Skip to main content

Table 5 Estimates from propensity score matching

From: Are migrants in large cities underpaid? Evidence from Vietnam

Matching scheme

Effect of “no registration book”

Effect of “migration since 2008”

Treated (migrant) (Y 1)

Controls (non-migrant) (Y 0)

Difference (Y 1 − Y 0)

Treated (migrant) (Y 1)

Controls (non-migrant) (Y 0)

Difference (Y 1 − Y 0)

Mean

Std. Er.

Mean

Std. Er.

1 nearest neighbor

2630.0

2539.6

90.4

113.8

2278.0

2312.6

−34.6

146.0

5 nearest neighbors

2630.0

2529.9

100.1

99.5

2278.0

2366.7

−88.7

93.5

Kernel, bandwidth 0.01

2630.0

2516.9

113.1

117.8

2280.3

2379.7

−99.5

109.1

Kernel, bandwidth 0.03

2630.0

2527.2

102.8

115.4

2278.0

2392.9

−114.9

106.1

Kernel, bandwidth 0.05

2630.0

2533.4

96.7

114.8

2278.0

2415.1

−137.1

104.5

  1. Note: Treated are overall migrants and short-term migrants. Controls are matched non-migrants. Standard errors are calculated using bootstrap with 500 replications (standard errors are corrected for sampling weights and cluster correlation)
  2. *Significant at 10 %; **significant at 5 %; ***significant at 1 %
  3. Source: Estimation from the 2009 UPS