We use newly available data from the Ontario Ministry of Labour to est
imate the effects of advance notification of permanent layoff on unemp
loyment durations. While notice is strongly negatively correlated with
unemployment in the raw data, most of this effect disappears when we
control for personal characteristics and when we correct for endogenei
ty of notice by using only its cross-firm variation to identify its ef
fects. Overall, while short notice intervals can have a substantial ef
fect on the probability of experiencing a short unemployment spell, we
find that even very long notice intervals have little impact on longt
erm unemployment.