Skip to main content

Table 6 Decomposition of changes in parameters of the distribution, Tajikistan 2007

From: Is there an informal employment wage premium? Evidence from Tajikistan

Percentiles

Raw gap

Predicted gap

Due to coefficients

Due to covariates

Residuals

10th

-1.050

-1.047

-0.443 [42%]

-0.604 [58%]

-0.003 [0%]

  

(0.0234)

(0.0352)

(0.0247)

 

20th

-1.058

-1.012

-0.390 [37%]

-0.622 [59%]

-0.047 [4%]

  

(0.0185)

(0.0292)

(0.022)

 

30th

-1.022

-0.974

-0.370 [36%]

-0.604 [59%]

-0.048 [5%]

  

(0.0184)

(0.0274)

(0.0199)

 

40th

-0.916

-0.892

-0.343 [37%]

-0.549 [60%]

-0.024 [3%]

  

(0.0174)

(0.0273)

(0.0206)

 

50th

-0.783

-0.792

-0.306 [39%]

-0.486 [62%]

0.009 [-1%]

  

(0.018)

(0.0265)

(0.0193)

 

60th

-0.758

-0.715

-0.268 [35%]

-0.447 [59%]

-0.042 [6%]

  

(0.0174)

(0.0263)

(0.0191)

 

70th

-0.664

-0.669

-0.239 [36%]

-0.429 [65%]

0.004 [-1%]

  

(0.0184)

(0.0269)

(0.0201)

 

80th

-0.632

-0.639

-0.221 [35%]

-0.418 [66%]

0.008 [-1%]

  

(0.0218)

(0.0314)

(0.0228)

 

90th

-0.652

-0.633

-0.216 [33%]

-0.418 [64%]

-0.019 [3%]

  

(0.0266)

(0.038)

(0.0267)

 
  1. Notes: Machado and Mata Oaxaca-Ransom decomposition. The bootstrapped standard errors in parenthesis. Bootstrap estimates are based on 800 replications. The differential is calculated by every 5th percentile. The main percentile levels are presented. The specification controls for total number years of education, potential experience, marital status, job tenure, Tajik ethnicity, dummies for female, being in a private sector, rural settlement and occupational dummies.