Economic consequences of ill-health for households in northern rural India

Table 6 Determinants of HHs’ borrowing behavior

Variable	Relatives	Friends, neighbours	SHGs	Moneylender	Doctor or hospital
Outpatient_chronic	−0.049*	0.002	0.012	0.045*	−0.006
Outpatient_acute	−0.079***	0.095***	−0.016	−0.029	0.026*
Inpatient	0.065**	−0.177***	0.006	0.124***	−0.122***
Proportion elderly	−0.122	0.193	0.076	−0.113	−0.074
Proportion children	0.029	−0.076	0.012	−0.017	0.037
Proportion reproductive	0.040	−0.075	0.024	−0.065	0.091**
Scheduled	−0.004	−0.043	−0.017*	0.032	0.025*
Hindu	−0.034	0.035	0.016	−0.025	−0.027
SHG	−0.031	−0.047*	0.077***	0.038*	−0.009
Kanpur	0.141***	0.260***	0.025*	−0.302***	−0.055***
Allahabad	0.088***	0.291***	0.070***	−0.417***	−0.017
Medium wealth third	−0.001	0.006	−0.009	−0.002	0.010
Upper wealth third	0.018	0.027	0.005	−0.055*	0.018
Observations	3406	3406	3406	3406	3406
Baseline probability	0.18	0.41	0.06	0.26	0.05
Pseudo R2	0.1469	0.1404	0.0956	0.1546	0.0309

Notes: Marginal effects from probit regressions on each of the main borrowing types. Data is at illness level. Regressions are only run on the sample of illnesses for which money was borrowed. Models account for clustering of observations on the household level and sample weights. Maternity care is the omitted healthcare use category. ***significant at 1%, **significant at 5%, *significant at 10%.

ISSN: 1472-6963