Feminization of poverty: an analysis of multidimensional poverty among rural women in China

The Sustainable Development Goals highlight the continued prevalence of extreme poverty and gender inequality, with women disproportionately affected. In developing countries, female-headed households experience significantly worse well-being than male-headed ones. This is due to factors like traditional gender roles, time poverty, and limited access to resources like education, employment opportunities, financial capital, and adequate living conditions. The feminization of poverty is a global concern, and China, despite significant poverty reduction efforts, still faces this challenge. Official data reveals a higher poverty incidence among women in poverty-stricken counties. While China eradicated absolute income poverty, multidimensional poverty (MP) remains a persistent issue, particularly affecting rural women who may experience a double disadvantage due to both gender and geographic location. This study aims to comprehensively measure and analyze the multidimensional poverty of rural women in China, using a six-dimensional system based on the Alkire-Foster (A-F) approach and the China Family Panel Studies (CFPS) data from 2010 to 2020. The study seeks to understand the current state of rural women's MP, its spatial and temporal characteristics, and potential strategies for reduction.

Traditional poverty assessments focused on income inadequacy, but the concept of MP recognizes poverty's multidimensional nature. Amartya Sen's capability approach emphasizes functioning and capability deprivation as key aspects of poverty, extending beyond mere income. The World Bank's recognition of MP as encompassing malnutrition, poor health, lack of opportunity, security, and empowerment further broadened this understanding. Several methods exist for measuring MP, including the A-F method, which is favored for its rigorous statistical foundation, flexibility, sensitivity, transparency, and incorporation of stakeholder perspectives. Existing studies on MP in China often focused on households rather than individuals, neglecting gender-specific resource allocation differences. Many also lacked longitudinal data, limiting insights into the dynamic nature of poverty. This study addresses these limitations by using individual-level data and a longitudinal approach, adjusting the measurement dimensions to reflect the specific context of rural women in China, aligning with the OPHI's recommendation for country-specific methodologies and incorporating factors like subjective well-being, social welfare, and income.

This research utilizes the CFPS dataset, a longitudinal survey covering various aspects of Chinese population well-being. The study focuses on rural women aged 16-97, drawing on data from 2010, 2012, 2014, 2018, and 2020. Data cleaning involved handling missing values on a case-by-case basis, leading to a final sample size of 133,557 across the five waves. A six-dimensional system for measuring MP was developed, encompassing income, health, education, social welfare, living standards, and subjective well-being. A total of 12 indicators were used, each converted into binary variables (deprived or not deprived) using specific cutoff values. These include household net income per capita compared to national poverty lines for income; years of schooling compared to the nine-year compulsory education standard for education; BMI, chronic diseases, and self-rated health for health; medical and pension insurance coverage for social welfare; housing asset ownership, access to clean drinking water, and cooking fuel for living standards; and life satisfaction and depression scores for subjective well-being. The Alkire-Foster (A-F) dual-cutoff approach, which includes assigning equal weights to dimensions and setting a deprivation cutoff (k) of 0.3, was used to identify and measure multidimensional poverty. The headcount ratio (H), average deprivation share (A), and multidimensional poverty index (MPI or M0) were calculated. Spatial comparisons were made across three regions (Eastern, Central, and Western), and temporal analyses tracked changes from 2010 to 2020. Additionally, the study classified individuals into three poverty groups (never poor, temporarily poor, and chronically poor) based on the duration of poverty (t). Sensitivity analyses were performed to assess the robustness of the results to changes in k and t.

The study reveals a high prevalence of multidimensional poverty among rural women in China. Compared to other subgroups, rural women exhibited a significantly higher risk of MP. Key findings include: **Demographic Differences:** Older rural women, those without spouses, and those lacking confidence in the future faced a significantly higher risk of MP. The deprivation rate of schooling, pension insurance exceeded 85% for older rural women. **Spatial Differences:** The Western Region showed the highest prevalence of MP among rural women, significantly higher than the Central and Eastern Regions. The Western Region also demonstrated higher deprivation rates in aspects like cooking fuel and clean water access. The Eastern Region, while showing lower overall MPI, had higher deprivation rates in medical and pension insurance and housing assets, potentially due to high population density and resource competition. **Temporal Differences:** Rural women's MPI showed a significant downward trend from 2010 to 2020. Deprivation rates decreased in several indicators (income, nutrition, medical insurance, pension insurance, cooking fuel, and life satisfaction), reflecting improvements in economic conditions and social welfare. However, other indicators (education, self-rated health, chronic diseases, drinking water, housing assets, and depression) showed increased deprivation rates. The contribution of non-material indicators (education, health, and subjective well-being) to MPI increased over time, suggesting a shift in priorities as absolute poverty is reduced. **Indicator Contributions:** The education dimension consistently contributed most significantly to MPI across all subgroups and regions. Social welfare and subjective well-being also played substantial roles, with subjective well-being having a particularly large contribution for women lacking confidence in the future. **Poverty Dynamics:** Around 74% of rural women experienced either temporary or chronic poverty (t=2). The proportion of chronically poor women was highest in the Western Region, and the proportion of never-poor women was highest in the Eastern Region. **Sensitivity Analysis:** The findings remained robust to variations in both the deprivation cutoff (k) and time threshold (t), although changes in k above 0.3 had a significant effect on the indices.

The findings confirm the existence of significant feminization of poverty in rural China, consistent with previous research. Patriarchal culture and social exclusion limit opportunities and resources for women, contributing to their increased vulnerability. The study's findings highlight the importance of considering various factors in addressing rural women's MP. Ageing poses challenges related to declining physical functioning; marital status impacts financial and emotional support; and confidence in the future influences motivational behaviors. Spatial disparities reflect environmental and socio-economic conditions, while temporal trends indicate that while poverty has decreased overall, improvements in certain areas are lagging and new challenges are emerging. The increasing importance of non-material factors suggests a shift in focus towards enhancing well-being as absolute poverty is addressed.

This study, using the A-F method on CFPS data, comprehensively examines rural women's MP in China. It highlights the significant vulnerability of rural women, particularly older women, those without spouses, and those lacking confidence. Spatial disparities exist, with the Western Region having the highest prevalence. While poverty rates decreased overall between 2010 and 2020, attention should focus on addressing persistent deprivations in education, health, and subjective well-being, especially in the Western Region. Future research should utilize primary data for better timeliness and explore the causal mechanisms in greater depth. Policy implications include adopting a gender perspective, tailoring regional interventions, and establishing dynamic tracking systems.

The study uses secondary data from CFPS, potentially limiting timeliness and depth of analysis. The findings are country-specific to China and might not generalize perfectly to other contexts. The study focuses on measuring MP but does not delve into the detailed causal mechanisms underlying the observed disparities. Future research could address these limitations through primary data collection, cross-national comparisons, and more in-depth analyses of causal factors.