Can a search model predict the effects of an increase in the benefit duration? Evidence from the Portuguese unemployment insurance reform

Table 2 Summary statistics and unconditional difference-in-differences

	15–24		25–29		30–34		35–39		40–44
	Treatment		Control		Treatment		Control		Treatment
	Before	After	Before	After	Before	After	Before	After	Before	After
Unemployment (days)	146.2	179.0	203.0	187.9	206.5	269.3	301.5	288.3	266.9	368.6
Differences	32.8		−15.1		62.9		−13.2		101.7
	(1.8)		(1.7)		(2.8)		(3.0)		(4.8)
Difference-in-differences		47.9				76.1		114.9
		(2.5)				(4.1)		(5.4)
Female (proportion)	0.49	0.57	0.48	0.60	0.43	0.56	0.41	0.52	0.35	0.47
Pre-unemployment wages
Average	401.1	443.5	513.0	599.4	581.2	645.1	662.3	660.0	689.5	679.4
Median	385.8	416.1	458.8	525.6	472.7	537.3	502.3	514.8	495.5	517.9
Age	22.0	22.2	27.0	27.0	31.9	31.8	36.9	36.9	41.8	41.9
Observations	5149	17,962	7084	30,254	5075	24,478	5358	17,827	3436	14,165

Mean values unless otherwise noted. The dataset and the estimation include region, month, and year of unemployment insurance claims. Standard deviations in parentheses