Unraveling the time-series dynamics between aggregate earnings and GDP

Abstract

This paper examines time-series variation in the relationship between aggregate earnings and GDP (AEG). Using quarterly data from 1979-2019, we document that aggregate earnings are positively associated with future real GDP after 2000 but not before. We further show that this variation does not result from an association between aggregate earnings and the corporate profits component of GDP. Instead, we discover that aggregate special items — not core earnings — drive the post-2000 AEG by conveying information about labor market outcomes. Specifically, aggregate special items relate to future GDP only when they predict wages and worker displacement. Our study contributes to the literature by documenting previously unexplored time variation in the AEG relationship and identifying its source, challenging conventional assumptions about how aggregate accounting measures signal future economic conditions.