Health system progress towards financial protection : China's progress and methodological improvements

Abstract

Financial protection (FP) is central to Universal Health Coverage (UHC), ensuring healthcare access without financial hardship (FH) from high out-of-pocket healthcare expenditure (OOPE). Global FP monitoring has revealed a critical trend: China has experienced negative FP progress over the past three decades. This thesis explores the reasons through four empirical studies, examining (1) determinants of FP shifts in China and (2) challenges in FP monitoring methodologies, offering insights for China and global FP monitoring practices.

Chapter 3 conducts a global systematic review (2008-2023, 85 studies, 1,308 analyses) of socioeconomic and health system determinants of catastrophic health expenditure (CATA) and impoverishing health expenditure (IHE), identifying rural residency, older household members, lack of insurance, and healthcare utilization as key FH risk factors. It informs variable selection in Chapter 5, highlights potential data and FP indicator limitations, and motivates Chapters 4 and 6.

Chapter 4 assesses China’s FP progress (1995-2022) using five nationally representative surveys, revealing persistent FH: CATA incidence (7.84%-32.03%), IHE (3.01%-18.77%), and the new official SDG 3.8.2 indicator (referred to as FH indicator, FHI), from 30.68% to 45.84%. Despite post-2017 improvements, pro-poor concentration and provincial disparities persisted. Variations in survey geographical scope, design, OOPE measures, and recall periods caused highly inconsistent FP estimates, underscoring data challenges in FP monitoring. OOPE per capita emerges as a robust monitoring metric.

Chapter 5 examines FP outcome determinants of FP shifts in China (2011-2021) using the Chinese Social Survey (CSS) data. Employing unconditional quantile regression, logistic regression, Recentered Influence Function decomposition, and Oaxaca-Blinder decomposition, it analyzes mean and distributional statistics (30th to 90th quantiles) of OOPE per capita and OOPE as a share of consumption. From 2011 to 2019, the rising OOPE was explained by income growth, population aging, and an increase in the number of young children. From 2019 to 2021, OOPE declines were linked to reduced social health insurance coverage (85.98% to 66.49%) during COVID-19. Heterogeneous quantile effects highlight the value of distributional analysis for OOPE.

Chapter 6 evaluates the performance of the new official SDG 3.8.2 indicator (FHI), which was officially adopted in 2025, against CATA and IHE indicators using Chinese Household Income Project (CHIP) data (1995-2018). FHI shows higher sensitivity to poor households’ FH (concentration index: -0.71 to -0.62) but lower specificity and health shock detection compared to conventional indicators, emphasizing context-driven indicator selection.

This thesis highlights methodological challenges in FP monitoring, showing how data source variations and FP indicator underperformance distort China’s FP narrative. To improve FP monitoring, it recommends selecting appropriate FP indicators based on specific purposes, standardizing survey designs, and including net and gross OOPE in surveys. To strengthen FP, the Chinese government should cautiously interpret CATA and IHE indicator estimates and enhance social health insurance benefits for vulnerable groups, including rural residents, older adults, and low-income households.