Skip to main content

Table 4 Difference generalized method of moments (GMM) estimation results

From: Growth, inequality and poverty dynamics in Mexico

Left-hand-side variable

System of equations comprising (1), (2) and (3)

Fixed-effect estimation of Eq. (4)

Difference GMM estimation

Fixed-effect est.

Consumption (t)

Gini (t)

Poverty (t)

 

Poverty (t)

Specification A: with year effects only

 Consumption (t − 1)

0.4226***

(0.0971)

− 0.1097***

(0.0221)

− 0.1287***

(0.0356)

Consumption (t)

− 0.4171***

(0.0157)

 Gini (t − 1)

− 1.7569***

(0.2277)

0.1411*

(0.0784)

0.2846**

(0.1125)

Gini (t)

1.0689***

(0.0494)

Wald χ2-test (Chi2(12))

379.57***

103.39***

   

No. of instruments

67

67

   

Arellano–Bond test (Order 2)

0.1642

0.7957

   

Sargan test (54)

0.9536

0.8414

   

R-squared

  

0.61

 

0.89

Specification B: with year effects and controls

 Consumption (t − 1)

0.4818***

(0.1187)

− 0.1048***

(0.0260)

− 0.1229***

(0.0409)

Consumption (t)

− 0.4349***

(0.0188)

 Gini (t − 1)

− 1.7196***

(0.2335)

0.1326

(0.0839)

0.2815**

(0.1058)

Gini (t)

1.0610***

(0.0438)

 Education (t − 1)

− 0.4117

(0.2881)

0.0140

(0.0846)

0.1221

(0.1285)

Education (t)

0.1576***

(0.0370)

 Urban (t − 1)

− 0.1266

(0.1230)

− 0.0226

(0.0487)

0.1223*

(0.0681)

Urban (t)

0.1637***

(0.0201)

 Tertiary sector (t − 1)

− 0.3273

(0.2193)

− 0.0302

(0.0784)

− 0.1577

(0.0886)*

Tertiary sector (t)

0.0034

(0.0419)

 Aged (t − 1)

− 1.1168

(0.7447)

0.0603

(0.2920)

0.3138

(0.2742)

Aged (t)

− 0.0022

(0.1382)

 Intercept

5.2651***

(0.8470)

1.2681***

(0.1898)

1.0952***

(0.3002)

Intercept

3.1240***

(0.1331)

Wald χ2-test (Chi2(16))

431.21***

144.99***

   

No. of instruments

71

71

   

Arellano–Bond test (Order 2)

0.1753

0.8226

   

Sargan test (54)

0.7747

0.8315

   

R-squared

  

0.63

 

0.91

  1. The number of observations is 320 for the difference GMM estimation, 352 for Eq. (3) and 384 for Eq. (4)
  2. Robust standard errors, adjusted for clustering on state, are shown in parentheses
  3. ***, **, and * represent significance at the 1, 5, and 10% level, respectively