Heterogeneity, Household Behavior, and the Transmission of Aggregate Shocks

[摘要]异质性、家庭行为和总体冲击的传导大多数用于研究货币和财政政策的模型都脱离了不平等，并假设家庭是理性的，并且完全了解经济状况。考虑到在消费、收入、财富和信息集方面具有异质性的家庭，对建模提出了挑战。本研究利用理论和计算的最新进展开发了一个异质决策者建模的代码库。开发的工具包将包括代码、笔记本形式的大量文档以及教材。开发的代码库将通过简化复杂宏观经济模型的计算和估计，为研究人员和决策者提供一个有用的工具包，以考虑异质家庭。本文对异构智能体模型的研究利用了序列空间平衡计算的新进展，通过联合求解未知时间路径，给定分布和其他状态的初始条件下的所有周期平衡方程。对于相对于聚合线性化到一阶的模型，序列和状态空间方法提供相同的脉冲响应，但是序列空间对于异构代理模型具有计算优势：它允许更广泛的冲击类别，易于扩展，并且可以实现较大的速度增益。开发的工具包将需要输入表征平衡的方程，包括异构代理块的规格。然后，该工具包将自动执行给定冲击的稳态和过渡动力学计算。扩展将允许估计以及最优策略的计算。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。

AbstractPI Last Name: AuclertTitle: Heterogeneity, Household Behavior, and the Transmission of Aggregate ShocksProposal ID: 1851717 Most of the models used to study monetary and fiscal policy abstract away from inequality and assume that households are rational and perfectly informed about the state of the economy. Considering households that are heterogenous with respect to their consumption, income, wealth, and information sets imposes challenges in modeling. This research develops a codebase in modelling heterogenous decision makers using recent advances in theory and computation. The developed toolkit will include codes, extensive documentation in the form of notebooks, as well as teaching materials. The developed codebase will provide researchers and policy makers a useful toolkit in considering heterogenous households by simplifying the computation and estimation of complex macroeconomic models. This research on heterogenous agent models leverages new advances on the computation of equilibria in sequence space by jointly solving the equations characterizing equilibrium across all periods for time paths of the unknowns, given initial conditions on the distribution and other states. For models that are linearized to first order with respect to aggregates, the sequence and the state space approaches deliver the same impulse responses, but the sequence space has computational advantages for heterogeneous-agent models: it allows for a wider class of shocks, is easily scalable, and can achieve large speed gains. The developed toolkit will require as input the equations characterizing equilibrium, including specification of heterogeneous-agent ?blocks?. The toolkit will then automatically perform computation of steady state and transition dynamics for given shocks. Extensions will allow estimation as well as computation of optimal policy.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.