美联储-迷失在聚合中_收入和支出的地理计量错误（英）-2025_36页_710kb

Summary of "Lost in Aggregation: Geographic Mismeasurement of Income and Spending"

When using median income from zip code areas as a broad indicator of individual household income, significant heterogeneity in consumption behavior is masked.
This study, based on comprehensive panel data analysis of 150,000 representative U.S. households, revealed that using such proxy aggregates as an alternative measure drastically simplifies the reality and yields misleading conclusions compared to self-reported income measures.
Specifically, analysis of real retail spending patterns during 2018-2024 showed that low-income households began reducing their spending while high-income households continued strong spending growth from mid-2021 onward.
However, when using median zip code income to categorize households, this divergence disappeared, indicating the strong limitation in the proxy approach.
Furthermore, table analysis quantified a significant discrepancy between self-reported income groupings and the implied income from zip code medians, ranging from 35 to 75%.
This underlines the critical need to rely on micro-level household income data for assessing heterogeneity in consumption, and highlights the dangers of over-relying on aggregated geocoded proxies in economic assessments.