Document Type : Original Article
Ach Medical University, Ulaanbaatar, Mongolia
Institute of Public Health, National Yang Ming Chiao Tung University, Taipei, Taiwan
School of Medicine, Etugen University, Ulaanbaatar, Mongolia
Mongolian Institute of Certified Public Accountants, Ulaanbaatar, Mongolia
National Statistics Office of Mongolia, Ulaanbaatar, Mongolia
High out-of-pocket (OOP) health expenditures are a common problem in developing countries. Studies rarely investigate the crowding-out effect of OOP health expenditures on other areas of household consumption. OOP health costs are a colossal burden on families and can lead to adjustments in other areas of consumption to cope with these costs.
This cross-sectional study used self-reported household consumption data from the nationally representative Household Socioeconomic Survey (HSES), collected in 2018 by the National Statistical Office of Mongolia. We estimated a quadratic conditional Engel curves system to determine intrahousehold resource allocation among 12 consumption variables. The 3-stage least squared method was used to deal with heteroscedasticity and endogeneity problems to estimate the causal crowding-out effect of OOP.
The mean monthly OOP health expenditure per household was ₮64 673 (standard deviation [SD] = 259 604), representing approximately 6.9% of total household expenditures. OOP health expenditures were associated with crowding out durables, communication, transportation, and rent, and with crowding in education and heating for all households. The crowding-out effect of ₮10 000 in OOP health expenditures was the largest for food (₮5149, 95% CI = -8582; -1695) and crowding-in effect was largest in heating (₮2691, 95% CI = 737; 4649) in the lowest-income households. The effect of heating was more than 10 times greater than that in highest-income households (₮261, 95% CI = 66; 454); in the highest-income households, food had a crowding-in effect (₮179, 95% CI = -445; 802) in absolute amounts. In terms of absolute amount, the crowding-out effect for food was up to 5 times greater in households without social health insurance (SHI) than in those with SHI.
Our findings suggest that Mongolia’s OOP health expenses are associated with reduced essential expenditure on items such as durables, communication, transportation, rent, and food. The effect varies by household income level and SHI status, and the lowest- income families were most vulnerable. SHI in Mongolia may not protect households from large OOP health expenditures.