Europe is heading into a super election year in 2027. Those elections will take place against a precarious security backdrop: war in Ukraine and Iran, a retreating US & Europeans increasingly worried about their own security. So shouldn’t this benefit the political centre? Not so fast. 🧵

What does this tell us? Safety at origin is not one factor among many. It determines when return becomes a real option. Until then, intentions/other predictors are weak signals of behavior. And worse conditions in host would not push refugees to rtrn while origin is unsafe. 8/8

Was this mainly about safety as we suggest? Respondents' assessments support this interpretation. Before Assad fell, only 26–37% described home as safe. The share jumped to 52% in the first post-fall wave and reached 75% by last wave 7/8

TWFE cannot rule out time-varying covariates (e.g. worse safety in Lebanon). So we compare areas that came under new Syrian government with Kurdish-controlled SDF areas, which were outside Assad’s control before and after he fled (thus less affected by regime change). 6/8

We use respondent and wave FEs to compare respondents with themselves over time Before Assad regime fell, most predictors were weak and insignificant. After Assad fled, several factors incl. intentions, economic conditions, and networks became much stronger predictors of return 5/8

Why? We argue refugee return happens in two stages. First, people ask whether home is safe enough for return to be a real option. Only once safety is cleared do other predictors (e.g. intentions, economic conditions, networks) shape return. 4/8

We asked respondents in 2019 if they intended to return. Did those intentions predict return behavior? Before Assad fell, those who intended barely return at higher rates than others Afterward, baseline intentions were much more strongly correlated with actual return years later 3/8

For 5 years before Assad fled, return barely increased Lebanon faced financial collapse, COVID-19, the Beirut explosion, & war with Israel. Most refugees lived in extreme poverty, yet only 1.6% returned per wave. After Assad fell, return jumped 4–7×, averaging 9.7% 2/8