A tale of inattentiveness and the loss function: A model for household-level macroeconomic expectations
本文提出一个理论框架，结合信息粘性和不对称损失函数，解释非专业代理人宏观经济预期的偏差与粘性，并用美欧消费者失业预期数据验证了预期均值与标准差之间的负相关关系。
This paper presents a novel theoretical framework that integrates information stickiness with asymmetric costs of overpredicting and underpredicting macroeconomic variables. We focus on non-professional agents who exhibit these traits. The asymmetric loss function leads to the presence of a non-null bias in expectations, and information stickiness generates a positive relationship between this bias and the stickiness of expectations. At the empirical level, this relationship translates into a negative correlation, at the level of socio-demographic groups, between the mean and the standard deviation of the time series of the expectations. Three decades of data on consumers’ unemployment expectations in the US and the EU support our results, validating the hypothesized negative relationship between the mean and standard deviation of aggregate expectations. The framework is also consistent with established patterns of consumer expectations. • The model combines an asymmetric loss function and individual agents’ inattentiveness. • The model leads to cross-sectional dispersion, bias, and stickiness of expectations. • More inattentiveness increases both bias and stickiness of expectations. • Three decades of data on consumers’ unemployment expectations provide empirical support. • We find evidence for the positive relationship between bias and stickiness.