Skip to main content

Table 14 Robustness exercise I

From: The impact of Syrian refugees on natives’ labor market outcomes in Turkey: evidence from a quasi-experimental design

Robustness exercise: treatment area vs randomly selected control areas

Refugee effects (R×T) are reported only

Region combinations (control area)

IE/P

LFP

U/P

FE/P

Alternative control area [1]

−0.0216***

−0.0186***

0.0007

−0.0037*

 

(0.0025)

(0.0023)

(0.0014)

(0.0022)

Alternative control area [2]

−0.0321***

−0.0266***

−0.0040***

−0.0016

 

(0.0025)

(0.0023)

(0.0015)

(0.0022)

Alternative control area [3]

−0.0392***

−0.0432***

0.0034**

0.0007

 

(0.0029)

(0.0028)

(0.0016)

(0.0026)

Alternative control area [4]

−0.0243***

−0.0277***

−0.0034**

0.0068***

 

(0.0026)

(0.0024)

(0.0015)

(0.0023)

Alternative control area [5]

−0.0163***

−0.0120***

−0.0009

−0.0035

 

(0.0026)

(0.0025)

(0.0014)

(0.0023)

Alternative control area [6]

−0.0181***

−0.0168***

0.0043***

−0.0057**

 

(0.0027)

(0.0025)

(0.0015)

(0.0024)

Alternative control area [7]

−0.0264***

−0.0222***

−0.0047***

0.0005

 

(0.0023)

(0.0021)

(0.0013)

(0.0020)

Alternative control area [8]

−0.0310***

−0.0299***

−0.0017

0.0006

 

(0.0025)

(0.0023)

(0.0015)

(0.0022)

Alternative control area [9]

−0.0161***

−0.0245***

0.0102***

−0.0018

 

(0.0028)

(0.0027)

(0.0016)

(0.0025)

Alternative control area [10]

−0.0031

−0.0060**

0.0024*

0.0005

 

(0.0027)

(0.0026)

(0.0014)

(0.0023)

Alternative control area [11]

−0.0218***

−0.0247***

0.0028**

0.0001

 

(0.0024)

(0.0022)

(0.0014)

(0.0022)

Alternative control area [12]

−0.0264***

−0.0173***

−0.0051***

−0.0040*

 

(0.0024)

(0.0022)

(0.0013)

(0.0022)

Log trade vol.

Yes

Yes

Yes

Yes

Year fixed effects

Yes

Yes

Yes

Yes

Region fixed effects

Yes

Yes

Yes

Yes

Other controls

Yes

Yes

Yes

Yes

  1. ***, **, and * refer to 1, 5, and 10% significance levels, respectively. Robust standard errors are reported in parentheses. Sample is restricted to the age group 15–64. Controls include gender, marital status, age dummies, education dummies, a full set of age-education interactions, region dummies, year dummies, log trade volumes, and urban versus rural area dummy. This exercise fixes the treatment area and compares it with 12 randomly selected combinations of control areas (excluding the original control area). Based on the NUTS2 region classification in Turkey, the alternative control areas are formed as combinations of 4 NUTS2 regions as follows: [1] 6, 9, 11, 17; [2] 4, 7, 14, 15; [3] 5, 6, 11, 16; [4] 4, 5, 16, 18; [5] 2, 8, 14, 19; [6] 7, 15, 16, 17; [7] 3, 4, 7, 9; [8] 4, 5, 8, 11; [9] 5, 11, 15, 17; [10] 2, 5, 17, 19; [11] 2, 3, 5, 9; [12] 6, 7, 9, 19. See Appendix 2 for the details of NUTS2-level regional division in Turkey
Back to article page