Skip to main content

Table 3 Mean (Absolute) standardized bias before and after matching

From: How sensitive are matching estimates of active labor market policy effects to typically unobserved confounders?

 

Before matching

After

  

Standard

specification

Extended specification

Typically observed characteristics X

 Socio-demographics

5.8

0.6

0.7

 Household characteristics

4.1

0.6

0.4

 Partner characteristics

2.2

0.6

0.6

 Regional characteristics

11.0

0.5

0.3

 Labor market history

10.3

1.1

0.7

 Total (X)

8.7

0.9

0.7

Typically unobserved characteristics U

  ttitudes towards work

3.6

3.3

0.9

 Job search behavior

14.9

8.5

0.5

 Concessions willing to make for a job

17.5

7.8

0.4

 Satisfaction in different domains

6.7

4.2

0.9

 Participation, social status and networks

3.6

2.1

1.1

 (Mental) Health

13.3

7.0

0.4

 Inter-generational information

5.0

2.9

0.2

 Total (U)

10.2

5.5

0.6

Total (\(X+U\))

9.0

1.7

0.7

Control variables

  Typically observed covariates X

 

\(\checkmark\)

\(\checkmark\)

  Typically unobserved covariates U

  

\(\checkmark\)

Number of participants (on support)

1009

1009

1004

  1. This table compares covariate balance across different blocks of covariates. Balance is measured using the mean (absolute) standardized bias (MSB, Rosenbaum and Rubin 1983)