1. Introduction
In Western societies, everyday experience suggests that in-kind gifts are generally more common than money gifts.Footnote 1 Indeed, several studies confirm that money is usually not appropriate as a gift (Webley and Wilson, Reference Webley and Wilson1989; Webley et al., Reference Webley, Lea and Portalska1983). However, the prevalence of in-kind gifts has puzzled economists because the giver may sometimes choose an in-kind gift that the receiver does not appreciate. Therefore, are money gifts not more efficient than in-kind gifts? Several studies argue that the latter do indeed create inefficiency (Principe and Eisenhauer, Reference Principe and Eisenhauer2009; Waldfogel, Reference Waldfogel1993, Reference Waldfogel2002). Granted that in-kind gifts are inefficient, why does the practice of in-kind gift-giving persist?
Several solutions to this puzzle have been suggested. One type of solution is based on the view that every observed institution is efficient (Leeson, Reference Leeson2019) and therefore, the inefficiency of in-kind gifts is only apparent. One such solution is based on the idea of signalling. It is argued that unlike money, in-kind gifts can signal some characteristics of givers and/or receivers (Camerer, Reference Camerer1988; Prendergast and Stole, Reference Prendergast and Stole2001). Another explanation for the prevalence of in-kind gifts postulates paternalistic givers, who use in-kind gifts to influence or educate receivers (Tremblay and Tremblay, Reference Tremblay and Tremblay1995). Yet another explanation involves imperfectly informed altruistic givers and asymmetric search costs between givers and receivers (Kaplan and Ruffle, Reference Kaplan and Ruffle2009). This explanation highlights that givers may be able to buy a gift that receivers desire at a lower cost than the receivers.
The justifications above show that in-kind gift-giving may be rational behaviour under certain assumptions about givers’ preferences and information and technological constraints. Alternatively, in-kind gift-giving can also be conceptualised as a social convention. That is, it may represent a rational response to given social constraints – to others’ choices and expectations. Such a convention may survive for a long time, even if it is inefficient, because of collective action problems (Camerer, Reference Camerer1988). Moreover, over time, individuals may develop a taste for in-kind gifts and stigmatise money gifts, which further increases the stability of the convention (Greif and Laitin, Reference Greif and Laitin2004). As a result, a society may consider money gifts inappropriate simply owing to path-dependency and lock-in effects (David, Reference David1994; Heikkila, Reference Heikkila2011; North, Reference North1990).Footnote 2
All these models are able to explain some facts about gift-giving. Nevertheless, most of them take unacceptability of money for granted. They largely overlook the fact that in some societies – in particular, in Chinese societies – money seems to be generally accepted as a gift.Footnote 3 On various occasions, such as Chinese New Year celebrations or weddings, the Chinese give each other red envelopes with money (so-called red packets; Mandarin: hongbao, Cantonese: lishi). In this study, we focus on Chinese gift-giving norms. In particular, we ask the following questions: When do the Chinese consider in-kind gifts appropriate? When do they consider money gifts appropriate? Can the existing models explain gift-giving norms of the Chinese? Do we find patterns that are common to both Chinese and Western societies?
To answer these questions, we conducted two surveys in China (n 1 = 796, n 2 = 795). In the first survey, respondents were asked to evaluate the acceptability of various forms of gifts on different occasions and to various receivers from the perspective of givers. According to Bicchieri's (Reference Bicchieri2017) terminology, this first study focuses on normative beliefs about gift-giving. From this perspective, the present study follows contributions such as those of Webley et al. (Reference Webley, Lea and Portalska1983) and Webley and Wilson (Reference Webley and Wilson1989). In the second survey, respondents stated appropriate values of in-kind/money gifts given on specific occasions to receivers in specific relationships.
We find that none of the existing models of gift-giving, apart from the model that considers gift forms as conventions, can fully explain the acceptability of in-kind gifts versus money in Chinese societies. We highlight the fact that not only the relationship between the giver and the receiver but also the gift-giving occasion matters for the acceptability of a particular gift form. Yet theoretical models generally overlook the influence of gift-giving occasions on the acceptability of money and in-kind gifts. Instead, they focus on the relationship between the giver and the receiver. Although our sample is mainly composed of urban households, our results confirm that in China, money, especially in the form of red packets, represents a generally appropriate form of gift. Even though this finding differs from the findings of studies on money versus in-kind gift-giving in Western societies, we do find some common patterns between Chinese and Western practices. For example, in both societies, money is usually given from elders to young people and children, especially within families. Therefore, while we emphasise the role of conventions in determining the gift form, our results are in line with the view that gift forms are determined by multiple factors. This multi-causal view is also supported by existing empirical literature (Waldfogel, Reference Waldfogel1993, Reference Waldfogel2002).
We also find that for agents in China, unlike in Western societies, money does not seem to be associated with a stigma. Therefore, givers who rely on money gifts do not feel obliged to compensate the receivers with higher value, arguably because money is not considered an inferior alternative to in-kind gifts. We interpret this result as further supporting the view that the choice of gift form is governed primarily by convention. This supports the view that red packets are used as part of an exchange ritual between individuals (Hermann-Pillath et al., Reference Herrmann-Pillath, Feng and Guo2019). Further, while in-kind gifts and red packets are both popular gift forms, gift vouchers are by far the least popular. This finding also contradicts studies on gift-giving in Western societies, which show that the acceptability of vouchers sits between money and in-kind gifts (Webley and Wilson, Reference Webley and Wilson1989; Webley et al., Reference Webley, Lea and Portalska1983).
The paper proceeds as follows. In section 2, we describe the Chinese practice of giving red packets from a historical perspective, and show how this practice fits within the broader framework of gift-giving. In section 3, we review existing theoretical models regarding gift/money-giving and derive hypotheses from these models. In sections 4 and 5, we present our data and results, respectively. In section 6, we discuss our results vis-à-vis stylised facts about Western gift-giving and the existing theoretical models. Section 7 concludes.
2. Red packets and gift-giving in China
The act of gifting has been at the core of social relationships in Chinese societies for centuries (Yan, Reference Yan1996). As a complex institution regulating inter- and intra-household relationships, gifting in China has attracted significant academic attention. Acting as an important instrument not only for networking and building social capital but also for signalling filial piety, gifting is used as the main means to express and maintain relationship-based ties specifically identified in the Chinese culture – including concepts such as guanxi (one's particularistic ties; Kipnis, Reference Kipnis1996) or renqin (favour exchange; Ruan, Reference Ruan2017). The flow of gifts in China permeates society and is advanced to follow existing norms of gift-giving (Yan, Reference Yan1996), regulating a wide range of exchanges, from the casual gifting of cigarettes to gifts to medical professionals during consultations. It is within this context that we assess the choice between money gifts (‘red packets’) and in-kind gifts.
Red packets are simply envelopes filled with money. They are commonly exchanged within, and between, households on various occasions, such as the Chinese New Year. Red packets have traditionally been associated with a wish of good luck. The tradition of giving red packets can be traced back several hundred years, when coins, then round with square holes in their centre (tongqian), were bound together with a red rope and given to children to protect them from evil (Siu, Reference Siu2001). In the late Qing dynasty, as coin shapes evolved into their modern form, red envelopes replaced red ropes. In contemporary Chinese communities, red packets contain bank notes rather than coins. They are still commonly exchanged on various occasions, ranging from birthdays to weddings, although it has been questioned whether the initial wish of luck still stands as a motive behind the gift (Siu, Reference Siu2001), as opposed to it being a simple gift in cash. It should be noted that the red packet has also been associated with corruption and bribery (Smart, Reference Smart1993), as it is a term also used in those settings. This aspect of gifting, however, will not be covered in this paper as we centre on gift exchange within personal networks in order to fit the purposes of existing theoretical models under discussion.
Interestingly, the exchange of red packets is codified and governed by social norms, with small regional variations. These norms indicate that age and status matter in determining the direction of the transfer. Commonly, children and young people receive from elders and the unmarried receive from the married. Outside the family as well, red packets are given according to a similar top-to-down pattern, such as managers giving it to their subordinates and individuals to their friends’ or acquaintances’ children. We identify some of these patterns in the data in section 3. Siu (Reference Siu2001) provides a detailed description of these norms as well as of the modern history of red packets in Hong Kong and Guangdong province.
With the emergence of online and mobile networking applications, such as WeChat, red packets have also undergone a recent adaptation to an online form (so-called e-hongbao). Although physical red envelopes are not included, they are exchanged according to social rules as previously described (Wu and Ma, Reference Wu and Ma2017) and are, for all intents and purposes, simple transfers of cash.
We argue that the existence of red packets, whether in physical or electronic form, marks an important difference between gift-giving norms in Chinese and Western societies. In particular, their existence suggests that unacceptability of money as a gift, identified by a large body of related literature, may not be a universal phenomenon. In the following section, we review theoretical models of gift-/money-giving and how red packets fit within them.
3. Theories of gift-/money-giving and the case of red packets
This section explores existing theoretical models that aim to explain why people in Western societies often give each other in-kind gifts rather than money. In this process, these models also yield predictions about factors affecting the acceptability of money vis-à-vis in-kind gifts. To add a dimension to the usual dichotomy of gift forms, we also introduce into this theoretical discussion the case of gift vouchers, conceptualised as an intermediary option following several of the theories discussed in this section.
According to the signalling model (Camerer, Reference Camerer1988; Prendergast and Stole, Reference Prendergast and Stole2001), in-kind gifts are a signalling device when an information asymmetry exists between the giver and the receiver. Specifically, in-kind gifts may reveal that the givers invested time and effort in discovering tastes of the receivers.Footnote 4 Therefore, by giving in-kind gifts the givers signal that they care about the receiver. From the perspective of the signalling theory, what matters to the receivers of the gifts are the costs incurred by the givers. In line with this prediction, studies show that the more costly gifts are, in terms of time and mental or physical effort invested, the more valued they are by the receivers (Robben and Verhallen, Reference Robben and Verhallen1994). The signalling model can also explain Flynn and Adams’ (Reference Flynn and Adams2009) finding that the receiver's valuation of a gift is not positively correlated with its purchase price. Unlike in-kind gifts, money cannot serve as a signal of knowledge of the receivers’ preferences for two reasons. First, money is not relationship-specific. To continue with the previous example, it does not signal knowledge of the receivers’ tastes. Second, money is valued by the receivers, which may attract dishonest individuals who will exploit the givers by collecting gifts without reciprocating (Camerer, Reference Camerer1988; Posner, Reference Posner1997).
From the signalling model, it follows that if no information asymmetry exists between the giver and the receiver, there is no reason to signal with in-kind gifts. Therefore, the signalling model predicts that the more familiar the giver and the receiver of the gift are with each other, the more appropriate money is as a gift. For instance, we should observe more money-giving among nuclear families than among other relatives and acquaintances. The signalling model also explains the 19th-century rule in some Western societies that money gifts should be exchanged only between relatives and closest friends (Zelizer, Reference Zelizer1994). However, some empirical studies on holiday gift-giving reveal that this is not always the case. Waldfogel (Reference Waldfogel1993, Reference Waldfogel2002) finds that money gifts are by far the most common from grandparents, followed by aunts and uncles. At the other extreme, friends, siblings and significant others have the lowest tendency to give money gifts.Footnote 5 Other studies on holiday gift-giving also confirm that grandparents have the highest tendency to give money (Caplow, Reference Caplow1982; Mortelmans and Sinardet, Reference Mortelmans and Sinardet2004).
The signalling model also offers predictions about the acceptability of gift vouchers. Unlike money, vouchers can, to some extent, signal knowledge of recipients’ preferences but not as effectively as in-kind gifts. At the same time, vouchers are more efficient than in-kind gifts but not as efficient as money. However, from the perspective of the model, it is unclear whether vouchers (i.e. the combination of characteristics of in-kind gifts and money) will generally be more or less preferable to in-kind gifts or money in their pure forms. Nonetheless, the signalling model predicts that if vouchers are appropriate, then it is for intermediate types of relationship in terms of closeness. For these kinds of relationships, givers and receivers are expected to be somewhat familiar with each other, and therefore there is less need for the givers to signal their care about the receiver.
In sum, the signalling model advances that the relationship between the giver and the receiver matters for the gift form, while it offers no predictions about the effect of the gift-giving occasion on the gift form. The reason is that types of relationships between the giver and the receiver differ in the degree of information asymmetry (and therefore, the need to signal), while gift-giving occasions do not affect the degree of information asymmetry.
The search-cost model of gift-giving with imperfectly informed altruistic givers (Kaplan and Ruffle, Reference Kaplan and Ruffle2009)Footnote 6 also emphasises the role of the relationship between the giver and the receiver in deciding the form of the gift. This model points out that although givers may have imperfect information about receivers’ preferences, they may have lower costs in obtaining the gifts that receivers desire than the receivers themselves. For example, givers and receivers may make their purchases in different shops. Consequently, givers will give money rather than gifts if they do not have a search-cost advantage over the givers and if they are unfamiliar with receivers’ preferences.
The search-cost model offers opposite predictions about the closeness of relationships and acceptability of money gifts to those of the signalling model. Namely, it predicts that we should observe more money-giving among more distant relatives and acquaintances. For example, grandparents will tend to give money not only because they are unfamiliar with their grandchildren's preferences but also because they are likely to have higher search costs. This view is confirmed by empirical studies showing that givers who are expected to give unappreciated in-kind gifts tend to give money and that grandparents have the highest tendency to give money (Waldfogel, Reference Waldfogel1993, Reference Waldfogel2002). Regarding vouchers, the model predicts that their acceptability depends on the familiarity with receivers’ preferences. In particular, givers will choose gift vouchers if they are somewhat familiar with receivers’ preferences (and therefore, they do not give money) but not sufficiently to give in-kind gifts. Consequently, the prediction of the search-cost model regarding the acceptability of vouchers is similar to that of the signalling model although the underlying rationale is different.
Paternalistic models of gift-giving assume that givers want to educate or influence the receivers (Tremblay and Tremblay, Reference Tremblay and Tremblay1995). Prominent examples of paternalistic gifts include encyclopaedias or sports equipment given from parents to their children. Therefore, the model predicts that we should observe less money-giving from parents to children than in other types of relationships. In line with the paternalism hypothesis, Trembley and Trembley (Reference Tremblay and Tremblay1995) show that the share of children in the population has a positive effect on per-capita Christmas spending. However, Waldfogel's (Reference Waldfogel1993, Reference Waldfogel2002) result that suggests money gifts are more common from individuals who tend to give unappreciated gifts suggests that givers care about receivers’ utility. This finding contradicts the model of paternalistic giving.
Finally, we consider the case according to which a particular gift form is a social convention. Convention is usually modelled as an equilibrium of a game with multiple equilibria (Burke and Young, Reference Burke, Young, Benhabib, Bisin and Jackson2011). In other words, givers choose a specific gift form because of receivers’ expectation. Simultaneously, receivers expect a specific gift form because givers typically choose this specific gift form. If the expectations were different, people would choose different forms of gifts. The model predicts the local conformity and global diversity of gift forms (Young, Reference Young1996). This global diversity is unlikely to be explained by differences in information and technological constraints or by differences in givers’ preferences. Therefore, we consider this model a residual explanation: If signalling, search cost, information asymmetry or paternalism do not explain the form of gift, then we conclude that the gift form is determined primarily by convention. The gift-form-as-convention explanation is in line with Waldfogel's (Reference Waldfogel1993, Reference Waldfogel2002) evidence that the gift form is not fully determined by cost–benefit considerations in a narrow sense. This model is also the only one that allows for the gift form to be occasion-specific. In particular, we may view various gift-giving occasions as distinct games with multiple equilibria. Each game may finally have a different equilibrium because of different initial conditions and different endogenous or exogenous shocks over time (David, Reference David1994; Young, Reference Young1996).
We derive several hypotheses from the aforementioned models. First, we focus on the influence of the relationship between the giver and the receiver on the gift form, proposing the following hypotheses:
H1:
The relationship between the giver and receiver affects the acceptability of gift forms.
H1a:
Money is more appropriate as a gift among individuals who are more familiar with each other.
H1b:
Money is more appropriate as a gift among individuals who are less familiar with each other.
H1c:
In-kind gifts are more common between individuals in a paternalistic relationship than in other relationships.
H1d:
Vouchers are most appropriate in an intermediate type of relationships in terms of closeness.
All the models of gift-giving are consistent with the observation that the relationship between the giver and receiver affects the gift form. The signalling model (H1a), search-cost model with imperfectly informed altruistic givers (H1b) and the model of paternalistic giving (H1c) all offer specific predictions about the ways this relationship affects acceptability of money versus in-kind gifts. The model of gifts as conventions does not offer specific predictions about the influence of the relationship between the giver and receiver on the gift form. In fact, it is consistent with any observation regarding this influence. The signalling model and search-cost model predict that vouchers are most appropriate in intermediate relationships in terms of closeness (H1d).
H2:
Gift-giving occasions do not affect the acceptability of gift forms.
The signalling model, the search-cost model with imperfectly informed altruistic givers and the model with paternalistic givers imply that only the relationship between the giver and receiver affects the gift form. By contrast, the model of gifts as conventions is consistent with the effect of a gift-giving occasion on the gift form, although it does not positively predict this effect.
H3:
The value of money gifts is not systematically higher than that of in-kind gifts.
The signalling model, the search-cost model with imperfectly informed altruistic givers and the model with paternalistic givers do not yield specific predictions about the relative value of money and in-kind gifts. Yet empirical literature finds that to compensate for unacceptability of money, more of it has to be given compared with the value of an in-kind gift (Waldfogel, Reference Waldfogel2002; Webley and Wilson, Reference Webley and Wilson1989). Since we expect money to be generally appropriate as a gift among the Chinese, we hypothesise that givers of money need not compensate receivers with a higher value of the gift. This finding is consistent with the model of gift forms as conventions.
4. Data
Data were collected in two consecutive surveys between October and December 2018 from the largest online panel service in China – www.wjx.cn. Online Panels emerged in recent years as a new data collection method, sourcing data for research in a range of fields in social sciences ranging from marketing studies to organisational behaviour (Smith et al., Reference Smith, Sabat, Martinez, Weaver and Xu2015; Woo et al., Reference Woo, Keith and Thornton2015). The first sample (n = 796) regroups respondents from provinces across the country and focuses on the socio-demographic group that has children and parents as part of the current household. This focus is to ensure that we are able to capture information on giving behaviour from professionally active individuals. In this context, the average age of respondents in the sample is 31 years, average monthly salary for the entire household is RMB15,072, 60.55% of respondents live with their partner and children and 79.27% are from urban areas. Further, 60.53% of the respondents are female. The second sample (n = 795) consists of different respondents but was collected through the same means and under the same criteria. Hence, the characteristics of the respondents in the second sample are quite similar, with 29.4 years the average age, RMB14,905 the average income, 74.21% of individuals from urban areas, 53.58% living with partner and children and 62.39% female respondents. While our sample is not representative of the Chinese population as a whole, it is similar in terms of age structure to samples that have been used in prior literature on gift-giving. Our sample is indeed mainly composed of respondents from younger urban households with income levels that are slightly higher than the national average.
In the first survey, we focused on gift-giving norms in general. We asked the respondents to specify the form of gift they consider on a particular occasion to an individual in a particular relationship. The forms of gift: in-kind gift, meal, voucher, red packets, electronic red packets and cash (other than in a form of physical/electronic red packets). Since we are primarily interested in in-kind gifts and money in general, we included meal in the category of in-kind gift and combined red packets, electronic red packets and cash into a single category labelled money. We kept vouchers as a separate category. The occasions considered: the Chinese New Year, birthdays, Christmas, weddings, visits, the birth of a child, graduation, Mother's Day, Father's Day and Valentine's Day. The types of relationship included were spouse/partner, parents (-in-law), grandparents (-in-law), own children, relatives, friends, friends’ children, co-workers, acquaintances and acquaintances’ children.
In the second survey, we focused on two gift-giving occasions: the Chinese New Year and birthdays. The Chinese New Year is the most important festival for the Chinese, and its importance is comparable to Christmas in Western societies. Birthdays were chosen because, just as the Chinese New Year, they are celebrated by all households in the country. Moreover, many of the previous researchers working on gift-giving in the Western societies also focused on birthdays. This enables us to compare our results with those of these earlier studies. In this second round, we asked the respondents not only what form of gift they consider for individuals in a given relationship but also the appropriate value of the gift. We selected two forms of gifts that dominate the gift-giving behaviour in China: in-kind gifts and red packets (including electronic red packets).
5. Results
General norms of gift-giving
We first examine the acceptability of money and in-kind gifts in general. Respondents were asked to answer the following question: If you give a gift on an occasion X to an individual Y, what do you give? The respondents were allowed to choose multiple answers. We asked about hypothetical choices, because individuals may have imperfect memory about actual gifts given and received. Moreover, hypothetical choices better reflect what kind of gift individuals consider suitable. We focused on the givers’ perspective as it is givers who chose gifts, presumably based on what they consider appropriate and what they expect receivers to consider appropriate. By contrast, receivers may or may not build expectations about what kind of gift they may receive. The results are shown in Table 1.
Table 1. General acceptability of different forms of gifts
a Gifts represent all in-kind gifts, including meal invitations.
b HB represents physical red packets.
c eHB represents electronic red packets.
d The first row shows the percentage of respondents who never choose this form of gift (i.e. regardless of the occasion and the receiver). The second row shows the percentage of respondents who at least sometimes consider this form of gift. The third row shows the percentage of respondents who always choose a particular form of gift (i.e. for all occasions and all receivers).
e The figures in the first row represent the percentage of times that this form of gift is selected by respondents relative to the total number of cases when a gift of some form is chosen. The figures in the second row show the percentage of times that a particular gift form is selected as a single option.
We find that money is generally appropriate as a gift in China. Specifically, 98% of respondents selected money in some form (i.e. red packets, electronic red packets or cash) as a gift at least for some occasions and in some relationships. Red packets (HB) dominate other forms of money; they are considered by 95% of respondents, compared with 65% respondents who consider electronic red packets (eHB) and 44% respondents who consider cash. Notably, the acceptability of money, measured in terms of proportions of respondents, is the same as acceptability of in-kind gifts. However, if we focus on the acceptability of money as a proportion of cases, it is lower than the acceptability of in-kind gifts. Money is considered in 67% of cases, while in-kind gifts are in 85% of cases. This result can be interpreted as follows: Nearly everyone considers money as a form of gift sometimes, but only for specific occasions and/or in specific types of relationships. Therefore, money is less ‘universal’ than in-kind gifts, which are considered for more occasions and/or types of relationships.Footnote 7 In section 5, we show that money is considered predominantly for the Chinese New Year and weddings, and is used relatively seldom for Valentine's Day or Christmas. Money is also considered as a gift to children more frequently than to grown-ups.
Table 1 also shows that money is generally more appropriate than vouchers. In fact, a voucher is the least appropriate form of gift among all the gift forms considered in our survey. This finding also holds if we consider various forms of money – red packets, electronic red packets and cash – separately. Specifically, 77% of the respondents would never select vouchers as a gift. Moreover, among the respondents, none considers vouchers as always appropriate gifts. Vouchers have also been selected the least times as a possible gift: only in 4% of all cases.
We now focus on gift-giving norms-based relationships between the giver and the receiver.Footnote 8 Table 2 summarises our findings. Overall, money is the least appropriate as a gift to acquaintances (47%), followed by co-workers (58%). This is in line with hypothesis H1a and it contradicts hypothesis H1b. That is, in-kind gifting may serve as a signal of knowledge of receivers’ preferences. In-kind gifts are the most appropriate gift form for each type of relationship except for gifts to friends’ children. For friends’ children, the most appropriate gift form is money (83% of respondents), specifically, red packets (HB, considered by 71% of respondents). Besides friends’ children, money is appropriate as a gift to one's own children (84%) and to one's spouse/partner (82%). However, there is a difference: To own and friends’ children, red packets are unambiguously the most appropriate form of money gift, whereas to a spouse/partner, red packets and electronic red packets are almost equally appropriate. The finding that red packets are appropriate as a gift to own and friends’ children contradicts hypothesis H1c, according to which in-kind gifts are more common as gifts to children. By contrast, this finding is in line with the fact that red packets traditionally emerged as a gift to children for the Chinese New Year, which illustrates path-dependency in gift-giving behaviour. However, gifting towards acquaintances’ children seems to break the pattern: ‘only’ 55% of respondents consider red packets as a gift to acquaintances’ children, compared with 67% who consider in-kind gifts. The reason for this result is that red packets are unpopular as a gift to acquaintances’ children for visits (22%). Just like for other children, red packets are popular as a Chinese New Year gift to acquaintances’ children (74%). Finally, in contrast to H1d, we find that vouchers are most appropriate as a gift to acquaintances (13%), followed by acquaintances’ children (10%) and co-workers (9%). In sum, we find weak support for H1a and we reject H1b–H1d.
Table 2. Gift-giving norms in China with respect to the relationship
a Those respondents who give nothing to an individual in a particular relationship are left out of the sample for this particular relationship.
b Gifts represent all in-kind gifts, including meal invitations.
c HB represents physical red packets.
d eHB represents electronic red packets.
e Excludes visits, Mother's Day and Father's Day.
f Excludes visits, the birth of a child, graduation and Valentine's Day.
g Excludes visits, the birth of a child, graduation, Mother's Day, Father's Day and Valentine's Day.
h Excludes visits, Mother's Day, Father's Day and Valentine's Day.
i Excludes Mother's Day, Father's Day and Valentine's Day.
Overall, the acceptability of various gift forms seems to vary across relationships, which supports hypothesis H1, albeit in a way that is not anticipated by existing theories. However, the results in Table 2 do not isolate the effect of relationships on the gift form, because giving to an individual in a particular relationship may occur on different occasions. To eliminate the possible effect of occasion on the gift form, we considered data only for occasions that are common to all types of relationships, namely, Chinese New Year, birthdays, weddings and Christmas. We then examined whether the proportions of each gift form are equal across relationships. The results are shown in Table 3. For each gift form, we reject on the 1% significance level the hypothesis that proportions are the same across all types of relationships, which is in line with H1. We conclude that although general patterns hold across various relationship types (i.e. in-kind gifts and red packets are the most popular gift forms, whereas vouchers are the least popular), relationship matters for the acceptability of a gift form. The effect of the relationship between the giver and the receiver on the gift form is difficult to explain through differences in preferences and/or constraints across relationships. The most plausible explanation seems to be that the acceptability of a particular gift form is determined primarily by social convention.
Table 3. Gift-giving norms across relationships
a HB represents physical red packets.
b eHB represents electronic red packets.
c CNY represents Chinese New Year.
*** indicates 1% significance.
Next, we examined gift-giving norms across occasions. Table 4 summarises our findings. Overall, the results suggest that the acceptability of gift forms varies across occasions, which contradicts H2. For example, weddings and the Chinese New Year are dominated by money gifts, while Christmas and birthdays are dominated by in-kind gifts. That said, we also observe that in-kind gifts are almost always appropriate, no matter the occasion. They are the least appropriate for weddings (‘only’ 68% of respondents consider an in-kind gift as a wedding gift). For the Chinese New Year, in-kind gifts are commonly appropriate (they are considered by 86% of respondents) but less so than red packets (considered by 93% of respondents). By contrast, money is least appropriate as a gift for Christmas (considered by 42% of respondents). Moreover, for five of the ten occasions, the acceptability of money is below 50%.
Table 4. Gift-giving norms in China with respect to the occasion
a Those respondents who give nothing on a particular occasion are left out of the sample for that occasion.
b Gifts represent all in-kind gifts, including meal invitations.
c HB represents physical red packets.
d eHB represents electronic red packets.
e CNY represents Chinese New Year.
f Excludes spouse/partner, parents, grandparents and own children.
g Excludes parents and grandparents.
h Includes only parents.
i Includes only spouse/partner.
The results presented in Table 4 do not isolate the effect of gift-giving occasions, because each occasion may involve givers and receivers in different types of relationships. Therefore, we considered only occasions that include all types of relationships, namely, Chinese New Year, birthdays, weddings and Christmas. Then we examined whether proportions of each gift form are equal across occasions. If yes, then the gift-giving occasion does not affect the gift form. As shown in Table 5, for each gift form, we reject on the 1% significance level the hypothesis that proportions are the same across the four occasions. Therefore, we conclude that acceptability of various gift forms varies across occasions. This result suggests that acceptability of money is occasion-specific. These findings contradict hypothesis H2, which further supports the view that the use of money or in-kind gift is primarily governed by convention.
Table 5. Gift-giving norms across occasions
a HB represents physical red packets.
b eHB represents electronic red packets.
c CNY represents Chinese New Year.
*** indicates 1% significance.
The value of gifts
In our second survey, we asked respondents to indicate not only the form of a gift they choose for a particular occasion but also the appropriate value of a gift to an individual in a specific relationship. We focused on two gift-giving occasions (the Chinese New Year and birthdays) and two gift forms, in-kind gifts and money (i.e. red packets, both physical and electronic).Footnote 9 Respondents chose the appropriate interval value starting from ‘RMB1–100’ to ‘RMB3,000 or more’. They made their choices separately for in-kind gifts and money and for each type and relationship. Then, we calculated the mean values of in-kind gifts and money by collapsing all intervals into single values. For example, the interval ‘RMB1–100’ was considered in the calculation of the mean value as RMB50, ‘RMB101–200’ as RMB150 and so on. For the interval ‘RMB3,000 or more’ we considered the value RMB3,500. Our results are shown in Table 6.
Table 6. The average value of gifts for the Chinese New Year and birthday
The t-test is used to test the differences in means; ***, ** and * indicate 1%, 5% and 10% significance respectively.
In line with H3, we find that the value of money gifts is not systematically higher than the value of in-kind gifts. For the Chinese New Year, we find that the mean value of money gifts is higher than the mean value of in-kind gifts for gifts to one's parents and grandparents and one's own children. Nevertheless, for gifts to one's spouse/partner and friends we find the opposite result. The difference between the mean value of in-kind gifts and the mean value of money gifts is statistically insignificant for more distant relationships.
As regards birthdays, the difference between the mean value of in-kind gifts and the mean value of money gifts is statistically insignificant for almost all types of relationships. The exception is the spousal relationship, for which the mean value of in-kind gifts is again larger than the mean value of money gifts. Not surprisingly, for both occasions and both gift forms, we find that closer relatives receive gifts of higher value. At the same time, the most valuable gifts are to one's parents and grandparents. Overall, while it is difficult to rationalise the differences in values for different types of relationships, our results unequivocally support hypothesis H3.
6. Discussion
We find that money is appropriate as a gift among the Chinese, although this acceptability depends on the relationship between the giver and the receiver as well as on the giving occasion. It does not seem possible to explain our findings with the specific preferences of givers emphasised by the paternalism model or information and technological constraints emphasised by the signalling and search models. Our findings are consistent with the view that in-kind/money gift-giving is governed primarily by social convention. This conclusion is further supported by our finding that in China, as opposed to Western countries, givers of money gifts do not compensate receivers with a higher value of the gift. We now discuss our findings in more detail vis-à-vis the literature.
The acceptability of money
In general, our findings suggest that money is more appropriate as a gift in Chinese societies than in Western countries. Using survey data from a US town, Caplow (Reference Caplow1982) finds that only 9% of all Christmas gifts were in the form of money. According to Waldfogel's (Reference Waldfogel1993) data collected from students in a US university, 11.3% of Christmas gifts were in the form of money or vouchers. Webley et al. (Reference Webley, Lea and Portalska1983) show that money is the least appropriate form of a birthday gift to one's mother among the British. Webley and Wilson's (Reference Webley and Wilson1989) survey among British students reveals that money is the least appropriate form of Christmas gift, regardless of the relationship between the giver and the receiver. Finally, Mortelmans and Sinardet (Reference Mortelmans and Sinardet2004) find that among the Flemish population the acceptability of money as a gift is limited, although this acceptability depends on the relationship between the giver and the receiver and giving occasion. Although these figures and results are not directly comparable with ours owing to the mismatch between questions in various studies, the difference in the acceptability of money between China and Western societies is sizeable and unlikely to be owing to the phrasing of the questions or representativeness of the samples. Our finding that money is appropriate as a gift in China is broadly in line with Waldfogel's (Reference Waldfogel2002) finding that Asians have the greatest tendency to receive money as a gift.
Empirical literature studying the acceptability of money in Western societies finds that if money is appropriate as a gift, then it is for downward close relationships, that is, from parents and grandparents to children and grandchildren. In contrast, a money gift to one's parents and grandparents seems inappropriate (Caplow, Reference Caplow1982; Mortelmans and Sinardet, Reference Mortelmans and Sinardet2004). Our results show that for the Chinese, these patterns hold only partly. Although red packets are mostly considered as gifts to children, they are not uncommon as gifts to one's parents and grandparents. Specifically, 62% of respondents consider red packets as a gift to their parents and 54% of respondents consider red packets as a gift to their grandparents. Moreover, more respondents consider cash as a gift to their parents and grandparents than to their children (see Table 2). Therefore, we conclude that money as a gift for upward relationships is also a common practice in China and is likely to be related to the cultural practice of filial piety (Ikels, Reference Ikels2004).
Several empirical studies find that acceptability of money depends on the type of relationship between the giver and the receiver (Caplow, Reference Caplow1982; Webley and Wilson, Reference Webley and Wilson1989), while others show that the acceptability of money is independent of the type of the relationship (Robben and Verhallen, Reference Robben and Verhallen1994). In our sample, the giver's relationship with the receiver matters, to some degree. Although the rankings of various gift forms are similar across relationships (e.g. money is more appropriate than vouchers for all types of relationship), the proportions at which various gift forms are considered differ across relationships.
One apparently similar aspect between the Chinese and Western money gifting is that in Western societies, money is usually wrapped in envelopes or self-made packages (Mortelmans and Sinardet, Reference Mortelmans and Sinardet2004). However, this packaging is not ritualised and self-made packages arguably signal care of the giver for the receiver. In this context, packaging helps to distinguish affectionate social relations from other social relations, such as market exchanges between strangers (Zelizer, Reference Zelizer1994).
By contrast, Chinese red packets are usually purchased in their traditional shape and colour and no extra effort is expected from the givers (Siu, Reference Siu2001). With the current wave of rapid digitalisation, electronic red packets have started to emerge in China, which removes any form of physical packaging, and only leaves the exchange rituals. At the time of writing both physical and digital red packets are commonly given. On the other hand, similarities, can be found in both sets of practices, as brand new bills for money gifts are commonly used in both China and the West (Siu, Reference Siu2001; Zelizer, Reference Zelizer1994), presumably to distinguish money gift from a market exchange.
Finally, we find that traditional red packets are relatively unpopular for ‘imported’ holidays, that is, Christmas and Valentine's Day. Our conjecture is that it is because the norms of gift-giving for these holidays are ‘imported’ too, and Western societies do not have the tradition of red packets. By contrast, electronic red packets are relatively popular for these two occasions. This finding may be explained by the fact that Christmas and Valentine's Day are celebrated by the young, among whom electronic red packets are more popular. However, more research is needed to test these preliminary conjectures.
The acceptability of in-kind gifts
Although in-kind gifts are generally appropriate for all types of relationships and occasions, we find that they are least appropriate for weddings. A plausible explanation, in line with our results as well as with common perception among the Chinese, resides in the convention of gifting red packets for weddings. Other forms of gifts would then deviate from socially expected gifting. Another possible (and perhaps complementary) reason for the relative unpopularity of in-kind gifts for weddings is that Chinese weddings involve numerous guests (sometimes several hundred), which would give rise to a coordination problem for the givers. When the number of wedding guests is large, they are more likely to choose to give similar in-kind gifts to the newlyweds. Moreover, when numerous in-kind gifts are received for an occasion, it may be difficult to store these, which imposes costs on the receivers. Our hypothesis is in line with Mortelmans and Sinardet's (Reference Mortelmans and Sinardet2004) finding that the larger the number of people gathered at a gift-giving occasion, the more appropriate money is as a gift.
An alternative hypothesis, implied by the search-cost model (Kaplan and Ruffle, Reference Kaplan and Ruffle2009), suggests that gifts vis-à-vis money are less popular for weddings because, first, parents are less familiar with their grown-up children's preferences as they grow older, and second, parents do not possess any search-cost advantage over their grown-up children. Yet another explanation is implied by the signalling model (Camerer, Reference Camerer1988). According to this explanation, weddings represent a weakening of the cooperative relationship between parents and children and, therefore, signalling is not necessary anymore.
Neither the search-cost model nor the signalling model is supported by our data. These two models may explain why parents give money to children at weddings, but they do not explain why children receive money from other wedding guests. In fact, in our sample, in-kind wedding gifts are more appropriate from parents than from others and red packets are more appropriate from others than from parents. Specifically, 48% of respondents consider an in-kind gift appropriate as a wedding gift to their children.Footnote 10 This figure is lower when the receivers are relatives (31%), friends (33%) or co-workers (25%). By comparison, 61% of respondents consider red packets appropriate as a wedding gift to their children. This figure is higher when the recipients are relatives (74%), friends (76%) or co-workers (81%). This pattern cannot be wholly explained by the fact that relatives, friends and co-workers do not know the preferences of the recipients well, because relatives, friends and co-workers give each other in-kind gifts on other occasions, such as birthdays.
The acceptability of vouchers
We find that vouchers are the least appropriate gift forms, even if we consider various forms of money separately. These results contrast with those of the literature using data on European or American respondents. In general, it is observed that the acceptability of vouchers is between in-kind gifts and money (Mortelmans and Sinardet, Reference Mortelmans and Sinardet2004; Webley and Wilson, Reference Webley and Wilson1989). A possible explanation of our finding is that in Western societies, vouchers are a more appropriate but less efficient substitute for money. Since, in China, money is generally appropriate as a gift, there is no need to substitute it with vouchers.
The value of gifts
We find that closer relatives receive in-kind gifts and money gifts of a higher value. This finding is the same as the findings about gift-giving in the Western societies (Caplow, Reference Caplow1982; Mortelmans and Sinardet, Reference Mortelmans and Sinardet2004). Yet there is an important difference. Mortelmans and Sinardet (Reference Mortelmans and Sinardet2004) find that in Belgium, the most expensive gifts are to one's children, grandchildren and spouses/partners. This is in line with Caplow's (Reference Caplow1982) earlier result for US households. In contrast, for the Chinese, the pattern is reversed: the most expensive gifts are to one's parents and grandparents for both the Chinese New Year and birthdays.
Webley and Wilson (Reference Webley and Wilson1989) find that closeness of relationships affects the appropriate value of in-kind gifts more than the appropriate value of money gifts. We do not find this pattern in our data. If anything, we find the opposite (see Table 6): Regarding the Chinese New Year, the appropriate value of the money gift (SD = 198) varies more across relationships than the appropriate value of the in-kind gift (SD = 133). To a lesser extent, this is also true for birthday gifts (money gifts, SD = 140; in-kind gifts, SD = 135). These results once again highlight our general point that with respect to gift-giving norms, the findings about Western gift-giving may markedly differ from Chinese gift-giving.
Chinese versus Western gifting
We have identified large differences between Western and Chinese norms regarding the form and the value of the gift. Although we have not directly analysed factors accounting for these cross-cultural differences, we believe that these are significant and can hardly be fully explained by variations in technological and information constraints or levels of paternalism across cultures. In particular, a greater acceptability of money as a gift in China than in the West is unlikely to be due to cultural differences in demand for signalling or search costs. In fact, signalling generally plays an important role in Chinese societies and, because of import restrictions, the within-country asymmetries in search costs maybe be more important among the Chinese than among individuals in Western countries. Therefore, if anything, one would expect more, rather than less, in-kind gift-giving in China than in Western countries. Our hypothesis that the gift form is primarily governed by convention is further supported by the observation that norms regarding gift forms in countries neighbouring China are relatively similar. This is in line with the model of conventions, which predicts local conformity and global diversity of behaviour. In sum, while we do not deny the importance of factors such as signalling, search cost, asymmetric information or paternalism for gift forms in some contexts, our results suggest that these factors are secondary to the social influences. In particular, in our analysis, these factors turned out to be relatively unimportant even within a given culture.
On a general level, our results are in line with the cross-cultural studies of behaviour in strategic games (Henrich et al., Reference Henrich, Boyd, Bowles, Camerer, Fehr and Gintis2004). These studies show that behaviour in games, such as the ultimatum, dictator and public-good games, differs across cultures. Consequently, this behaviour is unlikely to be explained with hard-wired panhuman preference structures. Further, since all subjects played similar games, these differences are also not attributable to differences in constraints. The results of these experiments are best explained by differences in social norms and conventions across cultures. We argue that a similar explanation applies to cross-cultural differences in gift-giving. In line with path-dependency models (David, Reference David1994; North, Reference North1990), our view is that in-kind objects were by chance (or simply because money was yet not established when a particular convention evolved) selected as gifts in the early history of Western countries and the tradition has mostly survived until the present, contributing to ‘cash stigma’. By contrast, in early China, small coins were selected as gifts on certain occasions. Over time, the amounts of money given have increased but money as such remained a prevalent form of gift. Therefore, the Chinese were ‘lucky’ to end up with a more efficient norm of gift-giving than Western societies, at least on certain occasions.
7. Limitations and further research
Our results call for further investigations on several fronts. Regarding the theoretical aspects of gift-giving, we did not attempt to construct a theoretical model that would account for findings of this and other studies. Although we believe that various models of conventions can account for many of our results, some of the cross-cultural similarities in gift-giving practices call for a multi-causal model.
Regarding the empirical analysis, the scope of the study in terms of sampling coverage could be extended to cover additional socio-demographic groups, which could reveal differences among various socio-demographic dimensions. Sampling in a country as vast and diverse as China is bound to be a critical element of any research conducted among its citizens.
In our approach, we have focused only on normative beliefs of givers about gift-giving and we did not distinguish between personal normative beliefs and normative expectations (Bicchieri, Reference Bicchieri2017), that is, we cannot ascertain whether the givers consider a particular gift ‘right’ or whether they use a form of a gift that the receiver considers ‘right’. On a related note, we do not know the form of a gift considered appropriate by receivers. Although most individuals switch between the roles of giver and receiver, an interesting follow-up question resides in whether agents transfer their experience from the role of a receiver to that of a giver (Flynn and Adams, Reference Flynn and Adams2009).
Acknowledgements
This project benefited from the funding from the National Science Foundation of China (ref# 71750410693).
