Collaborative Research: Meta-Analysis of Empirical Estimates of Loss-Aversion

损失带来的伤害大于收益带来的愉悦。这一原则被称为“损失厌恶”，是行为经济学的一个基本原则，它利用心理学来改进经济预测和建议。损失厌恶被用于理解一些政策问题，包括国际贸易和股票市场交易。它还被政府和企业用作设计“助推”的工具，以帮助人们做出更好的决策。“损失厌恶”的价值取决于它所适用的具体事件和环境。因此，存在几个损失厌恶的计算值，为其在研究和政策中的使用带来了混乱和不确定性。本研究将收集所有损失厌恶的计算值，进行分类和总结，并将其放在一个中央存储库中，以便于参考。这项研究活动的结果将提高损失厌恶在研究和政策制定中的效用。这将使美国成为全球损失厌恶研究和应用的领导者，因为这将是世界上唯一的此类研究。对损失的厌恶是用一个数字“lambda”来衡量的，这个数字被认为是收益价值的1.5-2倍。这个范围的值是基于一些突出的例子，但实际上有几百个已发表的研究测量了“lambda”。我们提议的荟萃分析包括找到并识别每一篇有“lambda”统计估计值的论文，记录所有这些估计值，然后总结它们的结果。这样，科学家们就可以第一次看到假设的“λ”=1.5到2的范围是否正确。拟议的研究还将测量“lambda”在不同人群中以及基于科学方法差异（例如，受控实验室实验与从日常决策推断）的变化程度。这些证据可以帮助行为洞察团队通过对“lambda”的适当值进行更精确的假设，为特定人群定制推动。荟萃分析还说明了如何有效地积累科学知识，以克服对科学可重复性的可能担忧。这项研究也有助于提高科学研究的可重复性。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。

Losses hurt more than gains feel good. This principle, known as "loss-aversion", is a foundational principle in behavioral economics, which uses psychology to improve economic prediction and advice. Loss aversion is applied to understand several policy issues, including international trade and stock market trades. It has also been used as a tool by governments and businesses to design "nudges" to help people make better decisions. The value of "loss aversion" depends on specific events and circumstances in which it is applied. As a result, several calculated values of loss aversion exist, creating confusion and uncertainty for its use in research and policy. This proposed research will collect all calculated values of loss aversion, categorize and summarize, and placed them in a central repository for easy reference. The results of this research activity will improve the usefulness of loss aversion in research and policy development. This will establish the US as the global leader in loss aversion research and application as this will be the only such study in the world. Aversion to losses is measured by a number "lambda", which is thought to be about 1.5-2 times higher than the value of gains. This range of values is based on prominent examples, but there are actually several hundred published studies measuring "lambda". Our proposed meta-analysis involves finding and identifying every paper that has a statistical estimate of "lambda", recording all those estimates, then summarizing their results. Scientists can then see for the first time whether the assumed range of "lambda"=1.5 to 2 is correct. The proposed research will also measure how much "lambda" varies for different groups of people and based on differences in scientific methods (e.g., controlled lab experiments vs. inferences from everyday decisions). This evidence can help behavioral insights teams customize nudges for particular groups of people, using more precise assumptions about the appropriate value of "lambda". The meta-analysis also illustrates how to cumulate scientific knowledge effectively, to overcome possible concerns about the reproducibility of science. This research also contributes to the efforts to increase reproducibility of scientific research.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.