Collaborative Research: Labor Market and Public Policy Preferences

人们想要什么（偏好）是经济分析和政策制定的核心。然而，没有一种简单或普遍可接受的方法可以从数据中衡量偏好。本研究提出了一种可移植、可扩展且易于测量的偏好启示（启发）方法，称为贝叶斯自适应选择实验（BACE）。BACE改进了在政策分析、管制和诉讼中使用的现有偏好诱导方法。该方法将用于研究重要的和与政策有关的问题：(i)对工作便利的偏好，如工作场所的灵活性，以及坚定地提供这种便利--面对全球技术和组织变化的重要考虑因素。这一点很重要，因为随着提供工作场所便利设施的成本随着时间的推移而变化，全球技术和组织变革有可能改变工作结构。（二）在公共政策偏好方面，研究提出了比较不同政策对不同人口群体福利效益的方法。除了这些应用之外，BACE还展示了新一代调查和实验方法在帮助研究人员和政策制定者评估非市场资源方面发挥更大作用的潜力。本文的研究结果将有助于改进经济分析，提高公共政策和商业决策的质量，从而促进经济增长。BACE为进行选择实验提供了一个有效的动态启发过程。它通过生成基于先验的选择场景的信息最大化序列来实现这一目标，该先验根据先前的答案进行更新，以获得个人层面的贝叶斯后验估计。该过程允许参数估计的更高精度，提供给每个受试者的选择场景更少，同时也克服了从常用的静态方法估计平均偏好参数的系统偏差。将该方法应用于对工作便利的偏好，所得的个人层面偏好数据使估计补偿差异的新模型成为可能，该模型扩展了经典的Rosen（1986）框架。将BACE应用于公共政策的支付意愿，可以产生有关政策偏好相关性的新证据，并导致对公共资金边际价值的新估计。在本研究中实施BACE是向前迈出的一步，允许这种程序被广泛采用，以衡量理解社会科学中广泛现象的重要输入。本文的研究结果将有助于改进经济分析，提高公共政策和商业决策的质量，从而促进经济增长。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

What people want (preferences) is central to economic analysis and policy formulation. Yet there is no easy or generally acceptable way of measuring preferences from data. The proposed research develops a portable, scalable, and easy to measure preference revelation (elicitation) method called the Bayesian Adaptive Choice Experiment (BACE). BACE improves upon existing preference elicitation approaches that are used in policy analysis, regulation, and litigation. The method will be applied to study important and policy-relevant questions: (i) preferences for job amenities such as workplace flexibility, and firm provision of such amenities---important considerations in the face of global technical and organizational changes. This is important as global technical and organizational changes have the potential to alter the structure of work as the costs of providing workplace amenities shift over time. (ii) in preferences for public policies, the research produces method of comparing the welfare benefits of various policies for different demographic groups. In addition to these applications, BACE shows the potential for a new generation of surveys and experimental methods to do much more in helping researchers and policymakers value non-market resources. The results of this research will contribute significantly to improve economic analyses as well as improve the quality of public policies and business decisions and thus increase economic growth. BACE provides an efficient dynamic elicitation procedure for conducting choice experiments. It does so by generating the information-maximizing sequence of choice scenarios based on a prior that gets updated with previous answers to obtain individual-level Bayesian posterior estimates. The procedure allows for a higher precision of the parameter estimates with fewer choice scenarios presented to each subject while also overcoming systematic biases in estimating average preference parameters from commonly used static approaches. Applying the method to preferences for job amenities, the resulting individual-level preference data makes it possible to estimate a new model of compensating differentials that extends the classic Rosen (1986) framework. Applying BACE to willingness to pay for public policies produces new evidence on the correlates of policy preferences and leads to novel estimates of the Marginal Value of Public Funds. The implementation of BACE in this research is a step forward in allowing such procedures to be widely adopted to measure important inputs for understanding a broad range of phenomena in social sciences. The results of this research will contribute significantly to improve economic analyses as well as improve the quality of public policies and business decisions and thus increase economic growth.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.