An Investigation into the Effect of Organisational Culture on Behaviour & Preferences

行为经济学和实验经济学成功地为主流经济理论带来了新的见解。这些文献已经确认了传统经济人模型之外的偏好(不)的存在，并调查了直接背景对行为的影响(Tversky&Kahneman，1981；Zizzo，2010)。然而，许多文献中缺少的是对是什么机制导致了这些奇异偏好的令人满意的解释。我们的研究问题不只是把偏好作为职业选择的事前决定因素(“使命匹配”假说)，而是关注组织(公共部门)的文化是否会随着时间的推移改变招聘人员的偏好。总而言之，我们的假设是，偏好不是外生的或个人主义的，而是部分通过组织成员身份习得的。在国家层面上的政治经济学和经济史文献中，已经考虑了文化的影响。我们认为，纳恩(2012)关于文化的概念可能与国家层面一样重要，他认为文化是“根据我们在复杂和不确定的环境中做出决定的需要而发展起来的启发式或‘经验法则’”。在最基本的层面上，不同组织中的个人被要求学习截然不同的技能和行为，并以不同的方式与不同的人互动(想想投资者和教师)。此外，组织特征可能会导致不同组织的成员以截然不同的方式经历类似的情况。正是这些规则与该角色的职责相互作用，才产生了“组织文化”。事实上，我们感兴趣的正是这些在组织内部演变的“决策启发式规则或‘经验法则’”(Nunn，2012)，以及它们对个人在工作场所内外行为的影响。我们提出的衡量偏好假设演变的方法是，在个人的职业发展过程中，每隔一段时间对他们进行激励性的实验室现场经济实验。由于可获得性差，使用观测数据获得可靠的估计将是有问题的。此外，该项目考虑了传统经济理论之外的行为，因此我们不能使用检验从基本微观经济模型得出的预测的技术。出于这两个原因，我们选择了一种实验方法。我们将首先在大学不同系的学生身上进行实验。其理念是，不同系的学生就像是不同公司的员工，他们拥有各自课程和/或部门独有的经历，也会因某些偏好而受到奖励或惩罚，方式将是他们自己的学科选择所特有的。我们的实验将建立在现有方法的基础上，以引发亲社会动机(Banuri&Keefer，2016)或风险偏好(Eckel&Grossman，2008)。然而，我们会在学生的学业进步过程中每隔一段时间进行实验。然后，来自试验组和对照组的数据可以被输入计量经济学模型，该模型测试随着时间推移行为的重大变化。大学的实验是一个很好的第一步：它们对于博士项目是可行的，而且本身就有额外的好处(它们很便宜，可以在一个校园里接触数百个职业)。至关重要的是，来自大学实验的证据也将被用来激励与外部组织的互动。如果我们能证明亲社会动机、腐败或其他行为直接受到大学内部组织文化的影响，那么我们相信外部组织将有动力让我们在其成员身上进行实验。

Behavioural and experimental economics has been successful in bringing fresh insights to mainstream economic theory. The literature has identified the (non-)existence of preferences that lie outside of the traditional model of economic man, and investigated the impact of immediate context on behaviour (Tversky & Kahneman, 1981; Zizzo, 2010). What is missing from a lot of the literature, however, is a satisfactory explanation of what mechanisms bring about these exotic preferences. Rather than only considering preferences as ex-ante determinants of career choice (the 'mission-matching' hypothesis), our research question concerns whether the culture of the organisation (the public sector) changes the preferences of recruits over time. In general terms, our hypothesis is that preferences are not exogenous or individualistic but are in part learned in through organisational membership.The effect of 'culture' has been considered in literature on Political Economy and Economic History on a national level. We believe that Nunn's (2012) notion of culture as "heuristics or 'rules of thumb' that have evolved given our need to make decisions in complex and uncertain environments" may be just as important at the level of the workplace as the nation. At the most basic level, individuals in different organisation are required to learn starkly contrasting skills and behaviours and interact with a variety of people in distinct ways (think of investors and teachers). Moreover, organisational characteristics may cause members of separate organisations to experience similar situations in contrasting ways. It is these rules, interacting with the responsibilities of the role, which generate 'organisational culture'. Indeed, it is these "decision making heuristics or 'rules of thumb' that have evolved" (Nunn, 2012) within an organisation, and their effects on individual behaviour within the workplace and beyond, that we are interested in.Our proposed method for measuring the hypothesised evolution of preferences is to run incentivised lab-in-field economic experiments on individuals at intervals in their career progression. It would be problematic to obtain robust estimates using observational data due to poor availability. Moreover, the project considers behaviour that lies outside traditional economic theory so we cannot employ the technique of testing predictions derived from basic microeconomic models. For both of these reasons, an experimental methodology has been selected.We would begin by running experiments on students within different University departments. The idea is that students within different departments are like employees of different companies in that they have experiences that are unique to their course and/or department, and are also rewarded or punished for certain preferences in ways that will be peculiar to their own subject choice. Our experiments would build on existing methodologies to elicit, for example, pro-social motivation (Banuri & Keefer, 2016) or tastes for risk (Eckel & Grossman, 2008). However, we would run the experiments at intervals throughout the students' academic progression. The data from treatment and control groups could then be fed into an econometric model which tests for significant changes in behaviour over time. The University experiments act as a good first step: they are feasible for a PhD project and have additional benefits in their own right (they are cheap, and provide access to hundreds of professions on one campus). Crucially, evidence from the University experiments would also be used to motivate engagement with external organisations. If we can demonstrate that pro-social motivation, corruptibility, or other behaviours are directly impacted by organisational culture within Universities, then we believe that external organisations will have an incentive to let us run experiments on their members.