issue.