Skip to main content

Table 2 Logit model - Factors associated with economic support in Tanzania 2016–2017

From: Explaining external economic support inequality among households affected by HIV/AIDS in Tanzania: an Oaxaca Blinder decomposition analysis

Variables

Model 1

Model 2

Model 3

Model 4

Model 5

Economic support

Economic support

Cash transfer

Assistance with school fees

Material support for education

HIV positive

0.209*

0.320***

0.182

0.704*

0.441

(0.107)

(0.110)

(0.348)

(0.370)

(0.338)

Wealth index

−0.629***

−0.817***

−0.567***

−0.470**

−0.349**

(0.053)

(0.059)

(0.177)

(0.235)

(0.159)

Female

0.384***

0.386***

0.380

0.708*

0.401

(0.090)

(0.091)

(0.307)

(0.403)

(0.304)

Urban area

−0.415***

− 0.467***

− 0.781***

− 0.567

− 0.290

(0.089)

(0.093)

(0.272)

(0.368)

(0.274)

Number of children

− 0.017

0.005

0.056

0.148***

0.208***

(0.014)

(0.015)

(0.046)

(0.051)

(0.037)

Education – (No education)

(ref.)

(ref.)

(ref.)

(ref.)

(ref.)

 Primary

−0.025

0.049

0.477*

0.112

0.745**

(0.074)

(0.077)

(0.264)

(0.339)

(0.290)

 Secondary

0.204

0.143

0.938**

0.214

1.034***

(0.129)

(0.133)

(0.420)

(0.590)

(0.392)

 More than secondary

−0.800

−0.665

2.233***

NE

NE

(0.724)

(0.727)

(0.833)

NE

NE

Marital status – (Married)

(ref.)

(ref.)

(ref.)

(ref.)

(ref.)

 Living together

−0.185*

− 0.078

0.137

−0.417

− 0.239

(0.104)

(0.106)

(0.344)

(0.545)

(0.367)

 Widowed

0.373***

0.436***

0.439

0.263

0.320

(0.109)

(0.111)

(0.365)

(0.486)

(0.374)

 Divorced or Separated

0.229**

0.278***

0.494

0.512

0.267

(0.104)

(0.106)

(0.349)

(0.450)

(0.366)

 Age

1.455***

1.411***

1.975***

0.824*

0.852**

(0.106)

(0.109)

(0.380)

(0.492)

(0.359)

 Constant

−7.308***

−6.859***

−11.719***

−7.864***

−9.149***

(0.453)

(0.494)

(1.702)

(2.145)

(1.678)

Number of observation

12,008

12,008

12,008

12,008

12,008

Pseudo r-squared

0.089

0.122

0.081

0.070

0.092

Chi-square

578.74

816.77

95.23

48.85

111.63

Akaike crit. (AIC)

7564.9

7347.36

1164.52

709.3

1168.0

Region effect

No

Yes

Yes

Yes

Yes

Mean dependent var

0.109

0.109

0.01

0.01

0.01

SD dependent var

0.312

0.312

0.09

0.07

0.103

Prob > chi2

0.000

0.000

0.000

0.022

0.000

AUC

0.71

0.75

0.76

0.76

0.77

HL GOF p-valuea

0.216

0.47

0.78

0.88

0.80

  1. Note: Standard errors in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01; Source: Authors calculation based on Tanzania HIV Impact Survey 2016–2017 (THIS) - 2016-2017
  2. Reading: Model 1 is the econometric specifications of economic support, including all explanatory variables except the region effect. Additionally, model 2 included the variable related to the geographical area (region effect) as a control variable. In models 3, 4 and 5, different components of economic support were estimated as dependent variables, respectively, cash transfer, assistance for school fees, material support for education
  3. The AUC represents the classification performance of households with economic support and those without economic support. When the AUC is near “1”, the model has a good measure of separability and “0” for a poor model meaning that it does not have a good measure of separability. “NE” stands for “not estimated” due to the lack of statistical power
  4. aThe Hosmer-Lemeshow goodness of fit test represents the quality of the model’s fitness with the p-value > 0.05; the model fits reasonably well on the validation sample
Back to article page

Health Economics Review

ISSN: 2191-1991

Contact us