The Role of Gender Inequality in the Obesity Epidemic: A Case Study from India

Figures & data

Table 1. Regressions of female bodyweight on gender inequality and covariates—whole sample

	BMI (kg/m²)		Underweight (dummy)		Overweight (dummy)
	Fixed Effects	Hausman-Taylor	Fixed Effects	Hausman-Taylor	Fixed Effects	Hausman-Taylor
Married when minor (dummy)		−0.140*** (0.057)		0.021*** (0.007)		0.002 (0.007)
Observations		40,746		40,746		40,746
Number of households		23,087		23,087		23,087
Within R-square		0.045		0.011		0.025
Between R-square		0.217		0.095		0.156
Overall R-square		0.189		0.075		0.126
Age difference with husband (in years)		0.053*** (0.010)		−0.002** (0.001)		0.005*** (0.001)
Observations		39,904		39,904		39,904
Number of households		22,375		22,375		22,375
Within R-square		0.046		0.011		0.026
Between R-square		0.220		0.096		0.159
Overall R-square		0.191		0.075		0.128
Restricted mobility index (0-to-3 score)	−0.019 (0.018)	−0.023 (0.018)	−0.002 (0.002)	−0.001 (0.002)	−0.004** (0.002)	−0.005*** (0.002)
Observations	40,720	40,720	40,720	40,720	40,720	40,720
Number of households	23,082	23,082	23,082	23,082	23,082	23,082
Within R-square	0.053	0.045	0.012	0.010	0.031	0.025
Between R-square		0.217		0.095		0.155
Overall R-square		0.189		0.075		0.126
Decision-making index (0-to-5 score)	0.052*** (0.018)	0.061*** (0.018)	0.000 (0.002)	0.000 (0.002)	0.008*** (0.002)	0.009*** (0.002)
Observations	40,702	40,702	40,702	40,702	40,702	40,702
Number of households	23,076	23,076	23,076	23,076	23,076	23,076
Within R-square	0.052	0.045	0.012	0.010	0.031	0.026
Between R-square		0.217		0.095		0.156
Overall R-square		0.190		0.075		0.126
Community acceptance of domestic violence (0-to-5 score)	0.067*** (0.013)	0.059*** (0.013)	−0.004*** (0.002)	−0.003** (0.002)	0.005*** (0.002)	0.004** (0.002)
Observations	40,536	40,536	40,536	40,536	40,536	40,536
Number of households	23,050	23,050	23,050	23,050	23,050	23,050
Within R-square	0.052	0.045	0.012	0.010	0.031	0.025
Between R-square		0.216		0.095		0.156
Overall R-square		0.189		0.075		0.126
Control variables	Yes	Yes	Yes	Yes	Yes	Yes
Individual FE	Yes	No	Yes	No	Yes	No
State FE	No	Yes	No	Yes	No	Yes

Notes: Robust standard errors in parentheses: ***p < 0.01, **p < 0.05, *p < 0.1. Standard errors were clustered at a household level for robustness.

Source: Authors’ calculations from IHDS I-II database.

Standard errors were clustered at a household level for robustness. Since HT model specification does not report R-squared values, we computed those obtained in a random model specification.

Table 2. Hausman-Taylor regressions of female BMI on gender inequality and covariates—by subsamples

Dep. Var.: Female BMI (kg/m²)	Rural	Urban	Poor	Not poor
Married when minor (dummy)	−0.128* (0.066)	−0.182 (0.111)	0.274 (0.209)	−0.180*** (0.071)
Observations	27,365	13,381	7,102	33,629
Number of individuals	15,585	7,502	6,080	20,984
R-squared^a
- Within	0.040	0.060	0.048	0.045
- Between	0.172	0.140	0.099	0.209
- Overall	0.149	0.126	0.095	0.187
Age difference with husband (in years)	0.050*** (0.012)	0.061*** (0.020)	0.015 (0.034)	0.052*** (0.013)
Observations	26,825	13,079	6,930	32,960
Number of individuals	15,117	7,258	5,912	20,399
R-squared^a
- Within	0.041	0.061	0.048	0.046
- Between	0.175	0.145	0.101	0.211
- Overall	0.151	0.129	0.097	0.188
Restricted mobility index (0-to-1 score)	−0.015 (0.022)	−0.040 (0.033)	−0.041 (0.070)	−0.032 (0.022)
Observations	27,346	13,374	7,099	33,606
Number of individuals	15,579	7,503	6,075	20,978
R-squared^a
- Within	0.040	0.059	0.047	0.045
- Between	0.172	0.140	0.099	0.210
- Overall	0.149	0.126	0.095	0.187
Decision-making index (0-to-1 score)	0.057*** (0.021)	0.070*** (0.034)	0.084 (0.066)	0.069*** (0.022)
Observations	27,342	13,360	7,097	33,590
Number of individuals	15,580	7,496	6,073	20,964
R-squared^a
- Within	0.040	0.059	0.049	0.045
- Between	0.172	0.140	0.099	0.210
- Overall	0.149	0.127	0.095	0.188
Community acceptance of domestic violence (0-to-1 score)	0.046*** (0.015)	0.086*** (0.024)	0.053 (0.046)	0.060*** (0.015)
Observations	27,238	13,298	7,064	33,457
Number of individuals	15,555	7,495	6,043	20,937
R-squared^a
- Within	0.040	0.060	0.050	0.045
- Between	0.172	0.139	0.100	0.209
- Overall	0.149	0.126	0.096	0.187
Control variables	Yes	Yes	Yes	Yes
State FE	Yes	Yes	Yes	Yes

Notes: Robust standard errors in parentheses: ***p < 0.01, **p < 0.05, *p < 0.1. Standard errors were clustered at a household level for robustness.

Source: Authors’ calculations from IHDS I–II database.

^aSince HT model specification does not report R-squared values, we computed those obtained in a random model specification.

Table A1. Summary statistics

	Married under 18 (binary)		Age difference with husband (years)		Decision-making power (0–1 score)		Restricted mobility (0–1 score)		Community acceptance of violence (0–1 score)		Overweight prevalence (%)		Obesity prevalence (%)
	2004–2005	2011–2012	2004–2005	2011–2012	2004–2005	2011–2012	2004–2005	2011–2012	2004–2005	2011–2012	2004–2005	2011–2012	2004–2005	2011–2012
All subsample	0.52	0.51	5.28	5.40	1.45	1.44	1.94	1.99	2.09	2.45	0.16	0.27	0.03	0.07
Age
18–30	0.52	0.55	5.11	5.24	1.22	1.27	2.06	2.08	2.16	2.54	0.09	0.17	0.02	0.03
31–40	0.53	0.50	5.33	5.24	1.55	1.44	1.90	2.02	2.09	2.45	0.18	0.26	0.04	0.06
40–50	0.51	0.51	5.52	5.62	1.70	1.51	1.81	1.97	1.95	2.43	0.24	0.31	0.05	0.08
p-Value (χ^2)	0.006	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.039	0.000	0.000	0.000	0.000
Urban
Urban	0.41	0.39	5.59	5.68	1.59	1.52	1.80	1.98	1.79	2.25	0.27	0.42	0.06	0.12
Rural	0.58	0.57	5.15	5.28	1.39	1.41	2.01	2.00	2.22	2.53	0.11	0.20	0.02	0.04
p-Value (χ^2)	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Education
None	0.68	0.68	5.06	5.14	1.42	1.42	2.03	2.02	2.31	2.59	0.10	0.18	0.02	0.04
1–4	0.57	0.57	5.71	5.71	1.49	1.48	1.91	1.92	2.27	2.49	0.14	0.22	0.03	0.06
Primary	0.54	0.52	4.99	5.16	1.44	1.47	1.98	2.04	1.97	2.36	0.19	0.30	0.04	0.08
6–9	0.42	0.42	5.64	5.87	1.46	1.46	1.90	1.97	1.96	2.40	0.19	0.32	0.04	0.08
Secondary	0.25	0.24	5.64	5.72	1.52	1.40	1.82	1.99	1.76	2.30	0.26	0.41	0.06	0.13
Higher secondary	0.14	0.16	5.61	5.57	1.42	1.47	1.78	1.94	1.57	2.11	0.26	0.42	0.05	0.12
Graduate 15+	0.05	0.06	4.84	4.90	1.61	1.57	1.62	1.90	1.30	1.96	0.36	0.51	0.09	0.15
p-Value (χ^2)	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Per capita income
1st tercile	0.64	0.62	5.12	5.12	1.37	1.37	2.05	2.01	2.43	2.61	0.08	0.16	0.01	0.03
2nd tercile	0.56	0.56	5.25	5.40	1.43	1.46	2.00	2.01	2.19	2.49	0.12	0.23	0.02	0.05
3rd tercile	0.41	0.40	5.42	5.60	1.53	1.48	1.83	1.97	1.77	2.29	0.24	0.38	0.05	0.10
p-Value (χ^2)	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Poverty line
Poor	0.64	0.61	5.04	5.11	1.34	1.40	2.01	2.00	2.19	2.55	0.08	0.14	0.01	0.03
No poor	0.49	0.50	5.35	5.46	1.48	1.45	1.93	1.99	2.06	2.42	0.18	0.29	0.04	0.08
p-Value (χ^2)	0.000	0.000	0.000	0.000	0.000	0.197	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Caste
High caste	0.42	0.40	5.32	5.56	1.45	1.42	1.82	1.94	1.82	2.23	0.22	0.35	0.05	0.10
OBC	0.54	0.55	5.39	5.44	1.43	1.43	1.96	2.04	2.18	2.55	0.13	0.24	0.03	0.06
SC/ST	0.60	0.60	5.17	5.28	1.49	1.44	1.98	2.03	2.22	2.53	0.11	0.22	0.02	0.05
Muslim (w/o caste)	0.57	0.54	5.47	5.57	1.40	1.49	2.02	1.96	2.20	2.60	0.18	0.32	0.04	0.08
Christian, Sikh, Jain (w/o caste)	0.15	0.10	4.44	4.59	1.49	1.46	1.78	1.94	1.52	2.07	0.36	0.51	0.09	0.17
p-Value (χ^2)	0.000	0.000	0.000	0.000	0.000	0.001	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000
Employed
Yes	0.61	0.58	5.25	5.31	1.54	1.46	2.01	2.01	2.29	2.58	0.10	0.19	0.02	0.04
No	0.46	0.46	5.31	5.48	1.39	1.43	1.90	1.98	1.95	2.33	0.21	0.34	0.04	0.09
p-value (χ^2)	0.006	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.000

Notes: Descriptive statistics for our sample: non-pregnant ever-married women between 18 and 49 years old in 2004–2005. Pearson’s chi-square test results are presented (H0 is the independence hypothesis across groups).

Source: Authors’ calculations from IHDS I-II database.

Data availability statement

This database is publicly available, anonymous, and does not contain personal information. Please see the website for more information about the data: https://ihds.umd.edu. The authors are responsible for all remaining errors. Data and codes are publicly available in the following data repository: Alvarez Saavedra, Levasseur, and Seetahul (2022).