Skip to main content

Table 5 Regression results (2SLS): Obesity, non-high-income countries

From: Is the development of obesogenic food environments a self-reinforcing process? Evidence from soft drink consumption

Dependent variable: Soft drink consumption, QSD Dependent variable: Prevalence of obesity, OBE
Independent variables   Coefficient Std. Error1 Independent variables   Coefficient Std. Error1
Prevalence of obesity OBE 1.776*** 0.270 Soft drink consumption QSD 0.058** 0.029
Soft drink price PSD −0.628* 0.324 Dietary energy supply DES 2.572*** 0.548
Income per capita GNI 2.780*** 0.481 Urban population URB 0.389*** 0.045
Population aged 65 and above AGE −1.652 1.185 Population aged 65 and above AGE 0.825*** 0.183
Economic globalisation GLO 0.213* 0.086 Economic globalisation GLO −0.062** 0.019
N. of obs. 578 N. of obs. 578
F-statistic, F(5, 530) 85.95, Prob. 0.000 F-statistic, F(5, 530) 193.15, Prob. 0.000
Underidentification test 77.41, P-val. 0.000 Underidentification test 31.36, P-val. 0.000
Weak identification test 167.47 Weak identification test 45.41
Sargan-Hansen J statistic 7.303, P-val. 0.007 Sargan-Hansen J statistic 0.164, P-val. 0.685
  1. Notes: Tables 4 and 5. 2SLS: Two-stage least square estimation, with fixed effects
  2. 1 Heteroskedasticity-robust standard errors. *, ** and *** denote p < 0.05, p < 0.01 and p < 0.001, respectively
  3. Underidentification test (Kleibergen-Paap rk LM statistic), Weak identification test (Cragg-Donald Wald F statistic),
  4. Sargan-Hansen J statistic (overidentification test of all instruments)