A1276
Title: Effects of recent and current income on life satisfaction: A functional regression model with endogeneity
Authors: Mirkan Oecal - University of Cologne (Germany) [presenting]
Sven Otto - University of Cologne (Germany)
Dominik Wied - University of Cologne (Germany)
Abstract: The aim is to propose a functional regression model to analyze both concurrent and historical effects of income on life satisfaction. A model is developed with a recent history operator to capture how past income affects current happiness. A finite-dimensional factor structure is imposed on the functional regressor, reducing the infinite-dimensional problem to a finite-dimensional regression and enabling valid statistical inference. The key contribution is addressing the endogeneity in the income-happiness relationship - particularly reverse causality where happier people may earn more - through an instrumental variables strategy adapted to the factor-augmented model. The methodology builds on recent work in related functional linear models, but extends these by implementing an IV approach for causal interpretation. Estimation proceeds by identifying factors as (left-)singular functions of the cross-covariance operator between life satisfaction and income history, followed by two-stage least squares regression. An empirical application using the GSOEP data demonstrates the practical implementation of the proposed methodology.
