The Longitudinal Unexpected Event during Survey Design

The Unexpected Event during Survey Design (UESD) has emerged as a preeminent identification strategy to ascertain causal effects using observational cross-sectional survey data (Muñoz et al. 2020). UESD capitalizes on the sudden occurrence of events, unrelated to the timing of survey fieldwork, to establish a natural experimental setting. Despite its advantages, the inherent characteristics of the data within UESD present significant challenges for establishing causality. In this article, we explore how to combine the UESD with panel data. This strategy opens new analytical opportunities that can significantly strengthen the credibility of causal effects. In particular, we claim that additional survey waves offer opportunities for more robust substantiation of UESD's primary assumptions; it allows adopting a Difference-in-differences approach in scenarios where underlying assumptions may not be substantiated; and it enables the examination of the durability of causal effects. To demonstrate LUESD's benefits, we delve into the "electoral winners" literature in the context of Chile's 2022 constitutional referendum. Utilizing an extensive panel survey, we analyze the referendum's influence on satisfaction with democracy. Given the growing number of panel surveys in the social sciences, we believe this design holds promising opportunities for researchers.