Labor Turnover before Plant Closure:'Leaving the sinking ship' vs. 'Captain throwing ballast overboard'

Involuntary job loss in administrative data is commonly identified by focusing on mass-layoffs or plant closures. However, such events usually do not happen without prior knowledge, which potentially leads to selection in the labor turnover of distressed firms. We find that workers separating from closing plants up to 2 quarters before closure are associated with significantly lower displacement costs and on average significantly higher pre-closure earnings levels as opposed to ultimately displaced workers. Furthermore, our results indicate that displaced workers with high pre-closure earnings experience significantly lower reductions in future employment probabilities. These findings suggest that compositional differences cause estimated displacement costs to differ between early leavers and ultimately displaced workers. Focusing exclusively on the latter group would lead to an overestimation of displacement costs.