This study examines the economic correlates of U.S. presidential approval from 1953 to 2023 using a systematic, data-driven approach. Rather than focusing on a narrow set of preselected indicators, the analysis draws on a large macro-financial dataset and applies penalized variable-selection methods; Elastic Net and Smoothly Clipped Absolute Deviation (SCAD); to identify economically relevant predictors. The selected predictors are then organized into six economic categories and analyzed in a Structural Vector Autoregression (SVAR) framework to characterize dynamic co-movement between macroeconomic conditions and presidential approval. Three descriptive patterns emerge. First, presidential approval exhibits strong persistence and retrospective dynamics, with economically relevant information loading most prominently at short horizons and around one year. Second, public and private sector activity displays the strongest and most persistent associations with approval, while monetary and labor market conditions contribute in systematic but heterogeneous ways. Third, housing and mortgage indicators are associated with delayed responses, whereas recession indicators exhibit short-run dynamics consistent with temporary rally-type patterns. Overall, the findings provide a descriptive characterization of how presidential approval co-moves with a broad set of macroeconomic conditions, illustrating how high-dimensional screening combined with dynamic analysis can complement parsimonious approaches in the approval literature.
Youssuf Abdelatif (Thu,) studied this question.