DIRECT AND INDIRECT TRUST AND RECIPROCITY^:直接和间接的信任和互惠.doc_第1页
DIRECT AND INDIRECT TRUST AND RECIPROCITY^:直接和间接的信任和互惠.doc_第2页
DIRECT AND INDIRECT TRUST AND RECIPROCITY^:直接和间接的信任和互惠.doc_第3页
DIRECT AND INDIRECT TRUST AND RECIPROCITY^:直接和间接的信任和互惠.doc_第4页
DIRECT AND INDIRECT TRUST AND RECIPROCITY^:直接和间接的信任和互惠.doc_第5页
已阅读5页,还剩39页未读 继续免费阅读

下载本文档

版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领

文档简介

DIRECT AND INDIRECT TRUST AND RECIPROCITY Nancy R. Buchan * Rachel T.A. Croson * Robyn M. Dawes * Abstract We examine the extent to which people trust and reciprocate in situations of direct and indirect reciprocity. Our first study shows that as the potential for direct reciprocity between the trustor and reciprocator decreases, levels of trust and reciprocation decrease also; yet, on average, these levels remain significantly above equilibrium predictions. Our second study extends these results to three Asian countries. We find not only that levels of trust and reciprocation vary across countries, but that the drop-offs in trust across direct and indirect reciprocity conditions vary across countries as well. These differences in drop-offs suggest a significant country influence on the boundaries of trust. We thank the participants at the Economic Science Association, the Society for Judgment and Decision Making, and the European Association of Experimental Social Psychology meetings along with seminar participants at Yale, Texas A LaPorta, et al. 1997. Psychologists maintain that trust plays a prominent role in the emergence of cooperation in problems of collective action Dawes 1980. Business researchers have claimed that being a “trusted cooperator” within a global network is requisite to achieving competitive advantage Morgan and Hunt 1994. Finally Alan Greenspan underscored the critical nature of trust to our economic way of life, “Trust is at the root of any economic system based on mutually beneficial exchange if a significant number of people violated the trust upon which our interactions are based, our economy would be swamped into immobility” 1999. This previous research demonstrates the aggregate benefits of trust to groups of people or to societies as a whole. In contrast, most of the experimental research on trust in economics has focused on trust and reciprocity at the individual level e.g. Fehr et al. 1993; Berg, Dickhaut and McCabe 1995, Glaeser et al. 1999. This individual level research demonstrates that trusting behavior and reciprocation are robust across a number of economic contexts. However, these studies all have examined trust in situations of direct reciprocity where the actions of the reciprocator (in whom one placed trust) directly impact the fortunes of the trusting individual. In contrast, in societal-level trust, reciprocity is likely to be indirect. What is needed is a link between our understanding of trusting behavior on the part of individual and our knowledge of 2 trust and its influence on society. Our paper begins constructing this link by experimentally examining trust in the presence of indirect reciprocity - where the actions of the reciprocator (in whom one placed trust) will impact the fortunes of someone other than the trusting individual. Some empirical research provides a clue to this link by demonstrating the crucial role of trust in sustaining successful export clusters e.g. Krugman 1995; Schmitz 1999. The key to achieving such trust among cluster members, suggests Sabel 1993, is a shift in thinking of trust as rational self-interest to trusting in a shared sense of community with a common fate. In this paper we examine the extent to which economic actors are willing to trust for the benefit of the group or community. We begin by examining the extent to which individuals trust in situations of direct reciprocity, and compare that to levels of trust and reciprocation displayed in two different situations of indirect reciprocity. In one situation a member of the trustors group will benefit from potential reciprocation, in the other, a random member of society will benefit from the trustors actions. To examine indirect trust we extend the trust game devised by Berg, Dickhaut, and McCabe 1995 by manipulating the target of potential reciprocation. In the original trust game two players, the sender and the responder, are each given an endowment. The sender is told he can send some, all, or none of his endowment to his anonymous partner, the responder. Any money sent is tripled. The responder then chooses how much of her total wealth (her endowment plus the tripled money) to return to the sender. Any money the responder does not return is hers to keep; thus the responder plays a dictator game with the pool of money generated by the senders actions. The unique subgame perfect Nash equilibrium for this game (for self- interested actors) is for the responder to return no money, and thus for the sender to send none. In our experiment we label this the direct condition because of the direct nature of the potential 3 reciprocation (the responder returns money directly to the sender from whom she received it). This basic trust game is similar to the economic setting of one-shot investing with imperfect contracts. The sender (investor) creates value with his investment, but cannot perfectly contract with the responder (agent) to ensure sharing of that value created. We extend the direct trust experiment by adding two conditions to examine different types of indirect reciprocity. In the group condition, the target of potential reciprocation is someone in the senders (experimental) group. In the society condition, the target of potential reciprocation is a randomly chosen person from (the experimental) society. In both indirect conditions the equilibrium remains for the responder to return no money and for the sender to send none. Although our experiment protects anonymity, the group condition is similar to the type of indirect trust found in exporting clusters; another member of a small group benefits from one players trusting behavior. The society condition is similar to the situation of contributing to a public good; another member of a large society benefits from one players trusting behavior. Our first study compares these three treatments using subjects from the United States. Our results demonstrate that as the potential for direct reciprocity between the sender and reciprocator decreases, levels of trust and reciprocation decrease also; yet, on average, these levels remain significantly above equilibrium predictions. In our second study, we extend this experimental design to three Asian countries: China, Japan and Korea. We examine the extent to which these individuals trust in situations of direct and indirect reciprocity. We then aggregate these results to the country level to discern whether, as some have suggested, there are differences across countries in levels of trust Arrow 1972; Knack and Keefer 1997; Fukuyama 1995. Our results reveal differences across countries in 4 levels of trust and reciprocity although in ways different than predicted by theories from economics or cross-cultural research. Perhaps the most interesting finding from this study is an interaction between condition and country. That is, across the four countries, we find different drop-offs in trust extended across experimental conditions; in the US and China, the drop-off in trust comes between the direct and group conditions (direct group = society). In Japan and Korea, the drop-off occurs between the group and society conditions (direct = group society). For a preview of these results, the reader is referred to Figure VI. These results imply not just that the extent of trust varies across countries, but that the reliance of trust on direct reciprocation varies across countries as well. Our paper continues as follows. In Section II we provide a review of the literature on indirect trust and reciprocation and our hypotheses. In Section III we describe our methodology, which is followed by a description of our results in Section IV. Sections V, VI and VII provide hypotheses, methodology and results for the second study. In Section VIII we discuss the implications of our results for economic behavior and conclude. II. Previous Literature and Hypotheses The trust game employed here has been widely used to examine trust and reciprocation in economic settings.1 Deviations from the self-interested equilibrium of the game are attributed to trust (on the part of the sender) or reciprocity (on the part of the responder). The first paper using the trust game Berg, Dickhaut, and McCabe 1995 reported significant levels of trust and reciprocation; 30 of 32 senders sent money in the game (exhibited trusting behavior), and 24 of those 30 decisions resulted in money returned. These results are also found in other settings. For example, the experiments of Fehr, Kirchsteiger, and Riedl 1993 and Fehr, Gachter, and 1 Other, related games have been developed and tested experimentally as well, e.g. Kreps 1990, Rosenthal 1982. 5 Kirchsteiger 1997 demonstrate that generous (trusting) behavior on the part of firms induces similar reciprocal responses on the part of workers in the context of labor markets with incomplete contracts. McKelvey and Palfrey 1992 observe similar trust and reciprocal behavior in experiments based on the centipede game where each player can increase the pie by trusting the other not to take the money and exit the game. In the peasant-dictator game of Van Huyck, Battalio and Walters 1995, when the peasant moves before the dictator (as is the sequence of moves in the trust game), peasants trust dictators more than predicted by economic theory, and display trust most at low levels of endowment. Finally, Van der Heijden et al. 1998 demonstrate the power of expectations of reciprocation in the poverty game by showing that average gifts by senders were 40% higher when the responder would learn the size of the gift, than when they would not.2 All of the above studies have experimentally demonstrated the existence of trust and reciprocity under conditions of direct reciprocity. Based on this previous research we hypothesize: H1: We will observe positive amounts sent and reciprocated in the direct reciprocity condition. However, we anticipate that trust and reciprocity will decrease in situations of indirect reciprocity. Williamson 1993 suggests that trust is based, at least in part, on expectations of reciprocation. It follows then, that if the chance of receiving direct reciprocation is absent, the motivation to trust is decreased. Additionally, if reciprocation is based in part on a desire to reward or punish a partner as suggested by Fehr et al. 1993; Fehr et al. 1997; Abbink, Irlenbush, and Renner 2000, then the propensity to reciprocate should be diminished in a context where the 2 Van der Heijden et al. 1998 also study a form of indirect reciprocation in their pension game. In this game, however, the self-interested motivation for trust is highly salient due to overlapping structure of the game; ie. I give to Player 1, but then Player 3, who gives to me, may reward or punish me for my actions to Player 1. This is in contrast to the group and society conditions in our experiment where the self- interested motivations for trusting and reciprocal behavior are clearly minimized. 6 ability to reward or punish a trusting or non-trusting partner is removed. Therefore we hypothesize: H2: Amounts sent and reciprocated will be less in the indirect reciprocity conditions than in the direct reciprocity condition. Finally, we suggest that trust and reciprocity are likely to be higher in the group condition than in the society condition. Glaeser et al. 1999 demonstrate that the degree of social connection between the sender and responder in trust games generally predicts the level of trust and reciprocation.3 In a related (dictator) setting, Hoffman, McCabe and Smith 1996 conclude that as social distance (isolation) increases, amounts given decrease. Similarly, experimental research studying ingroup/outgroup biases in social dilemmas Orbell, van de Kragt and Dawes 1988 and in a three person dictator game Frey and Bohnet 1997 demonstrate that behavior toward someone in ones own group will be more favorable than behavior to someone who is not a member. Based on this research, we believe that the degree of connectedness between the sender, the responder, and the target of reciprocation will have a positive influence on trust and reciprocation. Therefore we hypothesize: H3: . Amounts sent and reciprocated will be higher in the group condition than in the society condition. III. Experimental Design and Procedures In this section we describe the experimental design and procedures that we use to test the hypotheses above. A. Experimental Design One hundred and forty students from the University of Wisconsin-Madison participated in this study. Subjects were primarily sophomore or junior students in economics or business 3 Glaeser et al. define “social connection” as the “number of friends they have in common, being members of the same race or nationality and the duration of their acquaintanceship” p.4. 7 classes who were paid their actual monetary earnings from the experiment. We ran three sessions of the experiment, one for each of our conditions. Subjects signed up for a session without knowledge of any differentiation in experimental conditions. Once arriving at the experiment, subjects were randomly assigned to one of two rooms, the sender room or the responder room. An effort was made to assign friends who came to the experiment together to the same room. The reason was to minimize any perception on part of subjects that their partner in the other room (though identified anonymously) could be their friend. B. Experimental procedure The experiment was run using double blind procedures; thus, the experimenter did not know the identities of the subjects, and subjects identities were kept anonymous from one another. Subjects were instructed not to talk or communicate with any other subjects during the experiment. This “no communication” rule was enforced by monitors located in the experimental rooms. The basic procedure for the trust game is as follows. Subjects in each room received written instructions for the game, including a diagram of the game structure (Figures I, II, and III below), and a unique subject ID number. The instructions also included a quiz concerning the experimental procedure and the transactions involved, to make certain that all subjects understood the task. All experimental instructions are available from the authors upon request. Once subjects completed the quiz correctly, they were given a numbered envelope containing the experimental fee of $10. The number on the envelope represented the subjects own identification number. Senders removed their money and replaced in the envelope any money they wished to send to the responder. Monitors collected the envelopes and took them to the experimenter in a different room who recorded the amount sent, tripled it, and placed the tripled money into 8 different numbered envelopes for delivery by another monitor to the appropriate responder. Notice that this implements a double-blind procedure Hoffman et al. 1996. The monitor who saw the participant did not know the amount they sent, while the experimenter, who saw the amount sent, did not know the participant. The subjects, of course, did not know the identity of their partner. The responders then opened their envelopes, and decided how much of their own experimental fee plus any (tripled) money received, to return to the appropriate responder. They placed that amount back into their envelopes. Monitors collected the envelopes from the responders and gave them to the experimenter who recorded the amounts returned, placed the money back into the senders original envelopes, and forwarded the envelopes into the senders room for distribution by the monitors. At this point the experiment had ended. Subjects completed a post-experimental questionnaire, were paid their earnings privately and dismissed. C. Direct Condition The direct condition is depicted in Figure I. This is the most basic condition in the experiment, identical to that of Berg et al. 1995 and Glaeser et al. 1999, in which Sender A sends some, all, or none of his money to Responder B. On the way from Sender A to Responder B the experimenter triples the money. Responder B then has her own show up fee plus triple the money sent by A, and can return some, all, or none of her total wealth to A. Aside from explaining the experimental procedure, the wording on the instructions specific to this condition is “You will interact with a Responder for purposes of this experiment.” -Insert Figure I about here- D. Group Condition 9 The group condition is depicted in Figure II. In this condition, Sender A sends money (which is tripled by the experimenter) to Responder B. At the same time, Sender C sends money (which is tripled by the experimenter) to Responder D. Responder B then returns some proportion of her wealth to Sender C, while Responder D returns some proportion of his wealth to Sender A. Aside from explaining the experimental procedure, the wording on the instructions specific to this condition is “You will interact with three other people for purposes of this experiment.” As an additional reinforcement of the group manipulation, each group was given differently colored instructions. -Insert Figure II about here- E. Society Condition The society condition is shown in Figure III. In this condition, Sender A sends money (which is tripled by the experimenter) to Responder B. At the same time, Responder D receives money from a different, randomly chosen, sender in Sender As room. Responder D then returns some proportion of his wealth to Sender A, while Responder B returns some pro

温馨提示

  • 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
  • 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
  • 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
  • 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
  • 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
  • 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
  • 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。

评论

0/150

提交评论