The unhappily unemployed return to work faster

Table 12 Duration of unemployment when controling benefits, job search and income

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	All	All	All	All	All	Women	Men
Search		0.177***	0.190**	0.128**
		(0.0616)	(0.0796)	(0.0627)
Δ H a p	-0.0254***	-0.0302***	-0.0333***	-0.0285***	-0.0255***	-0.0198	-0.0290**
	(0.00712)	(0.00763)	(0.00947)	(0.00762)	(0.00894)	(0.0136)	(0.0123)
Log (Benefits)			-0.125**		-0.156***	0.0170	-0.288***
			(0.0512)		(0.0478)	(0.0703)	(0.0747)
Log (real yearly income)				0.260***	0.270***	0.187**	0.384***
				(0.0401)	(0.0527)	(0.0817)	(0.0727)
Observations	1344	1180	772	1160	842	343	499
Standard controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Region and Wave	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Source: BHPS, waves 1–16. Standard errors in parenthesis, *significant at 10%, **significant at 5%, ***significant at 1%.
All specifications above are from a duration model using the StCox procedure.
Standard controls include age, age squared, gender, number of children, household size, dummies for education and dummies for marital status. Region and Wave dummies are added in all specifications.
Diff-hap measures the initial change in happiness when becoming unemployed. Given that most observations of Diff-hap are negative, the results have to be interpreted as follows: if happier about losing a job, then one spends more time in unemployment.