How to Deal With Reverse Causality Using Panel Data? Recommendations for Researchers Based on a Simulation Study

Does X affect Y? Answering this question is particularly difficult if reverse causality is looming. Many social scientists turn to panel data to address such questions of causal ordering. Yet even in longitudinal analyses, reverse causality threatens causal inference based on conventional panel mode...

Full description

Saved in:
Bibliographic Details
Published in:Sociological methods & research Vol. 51; no. 2; pp. 837 - 865
Main Authors: Leszczensky, Lars, Wolbring, Tobias
Format: Journal Article
Language:English
Published: Los Angeles, CA SAGE Publications 01.05.2022
SAGE PUBLICATIONS, INC
Subjects:
Attribution Theory
Causal Models
Causality
Comparative Analysis
Criticism
Inference
Longitudinal studies
Monte Carlo Methods
Panel data
Research Problems
Simulation
Social Science Research
Social sciences
Social scientists
Specification
Statistical Bias
Time
exogeneity
causal inference
dynamic panel model
panel data
reverse causality
fixed effects
ISSN:0049-1241, 1552-8294
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Does X affect Y? Answering this question is particularly difficult if reverse causality is looming. Many social scientists turn to panel data to address such questions of causal ordering. Yet even in longitudinal analyses, reverse causality threatens causal inference based on conventional panel models. Whereas the methodological literature has suggested various alternative solutions, these approaches face many criticisms, chief among them to be sensitive to the correct specification of temporal lags. Applied researchers are thus left with little guidance. Seeking to provide such guidance, we compare how different panel models perform under a range of different conditions. Our Monte Carlo simulations reveal that unlike conventional panel models, a cross-lagged panel model with fixed effects not only offers protection against bias arising from reverse causality under a wide range of conditions but also helps to circumvent the problem of misspecified temporal lags.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0049-1241
1552-8294
DOI:10.1177/0049124119882473