CAREER: An accessible, dynamic model of pricing applied to wage and housing purchases

美国各地的雇主都面临着劳动力供应危机，部分原因是员工希望的工作条件和他们认为公平的工资发生了变化。与此同时，潜在购房者面临着不断上涨的成本和激烈的住房竞争--老年人、退伍军人、少数族裔和低收入家庭受到房价上涨的打击最大。这个项目试图了解人们如何在这些类型的交易(工资，购房)中分配价值，重点是个人如何对竞争和市场变化做出反应。该项目团队正在与有关资金管理和首次购房者的金融教育项目合作，以了解人与人之间的差异--例如对风险的容忍度或将购买推迟到未来某个日期的意愿--如何预测他们在就业和住房市场上的成功。这些发现将使我们能够更好地理解价值心理学，创造教育资源来帮助人们采用更好的找工作或住房策略，并确定可以缓解住房和劳动力短缺的公共政策干预。通过一系列实验，该项目测试了多选择、多属性环境下的定价计算模型。这些实验操纵模拟的市场竞争、延迟、风险和选择备选方案的属性，以考察个体之间的定价动态，并测试定价模型的认知机制。为了提高定价模型拟合和比较的可及性和效率，该项目将它们嵌入到神经网络中，训练神经网络将观察到的定价数据映射到生成模型参数。这些将被添加到在线工具、教学材料和研讨会中，以便广泛使用。定价模型将被用来在多选项选择实验中估计风险和延迟厌恶、偏差和定价动态。模型参数还将用于预测金融教育项目参与者的求职和找房结果。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Employers across the United States are facing a crisis of labor supply, driven in part by changes in the working conditions employees desire and the wages they perceive as fair. Simultaneously, prospective homebuyers are faced with rising costs and fierce competition for housing – with aging adults, veterans, minorities, and low-income families hit hardest by rising home prices. This project seeks to understand how people assign value in these types of transactions (wages, homebuying), focusing on how individuals respond to competition and market changes. The project team is partnering with financial education programs on money management and first-time homebuyers to understand how differences between people – such as their tolerance for risks or willingness to delay purchases until a future date – predict their success on job and housing markets. The findings will allow us to better understand the psychology of value, create educational resources to help people adopt better strategies for finding jobs or housing, and identify public policy interventions that can alleviate housing and labor shortages.Across a series of experiments, this project tests computational models of pricing in multi-alternative, multi-attribute settings. These experiments manipulate simulated market competitiveness, delays, risks, and attributes of the choice alternatives in order to examine the dynamics of price-setting among individuals and test the cognitive mechanisms of pricing models. To improve the accessibility and efficiency of pricing model fitting and comparison, the project embeds them in neural networks trained to map observed pricing data onto generative model parameters. These will be added to online tools, teaching materials, and workshops to make them widely available. The pricing models will be used to estimate risk and delay aversion, bias, and pricing dynamics in multi-alternative choice experiments. The model parameters will also be used to predict the outcomes of job and home searches among participants in financial education programs.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.