Sample balance test.

In the context of increasing population aging and decreasing birth rate, it is of great practical significance to explore the impact of social support on the health of elderly men, which is of great practical significance to smoothly promote the strategy of healthy China and actively implement the strategy of population aging. Using the newly released 2020 China Health and Retirement Longitudinal Study (CHARLS) data for a total of 1,510 study participants, this paper analyzes the impact of social support on the health of elderly men using the Oder Probit, OLogit, and PSM models, and then explores the variability across age stages and literacy levels, and eliminates the propensity score matching modeling by endogeneity problems caused by sample selectivity bias. The study shows that social support significantly improves the health of older men in the context of a deepening aging process, showing positive improvements in both self-assessed health and mental health. The robustness test (by replacing the econometric model) further confirmed the reliability of the findings. Heterogeneity tests, on the other hand, revealed significant differences in the impact of social support on the health of male older adults: its health-enhancing effect was more pronounced in less literate male older adults compared to the more literate group. The results of the endogeneity analysis showed that the PSM model effectively mitigated the endogeneity problem of the model. Failure to deal with endogeneity would lead to underestimation of two aspects: first, the health-enhancing effect of social support on self-assessed health of male older adults; and second, its improvement effect on mental health. Therefore, the Chinese government should pay more attention to the key role of social support in improving the health of male older adults, and take diversified and innovative initiatives to focus on enhancing social support in order to more effectively promote the goal of healthy aging.在人口老龄化加剧与出生率持续走低的背景下，探究社会支持对老年男性健康的影响具备重要现实意义，这对于稳妥推进健康中国战略、积极落实人口老龄化应对战略均具有重大实践价值。本文采用最新发布的2020年中国健康与养老追踪调查（China Health and Retirement Longitudinal Study, CHARLS）数据，共纳入1510名研究对象，运用有序Probit（Ordered Probit）、定序Logit（OLogit）以及倾向得分匹配（Propensity Score Matching, PSM）模型分析社会支持对老年男性健康的影响，并进一步探讨不同年龄阶段与文化水平下的效应异质性，同时针对样本选择性偏差引发的内生性问题，采用倾向得分匹配模型进行修正。研究表明，在老龄化程度不断加深的背景下，社会支持可显著改善老年男性的健康状况，在自评健康与心理健康两方面均呈现出积极的提升效果。稳健性检验（通过更换计量经济学模型实现）进一步验证了研究结论的可靠性。异质性检验则揭示了社会支持对老年男性健康的影响存在显著差异：相较于文化水平较高的群体，社会支持对文化水平较低的老年男性的健康促进效应更为显著。内生性分析结果显示，倾向得分匹配模型有效缓解了模型的内生性问题。若未对内生性问题加以处理，则会导致两方面的效应被低估：其一为社会支持对老年男性自评健康的健康促进效应；其二为其对心理健康的改善效果。因此，中国政府应更加重视社会支持在改善老年男性健康过程中的关键作用，采取多元化、创新性举措着力强化社会支持，以更高效地推进健康老龄化目标的实现。