Primary Submission Category: Longitudinal Data Analysis
The causal effect of volatility: Estimation by marginal structural models
Authors: Nanum Jeon, Ian Lundberg, Hao Liang,
Presenting Author: Nanum Jeon*
U.S. workers operate in a labor market characterized by a high level of both inequality and volatility. Inequality exists to the degree that economic well-being is unequally distributed across people at a single point. Volatility exists to the degree that the economic well-being rises and falls for a given person over time. Both may be consequential, yet the causal effects of volatility have been understudied because volatility is a complicated causal treatment: it is by definition a trajectory experienced over time. We conceptualize the causal effect of volatility in the potential outcomes framework and show how to estimate using marginal structural models (MSMs). We develop new structural models designed to directly estimate the effects of volatility. We illustrate the method by studying the causal effect of employment volatility on marriage, and we close with a discussion of how our approach could unlock new research about the causal effects of volatility.