Statistics
Labor Economics Analysis
Econometric analysis of 69K+ Census records quantifying the marginal return to work hours using log-linear regression.
The Problem
Context & Challenge
Policymakers need reliable estimates of how work hours affect earnings to inform labor regulations. The challenge was isolating this relationship from confounding factors in large-scale survey data.
The Approach
Architecture & Implementation
Analyzed 69K+ employed workers from 2023 CPS ASEC Census data. Compared linear, log-linear, and polynomial regression specifications with robust standard errors (HC1). Applied log transformation to address income skewness and improve model fit.
The Results
Impact & Metrics
Log-linear model (R²=0.307) showed each additional hour worked increases earnings by 5.6%. Residual diagnostics confirmed model validity for policy interpretation.
Key Result
Each additional hour worked increases earnings by 5.6% (log-linear model, R²=0.307)
Technologies & Methods
ROLS RegressionLog-Linear ModelsEconometricsRobust Standard Errorsggplot2