Is Health Care Spending Impoverishing? Evidence from Micro-lending Clients' Health Care Seeking Patterns in Karnataka, India

Monday, June 13, 2016: 3:00 PM
G65 (Huntsman Hall)

Author(s): Alison B Comfort; Esther Duflo; Abhijit Banerjee; Richard Hornbeck

Discussant: Yulya Truskinovsky

High out-of-pocket health care spending drives a large proportion of individuals into poverty each year in low-income countries. These individuals are particularly vulnerable to financial shocks due to major health events because they lack access to consumption smoothing devices, such as health insurance. Instead, they rely on informal mechanisms, such as drawing on savings and assets or reducing consumption on other essential commodities, which typically provide insufficient financial protection. There is a lack of rigorous evidence assessing whether health care spending can impoverish low-income individuals, after taking into account the role of these coping mechanisms and other factors such as delaying care, altering the choice of providers based on costs and quality, and accounting for lost income due to illness. This study takes advantage of a rich set of data collected from micro-lending clients in rural Karnataka, India, who were offered health insurance as part of a randomized-controlled trial (see Banerjee, Duflo and Hornbeck, 2014). The data cover over 27,000 major health events among micro-lending clients and their households over a 20-month period. The data include the type of health event, whether these individuals sought care and where, what services and treatment they received, how much they spent, and how they financed this care. We present descriptive statistics for these outcomes, estimate the extent of impoverishment from health spending and associated lost income, identify predictors of the major health events and related health spending, and assess whether sources of financing vary by amount of health spending. We also assess whether households that experience a major health event remain impoverished by endline compared to their baseline household consumption. Preliminary results show that almost all micro-lending clients or their household member sought care for a major health event. Most sought care from only one provider, but 12% sought care from two or more. Almost 90% sought care from private rather than public providers, consistent with other evidence from India on type of provider. They spent 789 Rupees (Rs.) at the first provider, of which 55% was on oral medications and 15% on injections. Health care spending was significantly higher at a second subsequent provider (Rs. 1,496) or a third (Rs. 3,029). Overall, health spending on a major health event was Rs. 929, equivalent to 1.1 times per capita monthly household consumption (Rs. 867). Among the top 5th percentile, health spending was Rs. 3,050 or more per health event, representing 3.5 times per capita monthly household consumption. When income lost due to illness is included, a major health event costs households Rs. 1,685, almost two times per capita monthly household consumption and 3.2 times the health insurance premium (Rs. 525) which few clients were willing to pay. Households primarily financed these catastrophic health expenses through savings (40%) and borrowing from family (32%) or moneylenders (22%). Overall, these findings suggest that the micro-lending clients, despite having not enrolled in the health insurance product, are at significant risk of impoverishment as a result of catastrophic health expenses and likely could benefit from health insurance.