Indirect Estimators of Intergenerational Mobility

TL;DR

This chapter reviews various indirect methods for estimating intergenerational mobility, emphasizing how different approaches infer family associations when direct income data are incomplete.

Contribution

It introduces a unifying framework for understanding indirect estimators and clarifies how their choice of instruments or imputation strategies affects what aspects of transmission they capture.

Findings

Different indirect estimators measure distinct transmission channels.

The choice of instrument influences the aspect of mobility being estimated.

Indirect estimators can reveal long-term persistence beyond parent-child correlations.

Abstract

This chapter reviews indirect estimators of intergenerational mobility, focusing on approaches that infer parent-child or other family associations when direct income data are incomplete or unavailable. We synthesize methods based on instrumental variables, imputation using observable characteristics such as education and occupation, surname-based estimators, and multigenerational linkages. To unify these approaches, we introduce a stylized framework in which socioeconomic status is transmitted through multiple pathways with heterogeneous persistence rates. Within this framework, both direct and indirect estimators can be interpreted as weighted averages of these underlying transmission channels. A central insight is that the choice of instrument or imputation strategy determines these weights, leading different methods to capture distinct aspects of the transmission process. We highlight implications for interpretation, showing that indirect estimators need not recover conventional parent-child correlations but can instead provide complementary evidence on long-run persistence and the mechanisms underlying persistent inequalities.