Creating a market in the presence of cultural resistance: the case of life insurance in China

Abstract

This article brings together two different conceptions of culture—a shared meaning system on one hand and a repertoire of strategies on the other—to understand the emergence of a market. Based on ethnographic data, it examines how a Chinese life insurance market is emerging in the presence of incompatible shared values and ideas acting as cultural barriers, and how these cultural barriers shape the formation of the market. The findings reveal a burgeoning Chinese life insurance market despite local cultural logics incompatible with the profit-oriented institutional logic of life insurance. This Chinese market, however, has developed along a different trajectory from what might be expected. It first emerged as a money management, rather than a risk management, market. I argue that the very cultural barriers that compose the local resistance to a new economic practice also necessitate the mobilization of the cultural tool-kit to circumvent this resistance. These dual processes, shared ideas composing the resistance and the cultural tool-kit circumventing the resistance, shape the trajectory and characteristics of an emergent market. I propose a theoretical model specifying the mechanisms through which the two forms of culture interplay to influence the development of the life insurance. I apply this model to extend Zelizer’s (1979) insights and discuss how culture matters in forging a new market in the global diffusion of capitalism.

In the case of life insurance the trick was to sell futures – pessimistic futures (Zelizer 1979:153).

[The Chinese] see accidents as being far removed from their lives. They’ve seen accidents happen, but insist that that kind of misfortune would not happen to them (A general manager of a foreign life insurer in Shanghai 2001).

This slogan about life insurance—“Guarantees for the if in life”—appeared in an advertisement in a recent issue of Newsweek magazine for the New York based Metropolitan Life Insurance Company.¹ This advertisement attempts to demystify the meaning of life insurance by defining it as “a plan for if.” But what if the if embedded in the meaning of life insurance is “the unthinkable” for a population? How can life insurance markets be created in places where the local cultural logics are incompatible with the assumptions on which commercial life insurance operate? Specifically, how can a life insurance market emerge in a Chinese society where death and fatal misfortunes are taboo subjects?

Commercial insurance was first brought by the British to some major cities in China in the early nineteenth century. Two features characterized the insurance industry in China during the nineteenth to the early twentieth centuries. First, life insurance was never popular at the individual level, and only one-fifth of the insurers carried life insurance policies (Wu and Zheng 1993; Huebner 1930). Second, the insurance industry was dominated by foreign insurers who controlled over 80% of the business (Xu et al. 2001). Then, during the Maoist regime of the 1950s-1970s, which spanned an entire generation, commercial insurance essentially disappeared from mainland China.

Commercial insurance re-appeared in mainland China toward the end of the twentieth century, at the intersection of China’s economic reforms and the insurance industry’s continuous global expansion. In 1992, American International Assurance, Ltd. (AIA), which is a subsidiary of American International Group, Inc. (AIG), was the first transnational insurer to gain a toehold in the potentially lucrative Chinese market. Creating demand for life insurance, however, has never been a simple process (Zelizer 1978, 1979, 1985; Clark 1999). Although urban China with its dramatic economic growth and diminishing state-provided benefits has provided a favorable economic and institutional context for life insurance to emerge, cultural conditions have not been similarly conducive. In 1998, then Chairman and CEO of the Hartford based Aetna Insurance Company, Richard Huber, commented that a bigger obstacle to the life insurance business than China’s bureaucracy was its culture, specifically, that the population had “very little concept of insurance.”² The General Manager of a Sino-German joint-venture life insurer in China complained that the Chinese still held, in his words, “a stupid superstitious belief” that the topic of death should be avoided at all costs.³ Likewise, the insurance sales agents in China were frustrated that people “don’t want to hear about misfortunes.”⁴

What is puzzling, however, is that the life insurance business in mainland China has been growing quite remarkably since the mid-1990s. The average annual real growth of revenue from the life insurance business (so-called life premium income) from 1995 to 2004 reached 30.7%, compared to 9.1% growth from property insurance. The proportion of revenue from life insurance sales compared to revenue generated by the entire insurance industry increased from 34% in 1995 to 75% in 2004. In Shanghai, the life insurance business expanded even more rapidly, with an average annual real growth of 33.6% during the same period.⁵ How could this phenomenal expansion of life insurance in China be possible given the difficulty of disseminating the idea of life insurance to the Chinese?

This article asks two complementary questions. First, how can a particular market emerge in the face of cultural barriers? Second, to what extent and in what way do cultural barriers affect the formation and the characteristics of a market? To address these questions, I propose that it is necessary to bring together two different paradigms of cultural sociology—one defining culture as a subjective and coherent meaning system, and another defining culture as a practical and yet fragmented tool-kit.

My theoretical claim is that culture in two different manifestations affects markets in different but interacting ways. When culture as shared ideas, beliefs, and values is unfavorable to the emergence of a market, economic actors seeking market creation mobilize the tool-kit properties of culture to circumvent cultural ideas prohibiting market development. Nevertheless, when economic actors circumvent cultural barriers, the mechanism is like “bricoler.” Bricoler is an old French verb introduced by Levi-Strauss (1966) to elaborate the concept of “bricolage,” which describes diverging movements from a direct course to avoid an obstacle.⁶ When insurance practitioners in China circumvent the local cultural resistance, their movements inevitably shape the course and features of the market’s development. Contra scholars of the fragmented notion of culture who argue that the content of shared beliefs and ideas have little effect on action (Swidler 1986, 2001; DiMaggio 1997), my findings demonstrate that shared symbolic content has steered the trajectory of the Chinese life insurance market through a consistent pattern of preferences. I argue that while culture as a repertoire of strategies makes the market possible, culture as a coherent meaning system shapes the characteristics of the market. This argument highlights the agency of economic actors in materializing symbolic elements into resources, and yet it puts equal weight on the structural constraints that culture exercises in the process.

In the sections that follow, I first lay out the theoretical concerns of this article. After detailing the method and data collection, I outline the key features of the life insurance market that has been emerging in contemporary China, posing a few empirical puzzles. Next, I offer my answers to these puzzles and my line of reasoning, examining how the local preferences, which are shaped by the Chinese concepts of life and death, are incompatible with the profit-oriented institutional logic of life insurance. I present the strategies adopted by foreign and domestic life insurers in dealing with this incompatibility, highlighting how they, particularly the domestic insurers, mobilize the tool-kit properties of culture to overcome it. Then, I propose a theoretical model specifying the mechanisms through which culture in two different forms plays a role in life insurance development, and apply this model to extend Zelizer’s insights. In the concluding section, I discuss the theoretical implications of this study for the role of culture in forging a new market and in the global diffusion of capitalism.

Theoretical concerns

The remaining puzzle about culture and life insurance

To examine how culture affected the formation of the life insurance market in China, I took as a starting point Viviana Zelizer’s seminal book, Morals and Markets (1979). Rejecting a purely economic model of explanation, Zelizer forcefully added a cultural factor to explain why the American life insurance market began to develop in the 1840s, but not earlier. Her major argument is that the societal cultural values in the United States before the 1840s were incompatible with the ideological logic of life insurance. Apart from fundamentalist religious beliefs and a superstitious fear of precipitating death, the most critical cultural obstacle to the development of life insurance was the societal value that human life was sacred and priceless. Such a value was in tension with the commensuration logic of life insurance (Espeland and Stevens 1998; Heimer 2001). After the 1840s, Zelizer reasoned, changes in societal cultural context made the public more receptive to adopting life insurance as a new form of risk management. These cultural changes included the emergence of a sympathetic liberal theology, an increasing rational speculation on economic risk and, more importantly, a growing awareness of the economic loss of death.

Zelizer convincingly argues that societal values had an effect on the adoption of life insurance. However, questions about the role of culture remain if we apply her insights to the global diffusion of life insurance. If the development of a life insurance market requires a particular set of societal cultural values as a prerequisite, does the global expansion of the life insurance industry imply the existence of a set of worldwide cultural values? To put it differently, if life insurance markets emerge in places where the favorable cultural conditions that Zelizer found in the United States are absent, does this mean culture does not matter? Given that commercial life insurance historically originated in culturally specific western contexts, how can this business be globalized and expanded to places with different cultural traditions?

Of course, it is not Zelizer’s claim that for life insurance to develop, the very same favorable cultural contents she found in the American case have to be present. However, she did emphasize the impeding power of certain moral values and the enabling force of certain rationalized values in relation to life insurance development, and argued that cultural values could suppress the emergence of a market (Zelizer 1979:34–39). As I illustrate through the Chinese case, a life insurance market can grow rapidly even in a context with unfavorable cultural values and with a population resistant to accepting the risk management concept of life insurance. This case study, however, is not to refute her “values matter” argument. Instead, it aims at extending her insights by incorporating a “tool-kits matter” perspective. In fact, in her book, Zelizer (1979) documented a number of strategies that American life insurers deployed to make use of the existing cultural and religious symbols in marketing their products. However, she argued that prior “changes in public receptivity” were a prerequisite for insurers’ success and, consequently, left insurers’ strategies under-theorized.

Culture, action, and economic practice: shared meanings versus tool-kits

Max Weber’s (1991) masterpiece The Protestant Ethic and the Spirit of Capitalism offers a compelling and yet controversial account of how culture shapes the economic behaviors of those who share a particular set of values and beliefs. In the Weberian tradition, the subjective aspect of culture that constitutes actors’ patterns of preferences and dispositions is central to both economic and non-economic action. Shared ideas that exist in individuals’ heads act like “switchmen” that “determined the tracks along which action has been pushed by the dynamic of interest” (Weber 1946:280). The Weberian legacy has been carried through a number of important works, which all assert the constitutive role of systemic cultural orientation in patterns of human action.⁷

This constitutive role has been challenged by some contemporary cultural sociologists, most noticeably by Ann Swidler in her ground-breaking article, “Culture in Action,” that appeared in American Sociological Review in 1986. Along the line of Bourdieu’s (1977) theory of practice, Swidler argued that culture influences human action more through “a style or a set of skills and habits” than “a set of preferences or wants” as proposed in the Weberian and Parsonian approaches (1986:275). Using the metaphor of a “tool-kit,” Swidler proposed that culture matters as a repertoire from which individual actors construct their strategies of action to solve various kinds of problems. Quite contrary to the classical conception of culture that highlights its shared and coherent nature, culture in Swidler’s (1986: 277) conceptualization contains diverse or even “conflicting symbols, rituals, stories, and guides to action”. Its incoherent and fragmented nature is grasped by DiMaggio’s (1997:267) bold analogy, a “grab-bag of odds and ends”. With this notion of culture, people who share similar values and beliefs can behave very differently, often depending on practical urgency and cultural competences, as well as institutional demands.

The tool-kit concept of culture has made a significant contribution to the theory of culture and action. It explains why people who share the same cultural values and face the same structural constraints differ in their strategies of action. Nevertheless, scholars who apply the tool-kit concept of culture often posit their theoretical position against the coherent paradigm. One of the most heated debates between the coherent and fragmented paradigms was presented by Alexander and Smith (1993, 1999) in one camp and Battani et al. (1997) in the other. While Alexander and Smith (1993) proposed a single consistent symbolic framework that underlies the seemingly contrasting themes in the discourse of American civil society, Battani et al. (1997) argued the opposite by changing the grammar into plural form: American public discourse draws from contingent sources of meaning and multiple interpretive frameworks. This debate has dramatized the chasm between the coherent and fragmented paradigms of culture. I contend that this theoretical and positional divide is unnecessary and unproductive for the field of cultural sociology. As Neil Smelser (1992) commented, cultural unity or disunity is largely “a function of the vocabulary and the theoretical presuppositions of the investigator (original emphasis).” Therefore, a firm identification with either position is likely to fasten one’s attention to partial empirical facts and analytic tools. The two concepts of culture actually presuppose each other. The important question is not which concept should be prioritized in relation to action, but how to conceptualize the articulation of different manifestations of culture (Sewell 1999). This article addresses this line of inquiry, with a framework bearing a close affinity with Sewell’s (1992) conceptualization of structure, where both systemic schemas and polysemous resources are cultural components. Based on an empirical case study, my focus is on how culture as shared schematic references, in the form of values and perceptions, and as shared resources, in the form of practical tool-kits, constitutes an economic practice independently and interdependently.

While my ambition is to synthesize these two particular forms of culture in understanding the Chinese market, I should make it clear that I do not intend to prioritize the coherent vs. incoherent, or value vs. tool-kit perspective of culture. As I elaborate below, it was during my ethnographic research that I discovered the role of coherent symbolic content in shaping market practices; and it was also during the research process that I found the tool-kit approach to culture useful for resolving the remaining puzzle left by Zelizer, and for extending her theory about the role of cultural values in life insurance development to the Chinese case. In China, the local concepts of life and death generate a configuration of folklore, values, and perceptions that constitutes a pattern of dispositions and preferences incompatible with the assumptions on which life insurance operates as an instrument managing unexpected misfortunes. However, a life insurance market has been emerging despite this incompatibility, as insurance practitioners capitalize on their local cultural knowledge and favorable cultural symbols and practices to circumvent incompatible cultural elements. Thus, a Chinese life insurance market is emerging without a prior value change or a public receptivity to the ideological logic of life insurance. Nonetheless, coherent shared cultural values and beliefs are important. I maintain that the tool-kit paradigm in its own right is insufficient to explain the characteristics of the Chinese life insurance market. While life insurance in Euro-American societies first emerged as risk management and gradually moved to risk-cum-money management, life insurance in China first emerged as money management. This trajectory, I argue, is shaped by the Chinese concepts of life and death.

The ethnographic case study

This project is based on an ethnographic case study and comparative analysis. The case method involves case selections at the country, city, and company levels. The making of an individual life insurance market in mainland China is used as a case to examine the role of culture in the process of introducing a new economic practice to a population.⁸ Mainland China is considered as an excellent case for this theoretical query because the life insurance market has only recently been in the process of forming. Culture in such a process, prior to the forging of an institutionalized market, is at the forefront in “living” form (Jepperson and Swidler 1994: 362).⁹

In choosing the city, Shanghai presented itself as an ideal place for data collection for three reasons. First, Shanghai represents the “lowest end” of how Chinese culture plays a role in the construction of a modern capitalist enterprise. With its distinctive historical background as an international settlement from the 1860s to the 1930s, Shanghai is especially receptive to foreign influences. In the 1990s, Shanghai was an experimental city for everything new, modern, and capitalist. Thus findings that the economic behaviors of people in Shanghai, a city that is relatively “westernized” and eager to be further westernized, are shaped by certain Chinese cultural elements makes it likely that these cultural elements are as influential, if not more so, in other Chinese cities. Second, Shanghai demonstrates the “highest end” of global-local dynamics. Shanghai has been described as “the Olympic stadium” for commercial life insurance,¹⁰ where competition between transnational and domestic life insurers has been most intense. This fierce competition provides rich data on global-local interactions. Lastly, as the majority of the transnational insurance firms first started their operation in Shanghai, establishing their headquarters as a base for future expansion to other cities, their operations and experiences in this city often serve as a model for other cities to follow. All these factors make Shanghai a good case for generalizing the findings to other urban areas in China.

To begin the case study, I conducted six-weeks of preliminary research in the summer of 2000. Next, I selected four life insurance companies that represented the widest range of variety for a subsequent year-long field study during 2001–2002, and a follow-up study in the winter of 2004. The companies included (1) the foreign-owned insurer that first arrived in Shanghai, AIA; (2) a local insurer that experienced dramatic growth, Ping An Life Insurance Company of China, Ltd. (Ping An); (3) a seemingly very localized Sino-American joint venture, Pacific-Aetna Life Insurance Company, Ltd. (Pacific-Aetna); and (4) an apparently less localized Sino-German joint venture, Allianz-Dazhong Life Insurance Company, Ltd. (Allianz-Dazhong).¹¹

In more than 14 months of field research in Shanghai between 2000 and 2004, I collected three sets of data at three levels of measurement: (1) the organizational level: insurance companies’ product development, marketing strategies, agency management; (2) the interactional level: insurers’ training for the sales agents and the sales agents’ selling strategies and interactions with potential buyers, and (3) the individual level: buyers’ consumption behaviors, preferences, and motives. Data were collected through formal and informal interviews, participant observation, questionnaires, and by referencing public documents and insurers’ publications. I conducted participant observations in the agency offices of each selected insurer for 2 to 3 months. I attended training sessions for the agents, joined in morning assemblies, and participated in small group meetings on a regular basis. With the consent of certain agents, I went with them to approach prospective buyers, to sell products, and to visit existing clients. From 2000 to 2004, I have observed 46 morning assemblies, 48 small group meetings or gatherings, 28 training sessions, and 43 occasions of agent-client or agent-prospect interactions. A total of 99 sales agents, 44 managerial staff, 96 clients, and 35 prospects in Shanghai were interviewed.¹² The observations and interviews were supplemented with 179 non-random sampled questionnaires (66 from sales agents, 63 from clients, and 50 from prospects). In addition, I spent 2 weeks in Beijing in August 2002 interviewing a total of 13 insurance practitioners. This brief study helped to assess the generalizability of the Shanghai case to urban China.

The data analysis was an ongoing process along with the data collection. I went to the field with the analytic frame that to create a life insurance market, insurance practitioners would have to create a sense of risk so as to create a feeling of need for the new commodity. Schneiberg and Clemens (2006) attested that cognitive consent can only be inferred from silences in texts, and normative consent can be discerned by the absence of alternatives or hypotheses. During the field research, it was the general absence of a risk discourse and the evasive attitude toward my hypothetical questions about misfortunes that informed me of the role of the cultural taboo in the process. These insights from the inductive data modified my initial analytic frame. Thus, my case analysis is a product of constant dialog between my analytic frame informed by theory and the images formed by the evidence, a process termed “retroduction” in Ragin’s (1994) research model. In identifying “typicality,” for example, the “typical” reasons for buying life insurance and the “typical” selling strategies, I mostly counted the frequency. Nonetheless, some of the analytical insights come from rare events, such as occasional slips of the tongue by the informants and inconsistencies between what they said in the interviews and how they behaved.

The emergent Chinese life insurance market and the empirical puzzles

The development of the life insurance market in China can be measured in terms of life premium income. As reported in the introduction, life premium income has been growing rapidly since the mid-1990s. Until the end of 1991, before AIA arrived, the life premium income in the entire country was only 4.14 billion yuan (US$500 million).¹³ In just 10 years, the amount rose 34 times to 142.4 billion yuan (US$17.2 billion) in 2001. In Shanghai, life premium income during the same period increased 54 times, from 260 million yuan (US$31.4 million) to 14 billion yuan (US$1.69 billion). This emergent Chinese market, nevertheless, displays several intriguing characteristics that must be explained.

The growth pattern

First, note that the pattern of life insurance business growth during its emergent stage is neither stable nor linear. Figure 1 indicates that the market, both at the national level and in Shanghai, experienced extraordinary growth between 1995 and 1997, followed by a setback in 1999–2000, before it reached another peak of growth in 2001–2002. Growth began declining in Shanghai in 2002, and nationally in 2003.

Fig. 1

Life insurance growth in relation to GDP growth and per capita disposable income growth 1994–2004. Sources: (1) Almanac of China’s Insurance 1998–2005 for life premium in China and Shanghai from 1997 to 2004; (2) Xujin Wang et al. 2003 for life premium in China before 1997; (3) Marketing and Research Department of a joint venture, Haier New York Life, in Shanghai provided information about the life premium in Shanghai before 1997 (note: official data on life premium income in Shanghai in 1993 and 1994 were missing. Haier New York Life estimated the figures by assuming constant growth rate based on the figures of 1992 and 1995 and some available figures of individual life insurers’ life premium incomes.); (4) China Statistical Yearbook 1998, 2006 for Shanghai’s GDP; (5) Survey data from the National Bureau of Statistics in Shanghai for the average per capita disposable income in Shanghai, cited in Shanghai Statistical Yearbook 2006

The most common explanation offered by economists for the sudden expansion of the life insurance business in China is its conducive economic environment, specifically the economic growth of the country and corresponding economic wealth of the people (Headey et al. 2002; Hwang 2003; Wang et al. 2003). From the late 1970s, the Chinese economy indeed experienced enormous growth, and the urban population in the early 1990s had strong purchasing power and substantial disposable income for life insurance. However, economic growth is only a necessary condition for the development of insurance in general and life insurance in particular; it is not a sufficient condition. Using Shanghai as a case, Fig. 1 also demonstrates that the growth pattern of life insurance sales has no observable, direct correlation with the growth of GDP, or with the growth of per capita disposable income in the city. Market booms took place during 1995–1997 and 2001–2002 when the growth of GDP and per capita disposable income either slowed or flattened. While the large gains in life premium income during 1995–1997 were partly due to its low original starting point, it is puzzling why this growth rate suddenly dropped in 1998, why it climbed up again in 2001, and why it slid continuously afterward.

The dominance of domestic players

Another feature of the current Chinese life insurance market is the dominance of domestic insurers, in contrast to the market in nineteenth century China that was controlled by foreign players. While foreign life insurance companies were most active in Shanghai and maintained a comparable number with their domestic competitors during 1997 through 2004, domestic insurers consistently captured around 80% of individual life premium income in this city. The dominance of domestic insurers was even more prominent in other cities where foreign insurers were either smaller in number or absent entirely.

To what extent can the domestic companies’ dominance be attributed to state policies that favored them? In the 1990s, the state did impose two major restrictions on foreign players: (1) restricting their operation to a few designated cities; and (2) limiting their business to individual life insurance so as to leave the group business to domestic firms. However, the above restrictions cannot explain why experienced foreign insurers lost over 80% of the individual life business to their domestic competitors in a single city. Furthermore, state favoritism fails to explain why the former state-owned life insurer, China Life Insurance Company, Ltd. (China Life), which was supposed to receive the same if not more favorable treatment from the state, lost market share to the newly established private insurer, Ping An.

Money management market

Another feature of the Chinese life insurance market that requires explanation is that it emerged as a money management market rather than a risk management market.

Insurance belongs to a category of commodities requiring a specific logic of operation to make profits possible. The basic principle on which insurance operates is a probabilistic calculation of risk. For traditional risk management insurance products (typically term life, accident insurance, health insurance, and critical disease insurance), profits mainly derive from the profit margins that are calculated upon death and casualty rates. For money management products (typically products with savings and investment functions, such as endowment or annuity insurance, participating policy, and variable life), profits principally come from the investment returns of premiums. The profitability of money management products, thus, relies more on the investment environment, which is often less predictable. Moreover, defining life insurance as primarily a money management instrument undermines insurance’s unique function of managing casualties that other financial instruments cannot provide. Therefore, according to the profit-oriented principle, it is in the best interests of the insurers to define life insurance primarily as risk management for a specific market niche, and secondarily as money management for diversification purposes. This has been happening in Britain and the United States, where the popularity of life insurance began with policies insuring against deaths.¹⁴ Nowadays life insurers in the United States and Europe carry a wide range of risk and money management products. However, traditional whole life and term life remain the most widely sold.¹⁵

In mainland China, it was even more important to sell life insurance as a risk management instrument early on, because a number of restrictions were imposed on how insurers could invest their premium income. Insurers’ investment returns were largely dependent on the interest rates of savings deposits.¹⁶ As the regulatory and investment environment was unfavorable for the development of money management insurance products, the experienced transnational life insurers, for reasons of profitability, all made efforts to sell risk management products with insured amounts payable only when the specified circumstance(s) such as death, injuries, or critical diseases occurred.

Indeed, the economic and institutional restructuring in China should have provided a fertile ground for the reception of risk management insurance products. China in the 1990s was at the height of its economic transformation from a planned to a market economy. While urban dwellers were reaping financial gains from the economic reforms, they were simultaneously facing increasing risks. The so-called “iron rice bowl” system, an analogy for lifetime employment, was shattered when thousands of state-owned and collective enterprises were replaced by private corporations. The collapse of state and collective enterprises also marked the end of the comprehensive socialist welfare system. Medical care was privatized, the pension scheme individualized, and housing and child care benefits cut back (Davis 1999; Zhao and Xu 2000; Guo 2003). In the meantime, traditional risk management mechanisms, mainly kinship support and inter-generational care, became increasingly tenuous with urbanization and decreasing family size.¹⁷ All these were accompanied by the one-child policy, launched in 1979, that substantially re-structured Chinese demography and further thinned kinship support networks.

Surprisingly, the emergent Chinese life insurance market thus far is predominantly a money management market. As systematic official data about the types of insurance policies sold are not available, I use the data collected from my interviews and surveys to explicate the money management character of the market. In this sample, a total of 128 clients in Shanghai bought 282 policies.¹⁸ Table 1 shows the types of policies they bought.

Table 1 Types of life insurance policies purchased as reported by 128 clients in Shanghai

One of the most striking features in this table is that more than two-thirds (72.7%) of the policies bought belong to the category of money management. Yanglao insurance, which refers to retirement insurance, was the most popular, with over half of the policies bought belonging to this category. Child savings insurance ranked the second. In addition, my findings indicate that 71% of the yanglao policies and 42% of the child policies bought carried a dividend component. Insurance with a dividend component is para-investment insurance, with a guaranteed insured amount plus some share of insurers’ surplus earnings given in the form of dividends.

The interview and survey data are consistent with some aggregate figures made available in Shanghai starting in 2001. Products with investment connotations, which were only introduced in late 1999, accounted for over 38% of total life premium income by 2001, with their share rising to 58% in 2003. These products include dividend insurance and unit-link, a British variant of variable life with an insured amount dependent on the investment returns of the premium. On the other hand, products intended primarily for risk management have had a consistently low share throughout. Personal accident insurance, for example, accounted for only about 2% of premium income during 2000–2004.¹⁹ The popularity of money management products was not confined to Shanghai. At the national level, unit-link and dividend insurance accounted for up to 65.2% of total life premium income in 2003, compared to personal accident insurance’s 2.5%.

I should, however, note that this Chinese market is a life insurance market, despite its money management character. All the products issued by life insurance companies must carry a life component with a payment for death, though the amount can be insignificant. The question is: Why were products that sold for primarily a risk management purpose far less popular than products that sold for a predominantly money management function?

Culture provides an answer

I propose that the characteristics of the emergent Chinese life insurance market arise through the interplay of two forms of culture—one as cultural obstacles and the other as cultural resources. This account is supported by the empirical materials presented in the next three major sections, and is derived from the following line of reasoning:

First, my fieldwork revealed a local resistance to receiving life insurance as a form of risk management. It was manifested in my informants’ denial of the possibility of fatal misfortunes and their unmistakable preferences for money management products. This resistance, I argue, arises from the cultural taboo on the topic of premature death, the notion of a “good life” as living well toward the end, the definition of a “good death” as dying after a full life, and a selective attention to risks. These values, moralities, and perceptions, as I elaborate below, are rooted in the Chinese concepts of life and death.

Second, local preferences for money management products are incompatible with the profit-oriented institutional logic of life insurance. Foreign and domestic life insurers dealt with this incompatibility differently. Foreign players adhered to a profit-oriented model, attempting to remove the local resistance by offering risk management products and preaching a new concept of risk. The domestic players, on the other hand, deployed a market-share approach. They simply yielded to the resistance by re-defining life insurance as a variant of money management. Furthermore, domestic insurers took the lead in strategically mobilizing the etiquette of renqing (interpersonal obligations), the child-centered ethos, the habit of saving, and the “stock fever” fad to facilitate their sales.

Third, a money management market dominated by domestic players developed through their unrestrained accommodation with the cultural obstacles and their aggressive mobilization of local practices, symbols, habits, and fads. However, this market was not sustainable for long, due to the profit structure of life insurance and given the investment restrictions imposed by the state. After suffering financial losses, insurance practitioners in China essentially turned their efforts towards creating a risk management market. However, this move substantially slowed the growth of the market, largely due to the presence of cultural resistance.

The cultural resistance

The taboo on premature death

The most powerful and overarching cultural barrier to the development of life insurance in a Chinese society, I suggest, is the taboo on thinking and talking about death, especially premature death. This taboo is manifested in an observable avoidance of the topic, particularly unexpected, accidental, or premature death. This cultural taboo is rooted in Chinese concepts of life and death that deserve some elaboration.

A this-worldly orientation towards life has been widely recognized as part of the Chinese cultural tradition (Weber 1951). While death has been a central theme of Western philosophy, it is rarely mentioned in the teachings of Confucianism, except as something that is unknown. The Confucian notion of death does not imply the possibility of eternity as it does in Christianity (Hui and Xu 2000; Hou and Fan 2001). Death for common people is already a mystery, and the avoidance of the topic further mystifies and intensifies a general sense of fear regarding death. Furthermore, the influence of folk Buddhism has resulted in this “something unknown” taking on a dreaded and terrifying image, breeding the idea of a cruel hell and a damp, cold, “dark world.” Hell is seen as reserved for those with bad karma, and exercises numerous cruel punishments. The dark world holds the majority of commoners, who have become ghosts awaiting reincarnation. These notions of hell and the dark world have become a folk belief continuing into the contemporary era (Hou and Fan 2001).²⁰

In addition, Taoism advances the idea that the degree to which a death is seen as terrifying depends on when and how the death occurs. A good death is dying when one is old and has lived a full life, that is, one with descendants and dignity toward the end of the life. A miserable death is dying without offspring or dying prematurely. The differentiation in the meaning of death is dramatized in the hierarchy of death rituals (Whyte 1988).²¹ Moreover, whether or not a death occurs expectedly affects the extent to which the death is viewed negatively. The most feared death is the unexpected premature death, as when a young person dies suddenly due to an accident or illness. This kind of death is spiritually dangerous, as the dead are “cheated out of their lives” and would therefore turn to “hungry ghosts.”²² In addition, the superstitious belief that thinking and talking about death will precipitate death is still loosely held by those in middle age and older.²³ Premature death has thus become a deeply ingrained taboo topic.

Although philosophical and religious traditions gave rise to the taboo, today this taboo has an independent power in shaping human action. In other words, one does not need to hold beliefs about hell, the dark world, ghosts, and precipitating death in order to observe the taboo. When something has customarily been a taboo subject, a violation of the taboo is not only socially offensive but out of the conceivable “normality” of being a human being (Durkheim 1965). Thus, what matters is that observing the taboo produces collective avoidance of the topic of premature death. Such avoidance results in “subjective immunity” (Douglas 1985) that puts premature death out of the cultural schematic frames of most Chinese.

“Good life,” “good death,” and risk perception

The concepts of life and death also define what a “good life” and a “good death” are. In both Confucian and Taoist worldviews, a good life involves improving one’s life quality as one grows older. A person who has a hard life as a youth but a comfortable life when older is considered by the Chinese as more fortunate and happier than a person who has a comfortable life early on but a hard life when old. In other words, to have a truly good life is to have a good life at the end, which is associated with the accumulation of virtues. This is why traditionally the Chinese place significant emphasis on yanglao (literally translated as cultivating or feeding the aged life, but today it refers to life after retirement). The focus on a good life in the end directs people to be extremely concerned about their quality of life during retirement. At the same time, when a “good death” is defined as dying after a full life, it may only occur when one’s children are already grown-up and financially independent. Therefore, the meaning of a “good death” promoted by commercial life insurers that involves posthumous financial arrangements for the living dependents does not make sense to the Chinese.

The importance of life quality during retirement, together with the taboo on thinking about premature death, organizes the selective attention of the Chinese to different kinds of risks. In Chinese cosmology, aging, illness, and death are the natural and inescapable fate of human beings. They are not only universal but predictable and, thus, not seen as a “risk.” On the other hand, as a good life is defined as improving one’s life as time goes by, the most commonly perceived risk is not having enough resources to live a comfortable life as one is getting older. This perceived risk, nonetheless, would not have diverted people’s attention away from fatal accidents that might happen during one’s pre-retirement age, if there were no taboo surrounding premature death. The collective avoidance of thinking and talking about premature death, along with disproportionate attention to life quality during retirement, results in moving fatal misfortunes during one’s pre-retirement age far away from the center of attention.

Taken together, the taboo on premature death, the definition of a “good life” and “good death,” and the selective inattention to fatal risks, generate a pattern of dispositions: a reluctance to discuss the possibility of misfortunes for the self and an inclination to save enough money for retirement. Saving has been a long-established habit and risk management practice of the Chinese that is unlikely to change in the near future. This habit has been so ingrained that China’s household saving rates have been continuously rising despite the emergence of a consumer society (Croll 1999; Davis 2000).²⁴ The disposition to avoid the topic of premature death and to save money, as I discuss below, constitutes a pattern of local preferences incompatible with the profit-oriented institutional logic of life insurance.

Cultural logics and local preferences

The Chinese cultural logics that define death, good life, and risk are not at all compatible with the concept of risk embedded in life insurance. This incompatibility is observable in the general public’s preferences and attitudes with respect to life insurance. Regardless of the specific details of an insurance policy, the crux of the matter for the informants’ receptiveness to an insurance policy is whether the insured will get the money back when they are alive. To quote one client:

Whether the products belong to the returning principal type or not is crucial. All the people I know, including myself, will buy only those with returning principal. The accident insurance doesn’t have any market…. Why? Accident insurance means the insurance company will pay only when you die…. You’ll get nothing in return.²⁵

The terms “returning principal” (huanben) and “non-returning principal” (bu huanben) were often mentioned by the informants to distinguish money management and risk management products. The former refers to insurance policies that have a guaranteed payment, whereas the latter refers to those that pay out only if the stated circumstances occur within the stated period. Furthermore, there are two kinds of “returning principal” products: one makes payments to the living insured (e.g., endowment or annuity insurance) and another makes payments to the beneficiary when the insured dies (e.g., whole life). When the Chinese talked about “returning principal” insurance, they all referred to the one where the insured could get their money back while still alive. They called the kind that makes payment to the beneficiary “death insurance,” which was not at all what they wanted. Using the term “principal” to refer to the premium is telling. It suggests that life insurance as a category and as a practice has been interpreted, understood, and treated as saving or investment.²⁶

The sentiment against purely protective policies, such as personal accident, critical disease, and term life, was noticeable, expressed by a feeling of being “a waste of money.” An informant explained why he would never consider buying critical disease insurance:

It’s useless! They won’t give you any money unless you are paralyzed, lying in the hospital awaiting death. They won’t give you any money if you are healthy…. What’s the point of buying this kind of insurance? It’s a waste of money!²⁷

In fact, the purchase of a purely protective policy has an invisible return – a sense of security, or as MetLife advertises, “a plan for if.” However, the purchase brings about a sense of security only if one feels insecure without such a purchase; it is perceived as a “plan” only if one sees a possibility for the if to happen. Since fatal accidents and premature deaths are “the unthinkable” for the Chinese, they feel that the premium for that impossible if is “a waste of money.”

The emphasis on benefits for living insured parties, rather than for beneficiaries, is found not only in clients who define insurance as money management. About one-third of the clients interviewed mentioned “protection.” A woman in her late 30s with a 13-year-old son said: “… buying insurance is for retirement. I want to have some protection when I am retired.”²⁸ Likewise, a 40-year-old father with a 9-year-old son also mentioned protection in his interview: “I think insurance is for future protection. When I can’t work in the future, it’ll be useful….”²⁹

Although both clients have a dependent child, the kind of risk that they both refer to is “having insufficient money for a decent life during their retirement.” “Protection” for them means preventing the occurrence of this risk. Even though they perceive life insurance as serving a protective function, they still prefer money management products.

However, as discussed above, it is in the best interests of insurers’ profitability to sell risk management insurance products in general and in China in particular. How did the insurance practitioners deal with the local preferences for money management products that may work to their disadvantage?

Divergent institutional logics between foreign and domestic insurers

Insurance companies, be they foreign or domestic, are supposed to hold the same economic interests, namely profit making and capital accumulation, as shaped by the institutional order of capitalism (Friedland and Alford 1991). However, when there is an obvious conflict between local preferences and insurers’ profitability, life insurance providers are confronted with the tough choice between market share and profits. As we see below, foreign and domestic insurers in China started with two disparate institutional logics of operation. While the former adhered to a profit-oriented model, the latter adopted a market-share approach.³⁰ As institutional logic constitutes a set of assumptions and values, usually implicit, about how to interpret organizational reality (Friedland and Alford 1991; Fligstein 1990; Thornton and Ocasio 1999), it organizes routines and choices (Douglas 1986; Biggart and Guillen 1999). Institutional logic, like cultural logic, embraces certain things and excludes others, and it sets the limits on the very nature of rationality and legitimacy.

Foreign insurers’ profit-oriented operation

Given AIA’s and other foreign insurers’ experience in Asian markets, especially in Hong Kong and Taiwan, they were certainly not ignorant about Chinese culture and local preferences.³¹ However, their profit-oriented institutional logic deterred them from fully localizing their products according to local preferences. A Pacific-Aetna actuary explained why his company had been insisting on promoting risk management products:

The most profitable product for an insurance company is the traditional type that reflects closely the meaning of insurance, mainly, risk management. For this kind of product, insurers can calculate the risks involved and set reasonable profit margins. Normally, protective products have higher profit margins than savings or investment products. To accommodate the prospects’ preference, we have added an endowment element into most products. However, the protective element should be the main feature. The endowment should be the secondary and at best be the rider.³²

Normally, market share yields profits. However, when the kinds of products the local people want are generally unprofitable, a large share in the market does not necessarily result in large profits. Thus, foreign players in China attempted to accommodate local preferences, but only up to the point where the profit margin of each product was not compromised.

Domestic insurers’ market-share operation

Facing the same resistance, as we discuss below, the novice Chinese insurers simply offered what the majority preferred. Ping An was among all most responsive to local demands and preferences and played a leading role in the domestic camp in shaping the definition and product development of life insurance. The operations of these Chinese insurers were based on an institutional logic of market-share that legitimized initial financial losses. One of the vice directors of product development at Ping An recalled the company’s rationale for launching the yanglao (retirement) and child policies in its early phase:

The Chinese don’t like to hear about misfortunes. They don’t like to think about those things. They don’t like to talk about those things. What should we do? We have to offer products that they like. The Chinese like savings. They are most concerned about their yanglao issue…. Now parents treat their only child as the king … or the queen. We knew that the child policy would be well received…. Any products that have a good market are good products.³³

Ping An’s innovative response to local preferences proved to be successful in boosting sales and capturing the market share. However, the large market share came with a price of financial losses, which I present in the next section.

Strategies in creating a Chinese life insurance market (1992–2004)

Introducing a new concept of risk management

Being the first commercial life insurer upon its entry to Shanghai in 1992, AIA carried the mission of “educating the public about the concept of life insurance.” It defined life insurance as a “protection against misfortunes” and offered a term life and two personal accident policies. To sell these policies, the agents were sent door to door to tell people stories of misfortune about families without insurance. They used the rhetoric of “love” and “responsibility,” as in the sales discourse proven effective at the inception of the American life insurance market (Zelizer 1979). However, a senior agent who joined AIA in 1992 described the difficulty of using this rhetoric:

Well, it was very difficult! It was very difficult to convince [the prospects] that accidents could happen to them…. Our job as an agent was to teach them that accidents were neutral, that accidents could happen to everyone. We told them that buying life insurance was a way to show love and care for their family…. But it was difficult. They didn’t want to hear about misfortunes…. [T]hey didn’t want to think about that.³⁴

Obviously, AIA attempted to raise people’s sense of risk so as to impose on them a new form of risk management. However, when the population at large refused to relate to the possibility of misfortunes like sudden death and dismemberment, it was extremely difficult to make people feel that they needed insurance. The AIA agents complained that even couples with young children refused to buy accident policies that cost a very minimal amount for a high insured amount.

Accommodating to the local preferences

In response to the difficulties of selling accident and term life products, AIA introduced a new whole life policy called LES that proved popular in Hong Kong. This policy served mainly a protective function for the beneficiary, because the insured amount was paid only upon the death of the insured. Surprisingly, AIA sold quite a lot of these policies. A senior agent of AIA explained why it was popular:

The LES carries an incremental dividend…. When we sold this product, we first talked to the prospects about the importance of protection and risk management. When we found that the prospects didn’t like to listen to what we said, then we talked about dividends. Very often, when they heard of dividends, they just liked it. They bought it as a kind of investment.³⁵

LES was the first life insurance product that brought AIA to the public’s attention. The dividend feature of this product gave the agents an inviting topic to broach with their prospects, allowing them to use specific framing language that better fit the prospects’ preferences. By highlighting the dividend feature of the policy, the sales agents were invoking a money management genre to legitimize this new commodity. This linguistic framing, which is quite commonly adopted by business organizations for legitimacy when introducing something new (Hirsch 1986; Fiss and Zajac 2006), contributed to a gradual growth of AIA’s business between 1993 and 1995.

AIA’s alignment with the local preferences was only rhetorical. Although LES carried a dividend component that sounded like an investment, it was indeed a risk management product. Nonetheless, its business growth was largely due to accommodating to the local interpretation of life insurance as money management. The mission of “educating the public about the concept of life insurance” had not been achieved. This unfinished mission in part doomed AIA to lose market share to the domestic insurers in 1996 and after.

An unexpected explosion in life insurance sales occurred in 1996–1997. Life premium income in Shanghai experienced 57.3% growth in 1996 and 98.5% in 1997 (cf. Fig. 1). The term baoxian, which is literally translated as “insurance” but is often referred to as “life insurance,” became popular and known to the general public. This sudden change was brought by a new, inexperienced domestic insurer, Ping An.

Ping An, founded in 1988, did not have an actuary or a trained underwriter when it joined the life insurance industry in 1994. This domestic company launched three policies that served primarily a savings function. They included a whole life endowment, a whole life annuity, and a child endowment-annuity policy. The term “yanglao xian” (retirement insurance) was used to characterize the whole life policies designed to cater to the penchant for savings and the concerns about life during retirement. The child policy, on the other hand, was offered to capitalize on the emerging child-centered way of life. More importantly, sales of these money management products did not have to address the troublesome topic of premature death and other misfortunes.

Exploiting the etiquette of renqing and the child-centered way of life

Defining life insurance as money management, nevertheless, created the problem of competition with other money management instruments. When Ping An launched its retirement insurance, the amount payable to the insured was based on an interest rate of 7.8%, which was obviously lower than the 10.98% interest rate offered by savings banks at that time. To sell its products, Ping An capitalized upon the etiquette of renqing that governs the interpersonal interactions among the Chinese to facilitate transactions.

Renqing, loosely translated as interpersonal obligations, is the core enactment of the norm of reciprocity among the Chinese. As a kind of “moral cement” reproducing and stabilizing social relations (Gouldner 1960), the etiquette of renqing demands that people with a good relationship ought to deliver asymmetric obligations to each other to maintain, strengthen, and reproduce the relationship. It is not uncommon for the Chinese to engage in a small economic transaction with a friend or relative in order to give face to or do a favor for this person instead of gaining economic benefits from it (Lin 2001; Hwang 1987). Therefore, when a new commodity like life insurance has to be marketed, existing relationships provide an entry point. To sell the retirement insurance that carried an unappealing interest rate, the agents of Ping An approached their siblings, relatives, close friends, old friends, ex-colleagues, and neighbors to ask them for face and favor. Under pressure to observe the etiquette of renqing for maintaining their relationships with the agents, prospects typically bought an insurance policy even if they did not feel the need for it (Chan 2003). “Doing my friend a favor” and “giving my friend face” were common responses from clients when asked what made them buy their first life insurance policy in 1995–1996.

When marketing the child policy, however, Ping An used a different strategy. The one-child policy in China has led parents to channel their energy and emotions into bringing up one perfect child (Gold 1985). Parents in Shanghai spent at least half of their household budget on their only child (Davis and Sensenbrenner 2000). To take advantage of this child-centered ethos, Ping An priced the child policy at 360 yuan per year, and its sales agents cited “saving just one yuan a day for your child’s future” to present the idea that any parents who cared about their child’s future would buy this policy. Many of the clients said “saving one yuan a day for the child” was what moved them to buy, not the death benefits that the policy covered. Nor did they think that the policy was appealing as a savings plan. They simply bought it as a trendy gift to show love for their child. Some parents even used the child policy to show off their financial status and the extent of their love for the child. A Ping An agent recounted the so-called “child policy fad” in 1996–1997:

We just intended to sell one [child policy] to each household. But later on many households bought ten policies or more. Why? Some parents asked how many policies the parents who lived next door bought. When they heard that their neighbors bought one, they said they wanted two. When they heard that their neighbors bought two, they wanted three or four…. They all wanted to show that they loved their child even more and to show off.³⁶

The idea of an insurance policy as a gift was so popular that some parents replaced the custom of giving “fortune money” to their child during the Lunar New Year with the new practice of buying the child an insurance policy. The popularity of Ping An’s child policy was not confined to Shanghai. A Ping An representative in Beijing reported that some 80% of the company’s life insurance policies sold in 1996 were issued to children.³⁷

Mobilizing the habit of saving and economic rationality

From 1996 to 1999, there were seven consecutive interest rate cuts in China. To capture the life insurance market, Ping An did not adjust its products’ guaranteed interest rates accordingly, making its products competitive savings plans. Ping An agents then mobilized an economic rationale that calculated economic gains and losses from an insurance policy. When meeting with prospects, they compared their products to the savings plans offered by banks and talked about “savings,” “returns,” and “good deals.” For the clients, their primary motives for purchasing life insurance shifted from fulfilling renqing or showing love for the child to saving and investing for future economic gains. During 1996–1997, whenever the interest rates in savings banks were about to drop, people queued up outside the headquarters of Ping An to purchase insurance policies, fearing a similar downward adjustment of the interest rates for Ping An’s products. As a result, Ping An’s sales volume exploded. In 1998, Ping An’s revenue from individual life policies reached 1.71 billion yuan ($206.6 million), whereas AIA’s was only 74.7 million yuan ($90.3 million). This newly emerged domestic insurer held over 46% of the total individual life premium income in Shanghai, the largest market share, just 4 years after it issued its first life policy.

The price of accommodating to the local preferences

Ping An took the lead in accommodating to local preferences by indigenizing the products and transcribing the definition of life insurance into money management. While other domestic insurers adopted Ping An’s model, AIA and other foreign insurers were at pains to defend the risk management definition of life insurance. In fact, accommodating to the local preferences and adopting a market-share institutional logic came with a price—most products that Ping An and other domestic insurers sold during 1996–1999 induced losses. By June 2000, the financial loss Ping An incurred from selling unprofitable savings products in Shanghai and other major cities amounted to 10 billion yuan ($1.2 billion). Other domestic insurers suffered from the same problem.³⁸

By the middle of 1999, the regulatory body mandated that domestic insurers stop selling products with high interest rates and set a ceiling rate for all life insurance products throughout the country. The disappearance of appealing savings products slowed the expansion of the market. The growth of life premium income in 1999 was below average at only 15.7% in Shanghai and 16.6% in China overall (see Fig. 1).

Capitalizing on the “stock fever” fad

In response to the state’s regulation, Ping An still refused to follow foreign insurers’ conservative but profit-oriented operation model. Instead, it insisted on its market-share strategy. Soon after it was mandated to stop selling high-interest-rate savings policies, it launched an investment product, unit-link, in October 1999 to capitalize on the “stock fever” fad. When the stock exchange was introduced to Shanghai in the early 1990s, it received an ebullient reception, as captured by Hertz’s (1998:3) ethnographic description that there was a “remarkable outpouring of enthusiasm for stock trading at the popular level”. In 2000–2001, the Shanghai stock index hit its highest point, which produced an optimistic atmosphere for investment. Catching the “stock fever” fad and the high stock index at the time, Ping An agents talked about “profits” and “investments” to their prospects. They exaggerated the profitability of the unit-link, and the prospects themselves were optimistic about the possible returns. According to the agents of Ping An, some prospects and clients actively approached them specifically for this product. A locally coined “unit-link fever” phenomenon was seen.

Although foreign insurers never liked the idea of defining life insurance as a new investment tool, they offered dividend insurance to compete with Ping An’s unit-link. The agents found it easier to talk to the prospects about “dividends,” “returns,” and “investments.” Even if their products served primarily a protective function and carried a small dividend element, they highlighted the dividend feature to make their products sound like an investment. The unit-link and dividend policies together brought another surge of insurance market growth to 52.4% in Shanghai and 42.8% in China overall in 2001 (see Fig. 1). The meaning of life insurance was re-defined from a savings instrument to a profit-making financial tool.

Getting around the cultural taboo to sell risk management products

The “unit-link fever,” nonetheless, came to a halt as a consequence of the downturn in the stock index toward the end of 2001. As the majority of unit-link clients did not really understand the risks involved in this product when they bought it, they blamed Ping An for misleading them. A portion of angry clients appealed to Ping An for refunds, resulting in a so-called “unit-link crisis” in Ping An.

Ping An’s sudden expansion was driven by three popular products: retirement insurance, the child policy, and the unit-link. However, increasing regulations imposed by the regulatory body and the unit-link crisis left Ping An without an option but to join its foreign counterparts to develop risk management products. Since 2002, Ping An and other domestic insurers have sided with their foreign counterparts in promoting risk management products that were more profit guaranteed.

Given that domestic insurers were used to selling money management products, how did their agents now begin to sell risk management products? Starting from mid-2002, Ping An concentrated on selling policies that combine life, accident, hospitalization, critical disease, endowment, and dividend components. While avoiding the topic of sudden death, the sales agents skillfully presented the policies as analogous to the long existing local practice of risk management, mainly, savings, and the widely accepted new form of money management, namely, investment. Lai Jiemin (aged around forty), a senior agent of Ping An, was quite successful in selling these policies.³⁹ I observed one of her sales to a woman in her early thirties. When Lai introduced the product, she started off by highlighting its money management function: “This is the newest dividend product of our company…. The dividend is given to your policy account annually....” When the prospect heard of the word “dividend,” she seemed curious. She asked if the dividend could be taken out annually. The answer should have been “no,” but Lai skillfully activated the usual concern about retirement to get around the question: “You don’t want to take the dividend out every year. You want to leave it in our company to let it accumulate…. You want to leave it in our company to generate profit. You can then use the profit as part of your retirement fund.” The prospect asked about whether the dividend rate was fixed. She wanted to have a fixed rate. Lai responded:

The dividend rate depends on the annual investment returns of our company…. An insurance company is a financial company…. The head of our actuarial department, Stephen Meldrum, was recruited from the United States. He is serving on the investment management committee of our company. Our investment returns must be good.

This time Lai framed the uncertain returns in terms of “investment” and drew on the notion that “American things must be good” to handle the prospect’s query.⁴⁰ The prospect seemed convinced. But she was not interested in the accident component because it had no cash value. Lai suggested that she cut down the insured amount for this component. The prospect finally agreed to get this policy. Throughout the hour-long meeting with the prospect, Lai did not mention death or other misfortunes. She talked about dividend, return, profit, and retirement, though the policy served various risk management functions covering sudden death, accident, hospitalization, and critical diseases.⁴¹

Facing a population that has an unyielding preference for money management products, insurance agents naturally craft their sales talks according to such a preference. In fact, the agents often share with the prospects the same perception of risk, even though their companies might have told them otherwise, or their products might have forced them to say otherwise. For instance, I observed an agent of AIA unintentionally slipping into her client’s definition of risk when interacting with the client. The following extract of ethnographic field note is illustrative of this point:

Chu (the agent) met with Zhang (the client) to sell him a new whole life policy that would benefit mainly the beneficiary…. She focused on the benefits that the policy could bring for Zhang’s 16-year-old son…. When Zhang refused, saying that he wanted his son to be financially independent…. Chu got stuck and didn’t know what else to say. She had managed only to leave him the product leaflet….⁴² On our way back to Chu’s office, I asked her if Zhang could consider putting his wife as the beneficiary if he preferred his son to be financially on his own. Chu responded, “It doesn’t make sense to put his wife as the beneficiary because they are about the same age.” What Chu implied was that Zhang and his wife would pass away at about the same time and so his wife would not benefit from it.

During an interview prior to the visit to Zhang, Chu, like many other agents of foreign insurers, emphasized that the main function of life insurance was for managing unpredictable risks and misfortunes, such as accidents and critical illnesses. However, when she was interacting with her client, the idea that Zhang might die long before his wife’s death did not come to her mind. Her instinctive alignment with the client was rooted in the deeply ingrained taboo on thinking about premature death and the subjective immunity that they share. This inconsistency between what the agent said in the interview and what she did when interacting with her client reflects the conflicting logics between what she has learned from her company and what she has in her cultural schemas. Using terminology from social psychology, what she learned from the company belongs to “deliberative cognition” that rarely, and only with conscious effort, can override what cultural schemas, or “automatic cognition,” offer (DiMaggio 1997). The possibility of sudden death is out of the automatic cognitive frame of not only the prospects but also the sales agents, unless the agents constantly make the conscious effort to expand their frame to embrace it. Thus, the agents who primarily sell risk management products have much greater difficulty convincing their prospects of the necessity of the products, because they themselves share the perception of risk with the prospects that mismatches the logic of the products.

It was not by coincidence that the growth of the market began to slow in 2002 when Ping An and other domestic insurers began modeling more closely on the risk management product development of their foreign counterparts. Life insurance sales experienced rapid growth because novice domestic insurers violated the profit-oriented institutional logic of insurance to accommodate to the local preferences. When Ping An and other domestic insurers finally succumbed to creating a risk management market, their market share dropped and the premium growth both in Shanghai and China overall declined substantially. During 1998–2001, the average annual growth of Ping An’s individual life business in Shanghai was 47%. The figure plunged dramatically to merely 1.26% during 2002–2004. Its market share fell from 46.5% in 2001 to 30% in 2004. At the same time, the growth of the life insurance business in Shanghai markedly declined from over 52% in 2001 to only about 20% in 2003. In 2004, it experienced negative growth (−0.6%) for the first time since its emergence (see Fig. 1).

The role of culture in life insurance development

This project brings together two different concepts of culture—one as a shared meaning system and one as a practical tool-kit—to understand the characteristics of the emergent Chinese life insurance market. Figure 2 specifies the mechanisms by which the two forms of culture shape the formation of the life insurance market in China.

Fig. 2

The interplay of two forms of culture in shaping a market formation

This case study demonstrates that the Chinese concepts of life and death produce a taboo on thinking and talking about premature death, define a “good life” as living well toward the end and a “good death” as dying after a full life, and filter selective attention to fatal risks. These values and perceptions are incompatible with the probabilistic assumption of risks and the commensuration logic on which commercial life insurance operates. Together they compose public resistance to receiving life insurance as an instrument of risk management (arrow A). This resistance is manifested by a preference for payments for the insured, adherence to savings as the primary mode of risk management, a feeling of purely protective products being a “waste of money,” and a denial of the possibility of fatal misfortune. This public resistance necessitates the invention of marketing strategies to remove or circumvent such resistance (arrow B). Foreign and domestic insurers in China adopted rather different strategies. The former attempted to remove the resistance by raising the local people’s sense of risk, and they accommodated the local preferences only to the extent that profits were not jeopardized. The Chinese insurers, on the other hand, endeavored to circumvent the resistance by re-defining the concept of life insurance according to local existing risk management practice and habit of money management, modifying product features to accommodate local preferences, deploying marketing strategies that took advantage of local etiquette, and designing sales talks that were appealing to their audience. It is through these circumventing strategies that life insurance sales were fostered (arrow C).

Thus, a Chinese life insurance market is made possible by the insurers and their sales agents’ strategic use of the local cultural repertoire to overcome the lack of public receptivity due to incompatible cultural elements. Nevertheless, the market has emerged with a trajectory and features different from the markets in Euro-American contexts. It first emerged as a money management market. The characteristics of this market, I propose, are shaped by the dual processes (the dotted arrows in Fig. 2) in which one form of culture composes the resistance and the other form of culture circumvents it. I argue that the role of culture as shared ideas in shaping human action is not as inactive as Swidler (1986) proposes. If the Chinese concepts of life and death did not matter, transnational life insurers would have succeeded in inculcating the risk management concept of life insurance into the local population and creating a predominantly risk management market.

The next question is: Does this Chinese case, with “life insurance” being defined, presented, and accepted as a variant of savings and an investment instrument, really demonstrate the creation of a “life insurance” market? Furthermore, to what extent has the new idea—life insurance as a modern form of risk management—really been introduced to the urban Chinese through mobilizing the cultural tool-kit? As reported in the last section, all life insurers, including the domestic ones, have begun concentrating their efforts on selling predominantly risk management products since 2002. Although this has slowed the growth of sales, the market, in effect, has been moving from a money management market to a money-cum-risk management market. While revenue from selling investment insurance (dividend and unit-link) in Shanghai decreased 12% and 8.5% in 2004 and 2005 respectively; revenue from selling personal accident insurance increased 9.9% and 18.8% in the same years. Does moving toward a money-cum-risk management market represent an increasing acceptance of the risk management concept of life insurance? Providing a valid and reliable answer to this question requires further research. Nevertheless, my data collected in late 2004 reveal that clients who had received payments from insurers or non-clients who witnessed close relatives or friends receiving payments were more likely to accept life insurance as a risk management instrument. Thus, life insurance as a new commodity and an alien concept was first re-framed and packaged as something familiar and culturally acceptable to the local Chinese. The purchase and consumption of this new commodity, inevitably, has had a material impact, which then became an agent of cultural change. Life insurance was first bought by Chinese clients primarily as a money management instrument, but the idea of risk management embedded in this commodity has been gradually discovered and accepted by these clients and other onlookers. Hence, life insurance is not only a product of the institutionalization of rationalization but also an instrument of such a process. The finding that cultural re-framing elicits certain actions likely to bring about new values supports Swidler’s (1986) theory of culture, though I have emphasized throughout that existing values have an impact on the use of the tool-kit for re-framing.

Moving beyond the case of China, the interplay of the two forms of culture as illustrated in Figure 2 can be used to reinterpret the making of the life insurance market in nineteenth century America. According to Zelizer’s (1979) documentation, Christianity-dominated socio-cultural values (including sacralizing human life and condemning speculating on death or collecting instant wealth) prior to mid-nineteenth century America resulted in widespread cultural resistance to life insurance (arrow A). The incompatibility between the socio-cultural values and the commercial logics of life insurance necessitated that life insurance providers adopt a religious discourse of life insurance from the 1840s to the 1860s so as to sanctify the nature of life insurance (arrow B). Existing cultural symbols and practices were mobilized by life insurers and their sales agents to undermine the commercial nature of life insurance and to bestow a quasi-religious function on the new commodity. Life insurance companies presented themselves as “human and benevolent institutions” and sales agents presented themselves as “the clergy” to convince the prospects that buying life insurance was a “religious duty” (Zelizer 1979:96–99). Furthermore, life insurance providers also capitalized on the traditionally male dominant culture by symbolizing the purchase of life insurance as a moral obligation of the breadwinners to provide “a form of economic immortality” for their wives and children. Life insurance was presented as an instrument that could prevent their wives from re-marrying so that they could control the sexual life of their wives upon their deaths (Zelizer 1979:58–61). All these strategies were deployed to define and package the nature, the meaning, and the appearance of life insurance in order to circumvent resistance (arrow C).

Zelizer attributed the sudden expansion of the American life insurance market to a change in cultural values. However, if a growing awareness of the economic loss of death and a rational calculation of economic risks preceded the development of the life insurance market, why did life insurers have to shun economic terminology in selling and deploy religious missionary rhetoric instead (1979:91–101)? Indeed, as Heimer (1980) noted, a number of quotations Zelizer (1979:57–63) cited to illustrate the newly emerged economic evaluation of a “good death” as proof of a change in cultural values came from insurers’ publications in the 1880s, 1890s, and even in 1964 when the life insurance market was already firmly established. There was little evidence in the data she provided that the societal cultural values had changed prior to the market’s expansion. This critique of Zelizer’s argument, however, does not deny the role of cultural values in shaping the development of life insurance. In both the Chinese case and the American case, the shared cultural values compose the resistance to accepting the ideological logics of commercial life insurance, which necessitate the strategic uses of the existing cultural tool-kits to circumvent the resistance.

In addition, the mechanisms of the interplay between shared values and cultural tool-kits can be applied to reinterpret the paradigmatic shift of life insurers’ marketing strategies in the American case. In Zelizer’s (1979) book, she further documented the change in life insurers’ marketing strategies from “moral persuasion” to “business enterprise” after the 1870s. Since that time, life insurers have made no secret of the commercial nature of their institutions and have presented life insurance as an investment for personal economic advantages. Zelizer attributed this sudden shift to the increasing scale and power of the insurance corporations and the consequent impossibility of hiding their commercial nature (1979:101–4). This explanation is partial, as Zelizer, surprisingly against her argument that “cultural values matter,” ignored another equally important factor at that moment—the disenchantment of the larger cultural context.⁴³ Secularization of life insurance, indeed, coincided with a larger rationalizing force, namely the secularization movements that spread from Europe to America (Durkheim 1964; Berger 1967; Glock and Bellah 1976; Dobbelaere 1981).⁴⁴ Therefore, a reinterpretation of the shift from a religious to an economic presentation of life insurance is that a change in societal cultural values (the rise of secularism) necessitated a change in the cultural symbols to be mobilized (arrow B). As a result, the content of the circumventing strategies after the 1870s was different from that of the earlier period, though the form remained the same (arrow C).

Conclusion: culture, market, and globalization

The aim of this article is not to criticize existing accomplishments in the literature on culture, action, and economic practice, but rather to extend Zelizer’s insights and to integrate the practical, tool-kit conception of culture with the symbolic, coherent meaning perspective. Deriving from inductive, ethnographic research, this article highlights the agency of both the globalizing agents and the local receivers, and yet it pays equal attention to the structural constraints imposed by local cultural logics and profit-oriented commercial logics, and how the actors interact with these structural forces. It sheds light on two related subjects: culture as it relates to the market, and as it relates to globalization.

This article may fall into the category of taking a “middle” theoretical position, as offered by Schudson (1989) who examined how culture works both as structure and as agency. It is, however, a little more (or less) ambitious than proposing a “middle position” theory.⁴⁵ It explores the mechanisms through which the two very different forms of culture interact. Moving beyond the life insurance industry, I propose three mechanisms through which the two forms of culture interact in shaping a new economic practice or the formation of a new market.

First, shared meanings and values necessitate the mobilization of the tool-kit dimension of culture to steer the direction along which the new economic practice or market develops. This mechanism recaptures and extends the Weberian theory of culture and economic practice. Culture as shared worldviews and beliefs underlying theoretical and ontological assumptions about the world (Campbell 1998) generates a configuration of folklore, values, moralities, and perceptions. This folklore, values, moralities, and perceptions in turn mold a pattern of preferences and dispositions that can be compatible or incompatible with the profit-making logic of a new economic practice. If this pattern is largely incompatible with the new economic practice, it necessitates the mobilization of the tool-kit dimension of culture to circumvent the incompatibility. The circumventing mechanism works like “bricoler”—extraneous movements made to avoid an obstacle. Such movements inevitably shape the path and the features of the market.

Second, the configuration of folklore, values, moralities, and perceptions, together with institutions, constitutes the symbolic contents of a cultural tool-kit and circumscribes the selective use of symbolic elements from the tool-kit.⁴⁶ This is why shared meanings and values in their own right rarely completely suppress a market embodying a contradictory set of values from emerging, precisely because they simultaneously constitute a repertoire containing numerous cultural elements that can be used to advance a practical agenda. The tool-kit elements, manifested as etiquettes, recipes, rituals, habits, styles, fads, skills, capacities, know-how, and practices, are constituted by various meaning systems and institutions. They are, therefore, heterogeneous and vast in number. However, as practices must be anchored in shared meanings (Sewell 1999; Swidler 2001), the shared meanings system prescribes a “matrix of possibilities” (Goffman 1967) by “setting the floors and the ceilings” (Douglas and Wildavsky 1982) or “blocking out certain possibilities” (Emirbayer and Goodwin 1994) for effective uses of cultural symbols and practical skills (Lamont 1992).

Last but not least, in circumventing the tension between the local cultural logics and the ideological logics of the new economic practice, it is economic actors who materialize the tool-kit into resources. Since the tool-kit dimension of culture draws its properties from various logics and institutional resources, it is indeed like “a bag of tricks or an oddly assorted tool kit” as portrayed by Swidler (2001:24). The contents of these elements can be incoherent and even in conflict, but they coexist because actors have bounded agency—the ability to pick and choose only those resources that are useful at a given time for a particular situation. As Swidler (1986, 2001) emphasizes, the cultural elements in a tool-kit have adaptable and flexible meanings for managing urgent, practical situations. I should add that, however, they are not a pool of resources in their own right. It is the actors who turn the normative rules and cultural symbols into resources. As rules and symbols can be contradictory and carry multiple meanings in different contexts, the economic actors must have enough local knowledge to mobilize judiciously the cultural elements needed to orchestrate the new economic practice.

In this model, coherent and incoherent cultural elements are not mutually exclusive. The role of the cultural tool-kit in constituting action does not exclude an equally, if not more important, role for shared ideas in the process. I maintain that the debates over whether culture is coherent or incoherent are not productive to the field of cultural sociology. What will be more productive, and here I agree with Jepperson and Swidler (1994), is to study the hierarchic order of different cultural elements affecting human action. However, theorizing about various cultural dimensions in a quasi-hierarchic order demands more empirical research, beginning with non-hierarchic treatment of different forms of culture in order to explore the many possibilities of their linkage and ordering.

This article also illustrates how the global diffusion of a particular capitalist enterprise is possible, despite variation in local cultures and cultural barriers. My findings point to the fact that localization is a necessary strategic move in creating local markets during the course of globalization. It is through localized interpretations and definitions that a particular capitalist mechanism can be grounded in a local context. These findings are in line with Dobbin’s (1994) argument that different moral perceptions shape different paths of railway development and support a moral construct perspective of the market (characterized by Fourcade and Healy 2007). In this perspective, the valuation and classification of particular commodities, or even of the market itself, is not fixed or predefined but is empirically variable. The Chinese life insurance buyers’ understanding of life insurance as primarily a money management category unmistakably illustrates this point. However, just as Fourcade and Healy (2007) noted that local variations should not blind us to the fact that unequal power exists between actors and between justificatory logics, my case study demonstrates that localization is not always a straightforward path. When pressure for localization is in conflict with the profit-oriented institutional logic of a particular business, negotiations and power struggles set in between transnational and local economic players, on one hand, and economic players, states, and local populations, on the other. Despite the Chinese insurers’ daring adoption of a market-share institutional logic at first, they finally succumbed to a profit-oriented strategy and began to imitate more closely their foreign counterparts’ product development. As different institutional logics have different connections to the economic performance (or profitability) of a market, the extent of localization is not without limits.

This brings us to the question of globalization and varieties of capitalism. This argument—that Chinese culture, no matter how different it is from western civilizations, is not an insurmountable barrier to the development of capitalism but shapes China’s path to capitalism—is not new (see e.g., Hamilton 1996, 2006; Hamilton and Biggart 1988). Along the lines of this argument, my study reveals the mechanisms and the micro-macro links among different levels of culture that produce the Chinese trajectory. The recent neo-instititutionalist literature has provided us with insights into the mechanisms by which varieties of capitalist practices are produced (see e.g., Fligstein 2001; Guillen 2001; Hall and Soskice 2001; Campbell 2004). Nevertheless, the role of culture in neo-institutionalism is construed either as a context in which institutional forces directly constitute the varieties of practices (see comments by Hallett and Ventresca 2006; Binder 2007), or as an intervening variable between institutions and practices. For instance, confining their concept of Chinese culture to the overarching Confucian ethic of obedience to authority, Hamilton and Biggart (1988) discredit a cultural account for national differences in market features and trajectories. However, as demonstrated through the Chinese life insurance case, culture in the form of overarching shared ideas can influence economic practices through a series of transmutations of cultural forms: from the concepts of life and death to values, moralities, and perceptions, and to preferences and choices. Culture on a very macro and high level may not shape action directly. Nonetheless, through transmuting its forms, culture travels from historically distant worldviews and ambiguous images to foreground preferences, perceptions, and dispositions that more directly organize actors’ strategies of action. The role of culture in shaping a market, therefore, is not as remote as Hamilton and Biggart propose. This article asserts that culture in the form of shared ideas can act as an exogenous factor and influence practices directly through a series of transmutations of cultural forms. The conditions and mechanisms through which cultural elements transmute and travel are yet to be explored.