点击蓝字 ● 关注我们
New Program
《公共行政》微信公众号开始向之前在《公共行政》上发表过文章的作者征集之前发表的文章简介，如果您有兴趣在本公众号上推广您的文章，您可以将您文章的简介发送至邮箱：larsonse@miamioh.edu。唯一的要求是需要作者本人提供文章简介。
我们期待您的投稿！
摘要
引言
INTRODUCTION
风险偏好是指决策者对风险的态度，对决策有着至关重要的影响。最近，公职人员的风险偏好特征引起了人们的密切关注。对于公职人员来说，冒险是一把双刃剑，它可能有利于治理创新，但也可能给公众带来灾难性的后果。心理距离理论认为，自我决策和他人决策涉及不同的心理距离，这可能导致风险偏好的差异。且根据社会价值理论，个人为自己和他人做决策的规则是不同的。在为他人做决策时，个人主要依靠感知的社会风险价值（大多数人都认可的社会规范），而为自身做决策时，社会价值以外的因素对其影响更大。而公职人员在为自己和为公众做决策时，是否会区别对待风险？由此受为自我与他人决策的心理距离差异和社会价值理论的启发，本研究将探讨公职人员在为自己和公众做决策时风险偏好的差异。
此外，前景理论提出了著名的收益和损失风险框架效应，该效应描述了个人风险偏好会因决策的收益或损失参考点而变化。在风险框架中，对于同一前景，个人在收益情况下更倾向于选择确定选项，而在损失情况下更倾向于选择风险选项。收益与损失的框架可能导致自我决策与公共决策的风险认知存在差异；特别是，公共利益的损失可能尤其面临来自公共部门和社会的问责。另外，除了决策框架外，决策金额的大小本身也会影响风险偏好。基于此，本文将进一步探讨公职人员个人决策与公众决策的风险偏好差异是否受收益与损失决策框架和决策金额大小的影响。
上述问题将在中国公职人员决策的基础上进行探讨。在中国，下级官员往往面临多重严格的问责压力。尤其是在过去的十年中，中国政府系统地强化了问责机制，以惩戒失职的官员。
Risk preference, which refers to a decision-maker's attitude toward risk, has a crucial impact on decision-making. Recently, the risk preference characteristics of public officials receives acute attention. For public officials, taking risks is a double-edged sword, as it may be good for governance innovation but may also bring disastrous consequences to the public. The psychological distance theory suggests the decision making for self and for others involves different psychological distances, which may lead to difference in risk preference. Furthermore, according to the social value theory, the rules for individuals to make decisions for themselves and for others are different. When making decisions for others, individuals primarily rely on the perceived social value of risk (a social norm recognized by the majority), whereas when making decisions for themselves, factors beyond social value influence them more. Do public officials really treat risk differently when making decisions for themselves and when making decisions for the public? Inspired by social value theory and psychological distance difference in self versus other decision-making, this study explores the difference between the risk preference of public officials when making decisions for themselves and the public.
Additionally, Prospect theory proposed the famous gain and loss risk framing effect which describes that individuals' risk preferences change depending on the gain or loss reference point of decision making. In risk framing with a certain option and a risky option for the same prospects, individuals are more likely to choose the certain option in a gain situation and more likely to choose the risky option in a loss situation. Additionally, the gain versus loss frame may lead to a difference in the risk perception of self versus public decision-making; the loss of public benefits may especially face strict authoritarian accountability and social accountability in the public sector. Besides the decision framing, the size of a bet itself also affects the risk preference. Thus, we further explored whether the risk preference difference between public officials' decision-making for self and the public is influenced by gain versus loss decision framing and bet size.
The above question is explored on the basis of decisions made by Chinese public officials. In China, lower-level officials often face strict and multiple accountability pressures. In the last decade, China's government has systematically intensified accountability mechanisms to discipline at-fault bureaucrats.
研究假设
RESEARCH HYPOTHESES
假设2：相比于收益框架，公务员在损失框架中更加偏好风险（假设H2a）。此外，在收益框架下，公务员在公共决策中的风险偏好低于自我决策中的风险偏好；在损失框架下，两者之间的风险偏好差异不显著（假设H2b）。
Hypothesis 2. Compared with in the gain frame, public officials have a higher risk preference in the loss frame (H2a). Additionally, in the gain frame, public officials have a lower risk preference in decision-making for the public than the self, whereas in the loss frame, their risk preference is not significantly different (H2b).
假设3：相比于较小决策金额，公务员在较大决策金额情境下更加偏好风险（H3a）。此外，相较于较小的决策金额，个人决策与公共决策之间的风险偏好差异在较大决策金额下更加显著（H3b）。
Hypothesis 3. Compared with a smaller bet size, public officials have a higher risk preference with a larger bet size (H3a). Additionally, compared to a smaller bet size, the difference in risk preference between self versus public decision-making is more significant in the larger bet size (H3b).
图1：研究假设模型
研究设计
RESEARCH DESIGN
我们为中国公职人员设计了一个2（自我决策与公共决策） x 2（收益框架与损失框架）x 2（大额投注与小额投注）的被试间调查实验。实验场景是关于在乡村振兴过程中开展生态项目的决策，这是中国一项众所周知的全国性举措。在为自己决策的情况下，公职人员通过操纵与他们的奖励直接相关的项目结果来做出有利于自己的奖励的决策；在为公众决策的情况下，公职人员为乡镇村民（不包括公职人员）的集体收入做出决策。在获利条件下，项目是盈利的，决策的参考点是获得一些收益；在亏损条件下，项目是亏损的，决策的参考点是损失一些收益。至于赌注大小，我们设置了两种条件：较大的赌注（约20000元人民币）和较小的赌注（约2000元人民币）。参与者被随机分配到8个条件中的一个。在每个条件下，他们必须在两个选项之间做出选择：一个确定选项和一个具有相同预期效用的风险选项。当参与者选择风险选项时，我们将风险偏好评分编码为1，选择确定选项时编码为0。我们计算了每个实验条件下的风险偏好平均得分；得分越高，风险偏好越高。
我们使用G*Power 3.1计算了在95%的置信度下,α=0.05时，小样本效应（0.15）的适当样本量。我们通过Credamo平台在线收集数据。我们收集了1250名参与者的信息。我们进一步将样本的人口统计数据与所有中国公职人员的基本数据进行了比较。方法上，我们将ANOVA（⽅差分析）⽤于分析主要效应和交互效应，回归模型则⽤于探讨决策对象（⾃我决策 vs. 公共决策）、决策框架（收益 vs. 损失）和决策⾦额（⼤ vs. ⼩）与风险偏好的关系。
We designed a 2 (decision-making for self versus decision-making for public) X 2 (gain frame versus loss frame) X 2 (large bet size versus small bet size) between-subject survey experiment for Chinese public officials. The experimental scenario is about the decision-making of taking ecological projects in the process of rural revitalization, which is a well-known national initiative in China. In the condition of decision-making for the self, public officials make decisions for their own reward by manipulating the outcome of the project that directly relates to their rewards; in the condition of decision-making for public, public officials make decisions for the collective income of township villagers (excluding public officials). In the gain condition, the project is profitable, and the reference point of decision-making is gaining some money; in the loss condition, the project is losing money, and the reference point of decision-making is losing some money. As for bet size, we set two conditions: a larger bet size (around 20,000 RMB) and a smaller bet size (around 2000 RMB). Participants were randomly assigned to one of the eight conditions. In each condition, they have to choose between the two options: a certain option and a risky option with equal expected utility. We encoded a risk preference score as 1 when the participants chose the risky option and 0 when they chose the certain option. We calculated the average score of risk preference for each experimental condition; the higher the score, the higher the risk preference.
We used G*Power 3.1 to calculate the appropriate sample for a small-size effect (0.15) at 95% power, given α=0.05. We collected data online through the Credamo (www.credamo.com) platform. We collected information on 1250 participants. We further compared the demographic data of the samples with the basic data of all Chinese public officials. We further used ANOVA (analysis of variance) to analyze main and interaction effects, and regression models to explore the relationships between decision-making objects (self-decision-making vs. public decision-making), decision-making frameworks (gain vs. loss), bet amounts (large vs. small), and risk preferences.
分析与结果
ANALYSIS AND RESULTS
Balance test, manipulation check, and correlation analysis
我们使用F检验来评估8个组之间的平衡性，结果表明各组之间的人口统计学变量没有显著差异，这证明了实验条件的随机分配。操作检查结果表明，我们在自我决策与公众决策、收益与损失框架以及小额与大额赌注方面进行的操作非常有效。相关分析显示，变量之间的相关性，自我决策与公众决策对象与风险偏好得分呈显著负相关，收益与损失框架以及小额与大额赌注与风险偏好得分呈显著正相关。也就是说，公众决策的风险偏好低于自我决策。在损失框架下决策的风险偏好高于收益框架，而大赌注的决策风险偏好高于小赌注。此外，在本次研究中，公共服务动机（PSM）、风险态度和社会责任与风险偏好得分呈正相关。因此，我们将这些变量作为控制变量纳入回归分析。
We used the F-test to evaluate the balance among the eight groups, and the results showed that there were no significant differences in demographic variables among the groups, which proved the random assignment of experiment conditions (see Table 2). The manipulation check results demonstrated the effectiveness of our manipulations for self versus public decision-making, gain versus loss frame, and small versus large bet size. Correlation analysis shows that the correlation between the variables, self versus public decision object correlated significantly negatively with the risk preference score, gain versus loss frame and small versus large bet size correlated significantly positively with the risk preference score. That is, decision-making for the public resulted in a lower risk preference than decision-making for oneself. The decision-making in the loss frame has a higher risk preference than in the gain frame, and decision-making in the larger bet size resulted in a higher risk preference than that in the smaller bet size. In addition, PSM, risk attitude, and social responsibility were positively associated with the risk preference score in this study. Thus, we included these variables as control variables in the regression analysis.
2）假设检验 Hypotheses test
我们进行了方差分析，以测试自我决策与公众决策、收益框架与损失框架以及决策金额的主要影响，以及它们对风险偏好的交互作用。结果表明，自我决策与公众决策、收益框架与损失框架以及决策金额大小的主效应均显著。在回归分析中，自我决策与公共决策、收益框架与损失框架以及决策金额大小对风险偏好的影响是显著的。具体而言，公职人员在为公众决策时比为自己决策时风险偏好更低；因此，H1得到支持。在损失框架中，公职人员比在收益框架中具有更高的风险偏好；因此，H2a得到支持。公职人员在较大赌注中比较小赌注具有更高的风险偏好，因此，H3a得到支持。
结果表明,自我决策与公共决策、收益框架与损失框架对风险偏好的交互作用显著。具体来说,在收益框架中，公职人员在为公众做决策时，风险偏好低于为自己做决策时, 而在损失框架中，自我决策与公众决策场景下的风险偏好差异并不显著。因此，H2b得到支持。
此外，自我决策与公共决策以及决策金额大小对风险偏好的交互作用显著。具体而言，在赌注金额较大的情况下，公职人员在决策时对公众的风险偏好低于对自我的风险偏好，而在较小的赌注金额中，他们对于自我决策和公众决策的风险偏好没有显著差异。H3b得到支持。
We performed the ANOVA to test the main effects of self versus public decision-making, the gain versus loss frame, and the bet size, as well as their interaction effects on risk preference. The results showed that the main effects of self versus public decision-making, gain versus loss frame, and bet size were significant. In the regression analysis, the effect of self versus public decision-making, gain versus loss frame and bet size on risk preference were significant. Specifically, public officials have a lower risk preference when decision-making for the public than the self; therefore, H1 finds support. In the loss frame, public officials have a higher risk preference than in the gain frame; therefore, H2a finds support. Public officials have a higher risk preference in the larger bet size than the smaller one, thus, H3a finds support.
The results showed that the interaction effect of self versus public decision-making and gain versus loss frame on risk preference was significant. Specifically, in the gain frame, public officials have a lower risk preference when making decisions for the public than themselves, whereas, in the loss frame, the difference between risk preference in self versus public decision-making scenarios is not significant. Therefore, H2b finds support.
Additionally, the interaction effect of self versus public decision-making and bet size on risk preference was significant. Specifically, in the larger bet size, public officials have a lower risk preference in decision-making for the public than the self, whereas in the smaller bet size, their risk preference between self versus public decision-making is not significantly different. H3b finds support.
结论
CONCLUSIONS
本研究关注的是个人决策与公共决策，研究发现了公职人员在为公众决策时比为自己决策时风险偏好更低。这一发现与社会价值理论相一致。在为公众做决策时，公职人员的亲社会倾向可能会被激活。他们的社会责任和社会规则也可能导致他们在为公众做决策时比为自己做决策时更倾向于低风险。此外，在为自己和为公众做决策时，风险偏好的差异意味着自我与公众之间的心理距离可能不同于先前自我与他人决策研究中自我与他人之间的心理距离。在这些研究中，“他人”通常是指特定的人（例如朋友或陌生人）。心理距离理论认为，随着心理距离的增加，个人更倾向于冒险。但公职人员与公众密切相关，他们的权利来自公众，所有这些都减少了他们与决策受益者的心理距离。因此，公职人员在为公众决策时比为自己决策时更保守。
公职人员的行为遵循安全第一的逻辑，以避免任何事故。在这种情况下，问责规避逐渐演变成公职人员之间⼀种隐性的共享信念和群体⾏为。以往的研究主要讨论了问责制对公务员问责规避⾏为的影响，这在⼀个风险⽇益增加的社会中成为⼀种普遍⽽强⼤的激励机制。研究揭示了公职人员在公共决策中遵循公共价值取向，表现出风险规避的倾向。由于公共价值的相对稳定，公务员在为公众决策时的风险偏好也较为稳定，较少受到决策情境（如决策⾦额）的影响，公共价值可能被⽤作公务员进⾏公共决策时的⾏为参照点。
Focusing on self versus public decision-making, this study finds that public officials have a lower risk preference when decision-making for the public than for the self. This finding is consistent with social value theory. Public officials' prosocial orientation may be activated when decision-making for the public. Their social responsibility and social rules may also cause them to have a lower risk preference in decision-making for the public than for the self. Additionally, the risk preference in self versus public decision-making implies that the psychological distance between the self and the public may be different from the psychological distance between the self and others in previous self versus other decision-making studies. In those studies, “the other”usually refers to a specific person (e.g., a friend or a stranger). Psychological distance theory suggests that, as psychological distance increases, individuals are more inclined toward risk. Public officials are closely related to the public, and their rights are derived from the public, all of which serves to decrease their psychological distance from the beneficiaries of the officials' decisions. Thus, the officials are more conservative when making decisions for the public than for the self.
The officials' behavior follows the logic of safety first and avoids any accidents. In this case, blame avoidance has gradually evolved into an implicit shared belief and group behavior among public officials. Previous studies have mainly discussed the impact of accountability on public officials' blame avoidance behavior, a prevailing and powerful incentive for public agents in an increasingly risky society. The research demonstrates that public officials are revealed to follow a social/public value orientation in public decision-making, namely exhibiting a tendency of risk avoidance behavior. Due to the relatively stable social value, public officials' risk preferences are relatively stable and less influenced by decision-making contexts (such as bet size) when making decisions for the public. This research implies the social value may be used as the behavioral reference point of public officials when making public decisions.
* 此推文部分内容及图片引自公众号：行为公共管理Lab
Part of the content and image of this post are quoted from: BPA_lab
文章来源：
Zhang, J., Wen, X., Mao, H., Xu, R., & Zhang, S. (2024). Does public officials' risk preference differ in self versus public decision-making? It depends on decision framing and bet size. Public Administration, 1–26.
原文链接：
https://doi.org/10.1111/padm.13037（或点击文末“阅读原文”查看）
《公共行政》微信公众号刚刚上线不久，我们期待来自各界的宝贵意见，您可将您的建议发送给后台。如果您对我们推送的文章感兴趣，也可在下方留言评论，与其他读者一起交流。
