Consumption plays a foundational role in economic development, and enhancing resident consumption is one of the key pathways to promote stable growth in consumption. According to the Seventh National Population Census, the proportion of empty-nest elderly households in China reached 43.55%, corresponding to approximately 145 million empty-nest elderly individuals. It is estimated that by 2030, the number of urban and rural empty-nest elderly will rise to 200 million. As the population of empty-nest elderly continues to grow, their consumption will exert an increasingly significant influence on the development of the resident consumption market.

This paper defines empty-nest elderly as single-living elderly and elderly couples whose children do not live with them, and divides the consumption structure of the elderly into four categories: basic living consumption, medical consumption, developmental consumption, and enjoyment consumption. Using five waves of data from the China Health and Retirement Longitudinal Study (CHARLS) covering 2011 to 2020, this study systematically analyzes the development trend of consumption structure and its influencing factors for empty-nest elderly in urban and rural areas in China. In order to distinguish individual consumption from household consumption, this study first employs the least squares method, following the analytical frameworks of Mankiw and Weil (1989) as well as Mi and Ren (2014), controlling for household size and structure, age composition of household members, and household income to estimate individual consumption expenditures for elderly couples living without children and the elderly co-residing with family members. Second, the Hierarchical Age-Period-Cohort (HAPC) model is applied to analyze the changing trends in consumption structures among urban and rural empty-nest elderly. Finally, the Fractional Probit regression model is utilized to examine the determinants influencing the consumption structure of the empty-nest elderly.

The result reveals that the consumption structure of urban and rural empty-nest elderly shows significant age, period, and cohort trends. Young-old and oldest-old empty-nest elderly have higher basic living consumption and lower medical consumption, and middle-old elderly have lower basic living consumption and higher medical consumption. Over the long term, the basic living consumption of the empty-nest elderly has decreased, while developmental consumption and enjoyment consumption have increased. And the medical consumption of the elderly couples and urban single-living elderly has decreased, while the medical consumption of rural single-living elderly has increased. Empty-nest elderly from earlier birth cohorts have higher basic living consumption and lower medical consumption; urban empty-nest elderly couples from later birth cohorts have higher developmental consumption, while the cohort difference in developmental consumption among rural empty-nest elderly is minimal. The age effect of developmental and enjoyment consumption, and the cohort effect of enjoyment consumption among empty-nest elderly are not significant. The economic ability, health status, social security, cognitive ability, and internet ability of empty-nest elderly all have significant impacts on their consumption structure through the social participation path.

The study suggests that increasing the economic income, advocating emphasis on disease prevention and health management, expanding social participation, advancing age-friendly internet adaptation, and devoting to strength digital skills training for the elderly will help improve and upgrade the consumption structure of empty-nest elderly.