Skip to main content

Table 5 Results for alternative matching models

From: Reemployment premium effect of furlough programs: evaluating Spain’s scheme during the COVID-19 crisis

Model

Group

Single quarter

Two quarters

On support

ATT (BS SE)

On support

ATT (BS SE)

(I)

(II)

(III)

(IV)

(V)

(VI)

k=1, cal=0.01, repl.

Untreated

1447

 

444

 
 

Treated

3102

0.294 (0.0287)

333

0.114 (0.0535)

k=1, cal=0.001, repl.

Untreated

1219

 

153

 
 

Treated

2920

0.290 (0.0275)

144

0.069 (0.0800)

k=10, cal=0.01, repl.

Untreated

1447

 

444

 
 

Treated

3102

0.304 (0.0279)

333

0.139 (0.0493)

k=1, cal=0.01, norepl.

Untreated

964

 

159

 
 

Treated

936

0.259 (0.0210)

155

0.110 (0.0532)

epanech., bw=0.06, cs.

Untreated

1493

 

479

 
 

Treated

3088

0.301 (0.0253)

333

0.139 (0.0515)

epanech., bw=0.01, cs.

Untreated

1436

 

434

 
 

Treated

3088

0.305 (0.0262)

321

0.116 (0.0496)

normal, bw=0.06, cs.

Untreated

1493

 

479

 
 

Treated

3088

0.298 (0.0240)

333

0.137 (0.0513)

normal, bw=0.01, cs.

Untreated

1493

 

479

 
 

Treated

3088

0.307 (0.0257)

333

0.141 (0.0494)

  1. (I) Model specs –k no. of neighbours, calliper, replacement, kernel type, bw bandwidth and common support–, (II) group, (III) and (V) remaining subsample on support, (IV) and (VI) Average Treatment Effect on the Treated with 500 bootstrapped standard errors in parenthesis